Documenti di Didattica
Documenti di Professioni
Documenti di Cultura
"
OPEN-CHANNEL
)
I
)
i
I
I
'{
r
i
I
HYDR~AULICS
'.
'T'1::'
U.E.T
(,
'i
l:
London
of
~anama
Sydney Toront9
KOGAKUSHACOMPANY, LTD.
.Tokyo
"
-.
, ;
.;
"
1fi
.I
-!
.1
},,
i
ce
{.
OPEN-CHANNEL HYDRAULICS
:i
6"it
To
Exclusive rights by Kogakusha Co., Ltd., for manufacture and export from
_. Japan. This book cannot be re-exported from the cou\ltry to which it is
consigiled by Kogakllsha Co., Ltd., or by McGraw-Hill Boak Company,
Inc., or any .of its subsidiaries.
~.
~.
~.
VIII
\
-.
o.
ttJe
',!!-
(;;.
'toano
PR,IN'tlNG
co'.
,
,
"t,
~i;"
~.
l~~
...
.j.
1
i
@-,-
I
1,
(~
tm
;.,
~
"
~
,,
(
\
~.
I
Gt.
':
i,
f?\'
t
.fi
r
t\at
eie
.}
',1
)
\
.,0.
de
Ii'
..h.
/;"
"
fftJ
~~'
r
, \1
I
)
\
.....
/i['
' Y'~:i' .
;f
~.
.,l
viii
PREFACE
T
i
i
to
l'REFACE
ix
con
(~
?4
it IS
ther
la~
COInj
in rn
a~ I
shou;
Tjpj
hel
. riar
ail'!
POl~ta;
in tJu
I) ,
RuL
i.i
I . Ii
.x
...
. PREFACE
i'
I
J'" ..
~,
.)~
on,
t.::,
e)
, r
hJ
;
.t
do
'-1
o ~lC
1an(
f-
)
npu3+.;on
~. ~'st
\
...
(I
I
{J'
I
(1,
I
tL,
(~
1
I
PREF~CE
..
xi
-,
.
l
xii
PREFACE'
University of Notre Dame, for h.is readi~g of the entire manuscript and
his valuable suggestions, Dr. Kolupaila' also helped in interpreting and
collecting information from the hydraulic literature written ill Russian,
Polish, Lithuanian, and several. other languages which are unfamiliar to
theauthol', . The author also wishes to express his warm gr[l.titucie t,o
those whQ have constantly shown a keen intei'est ill .his work, as this
iuterest lent a strong impetus toward the completion. of this volume.
CONTENTS
Ye.nTe Chow
Prefacs
.
PART I,
Chapter 1.
BASIC PRINCIPLES
1-1. D~scriptiori .
1-2. Types 'Of Flow .
1-3. State Qf FlQW
14. Regimes Qf Flow .
Chapter 2.
2-6.
27.
28,
2-9.
210.
,,
/.
",.
HI
19
Channel Geometr)~ .
Geometric Elements of Cha.nnel Section,
Velocity Distribution in a. Channel Section.
Wide Open Channel
Measurement of Velocity ,
Velodty-dbtribution Coeflicienu, .
DeterminaHon of Velocity-distribution Coefficients
PrllS<!ure Dh;tribution in a Channel Section.
Effect of Slope on Prcssui'e Distribution.
Chapter 3.
3
3
2-2,
2-3,
2-4,
2-5.
."
vii
20
22
,J ' ..
~J
24
26
27
27
28
30
32
39
39
41
42
43
46
49
5~
56
63
()
6(1
64
5f!
59
70 r )
74
xiii
\
I
xiv
CO!-T'l'ENTS
PART 11.
Chapter 5.
.\
i~
1....1
i
I()
I
-'
81l
89
89
91
93
94
98
10l
101
108
114
128
,12.8
131
134
136
140
142
144
146
148
157
A. NONERODIBLE CHANNELS
157
157
" )
158
158
B,
159
160
162
Do
NOT SILT
164
165
C. GRA8SEDCHANNELS
7-16. The Gra.ssed. Channel .
7-17. The Retardance Coefficient
7-18. The Permissible Velocity
167
168 .
170
172
175
liB
179
.179
184
192
194
128
6-1.
6-2.,
6-3.
6-4.
6-5.
6-6.
6-7.
184
184
UNIFOR.M FLOW
xv
CONTENTS
198
200
202
205
206
210
Chapter 10.
Methods of Computa.tion
101.
10-2.
10-3.
10-4,
10-5.
Cb,apter 11.
Practica.l Problems .
217
. ~ -i :
217
218
222
227
232
237
242
249
249
252
262
265
268
274
280
284
297
297
302
303
306
307
310
.317
319
320
321
327
327
329
..
l..i
12-3.
12-4.
12-5.
12-6.
Chapter 13.
IntroducUon.
333
341
346
347
357
3.57
'357
360
360
362
363
365
368
Chapter 17.
J(
xvii
CONTENTS
CONTENTS
xvi
461
'--:,....
17-1.
17-2.
17-3.
17-4.
1t-5.
17-6.
17-7.
17-8.
17-9.
17-10.
17-11.
17..12.
17-13.
17-14.
Sudden Transitions,.
Subcritica.l Flmv through Sudden Transitions
Contractions in Supercritical Flo'''''
Expansions in Supercritical Flow.
Constrictions
Sl!bcritical Flow through Constrictions
Backwater Effect due to Constriction
Flow through Culverts.
Obotructions.
ro'loW' between Bridge Piers
Flow through Pile Trestles
Flow through Trash Racks
Underflow Gates
Channel Junctions
461
464
468
470
475
476
490
493
499
501
506
506
507
512
\...-
.',
''''L ..
,
'-...-
\.,
(-
~70
370
374
380
382
384
385
393
3113
395
395
396
398
399
399
404
408
412
414
415
417
422
423
425
429
439
439
439
441
444
448
455
456
PART V.
UNSTEADY FLOW
l~.
Chapter 18.
18-1.
18-2.
18-3.
18-4.
18-5.
18-6.
18-7.
18-8.
525
Chapter 19.
.(
(
'--'
l_
554
,~
525
626
528
631
533
537
640
543
..
Flood Routing
-.'
554
557
559
566
568
572
575
578
580
C
'-...,'
l.
I..
i, __
586
(
20-1.
20-2.
20-3.
20-4.
20-5.
20-6.
Ii
,.
Routing of Flood
Method of Characteristics.
Method of Diffusion Analogy.
Principle of Hydrologic Routing
Methods of Hydrologic Routing
A Simple HydrologIC Method
Routing
of
"~._~ ' _ _ _ _ V~
y.'.----.-... ---
586
587
601
604
607
609
'-
\
~,
i
,--
CHAPTER
,1
1-1. Description.
The flow of water in a conduit may be either openThe two kinds of flow are similar in many ways
but differ in ohe important respect. Open-channel flow must have a
free surface, whereas pipe flow has none, since the water must fill the
whole conduit. A free surface is subject to atmospheric pressure. _:PiR~
, flow, being connnedin a .closed "Q~1duit, x~rts
irect atl!lo, heri
pressure but hydraulic pressure only.
., , "--,, ,,The t\VO kinds of flow are compared in Fig. 1-1. Shown on the left
side is pipe flow. Two piezometer tubes are installed on the pipe at'
sections 1 and 2. ,The water levels in the tubes are maintained by the
pressure in the pipe at elevations represented by ~he so-cag~a.lYJdr,guliq
g.rade line. The pressure exerted
by the water in each section of the
.
pipe is indicated in the corresponding tube by the height y of the watet
c;olumn above the center line of the pipe. The tIJt.al energy in the flow
of the section with reference to a datum line is the sum of the elevation z
of the pipe-center line, the piezoine'tric height y, and the velocity head
V2/2g, where V is the mean velocity of flow. l The energy is represented
in the figure by what is called the energy grade line or simply the energy
line. The loss of energy that results when w:ater flows from seqtion 1
to sectiori 2 is .represented by hi. A similar diagram for open-channel
flow is shown on the right side of Fig. 1-1. For simplicity, it is assumed
. that the flow is parallel and has a uniform velocity distribu,tion and that
the slope of the channel is, small. lP,..i,his c!1;,e, the 'Y~~!t~ ~rj~ce is ,the
.hldraulic ml~dine, and ~~~.1t.Qf..1~j~OIUjUg",,~
piezometric h~@t.2
'.,
.
.
~ Despite the similarity between the two kinds of flow, it is much more
cl.ifficult to solve problems of flow in open channels than.in pressure pipes.
~low conditions in open c~annels are complicated Py the ',fact that the
channel flotu or 'Pip~ flow.
---'-->
It is here assumed that the velocity is uniformly distrib\.ited across the conduit
othenvise a correc,tion ,,",ould have to be made, such Ilf is desaribed in Art. 2-7
fbr open c h a n r r e l s . !
,
: 2 If the flow were curvilinear or if the slope of the channel w~re large, the piezometric
height would be appreciably different from the depth of flow (l<ilts. 2-9 and 2-10). As
a result, the hydraulic grade Une would not coincide exactly with the water surface.
-"1"'-
: 1
s~ction;
;~
I" .,
,,--.. :
'-....
"
.~
xviii
CONTENTS
625
629
Appendix C.
640
Appendix A.
, Appendix D.
F(u.,N)
Appendix E.
r " I-UN
dlt
= Jo
and
F(u,N)_s,
r"
dtl
)0 l+'uN
,,
641
'~
PART I
BASIC PRINCIPLES
657
Name Index.
663
. Subiect Index
669
,- .
""
BASIC PRINCIPLES
position of the free surfa.ce is likely to ch~nge with respect to time .and
space and also by Lhe fact that the depth of flow, the discharge, and the
slopes of the channel bottom and of the free surface are interdependent.
Reliable experimental data. on flow in open channels are usually difficult to
obtain. Furthermore, the physical condition of open channels varies
-'much more widely than that of pipes. In pipes the cross section of flow
is fixed, since it is completely defined by the geometry of the .conduit.
; _I
The cross section of a pipe is generally round, but that of an open channel
may be of any
the circular to the irregular forms of natural
streams.' In pipes, the interior Surface ordinarily ranges in roughness
'VA
(I-I)
where V is the mean velocity and A is the flow cross-sectional area normal
to the direction of the flow, since the mean velocity is defined as the
discharge divided by the cross-sedional area.
In most problems of steady flow the discharge is constant throughout
the reach of the channel under cOllsideration;in other words, the flow is
continuous. rrhus, using
(1-1),
(1-2)
Pipe flow
Open-chann,1 flow
from that of new smooth brass or wooden-stave pipes, on the one hand,
t,o.that of old corroded iron or steel pipes, on the other. In open channels
the surface varies from tha.t of the polished metal used in testing flumes
to that of rongh inegular river bed'l. :.vloreover, the roughness in an
open channel varies with the position of the free surface. Therefore, the
selection of fdction coeffic~el1tS is attended by greater uncertainty for
~ open channels than for pipel$. In general, the treatment of open-channel
flow is 30mew~at morl:l~Ip.Eirical than that of pipe flow. The empirical
) method is the best available at present and, if cautiously applied, can
yield results of practical value.
The flow in a. closed conduit is not necpssarily pipe flow. It must be
" classified as open-channel flow if it has a free surface. The storm sewer,
. .1
for example, which is a closed conduit, is generally designEld for openchannel flow because the flow in the sewer is ekpected to maintain a free
surface most of .the .time.
'.
1-2. Types at Flow. Open-channel fiow call be classified into many
'\ types and descri:be'd in various ways, The following classification is made
according to the change in flow depth with respect to time and space.
...
.P. ....
I,
i
.~"
BASIC PRINCIPLES
is a practically
impossible condition. The term uniform flow" is ,
,
therefore, used hereafter to refer only to steady uniform flow.
Flow is varied if the depth of flow changes along the length of the
channel. Varied flow may be either steady or unsteady. ~nce unsteady
2l~fI.~~e, the t~,rm II unsteady flow" is used hereafter to designate
unsteady varied flow exclusively:-'-.
._ ...
Varied. flow roay be furt~; da.ssified as either rapidly or gradually
varied. The flow is rapidly; varied if the depth changes abruptly over a
comparatively short distance; otherwise, it is gradually varied. A
rapidly varied flow is also known as a local phenomenon,' examples are the
hydraulic jump and the hydraulic drop.
f'
I
I
I
'{'
I
r
I
Chong'e of depth from
time
time
to
G.V.F.
RV.F.
Sluice
Confraction
below the
sluice
----:---~-
~
-'.
//
R.V.F. - Bore
lJnsteqdy flow
t ~'.
The flow is turbulent if the viscous forces are 'weak relative to the
inertial fDrces. In turbulent flDW, the water particles move in irregular
paths which are ,neither smooth nor fixed but which in the aggregate still
represent the .forward motion of the entire stream.
Between the laminar and turbulent ,states there is a mixed, or transitional, state.
The effect .of viscosity relative to inertia can be represented by the
Reynolds number, defined as
Varied flow
'
A. Steady flow
L Uniform flow
.2. Varied flow
c. Gradually varied flow
b. Rapidly varied flow
B. Unsteady flow
L Unsteady uniform flow (rare)
2. Unsteady now (i.e., unsteady varied flow)
a.Gradually varied unsteady flow
b. Rapidly varied unsteady flow
Various types of flow are sketched in Fig, 1-2. For iIlu.strative purposes,
these diagrams, as well as other similar sketches of open channels in this
book, have been draw.l to a greatly exaggerated vertical scale, since
ordinary channels have small bottom slopes.
1-3. State of Flow. The state 01' behavior of open-channel flow is
governed basically by the effects of viscosity and gravity relative to the
in.ertial forces of the flow. The surface tension of water may affect the
behavior of-flow under certain circumstances, but it does not piay a significant role in most open-channel problems encountered in engineering.
,
Effect of Viscosity, Depeilding on the effect of viscosity relative to
inertia, the flow may be laminar, turbulent, or transitional.
The flow is laminar if t.he viscous forces are so strong relative to 'the
inertial forces that viscosity plays a signiftcant part in determining flow
behavior. i~ laminaLfl.mY...Jh~~arti.91esa,ppeal' to IAove in deE!lite
smooth paths, or streamlines, and infinitesimally thin layers of. fiid se~11l
To slia:e ove~\:n~Q~!.lt l~ye.r~: . - . - . - - - . - - . - - - - . - . - - -
RV.F.
1
i
I
I
II
j.
R == VL
JJ
(1-3)
\..J
I
"-
"BASIC PRINCIPLES
'
ill ft2/sec is equal ~o the dynamic viscosity, f.I (mu) in shlg/ft-sec divided
by the mass densl(;Y p (rho) in slug/ft,. For water at 68"F (2000)
5
f.I = 2.09 X 10- and p = 1.937; hence, 11
1.08 X 10-$.
"
An open~chan?,el flow is laminar if the ReYl10lds number R is small and
tUl'bt~lent if. R IS ,large. Numerous experinlents have shown that the
flow 10 n. pIpe . c~an~es, from laminar to tul'l)ul~l1t in the range of R
betwee~ the cntlCal val.ue~&Q.Q. ~n~ a valuethfl,t inay be as high as
50,000. . I~ th~se exp~nment.s the dIameter of the pipe was taken as the
chal'act~rIStlC. le~gth m defining the Reynolds number. When the
hYdra~11C radlUS 15 taken as the characteristic length, the corresponding'
range IS from ,500 to 11,50q, ,. since th.e~!.fl:,l!l~~~_ of '!..,Eipe is fo~r times its
hydraulic mdlUs.
'
--'--.-.-.--~--'-''''------
I
I
r
I
i
,,I
i
;t
(1-4)
,where 14, ~ the fl'1::Lional los; i.n ft. for flow in the pipe, i is the friction
fact~~. L L~ the l~ugth of the pIpe III ft, do is the diameter of the pipe in
V IS the velocIty of flow in fps, and g is the acceleration due to gravity
In ftfsec~.
'
:t,
1. .."
(1-5)
RO.~'
(1-6)
.1 t8 a result ~f Darc(s. stud.: [21: ~n flo~ in pipes, his name is commonly associated
wlthithat ~ WelS~o.ch [31 deslgna,tJng~hlS. equation which Weisbach firstformulated..
in
r'l
From Eq;S.
(1-3) and (1-5) it can be shown that
i
In
Act~IlYI dAubulSson [4] presenteji, prIOr to Darcy, a formula. that can be reducEid to
the form of Eq. (1-4).
i
.
, t tn. this equation, ~he hydraulic :radius is used as the characteristic lengtll in defin.. -.. mg the Reynolds .number. If the; diameter of pipe were used as ~he characteristic
length, the numerlllal constant of ~he numerator in this equationfwould be 0.316.
(1-7)
,!
+ 0.4 '
---
:'hls equation may a.lso be applied to uniform ahd nearly uniform flows
channels.
'
,
- - -_________
~/-m...0pen
2 log (R ...;'1)
I),
Nikuradse 8 ~
-'--~Eciuations (1-6) and (1-7) will be used i'n the follo\ving discussion as
~Th~-i~mi~~:, turbul~ntJ
,
I ,
I
Y
I
K = 8gR 2S
! .V'
(1-9)
11
10
BASIC PRINCrPLES
Since V and R have speeific ~'alues for any given chantiel shape, K is a
purely numerical factor dependent only on channel shape. For laminar
. flow in smooth cha.nllE)ls, the value of J( call be determined theoretically
[20} .. The pJot in Fig. 1-3 indica.tes that Kia approximately 24. for tht> reetanguJar channels and 14 fol' the triangular channel under consideration.
F: '.
(C
',{I td
a.o AI
R ., 4,01;:11
a til
a ..
R 0.0
to
..
R 3,0*
(I
..
4,
~4 tK
tloll
eta
-i.lli:vtRS;'TY n,= fLl!NOlS DATA
G
0.2
!~
lhjrro~,
.,..
FACE;
o
or
Sl'Flt;CTtiRAL
srUL
nU4NCAA...A.R CHA'HNl.,l'd" TO
fsd''V~Rrt)t ,t~I.ES,WlrH
0.08
0.1
om
O.Oi
0.06
().04
Q
100.0rt9fM
.. k.
0.02.831"
O.qz
0.02
O.OIl-----4----+-....:..--t---~'r-~
0.01
0.000
0.008
O.OO41--+-+++4-+-~-+-'f+-+-'-JHh-+-'-I--+-l-+I--!-++H
.J
~
$1'0'
3. The data in the turbulent region correspond closely to the Bllisius:j=lrandtl-vo1n Karman curve. This ~ndicates that the law for turbulent
fiowin smaoth pipes may be approximately representative of a.Usmooth
channels. The plot also shows that the shape of the channel does not
have an important influence 011 friction in turbulent flow, as'it does in
,"
laminar fl o:W.
The dataJor laminar flow obtained at the University of Minnesota [20J
and the da~ for turbulent flow cdllected individually by Kirschmer
l
.'
.
h channelS. . Bo.zin's channeis; No.4,
FIG. 1-4. The f-R. ~elatlonshlp for Row In rOU~ed wood; No. 14, 'unpolished wo~d
gravel embedded In aement; No. Il,. unpo m lon 10 mm high, andlO mm In
roughened by tra.nsverse wooden stripS 27.
sp.~in'" of
mm; No. 24, cem, ent
..
7
Ne 14. except WIt .1 a
te :
spacmg; No.1, same B.S.
'
d K' her's cha.nnel: smooth concre .
lining; and No. 26, u~p<Jhshed woo.
IrSC In
.
"
;n
R
;
"--'
10
\
I,
50
.Q
"
.
. .
(23] are shown in the diagram for fio~ in
(15,16}, Elsner [221, and Kozelnny
f the' data channel roughnes.~ is
(Fig. 1-4).
13Qme 0 '
f I _
.r
. h'
.
e8.SUre of the roughness fll1rttcles or;m
k
whic
IS
a
SIze
m
.
f
t
' s'.
Y , .
represen
d' .
illustrates the followmg ea ur.e
ing the channel surface. The lsgram
d fi d b Eq (1-8) In this
.
. th d til. can be e ne Y . .
. L In the lammat reg~on e ~ly h' h than it is for smooth channels
region, the value of K l.S genera
19 e r .
.
\,
I
'-.---
..
~-"
IJ.
)-J
(J.
.",
'ru
f'J
I'
)--
")
Ir}
~".
."-"'
12
~.-.-.---
-_._._-,_
t
j) -1.
[
and ranges bfltween 60 and 33, indicating the pronounced influence of the
channell'oughnesson the friction factor.
.
2. In the turbulent region the channel shape has a pronounced effect
on the friction factor. It is believed that, when the degree of roughness i~ const.ant the fdction. factor decreases roughly in the order
of rectangular, t:'iangular, trapezoidal, and circular channels. At the
suggestion of Prandtl, Kirschmer [15,lDj explained that the effect of
channel shape may be due to the development of secondary flow, which
is ~pparently more pronounced in rect(1ngular channels than in,say, triangular channels: The secondary flow is the movement of water particles
on /l, cross section normal to the longitudinal direction of the channel. A
high secondary flowrinvolves high energy loss andthns r.,ccounts for high
chaI!!lE11.J:{:!.i..frul.nce.
. . . .
3. In the turbl,1lent region most plots appear parallel to the Prandtlvon Karman curve, This curve serves as an approximate limiting position toward which a plot moves as the over-all resistance becomes less,
Accoi'dingtQ a concept advanced by Morris [241 (Art. 8-2), the rise of the
plots above the smooth-conduit curve may be explained as a resul~ of
additional energy loss generated by the roughlless elements. When the
Reynolds number is very high, some plots become essentially horizont,al,
reaching a stllte of SO-Galled complete turbulrnce. At this state the value
of fis independent of Reynolds number and dep~nds solely on roughness,
hydraulic' radius, .and channel shape .
. 4. The' plot of Varwick's data [16J for a give~ roughness,- hydraulic
radius, and channel sha~ star~ off from ll. curve parallel to thLPran..Q.tj~
Voill'{arman curve, then rises as the Reypolds n~mber incre~, and
final1y becomes horizo'ntal as a state of complete turbulence is reached.
The rise of the plot is a peculiar phenomenon which demands explanation, I
and, since this finding has not been verified by other data, mor.e experimental studies .seem necessary to substantiate it.
~-'-~It should be noted that the ab'Jve descriptions are limited to lowvelocity, 01' subcritical, flow (which will be defined later in this.-ar.ti.cl.e)_
and to Row onwhich 8.ll:.rface
tension
dOEls not have a 1IDlificant
influence.
.. _._._----;---,. .. - ....
-In most open channels laminar flow occurs very rarely. ,Th~!!l;~t tfu;.1
the surface of a streilr!!:J!:eE!~!g_~<JmlOoth and glassY_.JQ..Jtn ob.sru:.v.ru.2...by
no mea~__a..lL.i.ll.dicati2.~~inar; }llOID;' prohllJily, it
indicates that the sU11aqe vel.QQi!! is lower than that required for capillary
wav~Qnn. :J;.,.aminat open-cha~Jm9,!..n t'O exist, however,
~~::wJ1~~ th.in sheets...2f_ ~t?l' flow over the grq,,';1nd Dr where it is
created deliberat~ ~testlng ~~!lnels,
---
"
BASIC PRINCIPLES
13
(1-10)
where, V is the mean velocity of fiow in fps, g is the acceleration of gra Yity
in ft/sec z) and Lis a.characteristic length in ft. In open-channel fiow the
characteristic leng~h is made equal to the lmdraulic deptAJ2., which is
defined as the cross-sectional area of the water normal to the direction of.
flow in the channel divided by the width of .the fre surface. For rectangular channels this is equal to the depth of the flow' section,
When F is equa.l to unity, Eq. (1-10) gives
V
.../gD
(1-11)
v'{jD;
>IL
n-2
I O~ber dimensionless ratios used for the. sa.me purpoae.include (1) the lcinenc-flow
factor}.
VI/uL ... FI, first tlsed by Rehbock [251 and then by Ba.khmetefi' f26Ji
(2) the Bouuinesq number B "'" V / v'2UR, first used by Engel [27J; 8.nd (3) the kinelicity or velocitY-head ratio 11; = V'/2gL, proposed by Stevens [28] alld Posey [29J
respecti vaLr.
.
-I
itJ
'J
14
BASIC PRINCIPLES
.
designed for this effect ;that is, the Froude llumhflr of the flow in the model'
channel must be made "qual to tha,t of the flow in the prototype channeL
1-4. Regimes of Flow. A combined effect of viscosity and gravity may
produce anyone of four 'regimes of flo7JJ in an open channel, namely,
(1) 8ubcritical-larninar, when F is less than unity and R is in the 19,ininal'
range; (2) 8upel'cl'itical-laminar, when F is greater than unity and R is in
the laminar range; (3) ::mpercritical-turbulBnt, when F ia greater than unity
15
I
I
1
\........,;;
,_I
'1
'-J
~\
,-,'
~J
f,
i
I
C:--'\
\..-
~.~
Velocily, Ips
FIG.
1~5.
,_-
(Afler
'~
L
Fw. 1-6. Photographs showing four flow regimes in a laboratory cll!1nnel.
of H. Rouse.)
..
(Courtesy
photograph the direction of flow is from left to right. All flows are uniform except those on the right side of the middle and bottom views.
The top view represents uniform subcritical-laminar flow, . The flow is
su.b:ritical, since the Froude number was I',djusted to slightly below the
cntical value; and the streak of undiffused dye indicates that it is laminar.
Th~ middle ~i~w shows a uniform supercritical~laminar fl'ow changing to
v~r~ed subcntical-turbulent. The bottom view shows a uniform superc:'1.tlCal-turbulent flow changing to varied subcritical-turbulent. In both
cases, the diffusion of dye is the evidence. of turbulence.
{_.
16
\.
~l
It is'believed th~t gravity action may have a definitive effect upon the
flow resistance in cliurmels at the tut'bulent-flow range~ The experi,mental data studied by Jegorow [311 and Iwagaki [32J for smooth rec'tangular channels. and by Hom-:ma [33J for rough :channels have shown
that, ~n the supol'critical-turhulent regil;l1e of flo~, the friflj;ion fact<l!:...~
~Y~.lI!:_~!"~~~(L1:yj.ili increas~de n1Enbt:~. Generally, I;}le effect
of gravity is practically negligible where the Froude numbeds small, say,
less than 3. A further irivestigation by Iwagaki [34J indicates that, with
'increasing Froude number) the friction factor of turbulent flow in both
smooth and rough open channels becomes lal'ger than that .in pipes.
is possible that the presence of the free surface in (J~n~channel flow makes
. the channelrougw- than the pipe.--Whellmore data -and --:evidell(;e
oecome avaiTab-le, the'-Froude number, ,representing the gravity effect,
may have to be' considered as an additional faotor in defining the J-R
relationshipfm' supercritical-turbulent flow.
'
PROE\LEMS
, 1-1. 'With reference to Fig. 1-1, show that the theoretical dischl.i:ge of the openchannel flow may be expressed by
0-12)
(
I
where A 1 and A. are the cross-sectional areas of the flow at sections' 1 and 2, respectively, a.nd lJ.y'is the drop in water surface between the sections.
1-2. Verify Eq. (1-10).
1~3. Verify by computation the depth-velocity' relationships shown in Fig. 1-5 0:
the four flow regimes in a wide rectangular open channel. The temperature, o,f the
water is tnken as 68F.
.
1-4. A model channel is used to simulate a prototype channel 100 ft wi~e, carrying
;t disclmrge of 500 ds aL. a depth of 4 ft.
The model is designed fOI' gravity effect,
Itnd a: tUl'bu[ent-fiow condition is asaured. Determinet.he minimum size of the model
and the scale ratio, BSSuming the upper limit of the \'r1lnsitionru-flow region to be
R = 2,000. The sCII.le ratio is the ratio 'of the linear dimension of the model tot,ha.t
of the prototype. .
REFERENCES
'\
I
!
!I
,I
17
.j
18
BASIC PRINCIPLES :
b!lrt C. Nels01l, Trqnsaction8, American So.ciety o.J Civil Engineers, vol.. 123, pp.
707-712, 1 9 5 8 . .
.
22. FtlLnz Eisner:- Offene Gerinne (Open channeL), sec. 4 of vol. IV, "Hydro- und
Aerodynamik," in \V. Wien and F. Hanns (editors-in-chief): "Handbuch der
Experimentalphysik," Akademische Veda.gsgesellschaIt mbH, Leipzig, 1932,
p.298.
23. Josef l(ozeny: "Hydraulik" ("Hydraulics"), Springer-Verlag, Vienn9., 1953,
p.574.
.
24. Henry M. Mcirris, Jr.; Flow in rough condl,lits, Transactions, American Society oj
Civil Engi1J,ee1's, vo.l.120, pp. 373-398, 1955. Discussio.ns on pp. 3.99-410.
25. Th.R.ehbock: Zur Frage des Briickenstaues (On the prDblem Df bridge constrictions), Zenlralblail der B(11!verwaZtun!l, BerUn, vol. 39, ~o. 37, pp. 197-200, 1919.
26. Boris A. Bakhmeteff: "Hydraulics of Open Channels,"McGraw-Hill Book Com.
pany, Inc., New York, 1932, p. 64.
.
27. F .. V. A. E. Engel: Non-uniform flow of water: Problems a.nd phenomena. in open
channels with side cont.ractions, The Engineer, vol. 155, pp. 392-394, 429-430,
456-457, 1933.
.
...
.
28. 1. C. Stevens: Discussion Dn The hydraulic jump in sloping channels, by C. E.
Kindsva.ter, TramactiO'nll, American Society of Civil Engineers, vol. 109, pp: 11251135, 1944.
29. C. J. Posey: DiscussiDn on The hydraulic jump in sloping channels, by C. E.
Kindsvater, TransactioTLI/, American Society oj Civil Engineers, vol. 109, pp. 11351138, 1944.
30. J. M. Robertson and lIunter Rouse: On the four regimes of open-channel flow,
Civil E'lI.{Ji11eering, vol. 11, no.. 3, pp. 169-171, !vIa.rch, 1941.
31. S. A. JegorDw: Turbulente 'Oberwellenstromung(Schiesscnl in offenen Gerinnell
mit gla.tten Wanden (Turbulent 5up'<!rcritical flow in open channel with smoDth
wa.lls) , WGSS/!iTkraJt unci Wasserwirf.8chaJt, Munich, vol. 35, no.. 3, pp. 55-59, 1940.
32. Yuichi Iwagaki; On the laws of resistance to turbulent flow in .open smooth
channels, llfemoirs of: tJte Faculty oj Engin.eerill,f/, Kyo.to University, Jilpan, vol. 15,
no. 1, pp. 27-40, Janua.ry, 1953.
33. Masashi Hom-rna: Fluid resistance in water flow of high .Froude number, Proceedings oj the ed Japan National Congress Jor Applied Mechanics, pp. 251-254,
1952.
34. Yuichi Iwagaki: On the laws of resistarrce to turbulent flow in open rough channelS,
Proceedings oj the 4th Japan N a./irma!. C(mgres8 Jor Applied Mechanics, pp. 229-233,
1954.
.
II
\..:.,1
i,
CHAPTER
20
BA.SIC PRINCIPLES
I
I
i
i
r
1
I
!
.i!:
I;'.>
1 The side slope z; 1 of a pa.rl!.boli~ section at the intersection of the sides with the
free surface can be computed easily'by the simple formula z =
Russian e!lgineers [11 also use semielliptica! and para.bolic
of higher order:
y
azP with 11 .;, 3 Qr 4.
The constant CL is computed fl'om the side slope assumed
at the' free surface ..
21
22
BASIC PRINCIPLES
egg-shaped, ovoid, semielliptical, U-shaped, catenary, horseshoe, baskethandle, 'etc. The complete rectangle and square are also common for
large sewers. Dimensions and'properties of sewer sections may be found
in textbooks on sewerage.!
A special geometric section kilOwn as hydrostatic catenary or lintMrw
[4,5J is the shape of the cross section of a trough, formed of flexible sheets
assumed to be 'Iveightless, filled with water up to the top of the seCtion,
. and firmly supported at the upper edges of the sides but with no effects
of fixation~ The hydrostatic catenary has been used for the design of the
sectiolls of some elevated irrigation flumes. These flumes are constmcted
of metal plates so thih that their weight is negligible, and are :firmly
attached to beams at the upper
2-3, Geometric Elements of Channel Section.' Geometl"l:c elements are
pro'perties of a channel section tha.t can be defined entirely by the geometry of the section B.nd the depth of flow. These elemerits are very
importa.nt and are used extensively in flow computations.
For simple regular channel 'sections. the geometric element.s can be
expressed mathematically ill terms of the depth of flow and other dimensions. of the sectioll. For complicated sections and sections of natural
streams, however, ,110 simple formula can be 'written to express these
elements, but curves representing the relation -between these elements
and the depth of flow can be prepared for use in hydraulic computations.
The definitioris of several geometric elements of basic importance are
given below. Other geoIlwtric elements used in this book will be defined
where they first appear.
The depth of flow y is the vertical distance of the lowest point of a
/' 'chanilel section from the free surface. This term is often used inter-'
changeably with the depth offlow section d. Stridly sp~aking, the depth
of flow section is the depth of flow normal to the direction of flow, or the
height of the channel section containing the water. For a channel with
a longitudinal slop~ angle e, it can be 8een that the depth of flow is equal
to the depth of flow section divided by cos e. In the case of steep channels, therefore, the two tei'ms should be used discriminately.
The stage is the elevation or vertical distance of the free smfa-ce above
'a' datum. If the lowest point of the channel section is chosen as the
datum, the stage is identical with the depth of flow.
The top width T is the width of channel section at the free surface.
/
The
wafer al'ea A is the cro~s-sectional area of the fi~w norlD,al to the
/
direction of flow.
The wetted perimeter P is the length of the line of intersection of the
channel wetted surface' with a cross-sectional plane normal to the direction bf flow.
'
I Many typica.! seWer sections arEi described in [2J nnd [3].
23,\
The hydr.aulic radiu,$ R is the ratio of the water area to its wetted
perimet.er, or.
A
R
(2-1)
'1
The hydr~ulic de1Jth D is the ratio of ~le wa.ter area to the top width~ or
D
(2-2)
Z=A
(2-3)
.j
"
.!
~lements
'{
[
,\
\ '
\
--
24
)
), ,-
"
~',.#
, Example 2-1. Compute the'hydraulic radius, hydraulic dep~h, and section factor.?:
Of the~ra.pezoidal chUlluel section in Fig. 2-2. The depth of flow is' 6 ft.
,
'
~~'
, '
'
Q,?!/'
~
'
~~
.~'
C~b=20':, .1'
\
'~//l7~~0~
....&..
T:44"
P=46,8 -
25
BASIC PIIINCIPLES
1.5
t~.5'
'.
Tropezuidal chonnel
Triangular channel
Sholtow ditch
NarroW'
(@c!anglJlar
5~ction
),
\'I
Pipe
Natural irregular channel
I
r
r
I, .
the closer to the banks, the deeper is the maximum. Figure 2-3 illustrates
*e general pattern of veJocity distributio;l over \;al'ious vertical and
horizontal sections of a rdctangular channel sectio~ and the curves of
equal velocity in the cross section. The general patterns for velocity
distribution in several channel sections of other shapes are illustrated in
Ji]ig. 2-4.
;
,
I The velocity distributfoJ in a channel section depends also on other
f('l.clors, such as the unusu~l shape of the section, tlle roughness of the
26
BASIC PRINClPLES
(cleonlcur Jines
'or
(o)Con!our lines 01
equal ve c for (v)
Jr
ma9niludes of the
IOlerol currenlslvor '
FIG. 2-6. Distribution of the velocity components, fa.cing downstream a.t the midsection of a. straight flume. Voloeitiestl.re In em/sec (= 0.0328 fps); y/b = 1.0;
R "" 73,500; and Q = 701iters/s.ee (= 2.47 cfs). (Afler .A. ShlLkry [6J.)
27
the width. h,grea~.E 5 to 10 times. the depth <:>f flow ...4~IJe~on the
.2_'?~ion of l:lu):fac~ roughness. Thus, a wide Open channel can safely be
defined as a rectangular channel whose width is greater than 10 tiri1es the
dept.h of flow. For either experimental or anaJytical purposes, the flow
in the cen tral region of a wide open channel may be' considered to be the
same as the flow in !1 rectangular channel of infinite width.
2-6. Measurement of Velocity. According to the stream-gaging procedure of the U~S. Geological Survey,'lthe channei crosssecticH1 is-ciivid~d
-~rti6a1 strips by a number of successive vert~cals, and mean velocities
in verticals are determined by measuring t.he velocity at 0.6 of the depth
J!-l_~Gh vertical, or, where morc reliable results are required, by taking the
average of the velocities at 0.2 and 0.8 of the depth. When the stream
i.s covered with ice, the mean velocity is no longer close t.Q 0.6 of the water
dept.h, l;mUI!e average at 0.2 and 0.8 of the water depth still gives reliable
resul.1. The average of the ~ean velocities in any two adjacent verticals
multiplied by the are:1- between the verticals gives the discharge through
this vertical strip of the cross section.: ThG sum of discharges through all
strips is t.he total discharge. The mean velocit.y of the who.Ie section is,
therefore, eq ua! to the total discharge divided by the whole area.
It should be noted that the above methods are simple and approximate.
For precise measurements more elaborate methods must be used, which
are beyond the scope of this book.
2-7. Velocity-distribution Coefficients. As a result of nonuniform
9.istributiol1 of velocities over a channel section/the velocity hea.d of an
open-channel flow is generally greater than the value computed according
to the expression V 2 /2g, where V is the ri;lea.n velocity. When the,energy
principJe is used in computation, the true velocity head may be expressed
as a V 2/2g, where a is known as the energy coefficient or Co-riolis coe.fficient,
in honor of G .. Coriolis [12J who first proposed it. Experimental data
indicate that. the value of a varies from about 1.03 to 1.36 for fairly
straight pri!:imR.tic channels. The value is generally higher for small
channels and lower for large streams of considerable depth.
The nOllUniform di.stribution of '/elocities also affects the co.rnputation
of momen~umin open-channel flow. From the principle of mechanics,
the "momentum of the fluid passing through a channel section per unit
time is expressed by {JwQVj(J, where f3 is known asthe mom.entum coefficient or Boussinesq coe:ffi.cient, after J. Boussinesq [13] who fii'st proposed
it; w.is the unit weight of water; Q is the discharge; and V i.'3 the mean
velocity. It is ,generally found that the value of (3 for fairly straight
prismatic cha.nnels varies approximately from 1.01 to 1.12.
The two velocity-distribution coefficients are always slightly larger
than the limiting value of unity, a.t which the velocity distribution is
1
!9J
to [11 J.
_,. t:../
'I
) .
28
BASIC PRINCIPLES
strictly uniform across the channel section. For channels of regular cross
section and fairly straight alignment, the effect of nonuniform velocity
distrihution on the computed velocity head and momentum is small,
especially in comparison with other uncertainties involved in ~he computation. Therefore, the coefficients are often assumed to be unity. In
channels of complex cross section, the coefficients for energy and momentum Can easily be
great as 1.6 and 1.2, respectively, and can vary quite
rapidly from section to section in case of il'I'egular alignment. Upstream'
from weirs, in the vicinity of obstructions, or near pronounced irregularities in alignment, values of a greater than 2.0 have been observed. 1 ,
Precise studies or analyses of flow in such channels will require measurement of the'actual velocity and accurate determination of the coefficients.
In regard to the effect of channel slope, the coefficients are usually higher
, in steep channels than in flat channels.
For practical purposes, Kolupaila [16J proposed the values shown below
for the velocity-distribution coefficients. Actual values of the coefficients
for a number of channels may be found in {l7) and [18J.
as
"I
I!
Value of {j
Channels
Min
Av
'Max
lVIin
Av
Ma.x
'-"0-
I,
I
r
I
1.10
1.15
1.20
1.50
1.15
1. 30
1.50
1.75
1.20
1.50
2,00
2,00
1.03 1.05
L05 1.10
1.07 '1.17
L 17 1.25
1.07
1.17
1.33
1.33
29
corrected velocity head for the whole area as a V /2g, the total kinetic
energy 'is ,aw A/2g. Equating this quantity with 2:wv 3 AA/2g and
reducing,
J1]3 dA L;'~! LlA
(2-4)
V3A ,'"'"
va
The momentum of water passing AA, per unit time is the product of
the mass W<I AA/g and the velocity v, or wv 2 AA/g. The total momentum
is :Z1l!V2 AA/g. Equating this quantity with the corrected momentum
for the w'hole area, or {3wA V2 / (I, and reciuping,
, Jy 2 dA
{J
V2 A
2:v 2 LlA
"". V2 A
(2-5)
'p
= 1 + 3.% 1
f2
2t 3
(2-6)
(2-7)
where Ii = v.u/V
1, VM being the maximum velocity and V being the
mean velocity;
.
Computation of the velocity-distribution coefficients for. irregular
natural channels will be discussed later (Art. 6-5). In most pl'nctical
problems dealing with regular cha~nels it is not necessary to consider
the varia.tion of velocity throughout the cross section, since use of the
average veiocity will give the accuracy required. The expressions
V2/2g and wQ V j g a~e used extensively in this book with the understand:ing either that these items have been cOlTected for the effect of the non~
uniform velocity dis~ributionj Of that a value of unity is asiumed. 2
I These formula.s a.re obLained by I1ssuming a. logtLrithmic distribution Qf velocity
(Art. 8-5, Prob. '8-9). Assuming a lincar velocity distribution, Rehbock [201 obtained
c< = 1
f" I1nd.{j = 1
"/3.
.
For discussions on this subject, the reader may look into. 121J and [22]. However,
he should use' judgment in reading these references because they contain erroneous
BASIC PRINCIPLES
30
31
~.-~---
, ~l"tlow.
",/'
statements. Some authors have proposed the use of the mOluentUnl coeffic~en.tto ,
r'epJace the energy coefficient even in computations based on the energy PrJ?ClpJ~,
"
t
ec't V'Nhether the energy coefficient or the momentum
. , coeffiCient
. . o1 ,IS
d
TI 1IS IS no carr .
to be used depends on whether the energy or the moment?m prw::lple
-:,e .
The two coefficients are derived independently from baslca.lly different principles
(Art. 3-6). Neither of them is wrong and neither ca.n be replaced by t~e other; both
should be used in th~ correct sense.
'
"
'
I Specific qualifica.tions for parallel flow were clea:rly stated for the first tlme by
Belanger [23].
'
wd~~
g
(2-8)
IS Ill:r
.I
For computing the value ofc at the channel bottom, r is the radius of
cu::',rature of the bottom, d is the depth of flow, and for practical purposes
J,
. i
32
"
BASIC PRINCIPLES
/., i
I\
I
r.
~y}o
hI! dA
cv dA
sure head at any vertical depth is equal to this iiepth multiplied by a correction factor cos 2 e. . Apparently, if the angle (J is small, this factor will
not differ apprecill.bly .fl'Om unity. In fact, the correction tends' to
decrease the pressure head by an amount less than 1 % until e is nearly
, 6" i a slope of about 1 in 10. Since the slope of ordinary channels is far
less than 1 in 10; the correction foi: slope effect can usually be safely
_ _-:i"" ignored .. However, when the cha~l_slope is large and its eff5l9t becoroe~
appreciable, thecorrection should be made if Mcumte comput~ttion is
-'-~-.----.
-'
Pressure di.stribution
desired. A channel qf this type, say, with a slope gre~ter than 1 in 10,1.'.3
hereafter called a channel of large slope. Unless specifically mentioned,
all chMlnels descrlbeanereafte-r are ~onsidered to be channels of small
slope, where the slope effect is negligibie.
If a channel of large slope ,has a. 10ngitudin!11 vertical profil,; of appreciable. curvatul'e~ the pressure head should be cOl;rected fo]' the effect of the
curvature of streamlines (Fig. 2-9). In simple notation, the pressure
head may be expressed as g'y cos 2 Jt., """~~~~S==
In channels of large slope the
usu .
andhighel'
thl1n the critical velobity. When this velocity reaches a certain riHl.gni,
"tude, the flowing water ,vill entrain nil', produ'cing a swell in its volume:i
.and Ml increase in depth.l For this rel1son the pressUl'C computed uy
Eq. {2-11) or (2-12) 4,~n shown in several gases to b.e higher than th~
h = Y cos 2, e
h = d 90S e
'
(2-10)
}o '
where Q is the total discharge andy is the depth of flow. It can easily be
seen that a' is greater than 1.0 for concave flow, less than 1.0 for convex.
flow, and equal to LO for parallel flow.
. For complicated curved profiles, the total pressure distribution can be
determined approximately by the fjow~net method or DlOre exactly by
model testing.
,
In ra..mr!l.t.Y!1ried flow theghange in depth of fl.oJYJs so rapid and abl'Upt
.:that th<Ltrea!!J.1ines l,:!9ssess 8ubstfilliia.UIJD!J)J,m1Ul.nd di;v~nce. ~
seguently, the hy~ro8tatic law of pressure ,distribut~s not hold strictlJl.
i!!!.IJ!1riifl:U...Y!1:riedJto
.
- It should be noted at the outset .that throughout this boqk flow is
treated in general as either parallel or gradually varied. Ther~ore, the
effect of the curvature of streamlines, will not be considered (~hat is, it
will be assumed that a' = 1) unless the flow is specifically described as
either curvilinear or rapidly varied.
2-10. Effect of Slope on Pressure Distribution. With fPjerence to a
straight sloping channel of unit width and slope angle 6 (Fig. 2-8), the
weight of the shaded :"vater element of lengt;h dL is equaltq..:l:!:'lUlQ~.
The pressure due to this weight is wy cos 2 fJ dL. The unit pressure is,
therefore,equal to wy cos~ e, and thehead 1 is
or
=~
33
JAil' becomes entraine~.in ",,,,ter generally...!!! v"e19\l~ of !>~t go Ips and higher.[
Besides velocity, however. other factors such'as entrance condition, channel rough.,
ness, distance .traveled, cha.nnel cross section. volume
discharge, etc all have some;
bearing on a.ir entrainment.
'
.
;
M. Hasurrh has measured the distributibn of pressure along the slopihg faces of
we~ [241. The data. obtained from these ekperimenta have verified Eqs'. :(2 .. 11) and
(2-12) very sa~Jsf!l.Ctorily (25].
.
1
or
35
BASIC PRINCIPLES
struct curves showing the relationships between the depth y and the section elementS
A; R, D, and Z; and (b) determine, from the curves the geometric elements for y = 4.
34
Distance f,om
a reference point
neClr left bank, ft
Left bank: -5
Stage,
ft
5.6
4.6
4.0
-4
-2
1.9
0.8
0.2
0.3
0.2
0
1
3
5
'1
Distance fro~
a referenc(J point
near .left bank, ft
7
9
11
13
15
17
Stage,
ft
-0.1
-0.1
-0.4
-0.1
0.7
19
Right bank: 20
3.2
4.1
\\
2.6
j
\
2-6. The hydrostaLil" cll.tenary may be plott.ed for [Lny given depth l/ and slDpe angle
O. at its ends by the following two approximate equations:
:1:)=
fk [(1 -
%k' - 1)1.g,k4).p
Concave flow
the cross section wiII simplify cQmputation, with, the errors on the safe
side.
.[
PROBLEMS
2-1. Verify the formulas far geometrio elements of the seven channel sections given
in Tahle 2-1.
$-2. Verify the cUrY'es shown in Fig. 2-1.
2-3. Construct curves siinilar tQ those shown in Fig. 2-1, for a square channel
section.
2-4. Construct cu'rY'es similar to those shown in Fig. 2-1 for an equilateral triangle
with one side as the channel bottom.
9.-5. From the data g(ven below on the cross section 1 of a natural stream
con
. I
(2-13)
(2-14)
YI = if cos 10
Convex, flow
where :1'1 and YI, respectively, are the ordinate and abscissa measured from t,h.e midpoint of the free surface; k = sin (110/2); '" = sin- 1 f [sin (1/>/2)l!kl; and 11 is the slope
angle at the point (XI,l!I), varying from 0 at the bottom of the curve to 9, at the ends.
The above equations will define' the. cross section when the flow is at its full dept.h.
The slope angle at the ends of a hydrostatic catenary of best hydraulic efficiency is
found mathema.tically to be II, = 35'37'7". (a) Plot this section with' a depth
y = 10 ft, and (bl determine the values of A, R, D, and Z at the full depth .
2-7. Estimate the Ylllues of momentum coefficient (j for., the- given values of energy
c(lefficient ex = 1.00, 1.50, and 2.op.
,
2-8. Compute the energy and mo~entum coefficients of the cross st'ction shown in
Fig. 2-3 (a) by Eqs. (2-4) and (2-5), and (b) by Eq~; (2-6) and (2-7). The cross section and the curves of equal velocity can be transferred to a piece of drawing paper
and enlarged for deSired ll.ccuracy.
2-9. In designing side walls
steep chutes and overflow spillways, prove that the
overturning moment due to the pressure of the flowing water is equal to Yswy' cos' 9,
wherew.is the unit weight of water, y is the vertical depth of the flowing water, and 9
is t,he slope angle of the channel.
2 ..10. Prove Eq. (2-10).
2-11. A high-head overflow spillway (Fig. 2-10) has a 60-ft-radius flip bucket u.t
its downstream end. The bucket is not submerged, but acts to change the direction
of the flow from the slope of the lipillway face to the horizontal and to discharge the
flov1 into the air' between vertical training walls so ft apart. , At: a discharge of 55,100
ds, ;the water surface at the vertical section OB is at El. 8.52. 'Verify t.he curve that
represents the computed hydraulic ,pressure acting on the training wall at section DB.
The computatiQn is bailed an Eq,' (2-9) and on'the following assumptions: (1) the
velqcity is uniformly di~tl'ibuted across the section; (2) the vo.lu,e used for r, fQr pressur~ values near the wall base, is 'equal to the radius of the bucket but, for other
pre;isure values, is equal to the radius of the concentric flow lines; and (3) the flow is
entto.ined with air, and the density ,of the air-water mixtureca~ be estimated by the
(,
"
, i
of
/a)
j
)
.1
,.
!
36
BASIC PRINCIPLES
10
~0.2V:
. gR
- 1
(2-15)
where u is the percentage of entrained .air by voiume, V is the velocity of flow, and
$PilIW1!Y
Iraining wall,
eo II
cpO!1
;:;
C
.2
::l
.)
:OJ
t;j
i.
2
~
I
/,
i
,a
Un;l pressare, II 01 woler
\I
REFERENCES
1. S. F. Averillnov: 0 gidravlicheskom raschete rusel krivolineinoI formy poperech,
2.
3.
.
4.
nogo secheniia (Hydraulic design of channels with curvilinear form oithe crosS
section), lzvestiia Akademii Nauk S.S.S.R., Otdelenie ~'ekhnic"eskfk;h Nauk"
Moscow, no. 1, pp. 54-58, 1956.
Leonard Metcalf and H. P. Eddy: "American Sewerage Pra.ctice," McGraw-Hm
Book Company, 1M., New York, 3d ed., ,1935, vo!. 1.
Harold E. Babbitt: "Sewerage and Sewage Treatment," John Wiley &: Sons, Inc.,
New York, 7th ed., 1952, pp. 60-:.66.
H. M. Gibb: Curves for solving the hydrostatic oatenary, Engineering News,
vol. 73, no .. 14, pp. 668-670, Apr, 8, 1915.
I This iormull!. [26J is based on da.ta obtained from actual conorete and wooden
chutes, involving errOnl of 10%.
'
37
5. George Higgins: "Water Channels," Crosby, Lockwood &: Son Ltd., London, 1927,
pp.15-36.
.
.
6. Ahmed Shukry: Flow around bends in an open flume, Transactions, AmericilTl
Society of Civil Engineers, vol. 115, pp. 751-779, 1950.
'
7. A. II. Gibson: "Hydraulics and Its Applications,'" Constable &: Co., Ltd., London,
4th ed., 1934, p .. 332.
,
8.J. R. Freeman: "Hydra.ulic Laboratory Practice," Amedcan Society of Mecha.nical
,
Engineers, New York, 1929, p. 70: '
9. Don M. Corbett and ot.hers:8trealn-ga.ging procedure, U.S. Geologicnl SlI1vey,
Water Supply Paper 888, 1943.
10. N. C. Grover and A. W. Harri'ngtoo.: "S.ream FlOW," John Wiley &; 80ns, Inc.)
New York, Hl43.
11. Standards for methods and records of hydrologi~ measurements, United Natio7ls
Economic Comm.isslcn for Asia: and the Fa:r Ei.I$~, Flood Control Series, No.6,
Ba.ngkok, 1954, pp. 26-30.
,
12. G. CorioUs: Sur.l'etablissemellt de Ill. formule qui donne la figure des remons, et .sIU
12. ilorrection tiu'on doH y int,roduire POllr tenir compte des diffel'ences de vitesse
dans les diVers points d'une marne section d'un COUl'ant (On the ba.ckwater-curve
equation a.tid the corrections to be introduced to !lccount for the difference of the
velocitie$ at different points on the same cross section), Ivnmoire No. 268,
..,l,n'nalca du punts et chaw;sees, vol. 11, ser. 1, pp. 314-335, 1836.
13. J. Boussinesq: Esg's'i sur la theorie des eaux courantes (On the theory of flowing
waters), M~moire& ]fr/;sentes par diven savants ri l'Academie des Sciences, Paris,
1877.
. .
14. Erik G. W. Lindquist: Discussion un Precise. weir measurements, by Ernesf W.
Schader andT(ennethB. Turner, 1"7'(tllaac:l.ions, American Society of Civil Engineers,
vol. 93, pp. 1163-1176, 1929.
15. N. M. Shcha.pov: H Gidrometriia Gidrotelchnicheskikh SoorllllheniI i Gicir,omashin"
(" Hydrometry of Hydrv.lllic Structures and MacJ:Jnery ") I Gosenel'goizciat,
.
.
Moscow, 1957, p. 88.
16. Stcponas Kolupaila: Methods of determin!l.tion of the kinetic energy facto!', The
Port Engineer; Calcutta, India., vol. 5, no. I, pp. 12-18, Januo.ry, 1956.
17. M. P. O'Brien and G: H. Hickox: "Applied Fluid Mechallics," McGraw-Hill
Book Company, Inc., New York, 1st ed., 1937, p'.272.
'
18, Horace WilliamKing; i'Handbook of Hydraulics," 4th ed., l'evised by Ernest F.
Brater, McGraw-Hill Book Company, Inc., New York, 1954, p. '7-12.
19. Morrough P. O'Brien and Joe W. Johnson: Velocity-head correction for hydrau1ia
flow, Engineering News-Record, vol. 113, 0.0.7, pp. 214-216, Aug. 16, 1934.
.
20. Th. P..ehbQck': Die Bestimmung der I,age der Energielinie bei ftiessenden Gewfulsern mit HilIe des GeschwindigkeitshOhen-Ausgleichwertes (The determina.tion of
the position of the energy line in flowing water with the o.id of velocity-head
a.djustment), Der Bau.ingenieuT, Berlin, vol.. 3, no. 15, pp. 453-455, Aug. 15, 11122.
21. Boris A. Bak&meteff: CorioIis and the energy principle in hydraulics, in "Theodore
von !Urman Anniversary Volume," California. Institute of Teohnology, Pasadena,
1941, pp. 59-65.
22. W. S. Eisenlohr: Coefficient's for velocity distribution in open-Channel flow, Tra.nsac:I.ior/.$, American. Socie4/ of Civil Enginee7's, voL 11:0, pp. 633-644, 1945. Discussions, pp. 645-668.
23. J. B. Bela.nger: "Essai sur la solution numeriqne de ql,lelques problemes relatifs
au mou.-ement permanent des eaux courantes" ("Essa.y on tIle Numerica.l Solution of Some Problems Relative to Steady Flow of Wa.ter"), Carilian-Goeury,
Paris, 1828, pp. 10-24.
38
BASIC PRINCIPI..ES
24. It Ehrenberger: Versuche Iiber die Verteilung der Drucke an Wehrriicken infolge
des I1bsturzcnden '.Vassers (Experiments on the distribution 'of pressuresa\ong the
f~~e of w(d ..;; resulting from the impact of the fa.lling water), Die W IMJserwirtschaft,
Vienna, vol. 22, no. 5, pp. 65-72, 1929.
25. 'H&rald Lauffer: Druck, Energie und Fliesszustand in Gerinnen mit grossem
Gefiille (Pressure, energy, and flow type in channels with high gradients), Wasserkrafl, und Wasserwirtschaft, Munich, vol. 30, no. 7, pp. 78--82, 1935.
26. J. H. Douma: Discussion on Open channel flow at high velocities, by L. Standish
Hall, in Entra.inment of atr in flowing water: a symposium,T1'ansactions, American
Society of Civil Engineers, vol. 108, pp. 1462-1473, 1943.
LeI!.'foI.l-
.2.
h}jc/rct,;.d.i :/'t'--10
f? .
Yl
eN
c: c ,'" Y . 5 .
W/ibn
CHAPTER
(,
<
Y-'.c,tt/).Je1'
@
,
_ P
,~,> J:
..
::>..;t;.1".!/f"''' V"'r1.tl.-7
HI]~II tt'l"1
'I
~y.& o,C;...l."< ~
"'"'t
iJ
d
/,1 (f?)e.
'!, 6 rtf..~'( .
/J
'.
IS
r1.!1 /!.~Ci//.
bkrl~)~""'p1)
t'1' .
'7....
('/.
Saelioil 0
"
.) (
ftvtAP.
f
< .5q,-'Ui4'V"t
d.c;ot/..,
above a datum, the pressure head, and the velocity head. For example,
with respect to the datum plane, the totltl head H at a section 0 conta.ining point A on a streamline of flow in a channel of large slope (Fig. 3-1)
may be written
H
= ZA +
d A, cos
8+ V2;
(3-1)
where ZA is the elevation of pain t A above the datum plane, dA is the depth
of point A below the, water surface measured along the channel section,
8 is the slope angle of the channel bottom, and V.{2j2g is the velocity hea~
of the flow in the streamline passing through A.
In general, every streamline passing through a channel section will have
. 39
_. r"',
40
BASIC PRINCIPLES
U1
z + rl cos ()
.
S. = d cos
or, for 8: channel of small slope and a
.'
+ c ;P2g
.-
(3-3)
Consid~r !lOW a prismatic channel of large slope (Fig. 3-1). The line
r~pl'eSelltlltgJhL~_~~va~ion of the tota.Lhea..cLO-f.Jlow is the lill.&gy_line.
;rhe slope of the hne IS known as the energy g'radient, denoted by Sf.
The slope oLthe water surface is de;noted by Sw and the slope of the
channel bottom l by So = sin () . In uniform
10'"
- . ()
.
", S J-- S w-- S o-Sln.
According to the principle of conservation of energy, the total energy
head at the upstream section 1 should be eql.lal to the total energy head
at t~e downstream seetion 2 plus the loss of energy hf between the two
sectIons; or
...
I,
Zl
+d
V;2
+ al ~2g
.
cos /1
Z2
V"
+ d z COs /1 + az ---.l:
2g
.L
.
h
'f
(3-4)
ZI
V;
+ YI .+ al -2
=
g
Z2
.
+ Y2 + a2 -V~
+h
2g
.f
(3-5)
Wh~n
al
.
+
V I2
. . Yl+ -2
g
Zl
..
Z2
+y~
V 22
+ -2g
const
(3-())
, \
'JI
?y
V2
a 2g
,.
(3-7)
1,
E =y+-
(3-2)
H = z +d
V!
V
+ a --.:.
2g
41
(3-8)
. 2g
which illdic!l.tes that the specific energy is equal to the sum of the depth
of water and the velocity head. For simplicity, the following discussion'
will be based on Eq. (3-8) for a channel of small slope. Since V = Q/ A,
Eq. (3-8) may be written E = y + Q2/2gAz. It can.be seen that, for a
given channel section and discharge Q, the specific energy in a channel
section is a function of the depth of flow only.
When the depth of flow is plotted against the specific energy for a given
channel section and discharge, a specific-energy CUTve (Fig. 3-2) is obtained.
This curve has ,.two limbs AGand BG. 'rh~ .lifl.lb AC.~J)proache;3the
,horiz.ontall\-xis asymptoticallx toward the righ.t. The limb Be approaches
the line OD as. it extends l.lpward and to the right. Line OD is a line that
passes through the origin and !ul:,s _ananglfi Qf lr~eJlP.ati.QJUiQl!j11 to '!5~!-
For a channel of large slope, the angle of inclination of the line OD will
be different from 45. (Why?) At any point P on this curve, the ordinate represehts the depth, and the: ,abscissa represents the specific
energy, which is equal to the snm:6f the pressure head y a.nd the'velocity
head V 2 /2g.
.
The curve shows that, for a given specific energy, there are tW9 possible
depths,. for instance, the low stage YI and the high stage y~. The low
stage is called the alternate depth of the high stage, and vice vel'sa. At
point G; the specific energy is a n~inimum. It will be proved Inter that
this condition of minimllriJ. specific energy corresponds to the critical
state of fio.v. Thus, at the critical state the two alternate depths
apparently become one, which is known as the critical depth 1/e. When the
depth of ;fiowis greater than the critical depth, the velocity of flow is less
than the critical velocity for the given! discharge, and, hence, th:e fiow js
sub criticaL When the depth of flow is less than the critical depth, the
flow is supercritical. Hence, YI is the. depth of a supercritical flow, and
1/2 is the depth .of a subcritical flow. '
'
If the discrarge changes, the specific energy will be changed aucordcurves A' B' and A"
(Fig. 3-2) represent positions of
ingly. The
1wo
B."
id 1912.
42
BASIC PRINCIPLES
,the specific-energy curve when the discharge is less 'and greater, respectively, than the discharge used for the construction of the curve AB.
3-3. Criterion for a Critical State of Flow. The critical state of flow '
has been defined (Art. 1-3) as the condition for which the Froude number
is equal to uni~y. A more common definition is that it is the state of flow
at which the specific energy is a minimum for a given discharge. 1 A
ro----
VI
D
2g - 2
min~mum,
N\':;"
,~
or
(3-11)
j
Supercritica I
ct
Vl
2g
D cos
2
ct,
the criterion
(3-12)
where D is the hydraulic depth of the wateI: area normal to the channel
bottom. In this case, the Froude number may be defined as
,Fw. 3-2. Specific-energy curve.
(3-13)
theoretical criterion for critical flow may be developed from this definition
as follows:
Since V = Q/ A, Eq. (3-8), the equation for specific energy in a
channel of small slope with a = 1J may be written
Q2
217A -
E=y+--,
(3-9)
or
dE = 1 _ ~ dA = I _ VI dA
dy
gA3 dy
gAdy
. "
The differential water area dll near the free surface (Fig. 3-2) is equal
to T dy. Now dA/dy = T, and the hydraulic depth D:= A/T; so the
above equation becomes
'
dE
V2T
'V2
-=1--=1-dy
'gA ,
gD
I Th~ concept of critica.l depth bas'ed on the theorem of minimurh energy
introdl.).ced by BOss [3J.
'
Wall
first
,
,<
(j
(3-10)
This is the criterion for critical flow, which states that at the critical state
of flow, the velocity head is equal to half the hydraulic depth. The above
eq~ation may also be written V!...;gJ5 == 1,which means F = 1; this is
the definition of critical flow given previously (Art. 1-3).
If the above criterion is to .beused in any problem, the following conditions must be satisfied: (1) flow parallel or gradually val:ied, (2) channel
01 small stope, alld (3) energy coefficient assumed to be unity, If the
energy coefficient is not assumed to be unity, the criti~al-flow criterion is
y -----I
~~1
43
()
I,
1
"
i.,'
I'
,I
44
\I,
\
BASIC PRINCIPLES
ENERGY AND
appears, connecting the water surfaces before and after the 'drop.' The
point of inflection on the reverse curve marks the approximate position
of the cdtical depth nt which the specific energy is a minimum and the
flow passes from a "SUbcritical state to a supercritical state.
The free ove'rfall (Fig. 3-3) is a special case of the hydraulic drop. It
occurs 'where the bottom of a fiat' channel is discontinued. As the free
overfall entel'S the air in the form of a nappo, there will be no reverse curve
in the water surface until it strikes some object at a lower eleva.tion. It.
is the law of nature that, if no energy were added from the outside, the
y
v'{ijl,/::~LJ)
'7
f"'~'''1i
MOME~TUM
PRINCIPLES -
45
-,
-::!~
_ Ii
""c.
<U
'0
o "
"
""
o.tl:
/Theore.licOi water surfoce
Ir.ilial
deplh
Specific-ene'g)'
ju~np
CUrye
Hydraulic jump
Specific-force
Culye
'
46
BASIC PRINCIPLES
depth after the jump. The depth before the jump is called the initial
.depth y 1 and that after the jump is called .the .sequent depOt Y2; The initial
and sequent depths VI and Y2 are shown on the specific-energy curve
(Fig. 3-4). They should he distinguished from the altern!J.te. A~p~hs
YI and Y2'! whi(J~~e the two nossible depth~ fo!:....the same specific ener~l'
_.~~itial and s~lent depths are actual depths before and after a jump
in which ~rgy lOss b.E is invoh-ed. In other words, the specific
-energy E 1 at the initial depth Vl is greater tha.n the specific energy
at the sequent depth .y~ by an amount equal to the energy loss AE. II
there were 110 energy losses, the initial and sequent depths would become
identical with the alternate depths in a prismatic channel.
3-5. Energy in NQnprismatic Channels. In preceding discussions the
channel has been assumed prismatic so t.hat one specific-el~ergy curve
could be applied to evil sections of the channel. For non prismatic channels, however, the channel section varies along the length of the channel
and. hence, the specific-energy curve differs from section to section. This
cbm'plication can be seen in a three-dimensional plot of the energy curves
along the given reach of a nonpl'ismatic channel.
For demonstrative purposes, a nOllprismatic channel with variable
slope is taken as an example, in which agradually varied flow is carried
from a, stibcritical state to a supercritical state: (Fig. 3-5) .. The vertical
profile of the channel along its center line is plotted on the Hx plane with
the x axis chosen as the datum. For a variable-.slope channel, it is more
convenient to plot the total energy head H = z + y + V Z/2g, instead of
the specific energy, against the depth of flow on the By plane. For
simplicity, the pressure correction due to the slope a.ngle and curvature of
flow is ignored in this discussion. An energy line is then plotted on the
Hx plane below a line parallel to the x axis and passing through the initial
total head at the H axis. The exact position of the energy line depends
on the energy losses along the channel. Four channel sections are then
selected and four energy curves for these sections are plotted in the Hy
planes ~ shown. The initial section 0 is an upstream section in the
sUbcritical-flow region. The two depths corresponding to a given total
energy H 0 can be obtained from the energy curve. Shice this section is
in the subCl'itical-flow region, the high stage yo should be the actual depth
of flow, whereas the low stage is the alternate depth. Similarly, the
alternate depths in other sections can be obta,ined. In the downstream
sections rand 2, the low stages Yl and Y2 are the actual depths of flow singe
they are in the supercritical-flow region. The critical depth at each se ytion can also be obtained from the energy curve at the point of minimum
energy. At &ection C the critical flow occurs, and the depth y, is the
critical depth. On the H x plane, varioUs lines can finally be' plotted;
showing the channel bottom, water surface, critical-depth line, and
47
1
\
II
I
I
i
I
!
!
II
{
\
:l
48
BASIC PRINCIPLES
.
,
.
Sol1ttiDn. From the given data; the total'energy in the approaching flow meas~red
above the channel bottom i~ u: = 5
{100/(5 ?< 1O)1'/2g = 5.062 ft. This energy is
kept constant throughout l,lIC contractIon, since energy losses are negligible. A horizontal energy line showing the elevation'of the tobl head is, therefore, drawn on the
channel profile (Fig. 3-6).
.
"
49
., critical depth at t}Jis section is equal to the total head divided by 1.5 (Prob. 3-3), or
5.062/1.5 = 3.375 ft. By :r::q, (3-10), the critical velocity ts'eqllnl to V, = V3.375g =
10.4.5 fps. Hence, the width of this' critical section should be 100/(10.45 X 3.38) =
2.83 fi.
'
With the size of the mid-section determined, the side walls of the contraction can
bo drawn in ,with straight lines. The lm~ and high stages at each section are then
computed by the equation previously'given, As the flow upstream from the critical
section is subcritical, lts water surface should follow the high stage. Downstream
from the critical secti6n, the flow is sllpercritical and its Burface profile 'follows the
low-stage line.
The criti.::al-depth line is shown to sepai'ate the high from the low stage or the subCl'itical from the stipercritical region of flow. On the basis of Eq. (3-10), the critica.l
depth can be computed from the equation
(lOO/by,),
2g
y,
'2
. y = {/lO,OOO
,
"gb%
or
where' b is the width of the channel, which can be measured from the plan,
It should be noted :that the vertical scale of the channel profile is greatly exaggerated. Furthermore, the outline of the gradual hydraulic drop is only theoretical,
based on the theory of parallel flow. In reality, the flow near the drop is more or less
curvilinear, and the .actual profile would deviate from th/'! theoretical one..
.
This example also serves to demonstrate 11 method of designing a channel transition
(Arts. ll-5 to 11-7), The designer may fit any type of contraction walls he desires
to suit a given flow profile, or vice versa,
.
FIG. 3-6. Energy principle applied to a channel contraction (a) without gr:adual
hydraulic drop; (b) with gradual hydraulic .(hop. '
cfs,
The alternate depths for the given tot ..l energy. e..n be computed by Eq. (3-9) as
follows:
. '
"
100'
5.062 = Y + 2g(by)'
or
. , - 5 06'> "
y
. ~Y
+ 155.25
b'.
This is a cubic equtl.tion in which b is the width of the channel. At the entrance sec-'
tion, where b = 10 ft, its s61ution gives two positive roots: a low stage 'YI = 0.589 ft,
which is the altemate depth; and ahigh stage Y2 := 5.00 ft, which is the depth of flow.
At the exit section ,where 11 = 8 ft, this. equation gives a low ati\ge YI =. 0.750 ftiand a
high stage Y. = 4.964 ft.
!
When no gr~dual hydraulic drop is allowed in'the contraction (Fig. 3-6a), the1depth
of flow at the exit section.should be kept at the high stage, as shown, The high stages
for other intermldiate sections are then compute~ by the above equation, whicli giveB
the flow-surface ptofile. Similarly, the low stages are computed by the aboye procedure and indicated by the alternate-depth line~
;
When a grodua:~ hydrauiic drop is desired in the contraction (Fig. 3-6b), theldepth
of flow at the exit ~ection should be at the low stage, Since the point of inAection of
the drop or 11. critipal section is maintained at th~ mid-section of the c~ntracti<?n, the
'~l 0(
(3-14)
J(.. 1
.;.. "'>l~)(qt
cit!
)>1'
_, Yn
,I
I
!
!,
C(
, .
,
where Q, w, and 1:::' are :as.. previously defined, with subscl'ipts refe1'l'in'g to '
sectionr-l~nd
P! and P 2 are the resultants of pressures acting on the
two sections; W is the weight of water enclosed between the sections; and
F! is the total external force QL friE,tion and reslstanc~~ing-,ilong the
lLUliace of conta,Qt bet\V'een.the water and the cha;nl)el.The above equa-'
tion. .is known as the m;omentum
equation, l
I
,
2;
The application of the inomentum principle was first suggested by Belanger [5J.
50
r-
51
BASIC PRINCIPLES
PI
;'~Wb1l1:l.
P z = 7.wb yz2
FI = wh/by
. and
Assume
where h/ is the friction head and ii is the avera.ge depth, or (YI+ Y2)/2.
The discharge ~hrcillgh the reach may be taken as the produot of the
average velocity and the avera.ge area, or
~ ~i(Vl
+ V~)bii
Also, it is evident (Fig. 3-7) that eha weight of the body of water is
W = wbfjL
and
sin
{j
.\
(3-16)1
I
Co t1r
I
1.
1-::
fA
AZ)o
hdA= 1
+~
fA
AzJ~
cdA
(3-15)
where z is the depth of the centroid of the wa.ter area A below the free
surface, h is the pressure head on the elementary area dA, andc is the
pressure-head correction [Eq. (2-9)1, It can easily be seen th'1t pI is
!.Treater than 1.0 for ooncave flow, less than 1.0 for
and equai
=~~
to 1.0 f <2.r...1lllJ:allel flow:
It can be shown that the momentum e.quation is similar to the energy
equation when ~pplied to certain flow problems. In this case, a gradually
varied flow.is considered; accordingly, the pressure distribution in the
sections may be assumed hydrostatic, and (1' = 1. Also, the slope of the
. channel is ass~ed relatively small. 1 Thus, in the short reach of a
1 If the slope a.!l~le 8 is large, then PI = }iwdl~ cos 0 and P. = ;fW~1 cos O,where
ell and el, are the Beptbs of flow section and cos IJ is a. correction factor (Art. 2-10).
,
,a.
tI
/::}.
A
j'(,h .{- c)id JI. = " j.Ad4 . 1 <.l(cd
.4
A >(:i:.;.
! c dA
()
j:
IC!cr-'T(~ 0
'tl.
t'Yltkfr; ~t-Z,a:k.~
1.iJ" ~ e"" t Jt-l'!'
~.~~?}
r +~ (colA
1/ z
)
'"
52
~.
BASIC PRINCIPLES
53
qw
g
(!1.Y2 _!J_)
YI
= HWYI' -
MWy2 2
OIl
the top
Hwh(2y, - h)
.
V2Y (-4-h ) l~ H%
Y'T
(3-17)
//.
Qw (V2 -- 17 1)
g
PI - P2
PI = wi 1 A. 1
a.nd
P 2 = wi 2 A 2
where i l a~d i2 are the distances of the centroids of the respective wa.ter
areas Al and Az below the surface of flow. Also, Y I = Q/ Al a.nd
V 2 = Q/ A z Then, the above momentum equation may be written
p ..
Hwh[y,
+- (YI
- 11,)] =
Hwh(2YI - h)
The accuracy (If the last assumption has been checked ~xperimentally [61. If the
momentum equation (3-14) is applied to the body of water between the upstream
J
..
. '- 'Y)
c'Af:i
j .. /.::
.r-?
,rgf1lJ,o/,J<.
g.fi 2
'The v:llue of the coefficient actually depends on many factors: mainly, the rounding of the upstream corner, the length and slope of the weir crest, and the height of
the weir. Many experiments on bl'oad~crested weirs haye been performed. From
several of the wen-known experiments King [7J has interpolated the data and prepared tables for the coefficient uncler various conditions. A comprehensive analysis
including more recent data and a presentation of the results for practical applications
were made by Tracy [8J. The well"known experiments all broad-crested weirs are
(1) Bl].zin tests performed in Dijon, France, in 188.6 [9]; (2) U.B.D.lV.Bi Cornell tests
performed at Cornell University in 1899 by the U.S. Deep Waterways Board under
the direction of G. W. Rafter, and U.S.G.B. Corncllle.sts performedby the U.S. Geological Survey under the direction of Robert E. Horton in 1903 [101; (3) Michigan
tests performed at the University of Michigan during H)28-1929 [11]; and (4) MinneBola and Washington tests performed,'respectively, at the university of Minnesota and
Washington State University [6]. For some formulas and coefficients of discharge
developed in the U.S.S.R., see [12]. For an analytical treatment of the problem, lIee
'[13].
55
BASIC PRINCIPLES
The two sides of Eq. (3-18) are analogous and, hence, may be expressed
for ahy channel section by a general function
"
Qz'
54
- i t zA
dF
"(3-19)
u4'_
'
This function cOMists of two terms. The first term is the momentum of
the flow passing through the channel section per unit time per unit weight
of water, and the second is !:.l!.e for.J<.ELrillr....unit weigb.:Lm w~t.!lr. Since
, both terms al~e essentially force per unit weight of water, their sum may
, , be ,c~lled the 8PlJcifjE.l(jr.c~.1 ,Accordingly,
(3-18) may be expressfld
Q2 dA ,d(zA)
dy T --elY
gP
'
of
dF
_ Q! dA
gA2 dy
Since dA/dy
reduced to
c
o
~=-:-_.
45 for ()
= T,
Q/ A
V,and AfT
+A
=
i"
!i
Ez 'lEiE:,
channel of
,......,
zero or small
slope
(0]
l.bl
(a) Specific-
'
I,
I
I (
,I
= 0
This is the criterion for the criMeal state of How, derived e~rlier (Art. 3-3).
Therefore, it is proved that the dep~h at the minimum va-lueaf the specific
force is the critical depth.I It may also be stated that at the critical state
of flow the specific'jorce is a minim.'um for the given discharge.
'.. ,I
"'i
56
!'
BASIC PRINCIPLES
i
\
I
57
(3-18)
2,
Factoring,
Then, let
~V + 1l! _
( Y.:/
y,
(3-20)
2F = 0
I
For a. J:!.iv:? ,F'roude ?u~ber FI of the apprQaching flow, the xatio of the sequent. depth
to the IlUt1l3l depth IS given by the above equa.tion.
It should be understood that the momentum principle ill used in this solutiou because
the hydraulic jump involves a high amount of internal-energy losses whioh cannot be
evaluated in the energy equation., '
.The jOilit. UIlB of the specific-energy curve aud the specific-force curve helps to d~ter
~ne gra.plllcally the energy loss involved in the hydraulic jump for II. given appro8.chmgflow. For the given approaching depth 11., points P, ana. P,' are located on the
spe~itic-force curve and the spec~c~energy curve, respectively (;Fig, 3-4). The point
Pi'!gives the initial energy content E , . Dralv a vertical line, passing through the
P?int PI and intercepting the upper limb of the specific-force curve at point P" which
gives the sequent depth 1)2. 'Then, draw a horizontal line passing through the point
P a and intercepting the specific-energy curve a.t point P.", which: gives the energy content E, after the j';lmp. The energy loss in the jump is then equal to El - E"
represented by / l E . :
'
a-8.Momentum
priIicipl~
Applied to
Nonprisrnati~
Channels.
The
sp~cifj.c force, like the specific' energy, varies with the snape ofthechannel
FIG. 3-10. (mnergy and'momentum principles applied to II. channel expansion (a.) with
hydraulic jump; (b) without hydraulic jump. .
.:
o~ 4 n (Fir;. 3~lO). Determine the flow; profile in the transition if th~ frictional loss
through tl;\e tra.n&'ition is negligible. If, a hyqraulic jump occurs in the transition,
how C!l.n i~ be eliminated?
-:'
,
S"l'Uti"n~ From the given data, the tptal energy with respect to the channel bot~'
tom in the approaching flow is E ~ 0.5 + [IOO/(O.S X 8))'/2g = 10.207 It, and in the
dowiistrea.i:n,
E - 4,0 + (100/(4 X 10)Ji/2y
= 4.097 ft. ,It is appnhnt that this
,
,
58
'BA.SIC PRINCIPLES
IlMrgy difference 01 6.110 It must be dissipated through the transition by ~ome means,
since the frictlontll lru;s is negligible. Furthermore, the Fro.ude numbers 6.24 and
0.22 of the 8.pproaching a,nd downstrea.m flo'YS are, respectively, greater and less than
unity, indicll.ting IJ. change of the flow from supercritical to Bubcritica.l. Therefore, a
hydra.ulic jump can be expected to occur to dissipate the energy difference and to
effect a change in the flow st,ate, Whether this jump will occur within the transition
or in the upstream or. the downstream channel is, however, to be disclosed by fur~her
analysis.
.
for instance, by bolting orOM ,!limbers to the bottom. of the transition. It can be
!LSSumed ill this a'(ample that the energy diff-crcnce of 6.1 10 ft is dissipated uniformly
in tlte transition by artificial roughness. TllUB, the energy line in the tl'ansition is
simply a straight line joining the total heads of the two end sections (Fig. 3-lOb). For.
deSign p1l1-poses, it is convenient first to assUlne the lIow profile and then to proportion
the dimensions of the tranllition so 'that the jump can be eliminated, In proportioning the transition, the jump is eliminated either by varying the wid~h or by raising
the bottom of the transition. In this example. it is assumed that the bottom is to
be raised, 01' "humped" (Fig.3-10b). The subsequent procedure of the compuGlttion
is to (1) a,!lsume the flow profile; (2) compute the velocity head, which is equal to the
difference. between the total heo.d B.nd the water-surface eleva.tion, at a llumbe!' of
sel",cted sections; (3) compute the velocity and then the water area and depth of flow
foreayh section; (4) determine the elevation of the bottom or the transition, which is
equal to the devation of the water surfar.:e minus the depth of flow; (5) compute the
o.lternate depth, ~ince the bottom of the transition is fixed; and (6) compute F'l and
F, lines fo!' the low and high stages, an.d plot them on II convenient Sl'ale. It can be
seen that the two F' lines inter3ect and' become tangent to each other a.t a critical section, where the :flow cha.nges from low to high stage, that is, from 8upercritical to subcritical stale. If the critic31-depthlinc is plotted, it will intersect the alternate-depth
line and th~ water surface simultaneously at the critical section. Bl\Scd on the critica.l-depth line, a line of minimum speCific energy can also be construc,ted. This line
should be tangent to the total-energy line at the critical flection.
TABLE 3-1.
EXAMPLE
Section
width
b, ft
3-4
Low stage
'II" ft
PI
0.500
78.6
78.7
78.8
78.8
78.8
for E = 10.207
8.00
8,50
9.00
9.50
0.419
10,00
0.398
o 470
0.4'.1:3
DESCRIBED
IN
.,
High stage
1/2, ft
for E = 4.097
3.940
3.960
,3.979
3.987
4.000
F,
71.9
75.9
79.9
83.6
87.8
Take 'five sections of the transition with their widths shown in Table 3-1. For
the total approaching enel'gy of 1O.20i ft,.the low stage YI for each section call be computed by means of Eq. (3-8) or (3-9), or
"':"""-1::--=::'-
+ VI,:""
59
PRODLEMS
3-1. With reference to a channel of small slope !l.ud a section shown in Fig. 2-2, (a)
construct a family of specific-energy cllrves f(or Q = 0, 50, 100, 200, 300, and 400 ds,
(b) draw the locml of the critical-depth point on these curves, (c) plot a curve of the
critical depth against the'diilcharge, and (d) plot a family of curves of alternate depths,
Yl vs. V~, for the given discharges_
.
8-2. Construct the specific-energy curve for a 36-in. pipe carrying 3:n open-channel
flow of 20 cis (a) on It. flat slope, and (bj on a 30 slope.
S-B. Show that at the critical state of How tha specific-energy head in a rectangular
channeli;; equal to 1.5 times the depth of flow, assuming zero slope and a = 1.
3-4. Derive the equations for the lOCI15 of the critiMl-depth point on the specificenergy curve and for the curve of critical depth vs. discharge, as oota.ined in Prob. 3-L
3-6. Prove
(3~12).
3-6. PrC/ve Eq. (3-13).
3-7. Prove that at the critica.l state o( flow the discharge is II ma::dmum for a given
specific energy. 1
.
3-8. Show that 'the relation between the altc;rnate depths Y1,and 1/. in !1 rectangular
channel ca.n be expressed by
.
10.207
+ V2 =
4.097
Thelow- and high-stage linea are then construeted slong with tlle energy lines (Fig.
3-lOa). After these stage and energy lines are determined, the specific forces P! and
F. for low and high stages, respectively, at ea.ch section are computed and plotte~ to
any convenient scale and datum. The hydraulic jump must, occur where the specific
'forces for the low and IUgh stages are equal, or at the intersection of the F lines. At.
this <laction the water surface at low ,stage will jump to the high stage, as indicated
by a vertic,al line. (Fig. 3-10a.). Actuall:)" however, the jump will take place over a
'short distance,
shown by the dotted line.' The energy loss in the jump is represented bj' the vertical intercept between the upstr.9a.m and downstream energy lines,
which is equal to 6.110 ftl covering the energy; difference between the flows in the conneting channels. By varying tlie shape of the cross sections of the connecting chan:
nels the location of the intersection of theF lines, or the position of the jump, can be
altered. Changing the depth of flow in the downstrea.m channel will also change the
po~ition of tile jump. Generally, an increase in the downst~eam depth will mon
the. jump upstream, and a decreaSe in the depth will move the jump do:wnstream.
The hydraulic jump (lan be eliminated if the energy loss can be dissipated gradually
And smoothly. This can. be done by introducing proper roughness in the transition,
as
(3-22)
where Y. is the critical depth. Using values of y.!Yc as ordinates !tud of Yz/Yt as
abscissas, .construct a dimensiouleSiil graph for the above equation rand study its
characteristics.
'
The .concepG of critical depth based. on the theorem of maximum diacharge was
first introduced by Belanger [20J.
1
r
~
'1
I \'
.1
".!
60
!!ABIC PRINCIPLES
,
3-9. Solve the problem given in Example 3-1 (a) if there is a. total energy 108s of
0.60 ft unifol'Jnly diStributed throughout the lel!gth of the contraction, and (0) if
a gradual hydraulic drop is desired with its point of inflection at !l. diStance 20 it
upstream from tlie exit section.
3-10. Applying the momentumpl'inciple and the continuity equation to the analysis
of a submerged hydrllulic jump which occurs at the sluice outlet in a rectangular channel (Fig. 3-11), prove that
where y. is thesubme!"ged depth; y, is the height of sluice-gate opening; y, is th~ tailwater depth; und F,'
g~/gy!3, IJ being the discharge per unit width of the channel.
Neglect the channel-bed friction i,.
AE
()J2
-YI)' ./'
4j/lY2
_(3-24)*
3-12, If !I. hydrAulic jump is forllled on the ,horizontal floor at the toe of the spillway descl'ibed in Prob. 2-11, determine the sequent depUl and the energy loss involved
in the jllmp.
3-ill. ,With reference to a Ilha.nnel of small slope and Ll. section shown in Fig. 2-2,
(a) construct afamily of specific-force cUI'ves for Q = 0, 50, 100, 200, SOO, and 400 cfs,
and (b) plot a family of curves of initial depth against sequent depth for the given
discharges.
3~14. Construct the specific-force curve for a 36.-in. pipe c,nrying an open-channel
flow of 20 cfs on Ii small sIope.
3-lD. Prove Eq. (3-15).
3-16. Using the momentum principle, show th!Lt the Froude number of a parallel
or gradually Vllried flow in a cha.nnel of slope !Lngle IJ may be defined by
V
VgD cos 8/13
61
'
(3-25)
* This formula wliU, shown by Brcsse early in 1850 115J. At the slime time Bresse
illtrodu,ced the concept of critical depth, as a. dllpth at which the subcritical flow
chl>nges to BUpel'criticai, or vice versa.
where V is the mean veloility, D is the hydraulic depth of the section, and t3 is the
'momentum coefficif;nt for nonuniform velocity distribution. '
3-1'1. For eliminating the hydraulic jump in Example 3-4, the flow profile is
assumed to be composed of two Feversed circular curves tangent to el1ch other at the
middle aection of the transition and also to the water surfaces "in the connecting channels at the two ends of the transition. Verify the computation (shown in scale on
Fig. 3-100).
3-18. A frictional loss of 1.0 ft is p,Ssumed to be uniformly distributed along the, '
length of the tra.nsition in EXMllple 3-4. Determine the flow pro.file in the tmnsition.
REFERENCES
62
BASIC PRINCIPLES
17. Cll!irles Ja.eger: Contribution a l'etude des courants liquides a 8urf8.c.oe libre
(Contributil1nto the St11dy of Iree-surface liquid flo;,vs), RIWWi genlrale .de
L'hydrauliq1l.{l PI'tTis, yol. 0, no. 33, pp. 111-120; no. 34, pp. 139-153, 1943.
18. Charles Jae~e.l: De l'impulsion toto.le at de ses rapports a,yac I'energie totaled'un
cou:u,nt liquide a surface libre (The tota.l impulse and its relations with the toto.l
energy of a fl"ee-Ilurface liquid flow), RlWue gtntrale de l'hydravliq1l.e, Paris, voL 13, .
no. 37, pp ..12-1\); no. 38, pp. 86-87; no. 39, pp. 143-151; no. 40, pp. 191-197;
llO . 41, pp. 257-2iH, 1947.
HI. Etienne Crausse; "lIydr3ulique des canaux deco\lverts en regime permanent"
("Hyciro.ulio5 of Open Ch!lJlneL'! with Steady Flow"), Editions Eyrolles, Paris,
1951, pp. 111-112. .
,
20. J. B. Bela.nger: Notes sur Ie cours d ' hydraulique (Notes on the course in hydrl:lulies), llHmoire, gmt" Na.tionale de$ Pont8 e~ ChaU$88I3$, Paris, 1849-1850, pp~ 32-33.
CHApTER
I~-
.,)
I.
'.
64
I
, \
I
BASIC .PRINCIPLES
the change. It can be observed also that, when the flow is near the
critical state, the water sUl'face appears unstable and wavy. Such
phenomena are generally caused by the minor chnnges in energy due to
variations in channel roughness, cross section, slope, 01' deposits of
sediment or debris. In the design of a channel, if the depth is found at
or neart,he critical depth fOl" a great length of the chinnel, the shape or
slope of the channel should be altered,if practicable, in order to secure
greater stability.
The criterion for a critical state of flow (Ar~. 3-:3) is the basis for the
computation of critical flow/which will be explained in subsequent
articles. Two major applications of critiClll-fiow theory nre flow control
and flow measurement, which will also be discussed in this chapter.
4-2. The Section Factor for Critical-flow Computation. Substituting
V Q/ A in Eq. (3-10) and Simplifying,
11
. II
tJ
.
(4-1)
...g.
.<i
~~,..
. Q
z=v'g
"'1
..,
c
--:.~
0,-,
'"!"
-0
";l
'-'
:3
... 't
:5
..,... . ..c::..,"
'w .5
...
..,..':! .~...
Q
tj)
. Q
z=-v'g!cY.
('1-2)
"'0 ...
"OJ
'0
....o
Vg
fIl
-<
.j,
ci
~
(5
(4-3)
N;'
.s
OJ
Q=z:(2
1>
a '" ~...
dt5;;:.. ::>
a.
Q =Z
"...
'0
(4-4)
ci
OpI ~ puo' q/,( lO SGnlOI\
65
66
.
.
CRITICAL :FLOW: ITS C.OMPUTATION AND. APPLICATIONS
BASIC PRINCIPLES
This equation indicates that the value of M for the trapezoidal section is
a function of z and y/b. For values of z = 0, 0.5, 1.0, 1.5, 2.0, 2.5, 3.0,
and 4.0, a family of curves for 111 versus y/b are constructed (Fig. 4~2).
These curves in.dicate.that the value of JI.f varies in a range froni 3.0 to 5.0.
S.C , . . - - -.......--.,-----,---7""',--,"'7'1rm-"----,
5.0
4.0 I-----.... ~-'-.-.----- --'-----+.........,f'--I'--7'-HI+-+------j
will
1,0
o.a
0
~
,..
'0
c
0
......a,..
...
(4-6)
dOn Z)
3 l'
--ay- = 2
J.-
= JL
.A
.2
3[1
o,oe
O.os
0.05
0.04
vi A/T, and
0.03
(4-8)
0.02 .
25
3.0
3.5
4.0
Volues of M
4.5
5.0
5.5
,!
(4-9)
0.2
0.1
(4-7)
1 dT
2T
0.3
."
Equating the r~ght sides of Eqs. (4-7) an~ (4-8) and solving for M,
M
0.6
0.5
0,4
.,.
2.0
2y
5.0
Example 4-1. Derive arl equatioI! showing critic:u dillcharge through a rectangular
cha.nnel section in terms of the cha.nnel width and the total head.
Sol~tion. For the rectangular section, Tal?le 2-1 gives the sQcti6n fantor Z = by!.J;.
At the critica.l state'of flow, the depth y
HI1.5 (see Frob. 3-3). Substituting these
expressions in Eq. (4-3), using g = 32.16, and simplifying, we find that the cri~icar
discharge is
Q.
3.087bHu
. (4-5)
dif" "'"
67
. (4-10)*
I
z(1/(b) may be constructed. It is obvious that this curve:
be identical with the
1 in Fiilj. 4-2. Fo, convenience in application, however, a. family of
eurve for z
This equl).tion was a.lso developed indepe~dently by Chuga8v 13J. In this eq~e,
tion, M ca.n be regarded as a function of z(ylb):; a.ccordingly, a single curve aiM versus
I.
~
I
If.
as
&
pammeter.
58
BASIC,P~INClPLES
69
= 1[(10
10
+ 1/)
+ 2y
and
A = V(20
+ 2y)
The velocity is
Substituting the a.bove expressions for D and l'in Eq. (3-10) a,nd simplifying,
2,484(5
+ Y)
[!I(10
+ II)]!
Solving this equation for y by a trial-and-error procedure, Ye '"' 2.15 ft. This is the
critical depth. The corresponding area is A.
52.2 iV, !l.lld the critical velocity is
V< = 400/52.2 = 7.66 fps.
(4-11)
where Z 1 and
are section factors
for any two depths Yl and Yz of the
given section. This equation can
eesily be derived from Eq. (4-6).
In applying Eq. (4-11), a graphical m~thod is recommended instead of
direct computa.tion. This involves a logarit.hmic plotting of Z as ordinate against the depth as a.bscissa (Fig. 4~3). For most channels, except
for closed conduits with depth approaching !J, gradually closing crown and
some channels of peculiar shapes, the plot takes a more or less straightline fOl'm. The hydraulic exponent
equal to twice the slope of the
plot.ted stra.ight line. For a. depth approaching the gradually closing
ci'own of a closed conduit,the plot becomes a curve, and the hydraulic
exponent of a given depth is equal to twice the slope of the tangent to the
curve at that depth,
The hydraulic exponent M is described here only as a characteristic
v.alue of a channel section under the condition of critical flow. The
a:pplication of .this exponent will be further described in the computation
of gradually vari~d flow (Art. 10-2),
log y'
is
1:<'01'
f,
I
t
70
BASIC PRINCIPLES
71
If the channel has a critical slope (middle sketch in Fig. 4.5), then the
flow is initially uniform and critical throughout the channel. In the
presence of the dam, however, the flow through the pool Will be subcritical
and the pool surface will approach the horizontaL At the dowllstream
end a so-called drawdown ettrve will be developed, extending upstream
Flo'..,. oon<lilian controlled 01 the downstream e~
"I
rr---_
,
BOCkWO!il(
curve
1 Drowdawn
Zld~'"
= 0.226.
= 0.48
---.
---~
Depth of subcriticol
flow without dam
or
l.critIC<J~..s..2~..!.e_ctions;
4-5. Control of Flow. The control of fim:v in 3,11 open channel is defined
loosely in many v'ays. As used here the tertn means the establishment of
a definitive flow condit.ion in the channel or, more specifically, a definitive
relationship between the stage and the discharge of the flow. When the
contror of flow is achieved at a certain section of the channl?l, this section
is a control section. It will be shown later that the control section controls
tbeflow in such a way that it restricts the transmission of the effec~ of
changes in flow qondicion either in an upstream direction or in a downstream direction depending on the state of flow in the Chlll111el. Since the
control section holds a definitive st,age-discharge relationship, it is always
a suitable site for a gaging station and for developing the dis.charg6 raturtg
Gwrve, a curve representing the depth-discharge relationship at the gaging
station.
At the critical state of flow a definitive stage-discharge relationship
can be established and represented by Eq. (4-1). This equation shows
that the stage-discharge relationship is theoretically independent of. the
channel roughiless and other tlllcontrolled circumstances. Therefore, a
critical-flow section is a control section.
,
The location of the control section in a prismatic channel is generally
governed by the state of flow, wmchin turn is determined by the slope of
the channeL Take for an example a long straight prismatic channel in
which a pool is created bya dam across the chalUlel and the water flows.
over the dam through an overflow spillway (Fig. 4-.5). Three flow conditions in the channel are shown, representing the subcriliical, critical,
and supercritical flows, respectively. The slopes of the channel in the
three cases are, correspondingly, mild or ilubcritical, critical, and steep
or 8upercritical.
,\
Depth of supercrilicol
flow wilhoul dam
FIG,
!rom a section near the spillway crest and becoming asymptotic to the
pool level. .
If the channel has a sub critical slope (top sketch in Fig. 4-5), the flow
is initially subcritical. In the presence of the dam, the pool surface will
be fllrther raised for a long distance upstream from the pool in a so-called
backwater curve. The additional depth of water is required to build up
enough head to give the increased velocity necessary to pass water over
the spillway. This effect of backing up the water behind tb.e dam' is
.~
{l
72
BASIC PRINCIPLES
1J .~
i
~
Ii
,
><:
* ''.N
'
S;:-MMc<:)M:r.I~MP)(o)Mp')
,...r:OOOMp')~I'lO~tO<a\O
j.:!
r:- . . .
ae
Q)
co
QO Q
000'0
----roI.<;q~:1N~
..
..100C::OOOOQOOOOO
~_-:-l_."C"')C'}(t)~.:";ImMI.'I?
N-N
~ 0
;; " " 0 0
t: : : ;) ""'" -
C'l N
<:;I 0
C'1
0: 0
~'~
0- <:)
("1 cq
~oo(Oooooooooo
SECTION L-L
a:-t;~~
FIG, 4-6, Plan, elevation, ana dimensions o.f ,he Parsholl flume. (U.s. SlIil Conservation 8811>ilie [26J.} Plan and elevation of a concrete Parshall measuring fIl\me showing lettered dimensions as follows.:
W - size of flume in in, 01' ft; A = length of side wall of converging seetion;
,%A = dista.nce back"from end of orest to gage point: B
Ilxiallength of converging
section; (} = wid.th ofdownstl'ea.m end of flume; D = width of upstrea.m end of flume;
E = depth of flume; F
length ofthl'oat; G "" length of diverging section; K = difference in ele.... ation between lower eml oj' flume and crest; ,yI = length of approach /lOOI';
N ~ depth af depression in throat \:lelow crest; P = width between ends of curved
wing wa.lls; R
radius of curved willg wall; X = hol'izontn.l distanoe to E. gage
point from low point in tlu'Oll.t; Y "" verticil.l dilltance to. Eo ga.ge point from low
point in throl1t. See the tlible on the next page for actual dimellsions for various
sizes of flume."
(I'lC'lVlr')MC<')(?tnC")
. '"
q;)
~. ~
~~~~~~;::~~
'otI t--<,
~ CC .....
1:--
I~""~~~------------------I
'"
effect originlj.tillg from ~hepool will not extend far upstream. Instead,
the flow in the upstream channel will eontinue in:the downstz-eam direction at a supercritical state until the flow-s]lrface profile is actually below
the poolle'Tel;1 the~ .it will rise abruptly to the pool elevation in a hydrau-
.:~i:f
.=~M.~O
. O.
l~ .... ,......-tto':l(#)MM~"~"'')LI':l
ii~~~
~tO
:~
1 Itshoul:l: be noted that; the pool level in this case is ~ot. horizontal but curved.
The curved water surface has an 81 profile, which will be :described later (Art. 9-4).
73
:::fOOlQIt)OIOOf;!)Q
co
74
BASIC PRINCIPLES
lie jump. The backwater effect will not extend upstream through the
hydraulic jump. The fiowupstream from the jump.is governed entirely
by t.he Upst.lcam conditions.
The above example explains the important fact that on subcritical
slopes the effect of change in water-suI:face elevation downstream is
transmitted upstream by a ba:ckwater curve, whereas on supercritical
slopes the effect cannot. be transmitted far upstl:eam. The flow condition in a sub critical channel is affected by downstream conditions; but,
in a supercritical channel, the flow condition is dependent entirely upon
the condition upstream or at the place where water enters the channel.
Accordingly, tile control of flow is said to be at th,~ downst.ream end for
channels ,,,ith sub critical slope and at the upstream end 01' channels
with supercritical slope.
When t.he challnel is on a subcritical slope a control section at the
downstream end may be a critical5('1ction, sunh as that created on the top
of an overflow spillway. On a supercritic~l slope, the r;ontrol section at
the upstream end may also be a critical section, 2.S shown in the figure.
A sluice gate or an orifke or other control structure may also be used to
create a control section. It should be noted that whether the channel
slope is critical, subcritical, or supercritical will depend not only on the
measure of the actual slope.but also on the discharge 0; the depth of flow.
4-6. Flow Measurement. It was mentioned in the preceding article
that, at a critical control section, the relationship between the depth
and the discharge is defillitive, independent of the channel roughness and
other '.Incontrollable circumstances. Such a definitive stage-discharge
relationshipoffers.r. theoretical basis for the measurement of discharge in
open channels.
Based on the principle of critical flow, various devices for flow measurement have been developed. In such devices the critical depth is usually
created either by the construction of a low hump on the channel bottom,
such as a weir, or by a contraction 'in the cross section, such as a criticalflow fiwne. The use of a weir is a simple method,. but it causes relati vely
high head 109s. If water contains suspended particles, some will be
deposited in the upstream pool formed by the weir, result.ing in a ,gradual
change in the discharge coefficient. .These difficulties, however, can be
overcome at least partially by the use of a critical-flow flume.
,The critir.al-flow flume, also known as the Vent~ri flume, has been
designed in various forms.l : J+, is usually operated with an unsubmerged
,.
I'
ThrQat width
Equ.a.lion
3"
Q = 0.9921I.u4T
(4-12)
6"
Q=
Q=
Q~
Q=
{4-13}
(4-14)
9"
12" to 8'
10' to 50'
75
2.06H.ua
3.07 H.l.63
4WH.uuw''"
(3.6875W -I- 2.5)H.1.G
(4-15)
(4-16)
In the above equations Q is the free discharge in cfs Wis the width of
ft, ~nd Ha is the gage reading in ft. Whe:l the ratio of gage
readmg Hb (FIg. 4-6) to Ha exceeds the limits of 0.6 for 3-, 6-, and 9-in.
flumes, 0.7 for 1- to 8-ft flumes, and 0.8 for 10- to 50-ft flumes, the flow
becomes submerged. The effect of submergence is to reduce the discharge. In this case the discharge computed by the' above equatiollS
~ust be correc~~d by a negative quantity. The diagrams in Fig. 4-7
gIVe the correct1Ons for submergence for Pa.rshall flumes of various sizes.
The correction for the I-ft flume is made applicable to the larger flume's
by multiplying the correction for the 1-ft flume by the fador given
.
below for the particular size of the flume in use.
.thro~t in
Size of flu.me W, ft
1.5
2
3
4
6
8
Correction factor
1.0
1.4
1.8
2.4
3.1
4.3
5.4
Similarly, the correction for the 10-ft flume is made applicable to the
I Experin).ents on this type of mea.suring. device, then called the Ventu.ri flu rite, were
began by
M. Cone at,the hydraulic laboratory of the Colorado Agricultural Experiment Station, Fort Collins, Colo.' The initial studies were reported ill [20[ and {211.
~he name ." Parshall measuring flume" was adopted for the device oy the Execu~Ive. Commlttee of the !rrigation Division; American Society of Civil Eq.gineers, durmg 1~.Dec~mber meetIng of 1929. Further developments on the Parshall flume arc
d~scflbed by R. L. Parshall in [22] to [26J.
:
y.
The critical-flow flumes mentioned in the text are those developed and studied in
'the United States. Outstanding designs of critical-flow flumes were aIso developed
and tested by ,Jamcson [4,5J, Engel [ft,7], and Linford f8] in England; byCrump [9] and
Inglis [10] in India; by De Marchi [11,12], Contessini[ll], Nebbia [13-15], and Citrini
[16,17J in Italy; by Khafagi [18] in Switzerlandj and by Balloft;et [19J iIi. Argentina.
I
.i
..;.
~.
"
1
t
~l
76
BASIC PRINCIPLES
1.9 ZD
Discharge, cIs
77
Correction, cIs
(tt)
(OJ
Discharge, cIs
Correction. cis
( b)
(e)
Fro. 4-7. Diagrams ior computing Bubme.!"ged flow through ParshaU flumes of various
sizes. (Colorado Agricullural Ezperime:n.t Sta.tion [25J a.nd U.S. SlYil Conservation
Service [26).) (a) Diagram showing the rate of submerged flow. in cubic feet per
~econd,through a 3-in. Pa.rshall measuring flume.
(bl Dia.gram s!lOwing the rate of
submerged flow, in cubic feet per second, through a. 6-in. Pars!lall measuring flume.
(el Diagram showing the rate of submerged flow, in cubic feet per second, through a
9-in. P!lrshall measuring flume. (d) Dia.gra.m for computing the rate of $ubmerge.d
flow, in eubic feet per second, through a loft Parshall measl.lring flume,' (e) Diagram
for determining the correction in cubic feet per second Plll': 10 ft of crest for subDlergedflow discharge.
'" aa
c.
4.0
Oischarge, cIs
(~)
FIG. 4~7
4.5
5.0
5.5
6.0
BASIC PRINCIPLES
78
larger flumes by multiplying the correctioll for the lO~ft flume by the
factor given below for the paJ,ticular flume ill use.
CorrectiOlt factor
10
12
LO
1.2
15
1.5
20
2.0
25
30
40
2.5
3.0
4.0
5.0
50
;;
79
1
I
I
I
II
,
It is desiiable to set the crest of the Parshall flume so that free flow will
occur. If conditions do not permit free-floW operation, the percentage
of submergence Hb/H" should be kept, whenever possible, below the
practical limit of about 95 %, since the flume will not measure dependably
if the submergence is greater. The size and elevatiori of the crest depend
upon the discharge to be measured and upon the 5i;(;e of the flume and,
consequently, upon the loss of head through the flume. The loss or head
can be determined from ~he diagrams in Fig. 4-8. A practical example
(Example 4-5) will be given' to show the deterrcinatiol1 of the size and
elevat.ion of the flume crest.
Because of the contraction at the throat, the velocity of water flowing
through the flume is higher than that of the flow ill the channeL For
this. reason any sand or silt in suspension or rolled along the bottom call.
be carried through, leaving the flume free of deposit. When a heavy
burden of erosion debris is pl't;lsent in the stream, ho\yever, the Parshall
flume will become invalid like the weir, because deposition of ths debris
will produce undependable results. For use under such circumstances,
a modified Parsh.aLl flume known as the San DimGSfiume (27,28] has been
developed, which has the advantage of a self-cleaning mechanism for
heavily debris-laden flows in the stream.
For measuring open-channel flow in closed conduits, such as sewers and
covered irrigation canals, critical-flow flumes of special designs have been
proposed. Palmer and Bowlus [29-31J have developed several of these
flumes, including one'which is simplr a flat slab on the bottom and has no.
side contractions, one with a rectangular croSs section, and several witp
trapezoidal-shaped throats. Stevens [32], recommended a critical-flow
flume in which he used a blister-shaped' hump control on the bed of the
conduit to produce a critical flow over it. The frictional.1osses in this
design are believed to be very small.
Like. many measuring devices, the critical~flow flume has certain disadvantages. The flume CD.nnot be used directly with or, combined with
a head gate. "It is more expensive to build and requires more accurate
I,
II
l
.:~.
"\'..:,
~~:s'
4-8;cD()lio!l.Tgar;~sAf{)O~ del~ermalirunE'
'g th.s los,s h~!l.d t(hrough Parshall flumes of various
Seroi~e ~61.)
.ur
Slat~{)'11 251 and U.S. Sml Ccm.ssnJl1tion
of
lW
ell
,:z:pmmen
-\
"
80
BASIC PRINCl;l'LES
V.y, =%H.
V%i'if.
= 3.09H.1.S
Similarly, try 2" and 3-ft flumes. It is found that the respective crest elevo.tion~
are L53 and 1.23 ft and that the respective upstream water depths 1.1'(> 2.98 and 3.12 ft.
In deciding them,!st practical aize of flume to use, it will btl necessary to examine
the freeboard or the ch!LIInel and the effect of rise of the water surface upon the flow
through the h~adgate. If these conditions arc satisfactory, the 2-ft flume will be the.
I
j
4:-9.
FIG.
8.9ction of a Parshall Ilume illustrating ,th,e determination of the' proper crest
elevatlOn [26].
most eoonomical because of its small dimensions. However when the width of the
channel is considered, the final selection may be in bvor of th~ 3- or 4-ft flume because
moderat~ or long wing wall.s may be required for a smeoll structure. Usually, the
throat wldth of the flume Will be from one-third to one-h!'\lf of the channel width.
(4--17)
This is a theoreticru discharge equation i1) which H. is uncertain since the critical section is usually difficult to locate. For' practical purposes, however, the t\quation is
generally written q = CH'Jl, where H is the elevation of the upstream wa~er surface
iLbovethe weir c~est. This is the form described earlier (Example 3-2).
If an ...!'sated free overfall exists at the downstrea~end of the weir, the a\'ove equation ca.n be expressed in termso! the brink depth Y" which can easily be measured.
Since y, :. l.4y~ (Art. 3-4), the equ!l.tion required is
q
,
9.38y ol.l
PROBLEMS
. 4-1. Prove the following critical-discharge equations for the triangular trapezoidal '
aud circular' sections:
I
'
, ,
Channel 8e!lion
Triangular
(4-18)
,
Experiments h9.V~ shown that, when the head on the broad-crested w!:!r is greater
than about 1.5 times the length of the crest, the nappe of the free overfall becomes
detached and the weir is in effect a sharp-crested weir.
Example 4-6. Design ft. Parshall flume for handling 20 cfa of now in a.: chann'cl of
moderate slope when the water depth in the channel is 2.5 ft."
Solulimt. The discharge given can be m!"llsured by flumes of several si~es, but the
best selection' is the flume of most practical' a.nd economical size.
i
'
Assume W .. 4 it ,and H,/H.
0.7, For Q = 20 cfs, Eq, (4-15) gives' H. ~ L15
ft. Hence, H, = Q,81 f t . ;
,
'
At 70% submergence, the 1vater surface ih the throat, at Hb gage. is essentially level
with the surface of the tailwater. Under :this condition of flow, sllown lin Fig. 4-9,
1 The Vilhimonte weir sill consists or two; tri!LIIgular. tapered cotlvergin~ sills placed
on the culvert Hoor with an opening left between them (39}.
'
'
This example is sdopted from [26J.
~+~=~~.
Example 44. Using the theory of criticaillow, derive an equation for the discha,ge
over a broadcrested weir.
Solution. COllsiaer the section on the weir crest where critical flow occurs. At this
section, y, = 2(V,'/2g}
HJ1.5 or V, "" VUH,/1.5, where H. is the specific-energy
head at the section. The discharge per foot width of the weids, therefore, equal to
81
the tailwater depth D ~ 2.5 ft, and the elevation of the c~est above the channel b;t;..
torn is X
2.5 - 0.81 = 1.69 ft.
.
. '
From Fig. 4-8, the head loss corresponding to H~I H u = 0.7. Q = 20 cfs, and W =
4 rt is 0.43 ft. .Therefore, the depth of water upstrcuim from the flume will be
Equation
Q,
2.295zH..
Trapezoidal
Q _ 5.671{(b + zy)yJu
, - -(I) + 2zy)',$ -
Circular
Parabolic
Q.
(sin
Yz 9) 0,6
2.005TH.L5
(4-19)
(4-20)
U
n
(4-21)
(4-22)
4.-3. Compute .the hydraulic. exponeD.t M of a 36"1n. circular conduit having a,;
flow depth of 24 In. above the Invert, uSing (a) Fig. 4-2 and (b) the graphical method;
based on Eq. (4-11).
'
'
4-4. Prove that the critical depth and velocity for a rectangular channel are'
82
BASIC FRINCIFl;IllS
CRITICAL FLOW; ITS COMPUTATION AND ,APPLICATIONS
expressed by
83
(4-23)
and
(4-24)
where Q is the. discharge, b is the channel width, and "dll the energy coefficient.
4-5. A rectangular channel, 20 ft wide, carries g, discharge of 200 cfs. Compute
the critical depth and velocity.
4-6. Solve Example 4-2 by various methods if the discharge is 300 cfs.
4-7. Solve Example 4-3 by various methods if tr.e discharge is 15 cis.
4-8. An' a.pproxhnate but practiclJ.I fonnula for the critical depth of a circular section of <iiameter do, derived by BrILine [40] from ~.n equaticm equ.ivalent to Eq. (4.-21),
is
y.
= 0.325
(t +
[).083d.
(4:-25)
is
which accurate only when 0.3 < y./d Q < 0.9. ' Solve Example 4-3 and Prob. 4-7 by
this formula,
4-9. Referring to the naturalchaonel given in Frob. 2-5, construct a CUrve of critical
depth aga.inst discharge, ranging from 0 to 400 crs.
4-10. Prove that the section of a cha.nnel in which the flow is critical at any st'lge
takes the form expressed by
-{
!
(4-26)
whhle z is half the top width and II is the di5~ance of the water surface below the energy
line. Draw II skeLch of the section and d!lscripe its properties. Is this channel possible? If not, how could it be made pOllSible? Is this channel practicable ind the
flow stable?
4-11. Verify the computa.tions for the 2- and :HtPa.rshal! flumes tried in Example 4-5.
4-12. Determine the dischltrge through the 4i-ft Parsha.ll flume described in Example 4-5 if the percentage of submergence is .:l00/".
4-13. Determine the discharge measured by a 10-ft Parshall flume if the gage rea.ding H. is 3Al ft at a. free~flow condition.
.
4-14. Design a Parshall f!IJme to mell.sure 10 ers of flow in J1 channel ha.ving a depth
of flow equal to 1.5 ft.
'
4-16. A uniform flow of 300 cfs occurs'a.t a depth of 5 ft in a long rectan!!;lliar channel ,10 ft wide, Compute the mini,mum IH~ight of a. fiat-top hUmp that ca.n be built
on the floor of the channel in order to produce' a critical depth. What wilLresult if
the hump is lower or higher tha.n the computed minimum height?
4-16. If the critical depth in the above problem is produced by a contraction of the
channel, wha,twill be the maximum contracted width?
4-17. A low da.m 5 ft, high having a broad horizontal crest is built in a. t"6ct~ng\llar
channel 20 ft wide. Assuming that a depth df 2.5 ft measured on the crest is the
critical depth, compute the discharge and the depth of flow upstrea.m from the dam.
4-1B. On the basis-of the theory of critical flow, Stevens [32] has derived the rating
curves for the blister-lShaped critical-flow flume tlll~t he proposed for use in circular
conduits (Fig. 4-10). In the deriva.tion, it is as~umed (1) that there is no energy loss
from !ll to !It, (:~) that the !;I.pproachingvelocity in the pipe is eq ual to the discharge
I
'
1
FIG. 4-10. Rating curves of ;l. critical-flow flume proposed for a clos"d conduit.
J. C. Stevens [32J.) d. ""diameter of the cor.duit.
'
I
I
I
1
I
f
(After;
dh'ided lJy the water area. corresponding to th'e energy head instead or'the actual
area.: corresponding to Y" and (3) that the critical-flaw section i3 at the maximnm
~leight of. t~e cDntrol "hump." The second assu.mption elimina.te,s a. trial procedure
In determmmg the velocity hegd of the a.pproa.ching flow !l.nd, furthermore, tends to
, co,"?pensatc for the error involved in the first a,'55umption. Verify anyone of the
, ra tmg curves.
REFERENCES
1. Charles Jaeger: "~ngineering Fluid Mllchanics,~) trrulsu;,ted from the German by
P. O. Wolf, Bla.ckle & Son, Ltd., London and Glasgow, 1955, pp. 93-119. '
2. Ven Te. Chow: Integrating the eq1!aticns of gra.du!l.lly varied flow, paper 833,
Proceedtngs, Am.erican, Society of Civil Engi1l.llers voL 81 pp. 1-32 November
1955.
'
,
, '
,
.1
84
BASIC PRINCIPLES
4. A. H. Jameson: The Venturi flume and the effect of contractions in open channels,
1'ransactions, Institul-iolt of Water Engineers, vol. '30, pp. 19-24, June 30, 1925.
5. A. H. Jameson: The (il.rvelopment of the Venturi flume, Wawr and Water Engineering, London, voL 32, no. 375, pp. 105-107, Mar. 20, 1930.
6. F. V. A. E. Engel: Non-uniform flow of water: Problems and phenomena in open
ehannels with side contrn.c.tions, The Engineer, vol. 155, pp. 392-394, Apr. 21;
pp. 429-430, Apr. 28; pp. 45[':'457, May 5, 1933.
7. F. V. A. E. Eagel: The Venturi flume, The EngincliT, voL 158, PP: 104-107,
Aug. 3; pp. 131-133, Aug. 10, 19&4.'
8. A. Linford: Venturi flume flow meter, Civil Engineering and Publ'ic W01'ks Rev-iew,
London, vol. 36, no. 424, pp. 532-587, October, 19'11. An abstract is given in Journal, Americar;. Waler Works Association, vol. 34, pp. 1473-1475, September, 1942:
9. E. S. Crump: Moduling of irrigatioIl channels, Pt"dab Irrigation Brar4ch Publications, Paper Nos. 26 and 30A, Lahore, India, 1922 and' 1\)33.
.
lO. C. C. Inglis, NoteB on standing wave flumes and flume meter baffle falls, Pl:bUc
Works Department, Government of Bombay, Technical Papers, No. 15, India, 1928.
11. Giulio De Marchi (anthoT, pis. I and III) and FraMesco Contessi~i (author,
pt. II): Dispositivi per la misura della pOltatll dei canali con minime perdit,e di
quota: Nuove ricerche sperimentali slii misuratori a risalto !draulico (C'l.llnli
Venturi); Part,e I, Esame del pro(;esso idmulico; Parte II, Descri,.ione delle
esperienze; Palte III, Risultati delle esperienze [D.evices for measuring discharge
in cannls with minimum loss of level: New expaTimentnl researches on standing
wave flumes (Venturi flumes); pt. 1, AnnJysis of the hydralilic process; pt. II,
Description of the experiments; and pt. lII, Results of the experiments], L' Enfn'gia
. elettrica, Milano, vol. 13, no. 1, pp. 6-15, Jil.nnary, 1936; vol. 13, no. 5, pp. 236-244,
May, 1936; vol. 11, no. 3, p'p. 189-214, March, 1937. Reprinted as lstituto di
Idraulica e Co,~truz'io7'li JdJ'auliche, .Milano, Memorie e sludi Nos. 17, 25, and 26,
1936-1937.
.12. Giulio Dc Marchi: Nouvelles recherches experimentales sur Ie jaugeur !I. ressaut
hydra.uliqne, canal Venturi (New expel'imental researches' on standing-wave
.flume, Venturi flume), Ministry of Agriculture, Pa.ris, France, 1937. This is an
abstract of pt. I of \11].
.'
la. Guido Neuhia: Ventul'imetri-per canali a sezioni di forma generica (Venturi meter
for canals with cross sections of general forms), Acqua e gas, vol, 25, no. 11, pp.
2'rD-291, November, W36.
.
14. Guido Nebbia: VeJ,lturimetri per canali a sezioni di tipo monomio (Venturi 'meter
for canals with cross sections of monomial type), Acqua e gas, vol. 25, no. 12, pp.
326-333, December, 1936.
15. Guido Nebbia: Venturimetri per can ali a sezione di forma. generica: Primi rj;,'Ultat~
sperimentali (Venturi meter for cano.ls with cross sections of geneml forms: Preliminary experimental results), Aequo e (las, vol. 27, no. 5, pp. 155-181, May, 103S;
vol. 27, no. 6, pp. 199-214, June, 1938.
16. Duilio Citrini: Miauratori a risalto (Standing-wave flumes), L' Energia elettrica,
Milano, voL 16, nu. 10, pp. 758-763, October,1939; reprinted as Istitutodi Idra1,&lica
e Costruz'ioni Idrauliche, Milano, .Memorie e studi No. 35, 1939.
-17. Duilio Citrini: Modellatori a risalto: Guida al progetto (Standing-wave meters:
Direction~ for design), Centro studi per Ie. aPl?lica.zioni dell'ingegneria all'agricoltura, Sindaeato ingegn.eri di :A{ilano, Se;oorate Paper No.5, Milan,. 1941 ; reprinted
as I stituio Ji I draulica e Cos!ruzione' I drauliche, Milano, JIIlemorie e studi No. 44, 1941.
lB. Anwar Khafa.gi: Der Venturikanal: Them'ie und Anwendung (The Venturi flume:
theory and application), Eidgenossisehe Itchnische Hochscitu/e Zurich, Mitteilungen
der Versuchsallslaltfv.r Wasserbau und Erdbau, No. I, Zurich, 1942.
85
19. Arma~do Ball.offet: C:i~ical fl~w meters (Venturi flumes), paper 743, Proceedings,
Amencan Soc.ety of Cwtl Engmeers, vol: 81, pp. 1-31, JlIly, 1955.
20. V. M. Cone: The Ventnri flume, Journal of AgricultmaIResearch, voL 9, no. 4,
pp. 115-129, Apr. 23,1917. .
. . ,
.
21. Ralph L. Parshnll and Carl Rohwer: The Venturi flume, Coloraclo Agricultural'
Experiment Station, Bulletin No. 265 February 1921
..
22. R. L. Parsho.ll: The improved Vent~ri flume, Trans~clions, AmeJ";cr!.n SlJciety lJf
Ciuil.Enyineers, vol. B9, pp. B41~B51, 1926.
23.R. L. Parshall: The Parshall mell-suring flume, Colorado Agricultural Experimen l
Siation Btdlet'in No. 423, March, 1936.
'
24 R. L. Parshall: Measuring water in irrigation channels, U.S. Department lJf Ag"i.
~ulh!rB, Parmer's Bulletin No. 1883, January, 1932; revised, October, 1941. .
25. R. ~. Par.sl;al~: Parshall flumes of large size, Colorado A(Jricul/ural Experiment
Statton, BttLettn No. 3R~, May, 1932; revised as Bulletin No. 426A, March, 1953.
26. H. L. Parshall: Measurmg i.vater in irrigation channels with Parshall flumes and
s:nalt weirs, U.S. Soil Consenlation Senlicc, Circula?' 843, May, 1950. This
Circular supersedes [24],
27. H. G. Wilm, John S. Cotton, and H. C. Storey: Measurement. of debris-laden
stre~m flow with critical-depth flumes, Transactions, Ami-rican Society of Civil
Eng-meers, vol. 103, pp. 1237-1253, 1938.
.
.
2B. K. J. Bermel: Hydraulic influence of modifications to the San Dimas criticaldepth measuring flume, Transactions, American Geophysical Union, vol. 31 no. 5
pp. 763-768, October, 1950.
. '
,
2fl. Harold K. Palmer and Fred D. Bowlus: Adaptation of Venturi flumes to flow
measurements in conduits, Transactions, American' Society of CilJU E7tginee;s
vol. 101, pp. 1195-1216, 1936.
.
'
30. John H. Lu~wig and Russel G. Ludwig: Design of Pltlmer-Bowlus flumes, Sewage
~ndltst"al Wastes, vol. 23, no.,g, pp. 1096-1107, September, 1951.
31. Edwl~ A. Wells, Jr., I'.nd Harold B. Gotaas: Design of Venturi flumes in circular
condUits, Transactions, American Society of Civil ElIginee1's, vol. 123, pp. 749-771,
1958.
32.
C. Ste;ens: Discussion on Adaptation of Venturi flumes to flow measurements
III c~ndUlts; by Ha.rold re. Palmer and Pred D. Bowlus, Transactions, American
Society of Ciuil Enginee1's, vol. 101, pp. 1229-1231, 193fl.
N
33. Herbert Addison: "Hydmulic Measurements" JohnWile v & Sons In
. '
J
.
'
c., ew
York, 1941.
a;:d
!.
,
I
35. Horace William King: "Handbook of Hydraulics," 4th ed., revised by Ernest F
, . Brater, Mct;lraw-Hill. Book Company, Inc., New York, 1954.
' '..
;;6. F. T. MaVIS: Reducmg unknowns in small culvert' design, Engine~rin(l NewsRecord, vol. 137, no. 2, pp. 51-52, July 11, 1946.
37. W; O. Ree and F. R. C~ow: Measuring runoff rates with rectangular highway
culverts, Oklahoma AIlT!CUU'Ural Experiment Stat'ion, Technical Bulletin T-51
November, 1954.
. '
38. W. O. Ree ~nd ~. R. Crow: Culverts as Wo.teT rUIloff measuring' devices, AgriC1!/tural Eng~neenng, \"01. 35, no. I, pp: 28-31 and 39, January, 1954.
39. James R. Villemonte: New type gaging station for small streams; Enginee1'ing
News-Record, vol. 131, no. 21, pp. 748-750, Nov. 18, 1953.
40. C. D. C. Braine: Draw-down and other factors relating to the design of stormw2.ter outflows on sewers, Journal, Institution of CiviL Engineers, London vol. 28
no. 6, pp. 136-163, April, 1947.
. ' "
.PART II
UNIFORM FLOW
j
1
i.
1
I
'1
. \
. j,
CHAFTER
J
1
!I
,I
II
89
~,
I~~, '
'"
,.: ~..
.)
15-1. Qualifications for Uniform Flow. The uniform flow to be considert:d has the following mn.in features; (1) the depth, water area,
velooity, und discharge at every section of the channel reach I.1r,e constant;
and (2) the energy line, water surface, and channel bottom are allparallei; that their slopes are all equal, or Sj = S'" = Sn = S. For practica.l
pm-poses, the requirement of constant velocity may be liberally interpreted as
requiremel1t thtl.t the flow possess a constant mean velocity.
Strictly speaking, however, this should mean that thB flow possesses a
constant, velocity at every point on the clHmnel section within the uniform-flow reach. In other words, the velocity distributIon across the
channel section is uualtered in the I'each. Such fl. stable pattern of
velocity distribution call be attained when the so-called" boundary layer"
is fully developed (Art. 8-1).
'
Uniform flow is considered to be steady ouly, since unsteady uniform
flow is practically nonexistent. In natural streams, even steady uniform
flow is rare, for rivers and streams in natural states seanleiy ever experience a strict uniform-flow condition .. Despite this deviation from the
truth, the uniform-flow condition is frequently assumed in the computn:tiOD of flo\v in natural streams. The results obtained from this assumption are understood to be apPloximo.teallq. general, but they offer a
relatively simple and satisfactory solUtion to many prnctico.l problems.
As turbulent tmiform flow is most commonly encountered in engineering problems, it will be discussed e,.'(tensively il1 tbe following chapters.
Laminar uniform flow has limited engineering applications, and will be
described only in Art. 6-10.
" It should be noted that unifor:m flow cannot occur at very high velocities, usually described as ttltTarapid. 'This is because, when uniform flow
reache~ i certain high velocity, It becomes very unstable. At highel'
vetoci~ies' the fl~ "'1ill eventually entrain Edr and becbme unsteady.
The criterion fOl: instability of u,niform flow will be discussetl1i:1':A:rt:1f-8:
5-2. Establishment .of Uniform Flow. When flow OCCtll'S in an open
channel, resistu.nce is encountered by the water as it flO\vs downstream.
'
may again be exceeded by gravit:r forces .apd the flow may become varied
again.
For purposes of explanation. a long channel is showll with three different slopes: sub critical, critical, and supercritical (Fig. 5-1). At the
subcritical slope (top sketch in Fig. 5-1) the water surface in the transitory zone appears tmdulatoi:y. The flow is uniform in the middle reach
of the channel but varied at the two ends. 1 At the critical slope (middle
sketch in Fig. 5-1) the water surface of the critical Row is unstable.
Possible undula.tions may occur in the middle reach, but Ql1 the average
the depth is constant and the flow may bf'. considered uniform. At the
sl.lpc,rcritical slope (bottom sketch 'in Fig. 5-1) the transitory water surface
passes from the subcritical stage to the supercritical stage. through a
gradual hydl'aulic drop. Beyoml the transitory zone the flow Is approaching uniformity. The depth of a uniform flow is caned the nonnaJ depth.
In all figures the long dashed line represents the llormal-depth line,
abbreviated as N.D.L., and the short clashed or dotteclline represents
the critioal-depth line, or C.D.L.
The length of the transitory zone depends on the discharge and on the
phj"Sical conditions of the channel, Buch as entrance' condition, shape,
slope, and roughness. From a hydrodynamic standpoint (see Art. 8-1),
the length of the transitory zone should not be less than the length
required for the full development of the boundary layer under the given
conditions.
5-3. Expressing the Velocity of a Uniform Flow. For hydraulic computations the mean velocity of a turbulent uniform flow in open channels
. is usually expressed il,pproximately by a so-called unifoltn-flow formula.
Most practical uniform-fiow form1.lia,s ca:n be expressed in the following
general form:
90
Ivorled~w
Tronsitor
zone
un[f~rm
yt--I
I .. '
.
' I
r~~;::~1
t;"ifOftl1
fl 0 "'_ _ _-I.1
I
I
I
I
"
1
l
1
)
.\
t
(5-1)
;1'
91
UNIFORM FLOW
Il
long ohe.nne1.
If the water enters the channel slowly, the velocity and hence th(;l resista~ce are, small, and the resistance is outba,lanced by the gra;~ity forces,
resulting:in an accelerating flo~v in the upstr~am reach. The velocity and
. the resistance will gladually 'increase until a balance between resistance
aDd gravity forces is reached. At this moment alld' a.fterward the flow
becomes uniform. The upstream reach that is required for the establishment of uniform flow is known as the transilory zone;' In this zone the
flow is accelerating and varIed. If the channel is shorter than the transitory length required by the given conditions,. unIform flow cannot be.'
. attained. Toward the downstream end of the channel the resil5tance
1 Theoretically speaking, the varied depth a.t each end approaches the uniform
dept.h in the middle llSymptotically and gradually. For praotica\ PUfilvS\lS, however,
the depth ma.y be considez:ed constant if the variation in depth is within a certain
margin, say, 1 % of the lwerage uniform-fiow depth.
Z In. uniform flow f S :'" Sf =' S,. = So.
When the uniform-flow formula is applied
t.o the computation of energy slope in a gradually varied flow, the energy slope will
be denoted specifically by Sf instead of S.
.
. .l.
. i
-\
92
93
UNlF'ORM FLOW
A
.V
V ma
P
R
y
SUI
Q.
Qb
'IL
T
There have been dev,eloped and published a large number of practical
uniform-flow formulas, t but none of these formulas meets the qualifications of a good form'lIla as defined above. The best known and most
widely used formulas are the CMzy and Manning fOl'mulas, which will
be described in the' following articles alid used extensively in this book.
Theoretical uniform-flow formulas have also beell derived on the basis of
a theOl'etir.AI velocity distribution across the channel section, wl1i~h will
he discussed later (Art. 8-5).
A different approach to the determina.tion of the velocity in a natural
channel has been attempted by Toebes [5]. In this approach amultiplecorrelation analysis is applied to the following significant factors affecting
the velocity in a gi.ven alluvial channel: water area, maximum surface
velocity, wett~d perimeter, maximum depth,slope of water surface,
coefficient of roughness, and temperature of water. By this method it is
, possible to e"aluate the independent individual influence of each variable
on the magnitude of the velocity. When such an evahULtion is made,
the velocity under any givell condition of the variables is simply equal to
the algebraic summation of the individual contl'ibut,ions as <\ffected by
each variable. However, this method applies only to the streams in the
geographical region for which the analysis is made; hence, its applica.tion
.
cannot be generalized.
In British literature the term "l'llgosity coefficient IJ is used.
, A number of weH-known uniform-flow formulas are given Md discussed in [2]
to [5J.
.
1
(5-2)'
ChezY'$ G.
..
,FIG,
.
.\,.-,',
,
'
5-'.l. Derivation of the CI~ezy forml;!a. for uniform flow; in open channel.
The CMzy formula C.:l.n be'derived mathematically from two assumptions. The first assnmption was made by CMzy. It states that the
foi'ce resisting the flow per unit, area of the stream bed is proportional to
the square. of the velocity; that ls, this force is equal to X"V2 where ]( is a
constD.nt of proportionBlity. The surface of contact of the flow with the
stream bed is equal to the product of 'the wetted peri~eter and the length
of the channel l'efLClh, or PL (Fig. 5-2). The total foroe resisting the flow 2
is then equal to ](VSPL.
1 The source of this famous formula. is.not mentioned in most hydraullcs textbooks.
In fact, this knowledge has long been sought for. In 1876, the German engineer
Gotthilf Heinrich Ludwig Hagen mentioned in his 'work [7] that Geapard de Prony
, had stated tha.t Chezy set up this formula in 1775, on the occa$ion of a report that
Chezy made on the Canal de j'Yvette in conjunction with Jean-Rodolphe Perro net, .
"But," says Hagen, "I have sought in vain for further information on the' Bubject."
Then, in 1897, the American engineer Clemens Herschel through the assistance of a
friend in Paris traced the original Canal de I'Yvette report to its hiding place, then
translated the portion relating to the formula, IlInd published it in [8]. Chezy's report
revealed that the formula' was developed and verified by experiments made on an
earthen canal, the Courp!1let Canal, and on the Seine River in late 1769 ..
Thi~ channel resisting iorce may alao be ,explained by the principlea of fluid
dynamics. The open channel can be coneeived as a. ftnt plate warped into a oylinder
94
UNIFORM FLOW
v(wIK){AIP)S
= C vRS.
j
!
i
I
M~my
.41.65
C -
-n
n
+ ( 41.65 + 0.00281)
S
vB,
(5-3)
95
= __
157.t'__
1
+ m/vR
(5-4)
5-1.
Description of chanael
Bazin's.1n
0.11
0.21
0.83
1.54
2.36
The
formula was developed primarily from data collected from
small experimenta.l channels; hence, its general application is found to be
less satisfactory than the G. K. formula.
The Miami ConservancY'District [2] has made a study comparing the
v!l.riations ill Chezy's C, Bazin's m, and I{utter's n for Bazin's experimental data and several naturaL streams. The results based on this
study are showll in Table 5-2. The values of the average variation indicate that Bazin's formula is not as good as Kutter's even for his own
measurements.
This formula, an
(5-5)
-----------------------------------
the G . .K. formula be omitted in order to simplify the aP?lmranCe of the formula. a.nd
even to make -the general results more satisfactory.
1 From 1855 and 1862 an exten.~ive seiie:s of experiments on open-ch,mnel flow were
first begun by H. Darcy and then completed by Bazin. The results were published
by Bazin in 1865 (12). On the basis of the accumulated date., Baziu finally proposed
the formula in 1897 (13).
-\
1
. I
J
j
OJ
.S.
tQ
.>
,..
"'
'0
u..
'"s<
<f
i'l
1l
'tl
.~
"'"
.S
0
t1.
t:i
A
Q)
r:I
'"
A,
j;:
il:
~-
p"
..,:il
'"
...'"
til
T.tUILE
'"
Q)
.c:
.,
c
:=
;i!
OJ
..,
1:j
,S
tQ
:S'"
Average vaiue3
C
"
"'
.~
....
Q)
., c
oil
~g'O
~ l:t:1
ill' ",.",
'"""""
:El~
en
.q
'"
'0
.!!l
'"
:El
S.AZ~H:>
2~!!!~g.~E~i.~ii!~~
. . . . 0.185 O. 0121 ..
~~]
varia.tions, %
"I
,~
m,
t:i
~~
'"
BAZIN'g
Measurements
C,
AND KTJT'J'ER'S 11
g tJ"
is
...
"'-~
.,.,..,
<l " '"
'"~ g.~ ~ ~
Jl
.~. '0>.gj
.J:!
;0;:
III
<J
5-2,
oil
oil
.. :a
:a
.... ... !5
.., .i1III ....'" .~'"
... .!!l
0-
97
.8
g :El'"
...,'"
'"
ill
U
en
"'-
...
'A
'"
.,m-
.,
0;
"
5.,'"
...
..::;
'"~
'0
'"
'-'
Q)
I
lI
.,
!J
0,1420.0116 .... .
0.1990.0130 ... ..
0.1440.0117 .... ,
0.129 O.OU3 .... ,
0.324,0.0151.
1.6
OJHI0.0148 .... :j2.7
0.3210.0150 .... 4.4
0.'7150.0209 . . . . 4,2
0,7110.0212 ..... ' 5.7
0.7210.0215
16.1
0.4240.0'
1.8
0.444 0.0
3 .1
1 1
1.0
2,5
1.2
1.4
3.8
1.0
1.2
1.8
1.2
1.5
2.2
0.4
L2
8,8
5.7
gg
fL~.tG,~I~,LIW.!~I~!~I'L1UT~.ll~ttLI~I~.llk'~~~~~~WW~~WU~~~~o
S3l'iNlmlO .:Ici SIXV
C = 42 log (R/e). For smooth channels, the iurface roughness may be: .
so slight that e becomes negligible compared with R; then the formula
approaches the fbrm,C = 42 log (4R/C). Sinoe Chezy's is expressed
implioitly in the Powell formula, the solution of 'the formula for C requires:
!l. trial-and-error procedure.
.
.
The Powell formula was developed from limited laboratory experiments:
on smJoth and rougn channels and from the theoretical velooity distri-;
, (
i'
96
.
:1\
."
,',
.;
98
UNIFORM, FLOW
5-3.
Powell's.
Description of channel
New
Old
0.0002
0.0010
0.004
0.04
0.10
0.0004
0.0017
0.006
Example 5-1. Compute the velocity and discharge in the trape~oidal channel
described in Example 2-1, hONing a bottom width of. 20 ft, side slopes 2: 1, and a depth
of wat,er 6 ft. Given: Kutter's n = 0.015, and S = 0.005.
Solution. From Example 2-1, A = 192.0 it' and R = 4.10 It. Using the G. l{..
formula, the value or Ch6zy's C i"
0=
Then, by the
Chc~y
41 65 + 0.00281 + 1.811
0.005
0.015
. .
'41 6- + 0.00281) 0.015
+( 0
0.005
V4J.O
1242
.
formula,
Therefore,
Q = 192.0 X 17.8
3,420 ds
99
well-known form
(5-6)
-1
j
l.
100
UNIFQRMFLOW
,
-I
(5-7)
. where
"
basis of other later studies, some authors Stiggested ll. value oi % [20], and others suggested a variable depending on Rand 11. [21J. .
l The Manning formula was suggested for intern~tionaJ use by Lindquist [3j at the
Scandinavie:Sectional Meeting oi the World Power Conierenee in 1933 in Stockholm.
The final recommendation for such use was mlide by the Executive Committee at the
3d World Power Conference in 1938 in Washington, D.C. .
2 On account of this relationship, the Manning formula is sometimes oonsidered a
variation'oi the CItezy fortimla with Chezy's C defined by Eq. (5~7).
*' The Pa.vlovsIl.il formula. was published in severa! editions of Pll.v!ovskiI's "Ha.ndbook of Hydraulics" [211_ An article o.bout this formula. entitled Ft)1'l1m!a dlia
kt)effitlrien/.ll Chazy (Formula fur a Chtzy coefficient) is given in pp. 140-149 of the 193T
edition or the book. A footnote in this article reads: "'1'.he formula. was proposed in
1925."
101
for n between 0.011 and 0.040. For practical purposes, the following
approximate forms of Eq. (5-9) are generally suggested for use:
Y= 1.5
Y"" 1.3
vrn
vn
for R
for R
<
>
1.0 m
1.0 m
UNIFORM FLOW
102
clay, loam, or silt, ,the retardiI).geffect is much less tha;n whe:e t?e material
is coarse such as gravel;; or boulders. When the materIal IS fine, the
value of ~ is low and relatively un.affected by change i.n flow stage, 'Yhen
the material consists of gravels and boulders, the value of 71 is generally
high, particularly at low or high stage. Larger boulders usually collect
at the bottom of the st.ream, making the channel bottom rougher than the.
banks and increasing the value of 71 at low stages; At high stages, a
. portion of the energy of flow is used in rolling t.he boulders downstream,
thus increasing the value of n. A theoretical discu,ssion of surface roughness will be 'given in Art. 8-2.
B. T1 egetation. Veget.ation may be regarded as a kind of surface
roughness, but it also markedly reduces thecapacii~ ~f the c~annel. an~
retards the flow. This effect depends mainly on heIght, denmty, dlstnbutiol1, and type of vegetation, I;1nd it is very important in designing
.
small d[ainl;1ge channels.
At the University of Illinois an investigation has been made to determine the effect of vegetation on the coefficient of roughness [22]. On one
of the drainage ditches in central Illinois under investigation, an avera~e
n :value of 0.033 was measured in March, 1925, when the channel was ill
good condition .. In April, 1925, there were bushy willow? a~d dry w~eds
on the side slopes, and n was found to be 0.055. ThIS ll1crease m. 71
represwts the result of one year 's growt~ of vegetation. D~ring the
summers of U)25 and 1925 there waS a thICk growth of cattaIls on the
bottom of the channel. The n value at medium summer stages w~s
about 0.115, l,tnd at a nearly bankfull stage it was 0.099. The cattaIls
inihe channel were washed out by the high water in September, 1926; the
average value of 71 found after this occurrence was 0.072 .. The conclusions.
drawn from this investigati'on were, in part, as follows:
1. The minimum value of 11, that should be used for designing drain~ge
ditches in centml Illinois is .0.040. This value is obtainable at high
stages duri!1g the summer months in th~ most carefully n:aintained ch~n
nels where the bott.om of the channel IS clear of vegetatiOn and the. Side
slop~s are covered with grass or low weeds, bU,t no bushes. This low
value of n shonld not be used unless the channel IS to be cleared a:nnuaUy
of all weeds and b u s h e s . .
.
2. A value of n = 0,050 should be used if the channel is to be cleared
in alternate years only. Large weeds and bushy willows from 3 to 4ft
high on the side slopes will produce this value of n.
3. In channels that are not cleai'ed for a number of .years, the growth
may become so abundant that values of n ~ 0.100 may. be found.
,.
. 4. Trees from (3 to 8 in. in diameter growmg on the Side slopes do not
impede the flo\v so much as do small bushy growths, provided overhang.
ing branches are cutoff,
'.'
i
",1\.,'.,..
."'
'"
103
The U.S. Soil Conservation Service has made studies on flow of water
in small shallow channels protected by vegetative linings (Chap. 7,
Sec ..0), It was found that n values for these channels varied with the
shape and cross section of the channel, the slope of the channel bed, and
. the depth of flow. Comparing two channels, all other factors being equal,
the lesser average depth gives the higher n value, owing to a larger
proportion of affected vegetation. Thus, a triangular channel has a
higher n value than a trapezoidal channel, and a wide channel has a
lower n value th11D. a narrow channel. A flow of sufficient depth tends to
bend over and submerge the vegetation and to produce low n values. A
.steep slope causes greater velocity, greater flattening of the vegetation,
and low n valu,<::s.
.
The effect of vegetation on flood plains will be discussed later in item H.
. C. Channel Ir:regularity. ChanneUrregularity comprises irregularities
in wetted perimet.er and variations in cross section, size, and shs.pe along
the channel length. In na.tural channels, such irregularities are uSl,lally
introduced by the presence of sand bars, sand waves, ridges and depressions, and holes and humps on the channel bed. These irregul9.rities
definitely introduce. roughness in addition to that caused by surface
roughness and other factors. Generally speaking, a gradual and uniform
change in cross. sectioll, size, and shape will not a.ppreciably affect the
value of n, but abrupt changed or alternation of small and large sections
necessitates the use of a large value of 71. In this case, the increase in n
may be 0.005 or more. Changes that cause sinuous flow from side to
side of the channel will produce the same effect.
D. Channel Alignment . . Smooth curvature with large radius will give
a relatively low value of 71, whereas sharp curvature with severe meandering will increase 71. On the basis of flume tests, Scobey [23] suggested
that the value of 71 be increased 0.001 for each 20. degrees of curvature 111
100 ft of channel. . Altholagh. it is doubtful ,..hether curvatllre ever
increases 11, mOfe than 0.002 or 0.003, its effect should not be ignored, for
curvature may induce the accumulation of drift and thus indirectly
increase the value of n. Generally speaking, the increase of roughness
in unlined channels carrying water at low velocities is negligible. An
increase of 0.002 in n value would constitute an adequate allowance for
curve losses in most flumes containilig pronounced curvatures, whether
built of concrete or other materials. The meandering of naturai streams,
however, may increase the 71 value as' high as 30 %.
E. Silting and Scouring. Generally speaking,.8ilting may change a
very irregular channel into a comparatively uniform one and decrease n,
'whereas scouring may do the reverse and increase 71 . However, the
dominant effect of silting will depend on the nature of the material
deposited. Uneven deposits such as sand. bars and sand waves 'are
-1
I
1
104
UNIFORM FLOW
l'
VALUES OF
!I
i
105
of .inundfltion, as obseryed in the Nishnabotna River, low-a,fof the average gfowing season [24]. It should be noted, however, that vegetation
has a marked effect only up to certain stage and that the roughness
coefficient can be considered to remain constant for pl'actical purposes in
determining overbank flood discharges.
50
o
i!:
~ ~O
.'2
g"
'"
20
;:;:
O.k02;;;5.----:O':::-:03~O"..----;:O,..,.Oi,.3.",5-.---,;,.,!-,=---=~::,!
n value
0.045
n value
Flood-pilli~
Depth of
water, it
Channel
section
Under 1
1 to 2
2 to 3
3 to 4
Over 4
0.03
0.03
0,03
0,03
0,03
Corn
Pasture
Meadow
0,10
0,06
0,06
0.07
0,07
0,05
O,O~
0.04
0.04
0,04
cover
0.07
0,06
0.05
Small
gt'ains
Brush and
waste
0,10
0.09
0,12
0.11
0,10
0.09
0,08
0,07
0,06
O,Oi
I
I
Curves of n v!\lue versus stags (Fig. 5-4) in streams have been given
by Lane (25\; showing how vahle of n varies with stage
three
river channels. For the roughness of large canals, it study in connection
with the design of the Panama Canal ,yas made by IVley~r& and .Schultz
[26).1 . The two most important conclusions reached from this study were
(1) that the n value fol' a river !channel is least when the stage is at 01'
somev;rhat above normal bankfull stage, and tends to in~rease for both
1 A tJble of ; vJlues for eleven lar~e cbanneJs at tbe most efficie[lt depths and tlle
curVel, showing th.e variations 01 11 value with hydra.ulic radius in eight river channels
are also given in this reference.
UNlFORM FLOW
:106
;higher and lower stages; and (2) that the bankfulLn values do Dot vary .
greatly for rivers and canals in different kinds of material and in widely.
separated locfl.i;ions.
For circular conduits, Camp [27,28] was able 'to show that the n value
for 11 conduit flowing partially fuBis greater than that for a full conduit.
Using measurements on clean sewer pipe and drain tile, both clay and
concrete, from 4 to 12 in. in size, he found an increase of about 24% in the
n value at the half-depth (Fig. 6-5).l The n value for the pipe flowing
full was found to vary from 0.001:15 to 0.011. Taking an average value of
0.0103, the nvalue at half-depth should be about 0,013. This is identical
with the usual design value,which is based largely on measured values in
sewers Hov\e'ing partially fulL
.
I. SeasonoJ Change. Owing to the seasonal growtih of aqus,hicplants,
grass, weeds, willow, and trees in the channel or on the banks, the value
of n may increase in the growing seasOl]. and diminish in the dormant
season. This seasonal change may cause changes in other factors.
J, Suspended Material and Bed Load. The suspended material and
the bed load, whet.her moving or not moving, would consume energy and
cause head loss or increase the apparent channel roughness.
All the above factors should be studied I1nd evaluated wit,h respect to
conditions regarding type of channel, state of flow, degree of mainten~~ce,
and other related considerations. They provide a basis for determmmg
the proper value of n for a given problem. As a general guide to judgment it may be accepted that conditions tending to hldtlce turbulence
and c~use retardance will increase 1. value and that those tending to reduce
turbulence and retardance. will decrease n value.
Recognizing several primary actors affecting the roughness coeffic,el1~, .
Cowan [32J developed a procedure for estimating the value of n. By thIS
procedure, the value of n may be computed by
<
(no
",1
+ n2 + nl + n()m5
(5-12)
.,
,j
I
!
~""---'"~--'.".-~.-.-,'~
r-
,-I
J
--\
\'
-\
,1
~!
:~'
,:i:; 'r.~.'
107
'.
~~.
'I
'\
4;1
I'
UNIFORM: FLOW
I,
grown with .some v,l:eds and brush, none of thp- vegetation in foliage,
where the hydraulic radius is greater than 2 ft, and (c) growing seasonbushy willoW's about 1 year old intergrown with some weeds in full foliage
along side slopes, no signifieant vegetation along channel bottom, where'
hydraulic radius is greater than 2 ft.
.
(4) Very high for conditions comparable to the following: (a) 'turf
. grasses where the average depth of fio~ is less than one-half the height
of vegetation, (b) growing season-bushy willows about 1 yellr old, intergrown with weeds in full foliagealon.g side slopes, or dtmse growth of
eat tails along Ghannel bottom, with any value ofhydratilic radius up to
1001' 15 ft, aIld (c) growiIlg season:-trees intergro,vn with weeds and brush,
all in full foliage, with any value of hydraulic radius up to 10 or 15 ft.
Ir:. selecting the value of ms, the degree of meandering depends all the
ratio of th.e meander length to the straight length bf the channel reach.
The meandering isconsidel'ed minor for ratios of 1.0 to 1.2, appreciable
for mtios of 1.2 to L5, and severe for ratios of 1.5 and greater.
In applying the above method for determining the n value, several
things should be noted. The method dQes not consider the effect of
suspended and bed loads. The values given in Table 5-5 were developed
from a study of some 40 to 50 cases of small ftnd moderate channels.
Therefore, the method is qU<jlstionable when applied to large channels
whose hydraulic rn,dii exceed, say, 15 ft. The method applies only to
unlined natural streams, Rood ways, and drainage channels and shows.il.
rninimum value of 0.02 fol' the 11, value of such channels. The minimum
value of n in general, however, may be as low as 0.012 in lined channels
and as 0.008 in artificiH1 laboratory flumes.
6-9. The Table of Manning's Roughness Coefficient. Table 5-6 gives
a list of n values for channels of various kinds. I For each kind of channel
the minimum,. nOimal, and maximum values of 11, are ShO\VIl. The normal values for artificial channels given in the table are recommended only
for channels with good maintenance. The boldface figures are values
generally recommended in design .. For the case in which poor maintenance is expected in the future, values should be increased accordiQ.g to
the situation expected. Table 5-6 will be found very usp-Iul as a guide to
the quick selection of the 11. value to be used in !\, given problem. A
poplllur table of this type was~prepn.red by Horton [34) from an examination of the best available experiments at his time. 2 Table 5-6 is compiled
fi
108
"
)
L The minimum value lor Lucite was observed in the Hydra.ulic Engineel'ing La.bol'aLory nt the University or Illinois [33). Such a low n value Inay perhaps be obtained
also for smoot\) brass o.nd glass, but no observntions have yet been reported.
t A table showing n values and other elements hom 269 observations made on many
existinga.rtifiCia.l cha.Ilnels is also given by King [35J.
5-5.
:BY
I,
EQ. (5-12)
Channel conditionB
Values'
Earth
M.ateriul
in vo I.... ed
.,
109
Degree of
irregularity
0.020
Rock cut
0.025
nQ
Fine gravel
0.024
Coarse gravel
0.028
Smooth
0.000
Minor
r.,
0.005
Modernte
0.010
Severe
0.020
Gradual
Va.ria.tions. of
channel cross Alternating oceaSionally
section
Alternating irequently
0.000
n2
O.010~.015
Negligible
Relative
effect of
obstruutions
0.000
Minor
1\...."
. 0.005
O. 01O-{}. 015
..
'11..
~
0.030-{}.030
.~
0.040-0.080
Low
0.005-0.010
Medium
0.010-0.025
High
0.02S-{}.050
Very high
0.050-0.100
VegeLfLtion
Minor
Degree of
meandering
Appreciable
Severe
1.000
m.
1.150
1.300
)
110
UNIFORM FLO>'i
I Minimum I Normal
----~~~~~~~~~-I
II
Maximum
. I
A-I. Metal
a. Brass, smooth
b. Steel
1. Lockbar and welded
2. Riveted and spiral
c. Cast iron
1. Coated
2. Uncoated
d. Wrought iron
1. Black
2. Galvanized
e. Corrugated met.al
1. Sub drain .
2. Storm drain
A-2. Nonmp.tal.
a. Lucite
b. Glasll
c. Cemen't
1. Neat, surface
2. Mortar
d. Concrete
1. Culvert, straight and free of debris
2. Culvert with bends, connections,
and some debris
3. Finished
4. Sewer wit.h manholes, inlet, etc.,
strl!-ight
5. Unfinished, steel form
6. Unfinished, smooth wood form
7. Unfinished, rough wood form
e. Wood
1. Stave
2. Laminated, treated
f. Clay
1. Commondrain8.ge tile
2. Vitrified sewer
3. Vitrified.sewer with manhol,~s, illlet,
etc.
4. Vitrified subdrain with open joint
g. Brickwork
1. Glazed
2 .. Lined with cement mortar
h. S'anitary sewers coated with sewage
slimes, with bends and connections
i. Paved invert, sewer, smooth bottom
j. Rubble masonry, cemented
111
A.
-(
0.009
0.010
0.013
O.OlD
0.013
0.012
0.016
0.014
. 0.017 .
0.010
0.011
0.013
0.014
0.014
0.016
0.012
0.013
0.014
0.016
0,015
0.017
0.017
0.021
0.019
0.024
0.021
0.030
0.003
0.009
0.009
0.010
0.Of3 .
0.011
0.011
0.013
0.013
0.015
0.010
0.011
0.011
0.013
0.013
0.014
O.OlD
O.OlD
I
!
f
iI
0.011
0.013
0.012
0.015
0.. 014
1).017
0.012
0.012
0.015
0.013
0.014
0.017
0.014
0.016
0.020
O.OlD
0.015
0.012
0.017
0.014
0.020
0.011
0.011
0.013
0.013,
0.014
Q.015
0.017
0.017
0.017
0.014
0.016
0.018
0.011
0.012
0.012
0.013
0.015
0.013
0.015
0.017
0.016
0.016
0.018
0.019
0.025
0.020
0.030
J.
f
-----D.
Minimum
(continued)
Normal Maximum
. B-l. Meta!
a. Smooth steel surface
1. Unpainted
2. Painted
b. Corrugated
B-2. Nonmet.al
a. Cement
1. N el1t, surf!lce
2. Morta-r.
b. Wood
L Planed, untreated
2. Planed, creosoted
3. Unplan(ld
4. Plank with battens
5. Lined with roofing paper
c. Concrete
1. Trowel finish
2. Float finish
3. Finished, with gravel on bottom
4. Unfiliished
.
5. Guuite, good section
6. Gunit8, wavy section
7. On good excavated rock
8. On irregular excavated rock
d. Concr'lte bottom float finillhed with
sides of
1. Dressed stone in mortar
2. Random stone in mortar
3. Cp,ment rubhle masonry, plastered
4. Cement rubble maSQnry
5. Dry l'Ubble or riprap
e, Gravel bottom with sides of
1. Formed cOllCrete
2. Random stone in mortar
3. Dry rubble or riprap
f. Brick
1. Glazed
2. In cement mortar
11. Masonry
1. Cemented I'Ub ble
2. Dry rubble
h. Dresse.d ashlar
i. Asphalt
1. Smooth
2: Rough
. j. Vegetal lining
0.011
0.012
0.021
0.012
0,013
0.025
0.014
0.017
0.030
0.010
0.011
0.011
0.013
0.013
0.015
O.OlD
0.012
0.012
0.013
0.015
0.014
0.014
0.015
0.015
0.018
0.017
0.013
0.015
0.017
0.017
0.019
0.022
0.015
0.016
0.020
0.020
0.023
0.025
0.011
0.011
0 .. 012
0.010
0.011
0.013
0.01.5
0.014
0.016
0.018
0.017
0.022
0.027
0..015
0.017
0.015
0.020
0.020
0.017
0.020
0.020
0.025
0.030
0.020
0.024
0.024
0.030
0.035
0.017
0.020
0.023
0.020
0.023
0.033
0.025
0.026
0.036
0.011
0.012
0.013
0.016
0.015
0.018
0.017
0.023
0.013
0.025
0.032
0.015
0'.030
0.03.5
0.017
0.013
0.015
0.030
0.013
0.016
O.OZI)
'-,
0.500
fI
,
.'h
I:
t,
:.::."
,.,,,""::".:
~;
I
112
UNIFORM FLOW
i'
Minimum
C. Exe.WATIilD OR DREDGED
a. Ea.rth, straight and uniform
.1. Clean, recently completed
0.016
2. Clean, e.fter weathering
0.018
3. Gravel, uniform section, clean
0022
4. With short gr!l.8S, f.ew weeds
0.022
b. Earth, winding and sluggish
1. No vegetation
0.023
2. Grass, some wGeds
0.025
3. Dense weeds or aquatic plants in
0.030
deep channels
4. Earth bottom lind rubble sides
0.028
5. Stony bottom and weedy banks
0.025
6. Cobble bottom nnd clea.n sides
0.030
c. Dritgline-eltcavated or dredged
1. No vegetation
0.025
2. Light brush on banks
0.035
d. Rock cuts
1. Smooth and uniform
0.025
2. Jagged and irregular
0.035
e. Channels not maintained, wc.eds and'
brush uncut
. l.Dense weeds, high as flow dept.~
0.050
: 2. Clean bottom, brush on sides
0.040
3. Same, highest stage of flow
0.045
4. Dense brush, high stage
0.080
D. NATuML S'IREAMS'
D-l. l\llnnr streams (top width at flood stage
<100 it)
a. Streams on plain
Clean, straight, fun stage, no rifts or
0.025
deep pools
I
2. Same as ..bove, but more stones and
0'.030
I
weeds
3. Clean, winding, some pools and
0.033
shoals
.
0.035 .
4. Same ail above, but llome weeds a.nd
stones
. ,
1
5. Same as above, lower stages, more
0.040
ineffective slopes and sections
6. Same as 4, but more stones
0.045
7. Siuggish reaches, weedy, deep pools
0.050
13. Very weedy rea.ches, deep pools, or 0.075
ftoodways with heavy stand of tim- 1
ber a.nd underbrush
n (continued)
Normal Maximum
I:
I
I
I
II
I
J
!
I
I'
0.018
0.022
0.025
0.027
0.020
0.025
0.030
0.033
0,025
0.030
0.035
0.030
0.033
0.040 .
0.030
0.035
0.040
0.035
0.040
0.028
0.050
0.033
0.060
0.035
0.040
0.040
0.050
0.080
0.050
0.071iJ
0.100
0,120
0.080
0.110
0.1<10
0.030
0.033
0.035
0.040
0.040
0;04:>
0.045
0.050
0.0'48
0.055 .
0.050
0.010
0,100
0.060
0.080
0.150
0~050
TABLE5c6. V,u,UES
ll3
(continued)
description
~inimum Normal Maximum'
--~--------------~---I----~I
b. Mountain strea.ms, no vegetation in
cha.nnel, ba.nks usually steep, trees
and brush along banks submerged at
high sta.ges
.
1. Bo~tom: gra.vels, cobbles, and fetv
0.030
0.040
0.050
boulders
2. Bottom: cobbles with large boulders
0.040
0.050
0.070
D-2. Flood plains
<l. Pasture, DD brUsh
1. Short grass
0.025
0.030
0.035
,2: High grass
0.030
0,035
0.050
b. Cultivated areas
1. No crop
0.020
0.030
0.040
2. Mature row crops
0.025
0.035
0.045
g. !vIa ture field crops
0.030
0.040
0.050
c. Brush
1. Scattered brush, heavy weeds
0.035
0,050
0.070
2. LIght brush and trees, in winter
0,035
0.050
0~060
3. Light brmlh and trees, in sumInar
0.040
0.060
0.080
,4. Medium to densl'l brush, in winter
0.045
0.070
0.110
5. Medium to dense brush, in summer
0.070
0.100
0.160
d. Trees
1. Dense willows, summer, straight
0.110
0.150
0200
2. Cleared land with .tree stumps, no
0030
0.040
0.050
sprouts
3. Same above, but with heavy, i 0.050
0.060
0.080
growth of sprouts
I'
4. Heavy stand of timber, f.l few down
0.080
0.100
0.120
trees, little undergrowth, flood stage
below branches
6. Same as aboye, but with flood stag"
0.100
0.120
0.160
reaching bra.nohes
D-3. Major streams (top width at flood stage
>1.00 it). The 11. vallie is Jess than that
for minor streams of similar ,description,
becll.use banks offer less effective resistance.
a. Regular section with no boulders or /1 0.0215
0.060
brush
114
fl'
UNIFORrk FLOW
-l
1
-<I--
I
1
5-5. Typical channels ~howing different n values. (These photo
h
'
ft~m [3~J and [38J with the permi,nion of the U.S. Departme~~~f ~ ~~l~!t~:,:
~1~1~rtlj.m~1 rc~ure~ used Jo~ reprodu.ctil]ll purposes were supplied thTUugh th~ courtes;
~, .rr" C' E' ";0 ey tofr ph,otogrhrkphs 1 10 14 rkncl phol()gl'aph 19, rknd lhrou(lh the courtesy
FIG.
duced
~
,,,.
-'
.)
't '.
: .~
115
,-1
~l'
!I
I
1
i
j
I
(1)
(4)
1i
I
i
I!
!
\'
I
;
.\
~ii)
(6)
ll6
117
(8)
(9)
IL
:;
,. '\.
~
~1
'I
7. 1'1 =
8. 1'1 the earth
9. 1'1 =
(13)
(16)
(17)
(14)
l
i
(18)
I,
t
(15)
t.
I,
13. 11 = ().029. Ea.rth c~nal excavated in alluvial silt 'soil, with deposits of sand
on. bottom and growth of gi-ass.
14. 11 = 0.030. Canal with large-cobbleswne bed.
15. 11 = 0.035. Natural channel, somewhat irregular side slopes; fa.irly even, clean
a.nd regular bottom; in light gray silty cla.y to light tan silt loam; very little varia.tion .
in cross section.
12Q
(22)
(19)
./
(20)
(23)
(21)
(24)
I
~
.'{1
I
i,
FIG.
.22. n = 0.110.
5-5 (22--241
8~me as (21), but with much foliage and covered for about 40 It
123
.;;."
Jt
124
UNIFORM FLOW
PROBLEMS
5-1. Expiain why a uniform flow cannot occur (a) in Il frictionless channel, and (b)
in a. horizontal channel.
'
,
'
6-2. When Chezy's C determined by the G. Ie formula becomes independent of
the slope S, show that the value of R
3.28. Find the corresponding relation
between C all.:! Kutter's n.
.,
li-S. For tbe conditions given in Example 5-1, compute the values of Ba.zin'i! m and
Powell's ~.
6-4. Compute the velocity and discharge of flow in a new etu'th canlll having the
same shape, si;;e, slope, and depth of flow as the channel given in Example 5-1. Use
(a) tile G. K. formula, ,assuming'Kutter's n "" 0.022; (b) the Bazin formula, selecting
!l p'roper value of 11>; and (e) the Powell formula, selecting a proper value of .
5-6. Taking Manning'S n as the given value of Kutter's n, solve E):ample 5-1 by
the Manning formula.
,
'
5-6. If the coefficient of l'oughness 11 is unknown for the channe.l ill Example 5-1,
but a discharge of 2,000 cfs is observed under the given conditions, compute the v!Liues
of l{utter's 1~ and Manning's 11.
5-7. From the MllJlning formula (using a consta.nt of 1.486 instead of 1.49 for
theoretical a~curacy) and the Chazy formula, determine the reilltion betweeu CMzy's
C and Manning's n for the condition de:;ocribed in: Plob. 5-2. This \\'iii show that the
G. K. formula and the Manning formula are theoretically identical at the condition
when ,Cl,ezy's 'C is independent of the slope S.
, 6'-8. Prov~ that the friction factor f in the Darcy-Weisbach formuir.,Eq. (1-4), is
related to .Manning's n by f = 116n~/R~~.
1i-9. Run 12-4 of Bll.zin's teats [12J was made on a rectangular plank flume ,fi.4 ft
wide with wooden strips 1 em thick and 2.7 cm wide nailed crosswise on the bottom
and ~ides at a spacinll; of 3.7 em center to center
strips. This fiume gave, a mean
velocity 0[3.33 fps at a flow depth of 1.02 ft and a slope of 0.0015. The t.emperature
rending WIl.S 8.5C. Determine Manning's n, and compute (a) CM.2y'S C, (b) Kutter's
n, (c) ,Bazin's m; and, (d) Powell's .
"
6-10. Run 15-1 of Bazin's tests was the san1e as run 12-4, described in the preceding
problem, ex~ept that the spacing of stripa was increaaed to 7.7 ern. 'Using the same
disc,harge as that of run '12-4., the depth of flow wn.s found to be 1.33 :it. Determine
Mnnning's IL, and compute (a) Chez)"s C, (b) Kutter's 7'" {~} Bllzin's m, and (d)
Powell's ..Oomptu'e the values of obtained from runs 1274 and 15-4 with the height
of the strips, and explain I,he effect of roughness in both cases.
'6-11. U!iing the Manning formula., construct 0. discharge-rating curvel for the
natural chlll\uel section given in Prob. 2-5. The slope ii! 0.0016, and, 11; ;= 0.035,
Extend the sides of the channel by ~traight lines at high stages if necessary.
'6-12. Tile actual rating curve of the channel section in Prob. 2-5 is described below.
Construct a CUTve showing the' variation in Manning's r. with respect to the stllg'l
above the datum.
'
Discharge, cis
0.3
0.4
0.5
1.'0
2.3
4,.6
7.8
1l.0
15.0
20.0
25.0
'0.6
0.7
0.8
0.9
1.'0
1.25
Stage,
It
Discharge, cfs
1.50
1. 75
2.00
2.25
2.50
3.00
50.0
62.0
75.0
88.0
10'2.0
132.0
164.0
199.0
3.50
4.00
38.0
!l.
REFERENCES
Schnack~nberg: Slope discharge formuiae for alluvial streams and rivers,
Proceedings, New Zea.landlnsWt!lion of Enginesrs, vol. 37, pp. 340-409, Wellington, 1951. Discussions, pp. 4.10-4.4.9.
IVllJl E. Houk: Calculation of flow in {)pen channels, Miami Conservancy District,
'
Technical Report, Pl. IV, Dayton. Ohio, 191B.
Erik l,indquist: On velocity formulas for open channels and pipes, Transactions
of the World Power C01~feTimce, Sedional Meeting, Scand'inavi(l, Stockholm; vol. I,
pp. 177-234, 1933.
'
Philipp FOfchheimer, "Hydraulik" (" Hydraulics JI), Teubner Verlagsgesellscha.ft,
Leipzig e.nd Berlin, pp. 139-163, 1930.
Zivko Vladislavlievitch: Apen;u critique sur le~ formules pour la predetermination
de Ie. vitesse moyenne de I'ecoulement uniforme (Critical surveyor the formula()
for predetermination of mea.n velocity of uniform flow), Transacli(]11)1 of th~ lst
Congress, InternationaL Com'mislrion on Irrigalion and Drainage, New Delhi,
voL 2, rept. 12, question 2, pp. 405-4.28, 1951.
Oornelis. Toehes: Stres.mfiow: Poly-dimensional treatment of vari!).ble factors
affecting the velocity in alluvia.l streams and rivers, Proceedings, Inslittttion of
Civil Engineers, Lut,do'/1., vol. 4., no. 3, pt. Ill, pp. 900-93S, December, 1955. ,
G. H. L. He.gen: "Untersuchungen fiber die gleichformige Bewegung des Wassers"
("Research'!S on. Uniform Flow of Water"), Berlin, l~7f.L
,
'
Clemena Herschel: On the origin of the Chezy formula, J(l,.rnal, Association of
Engineering Sooel:ie3, vol. 18, pp. 363-368. Discussion, pp. 368-369, JalluaryJune, '1897.
A. Brahms: "Anfangsgriinde dcr Deioh- und Wasserbaukunst" ("Elements of
Dam and Hyd!,aulic Engineering"), Aurich, Germa.ny, 1754 and 1757, vol. I,
L E. C.
2.
3.
4.
5.
or
1 It should be noted that the synthetio rating curve thus obtained is very approxima.te, parti~ularly for a natural channel, beoause the n vDJue is a.ctually not !l. constant
but a function of th.e depth (see Art. 5-8).
'
Stage, ft
125
6.
7.
8.
9.
p.l05.
'
10. E. Ge.nguillet llJld W. R. Kutter: Versuch zur Aufstellung einer neuen allegemeinell.
Formel fUr die gleichfllrmige Bewegung des Wassers in CaniLlen und Fliissen (An
investigation to establish a new general formula for uniform Bow of water in canals
a.nd rivers), Zeitsclirift cIes Oeslerreichischel~ Ingenieu.r- u.nd Architekten Vern1les,
vol. 21, no. 1, pp. 6-25; no. 2-3, pp. 46-59, Vienna, 1869. Published as e. book in.
Bern, Switzerland, 1877; translated into English by Rudolph Hering a.nd John C.
126
UNIFORM FLOW
Trautwine; Jr., 1.\5 "A general Formula for the Uniform Flow of Water in Rivers
and Other Glumnais," John 'Wiley &; Sons, Inc., Nevv York, 1st ed., 1888; 2d ed.,
18!H and i\)Ol.
11. C:.\pt:.\in A. A. Humphrey.s and Lieut. H. L. Abbot, U.S. Army Corps of Topographical Engineers: "Report upon thB physics and hydraulics of the Mississippi
River; U!Jon the protection of the alluvi!>l region a,gainst overflow; and upon the
deepening. of the mOllths; based upon surveys and investigations . . . ," J. E,
Lippincott Company, Philadelphia, 1861; reprinted in Washington, D.C., in 18157,
and as U.B. Army Corps (Jf Ellllineel's, PtofC8sionai Pap.,. No. 13, 1875.
12. H. DI.\!'cy a.nd H. Ballin: "Recherches hydra.uliques,' lre partie, Recherches
experimenta.les sur l'ecoulement de Feau dans ies canaux decouverts; 2e partie,
Recherches experiment&les rel().tives aux ramous et il. In. propaga.tion des ,andes
("Hydralllic Resea.rches," pt. 1, 'Experimental research on flmv of water in open
channels; p~. 2, Experimenta.l research on 'bMkwater and the propagation of
'wave:;), ACMiemie des Sciences, Paris, 1865.
13. H. Bnzin: li:tude d'one nouvelle fonnuls pour ca.lculer Ie debit des canaux deoollverts (A llllW formula.' for the calcula.tion of discharge in open channels), Mtmoire
No. 41, Allnal6~ deB ponti I~t cha!,~~ees, vol. 14, seT. 7, 4me trimestrs, pp. 20-70,
1897.
.
14. Ralph W. Powell: Resi~tance to flow in l'ough cha.nnels, 'l',u,naadion'5, American
Geophysical Union, vol. 31, [lO, 4, pp. 575-582, August, 1950.
1.5. Hobert Manning: On the flow of water in open channels and pipes, Transactions,
I1/.~titution of Civil Ellgine(Jl'8 'of Ireland, vol. 20, pp. 161-207, Dublin, lSIH;
supplement, voL 24, pp. 179-207, 1895,
15. Ven Te Chow: A note on the Manning formula, Transadidns, American Geophysical U'~ion, vol. 36, no. 4, p. 688, August, 1955.
17. Allen J. C. OUlmingham: Recent bydrl'luIic experiments, Proceedings, Institution
. of Civil Engineel's, London, voL 7l, pp. 1-36, 1883.
.
18. Ph. Gauckler: Du 1l10UVeme[lt d~ l'eau da.::ls les ()onduites (The flow of water
conduits), Annates de~ pOJ~ts .1 chriu.ssees, vol. 15, ser. 4, pro 229-281, 1868.
10. A. St.ridder: J3eitriige zur Frage der Geschwindigkeitsformr.i und der Rauhi,gkeitsia.hlen fur Strome, !{a,nii.le und ge.schbssene Leitungen (Some contributions
to the problem of velocity fonnul" and roughness c(lI;lfficient for rivers, canals, and
dosed conduits). Jfillcil!l.nge'l des eidgen6:isischen Amteofiir }Vasserwirtachajt, Bern,
Swit~erland, no. 16,1923.
20. Thomas Blench: A new theory of turbulent fto,v in liquids of small viscosity,
Journal, b'stittlt'ion of einil Engineer., Lolkion, vol. 11, no. 6, pp. 611-612, April,
I~~
.
21. N. N. Pavlov'skiI: "GidravlkheskiI Spravochnik" ("Handbook of Hydraulics").
This book ha.~ many editions: (1) "Giclrl\vlicheskiI Spravochnilc," Put, Letlingrad,
1924, 192 pp.; (2) "Uchebnyl GidravlicheskiI Spravoclwik'" (for schools),.!~ubuch,
'Leningrad, lng, 100 pp.; 2d ed, W3l, 168 pp.; (3) "Gidrnvlicheskii Spravochnik,"
Onti, Leningrad and Moacow, H}37, 890 pp; !lnd (4) "Kratki! GidravlicheskiI .
Sprflvochnik," (concise version), Gosstrolizd'at, Leningrad and Moscow, 1940, .
314 pp.
,
22. George W; Pickets: Run-off investigat.ions in central IllinQis, University of Illinois,
Engineering Experim.ent Sta.iion, Bulletin 232, ,vol. 29, no. 3, September, 1931.
23. Frederick C. Scobey: The flow of water in flumes, U.S. Departm.ent of ;Agriculture,
Technical Bulletin No. 393, December, 1933.
24. Methodology for cmp and pasture inundation damage appraisal: "Training
manual for hydrologists on watershed protection and flood preventiQn work pIau
127
parties," prelimi~ary ~raft, U.S. Soil Conservation Sel'vice, Milwaukee, Wis., 1954.
25. E. W. Lane: DISCUSSIon on Slope discharge formulae for alluvial streams a.nd
by E. C. Schnackenberg, Pr(JcllJldin(J~, New Zealand 1~lItitu.l.ion of Enqillee!'s
vol. 37, pp. 4.35-438, Wellington, 1951.
'.
)
25. J. S. ~ey~!,~ :'I,nd E. A. Schultz: P~na;Ha Canal: The sea-level pI'oject, in A
symposIUm. Tlda.l currents, Tran,~act!OnB: Ametican Society of Civil Engineers
voL 114, pp. 668-571, ~g49.
'
27. Thomas R. Camp: Design of sewers to facilitate flow, Sewage 'Work.:; Joumal, voL
18, pp. I-HI, .January-December, 194.6.
28. 1i"lomas It <A1.tnp; Disoussion Oil Dete.rmination of Kutter's 11 for sewers pal.t!y
filled, by C. FI ank Johnson, Trans(lcl~ons, American Societ!! of Civil Engineers
V<ll. 109, pp. 240-243, 1944.
29. ~, ~, Wilcox: A compara.tive test of the ftow Gf water in 8~inch concrete and
vltTl.Hed clar .sewer pipes, Ur;.ilJersill1 ,of W'ashil1(Jlon, ElI(Ji'l'tee:ring Expariment
Slatton, Bulwt.ll 27, Mar. 1, 1924.
30. D. L. Yarn~ll and 8. M. Woodward: The flow of water in drain tile, U.S. Departmenl of A.gnc!tll.u.re, Bullelin No. 854,1920.
'
31. C. }'rank Johnson: Determination of Kutter's n for sewers partly Hl'led T
ti
A'S'
, .
, ransa;c
ons, 'mcnoon oClely of Ct~ll
vol. 109, pp. 223-239 1944.
32. Woody ~. Cow~n: Estimating hydraulic roughne;;s coeffici:nts, Agricultural
EngmeerL1l{/, vol. .3(, no. 7, pp.
July, 1956.
'
33, Donal~ Schnepper B,nd Ven Te Chow: Full scale toe-of-slope gutter model
unp~bI1SI:ed rep~'rt ~f an ~n",:,es~ig~t,ion conducted by the Department of Civii
Engmeenn.g, ,?nlverSlty of ~l1mols, In coopera.tion with the Division of Highways,
State of IllinOIS,. and the Bureau of Public Roads, U.S. Departn\ent of Commerce
May, 1954 (ava.tlable at the University of Illinois library).
'
34. Robert E, Horton: Some better Kutter's formula coefficients Eng'
: N
1 75""
( ,
tneer"ng ews
'1'0,
" no. 8, pp. 3r3-374, Feb. 24, 1916.
Discussions. by Fred C. Scobey and1
Robert E .. ~orton., vol. 75, no. IS, pp, 862-863, May 4, 1916.
35 . .Romce Wilham I{i?g, "Handbook of Hydraulics," 4th ed., revised by Ernest F.
~rate~, M?Gra.w-Htll Book Company, Iuc., New York, 1954, pp. 7-102 to 1-111
3a. 'Engmeer,mg ~Handbook: Hydra.ulics," U.S. Department of Agriculture, Soii'
Oonservation uervice, 1955, sec. Ii.
37. F. ~. Scobey; Flow of water ip irrigation and similar canals, U.S. Department of
Agr::cu1tw'~, Technical BuUeiin No. 852, February, 1939.
38. C;.t.. Ramser; Flow of water in drainage channels U.S. DSDartmeni 01 Anr' It
I B UL'1 etm
. N o. 129, November, 1929,
' . . ten u.re,
T ec11.n~ca
..~
. r
CHAPTER
(6-1)
(6-2)
where
The term J{ is known as the comeY(1,nce of the channel section; it is a
measure of the carrying capacity of the channel seDtion, since it is directiy
proportional,to Q . .
. .
.
I
When either the Ch~zy formula or the Mmmmg formult\ IS used D.s t"le.
uniform-flolv formula, i.e., when y = 72, the discharge by Eq. (6-1)
becomes
Q= K',IS
(6-8)
(6-3)
Primarily, Eq. (1)-8) applies to a channel s~ction when the flow is uniform. The right side of the equation contains the values of n, Q, and
S; but the left side depends only on the geometry of the water area.
Therefor'}, it shows that, fora giv~n condition of n,Q, and S, there is
only one possible depth for maintaining a uniform flow, provided that the
value of AR~i always increases with increase in depth, which is true in
most. cases. .This Q~h is the normal depth. Wyen nand S are known at
. a .()1~~!l!!!)1 section, it can 'be seeii1l1ilXl.f1Ji}]6-8) that there can beomyr!ne
discharge for maintainin a unjform flow through th~ sectlOD.~provlaect
..-:Ui8:tARW'a ways increases withincrease of depth.l ThTs discharge 18
. 'tfi'e-normal d~8~'i.a.r.gr---- .,._ ....,-.--.. :-_._-.... ,._-,-.' ...------.--:.-
Q.yS
129
Q = 1.49 AR%
(6-4)
. n
v'S
(6-9)
'
Thisis essentially thc product of the water area and the velocity defined
by the Manning formula. The).'UP-.?cript ~_g,--.ometill!!lL1L~.d t<LQecify
ihe_c..Qpdition of uniform fl9,F-.
In order to simplify the computation, dimensionless curves showing
the relation between depth and section fuctor ARH (Fig. 6-1) have been
prepared for rectangular,. trapezoidal, and cirpular channel' sections.
These self-ex:planatory curves will help to determine ,the depth for a
given section factor A.8~, and vice versa. .Ihe A R~ val ue~ for a circular section can also be 'found from the table in AppeEi!ix A.
The above two equations are used to compute the Conveyance when
the geometry of the water area and t1')e resistanc~ factor or rou~hlless coefficient are :given. Since the Manning formula IS used extensively, mos~
of the folldwing discussions and computations will be based on:Eq. (6-6).
6-2. Th~ Section Factor for Uniform-flow Computati<:n. '11h: ex~re:
sion AR~i is called the section factor for uniform-flow. eomputatwn; It IS
an important element in the computation of uniform flow. From Eq.
This is true for channels in which the v'alue of AR3, al'ways increasllII. with increase
of depth. since Eq. (6-8) will give one value of AR~,. which in turn gives only one
depth. In the CMe of a closed conduit having a gradua.ily closing top, the va.lue of
Ami will first increase wit~ depth and then decrease with depth when the full depth
is approached, becaUse a n;taximum value of AR3i usually occurs in such a conduit at
a depth slightly less than :the full depth. Consequentlj, it is possible to have two
depths for the 'lame value pf ARH, one greater and the either le5J3 than the depth for
the maximum value of Ami. For further discussion onthie subject see Art. 6-4.
This equation can Qe used to compute the convey!mce when the discharge
and slope of the channel are given.
.
When the Chezy formula is used, Eq. (6-2) becomes
K
CAR~~
(5-5)
1~8
131
(6-10J
dy
2y
(6-11)
..;
...
..s... .
Cl,
d(1nI{)
dy
= .!. dA:
A dy
+~!
dR
3 R dy
(6-12)
d
'0
c;
Q
co
.0
....'"
~(111 K) =
hO
~=
dy
.S
?- s
...
,..;
I
<0
ci
I!i;
lI
I This equation (IJ was "Iso developed independently by Chuga.ev [2) through the
use of the Chez.y lormula.
.
S Similar curves to tho'le in Fig. 6"2 for trapezoidal channels were constructed by
.Kirpich [3J and al~o prepared iudependently by Pa.vlovskil [4] alJd R~khml\noff (5).
N.
130
1
l>.
'A
- - -~---:---~~~----~--'------~~"--.,
...-..-,-"..---.c~---.~=...~-=-,_....-:
....=_
.. _... '-" .-.~~-----.
~
-,.~-.,-.~.'
':"
(6-14)
!:l
<3
(6-13)
Equating the right sides of Eqs. (6-11) and (0-13) and solving for N,
:>
2R
'0
l<J
(5T _ dyd!.)
.,.
"CJ
.,'"
J.-
3A
'"tt:
<t
.""-~-.-"--"
132
UNIFORM FLOW
or
I~+-'-----J
2.0
N' -1
0.3
:::
"
0.2
'J
log (!(dK~)
log (Yl/V2)
III
0.<1
<;
.-
(6-16)
+
I
1,0
0.8
.0.6
0.5
133
{)' Kll
./ The cutved'
plot when 'h.
depth opprooches
1 Ille
I
I
~I
crOllln
ot 0
closed conduit
0,;
0.06
0.06
0,q5
0.04
log Y
0.0 3
I
.1,
0,02
2.0
logarithmic plotting.
2.5
4,0
4.5
5.0
5.5
VOlues of N
':I
C}~ugaev
[2].)
i34
!
i
i,
J.
I
I
(
UNIFORM FLOW
the curve may' be divided into 11 number of short segments', arid eachsegment may be considered a straight line having,a constant slope orhydraulie exponent.
N ow take the circular section as an example, The dimensionl&s
log:l.rithniie plot of ARH against depth is shown in Fig. 6-1. Assuming a
const,ant '('Llue of n, this curve will show the same characteristics as if the
depth were plotted against K. As the depth' increases, the curve deviates
gradually from a straight line and finally reaches a prononnced curvature
at Yldo = 0.938, where the value of A RH/dij~'J is a maximum. Since the
n value is assumed constant, t,his ratio yl d Q = 0.938' also corresponds to
the maximum value of the conveyance K. The slope of the tangent to
the curve at this depth, according to the gl'aph in which the ordinate and
abscissa a,re interchanged, is hQrizontal, and thus the hydmulic exponent
N is equal to zero. For depths with ratio greater than 'lild o = 0.938,
the curve shows a decrease in t.he value 'Of A RH!do'~ and, hence, a decrease
in the conveyance K if r:. is ai$Bumed constant. The slope of the tangent
to the. curve and with it the hydraulic exponent will thus beconie
negative.
6-4. Flow Characteristics in a Closed Conduit with Ope-.l~channel Flow.
Taking the circular section as an example, the dhnenslol1less curves for
ARHI AoRoH and R 3i l Roy" are shown by the full lines in Fig. 6-5. The
subscript zero indicates the full-flow condition. 'If the n value is assumed
constant or independent of the .depth variation, these" two curves will
represent the variation of the ratios of the discharge and velocity to
their corresponding full-flow values (Le., QI Qoand YITT 0)' Both the
discharge and velocity curves show maximum values, which occur at
about Q,938d o and 0.81d o respectively. Mathematically, the depth for
the maximum discharge, or O.93ado) can be obtained oimply by equat.ing
to zero the first derivative' of ARH with respect to y, since the discharge
computed by the Manning formula is proportional to ARH for consta,nt
nand S. Similarly, as the velocity by t,he Manning formula is proportional to R*, the depth for the maximum velocity, or 0.81d o, can be
obtained by equating the first oerivative of R"" to zero. Flu;thermore,
the dimensionless curve of QIQo shows t.hat, when the depth is greater
than about 0.82do, it is possible to have t'l'IO different depths for the same
discharge, one !J.bove and one below the value of O.938d o. Similarly, the
cut've of V /1'0 shmvs that, when the depth is gn:&ter than the half-depth,
it is possible to have two different depths for the same velocity, one above
.
and one below the value of O.81do.
The above discussion is based on the assumption that the roughness
. coefficient remains constant as the depth che.nges. Actually, the value of
n for average clean sewer pipes and drain tiles, both. clay and concrete,
135
for example, hilS been shown to increase by as much as 28% from 1.00d n
to O.25dc, where it appears to be a maximum (see Fig. 6-5 and the disas a factor affectiIlg n value).
cussion in Art, 5-8 regarding t.he
This effect callses the actual maximum discharge and velocity to occur
. at depths of about O.97d o and O.94do, respectively. The corresponding.
curves of Q/Qo and ViTr Q are shown by the dashed lines in
6-5.
According to the assumption of constant n value, the velocity would be
the same for a half-full pipe ru; for a full pipe; whereas, if the n value is
. Subscri:>t "d' indicolas the lull tlow condition
r----:r:....-.-,---,--.-----r-- . . r--r~-.-__.~._-r--~
I
II
0.6
0.7
0,8
0.9
taken to vary with the depth, ru; shown, the veloc.ity at the half-depth is
only 0.8 the full velocity.
The discussion for the circular conduit applies also to any l:losed C011duit with a gradually closing top.. The exact depths for maximum
discharge and velocity, however, will depend 011 the shape and roughness
variation of the specific conduit section. Since the ma.x;imurn discharge
and velocity of a closed conduit of gradually closing top do not occur at
the full depth, this means that the conduit WillllOt. flow full at the maximum capacity as long as it maintaimr an open-challlHil flow 9n a uniform
grade free from obstructions. For praetical purposes, however, it may
sometimes be assumed that the maximum discharge of a cir('.ular cOllduit
or similar ,closed conduit with gradually closing top does occur at the full
I
136
UNIF,ORM FLOW
137
depth, because the depth for maximum discharge is so close to the top
. that there is always.a possibility of slight backwater to increase this depth
closer to and eventtu:IUy equal to the full depth. '
. 6-0. Flow in a Channel Section with Composite Roughness. In simple
chann81s, the roughness along the wetted perimeter may be distinctly
different fl'OJU part to PE1l't of the perimeter,but the mean velocity can
stiil be computed by a uniform-fl,uw formula without actually subdividing
the section.. For example, a rectangular channel built with a wooden
bottom and glass walls must have different n values for the bottom and the
walls. In applying the Manning formula to ~uch channels, it is sometimes'
necessary to compute an equivalent n value for the entire perimeter and
use this equivalent value for the computation of the flow in the whole
section.
.
. For the determination of the equivalent roughness, the water area is
divided imnginatively into N parts of which the wetted pel'imeters
P 2,
... , PN and the coefficients of roughness nl, n2, . . ) nN are known.
Horton [6] <Lnd Einstein [7,8] assumed thateachpal't of the area has the
same mean velocity, which at the same time i15 equal to the mean velocity
of the whole section i that is, VI = V ~ = . . . = V N = V. On the
basis of this assumption, the equivalent. coefficient of roughness may be
obtained by the following equation:'
.' .
where ~l' Hi, ... ,R/'f are hydraulic radii of the subdivided areas.
simple c~!\.nnel sections, it may be assumed that
For
RI = R2 = . . . = RN = H .
Raugnne88 of Ice-covered Channels. When a. channel is covered with
the wetted perimeter of the flcrw is greatly increased. . The bottom
surface of the ice cover may be either as smooth as a finished concrete
sUl1fa:ce or as rough as the natur&'l channel bed v,rhenclrifting ice blooks
.exist. Tabb 6-1 gives the 11, values for drcdg;ed channels covered with
ice, as proposed by Lotter [13].
.
.lee condition
n Value
1.3-2.0
>2.0
0.01D-0.012
0.014-0.017
Smooth iee:
Withol\t drifting iee blookB
1.3-2.0
0.OI6-0.01E
>2.0
.0.017-0.020
0.023-0.025
Let nand n1 be the roughness coefficients for channds with and without
ice cover, respectively. By means of Eqs. (6-17) to (6-19)' it is possible
to compute the roughness coefficient 7l-1 of the ke cover. However, the
coefficient thus computed may sometimes be a negative ,falue
. , which is,
of course, unrealistic,
.
In order t.o develop a realistic approach to the problem, Pavlovskil [14)
assumed that the total force resisting the noW' is equal to the sum of the
resisting forces due.to the channel bed and' the ice cover. Thus, from
Art. 5-4,
(5-20)
(6-17)
There are many other assumptions for the determination of all equivalent l'Oughness.Pavlovski! (9] and also Muhlhofer [10] and Einstein
and Banks [111 assumed that the total force resisting the flow (that is
JCV'"PL; see Art. 5-4) :is equal to the sum of the forces
the flo~
developed in the subdivided 3reas.By this assumption, the equivalent
roughness coefficient L'3
where the subscript 1 refers to the channel bed and 2 to the ice cover.
Since Chezy's C = -vu;jR OJ' J( = W/C2, the above equation b~comes
(Plnl 2
,I
J
(6-18)
Lotter [12] assumed that the total discharge 'of the flow is equal to thei
sHm of the discharges of the subdivided areas. Thus, the equivalent
roughness coefficient .is
PRH'
n = ...-----'-.-
(lH9~
C1
II
!
i
P1' P!
C1 2
Czl
+ P2
(1
+ a)P1i
(6-22)
t'/n,
1. 49R ,
(1
a),n2
(6-21)
RH!
(6-23)
138
UNIFORM FLOW
(1 + a)~2
11,2
(1 + ~)H
(E22
+ aEif,)
Tt2
VT+li
(a'' (T'
0Z''S)2~
7
:i
channel section is equal to the total discbirge divided by the total water "1
area.
Owing to the differenr.es that exist among the velocities of the subsections, the velocity-distribution coefficients of the whole section are
different from those of the subsections. The values of these coefficients
may be computed as follows:
Let Vt, V2, : , vI! be the mean velocities in the subsect,ions; let
al .. a2, . . . 1 aN and fJI I fJ~, ... ,fJN be the velocity-distribution coefficients for the corresponding subsections; let .6.A 1 , .6. A 2, " .6.AN be
(6-24)
139
(6-25)*
=
a = 1.
Tt =
..72 (1 + ~2~l)%
Fw. 6-6. A channel con8isting of one main section and two' side sedions.
(6-27)
(1.68n n -
nl~I)~l
(6-28)
Now let the discharges wit-hand without ice covel' be Q and Ql1
respectively. Then, using the Manning formula and assuming R = RI/2,
where Rand R 1' are the hydraulic radii with and without ice cover,
respectively, the discharge of an ice-covered channel is
Q
nl
= 0.63Ql
11,
VI =
Q = VA
(6-29)
Eq. (5-7).
Sf~
_ 1(/'1 SH
VN -
.6.AN
+ VN dAN
+ ](z + .. + ](N)S~~
0:: IC.r)
N ,
Sfi
and
Incorporating the above expressions with Eqs. (2-4) and (2-5) and
simplifying, the velocity-distribution coefficients of the entire section are
N
I
a
,~
I
f
~
i
!
X.I,
. .t .
(aNKNJjdA N')
(J
(6-30)
0; [(NY/ A'
l
and
1.
~-N~--~--~-
"
us~d
1(2
Ll.,A z
= VI .6..4. 1 + Vz dA 2 +
(6-26)
{Kl
V2 -
[(~ Sf.
.6.A. I
(f3N[(N
jdA N )
= -J-'N'-~--~--
(Il(NY/ A
I
"
(6-31)
:' ' I
140
UNI~ORM
FLOW
A, ft'
Subsection
Main section.
. - . - - I --,3-6-0-1\.
S
5,710
{J
I
~5 11~0-.-O-3-5~JII-1-.-1-0-'--1-'0-4-
P, ft
Side section. .. .. .. ..
Sol'u/'i01l:.
405
0.040
1. 11
1. 04
S"b~"tio"
AA
I-~
J. "
I /3K2/I!A
..
-I---!
",K'/I!.tl'
10"
32.49 X 10'0,
(3.136 X 10')'/11,070' = 1.29
9.76 X 10'
{J = (3.136 X 10)'/11,070 =1.10
and
, !
+ y)_
+ y vS
y(lO
10
and
A = y(20
+ 2y)
The velocity is
, II
'100
./
y(20 + 2y)
./
200
y(10
y)
."
Substituting the given quantities and the ILbove expressions in the Manning formula
Besides the methods described here, there are other methods for the computation
of uniform flo~, suc.h as the use of hydraulic tables., Popular tables for this purpose
can be found in [16] to [20].
1.49
0.025
[Y(lO
10
+ Y)-J~~ O.0016H
+ y V5
+ !:::)J2.5
Solving this equation for Y by trial and errol', V. = 3.36 ft. This is the normal depth.
The corresponding arelL is An = 89.S ft' alld the normal velocity is V. = 400/89.8 =
4.46 fps. From EXlLmple 4-2, it is known that the critical depth for the same discharge
in the channel is 2.15' ft. ,Since the normal depth is greater than the critica! depth,
, the flow is subcritical.
Soh;/ion 2: The Trial~and-e,.~or Approach. Some engineel's prefer to solve this
type of problem by trial and error. Using the given data, the right side of Eq. (6~S)
is 1IQ/1.49 ...;s = 167.7. Then, assume a value of y and compute the section fnctor
''AR%: Make several such trials until. the computed value of AR3; is very' closely equal
......t.u-liiJ....7.; Lhen the a.ssumed ii-i~;~ th~-~l~sest trial :e. the.llormaCdepth:-This tl'ial-and=
error computation is shown as follows:
0'=
'.rT"
,',
[md simplifying,
/71
or
n value
of
/r
-----1
3.00
3.50
3.30 I
3.35
, 3.36
78.0
94.5
87.7
89.5
89.8
-~II
II
2.34
2.65
2.53
2.56
2.56
Ii
I
RH
Am~
Remarks
1.762
1.915
1.852
J .870
1.870
137.4
181.0
162.6
167.2
158.0
Y too sll~all
Y too large
The closest
-_.,
142
UNIFORM FLOW
COMPUTATiON
UNIFORl\I FLOW
143
b. Determine the critic.!tl slope a.nd the corresponding norma.l depth when the dis"
charge is 400 cis.
'
'
c.D~term!ne the critir.;a.1 slope at the n01'l:11al depth of 3.35 ft, a.nd compute the ~or
respondmg dlscharge.
So:1ttion.
' . (v.)
h From the given data it is found that R = 2.56 it and V '= 446
' f ps..
Su bstltutmg
t ese va.lues; in the Manning formula and .~ol-dng for Sn,
yla., =
OF
il.72,
A,
Ad -
1.49 256"'8
I"
0.025
. ~.,,'
s.
0.0016
C'
or
This is the slo~e that will maintain a unifO~'m flow i~ the gh'en channel at a depth oI
3.36 ft and a discharge of 400 cfa (see Example 5.2),
b. From the given data t.he critl!:al depth is found to be 2.15 ft (see Example 4-2)
The corresp:ndin g. v:J.lues of R ~nd V are R
1.97 f~ and V
7.60 Ips. Substitutin~
the known ~ A.lues m the MnnnlIlg formula and solvmg for Sa, '
1. 97HB.l."
7.66
or
S.
J
I
'l.
,I
I
0.(1067
Th~s is the slope that will maintain a. unifornl and critical flow in the givell cha.nnel for
This is ~he slope that will maintain a uniform and critical flow in the given channel
at the glvp,n normal clepth of 3.36 ft. The
disoharge is equal to 9.3 X
89.8 = 835 ds.
Example 6-6. petermine the limit slope of !I. rectangular channel (Fig. 6-8) with
10 ft and n = Q.0l5.
.S~ll'tion. .Since the limit slope is the :;mall~.st critical slope, its value' maybe determmed gra.phlc.\J.lly.rro~ a Cll::ve ~f.the critical slope plotted aga.inst discharge.
~or the determmatJOn of a cntlcal slope, the following two cOlldition~ should be
satisfied:
L The first condition, from Eq. (6-3), is
S.n = 0.0070
.Q = K
vB.
(6-32)
(6-33)
i
I.
,t .
(f}-34)
'I
Q =Z,...;g
or, for the .ectangulul' chmnel,
Q=
1O....;g y'"
(6-36)
By using Eqs. (il-34) and (6-36) .iLnd eliminating 11, the relation between Q and S,
can be established. This relation is expr;lssed, however, as an implicit function, and
Supercrltlcal
flow
2.
Slope in 10-3
(al
F!G~
145
UN)FORM FLOW
144
:3
4
5
Slape In 10- 3
. (b)
When any four of the above six variab'lesare given, the remaining two
unknowns can be determined by the two equations. The following are
some types of problems of uniform-ftow computa~ioll:
A. To compute the normal discharge. Inpractical applieD.tlons, this
comput.ation is required.for the determination of the capacity of a given
channel or for the construction of a synthetic rating curve of the channel.
B. To: determine the velocity of. flow. This computation has many
appli.cations. For example, .itis often required for the study of scouring
and silting effects in a given channel.
C. To compute the normal depth. This computation is required for
the determination of the
of flow in a given channel.
D. To determine t.he channel roughness. This computation is used
to ascerta.in the roughness coefficient in a given channel; the .coefficient
thus determined may be used in other similar channels.
E. Ta compute the channel slope. This computation is required for
adjusting the slope 0(0, given channel.
F. To determine the dimensions of the channel section. This computation is required mainly for design purposes .
Table 6-2 lists the known and unknown variables involved in each of the
six types of problems mentioned above. The kllown variables are indicated by a check mark (v') and the unknowns required in the problem
by a question mark (?). The unknmm.val'iables that can be determined
The last
from .the known variables are indicated by a dash
TABLE
Type of
problem
t>-2.
Discha.rge
Q
v
v
v'
v
D
E
F
VeIoeity
V
.-
.-
I
De th
p
11
v
v
1
v
v
v
Roughness
v
v
v
?
v
v
Geometric
elements
Slope
S
v
v
v
v
? ,
v;
v
v
v
'I.
v
?
. Exa.mple
:
Prob. 5-5, (Ex, 5-0
Prob. 5-5, (Ex. 5-1-)
Example 6-2
Prob.5-6
Example 6-4a
Example 7-2
145
UNIFORM FLOW
column of the tabie shows the example given in this book for each type
of problem. The examples shown. in parentheses are solved by the. use
of the Chl:z;y formula. It should be noted, however, that Table 0-2
does not include all types of problems. By varying combinations of
various known and unknown variables, more types of problems can be
formed. In design problems, the use of the best hydraulic section and
of empirica.l rules is generally introduced (Art. 7-7) and thUl'J new types
of problems are created .
. 6l.9. Computation of Flood Discharge. In uniform-flow computation
it is understood, theoretically, that the energy slope Sf in the uniform-flow
formula is equal to the slope of the longitudinal water-surface profile and
also to the slope of the channel bottom (Art. 5-1). In natural streams,
however, these three slopes are only approximateiy equal. Owing to
inegular channel conditions, the energy line, wB,tel' surfa.ee, and channel
bottom cannot be strictly parallel to one another. If 1;he change in
velocity within the channel reach is not appredable, the energy slope
may be taken roughly equal to the bottom 01' t.he surfs.ce slope. On the
other hand, if the velocity varies appreciably from one end of the reach.
to the other, the energy slope should be taken as the difference between
the total heads nt the ends of the reach divided by the length of the
reach. Since the total head includes the velocity head, which is unknown,
a solution by successive approximation is necessary in the discharge
computa~ion.
.
During flood stages, the velocity var~eS greatly,. and the velocity head
should be included in the total head for defining the energy slope. Furthermore, flood flow is in fact varied and unstean'y, and use of a uniformflow formula for discharge computation is acceptable only when the
changes in flood stage and dischurge are relatively gradual.
The direct use of a uniform-fl.ow formula for the determination of flood
discharges is known as the slope-aJ'ea method. The flood discharge may
also be determined by another well-kno".,n method called the contractedopening method, in which the principle of energy is applied directly to a
contracted opening in the stream. Both methods l require information
about the high water marks that are detectable in the flomded reach.
Good locations for coll~cting such information may be found not only
on main streams but also on smaller tributaries; but they ml.\st be either
comparDtjvely regular valley channels free from bends and thus well
suited to the slope-area method or else contrll-cted openings wiyh sufficient
constriction to produce definite indrease in head and velocity and thus
suited t.o the contracted-opening method.
The foilowingis a description of the slope-area method. 2 The conI
COMPUTATION OF UNIFORM
:now
147
3. Assuming zero velocity head, the energy slope is equal to the fall F
of water surface in the reach divided by the length L of the reach, or
S
F'
(6-38)
vB.
(6-3)
s=!!:!.
L
where
j
I
I
!
I
i
J1,
(6-39)
(6-40)
and kis a factor.' When the reach is (;ontracting (Vu < V d ), k = 1.0.
When the reach is expanding (V" > V.), k = 0.5. The 50% decreru;e
in the value of Ie for an expanding reach is customarily assumed for the
recovery of the velocity head due to the expansion of the flow. The
corresponding discharge is then computed byEq, (6-3) u.'ling the revised
flow, but it is believed that at this stage of reading the rcad~r should be able to follow
the procedure.described here. This method shows how the uniform-flow formula can
be applied to gradually vllded flow and thus paves the way for a more cOinprehensive
treatment on the subject of gradually varied flow in Part Ill.
.~
148
149
UNIFORM FLOW
surface flo~ occurs mostly as a result of natural runoff, and is called oo'erland flow.
Uniform flow may be turbulent or laminar, depending upon such factors
11..'> discharge, slope, viscosity, and degree of surface roughness.
If
velocities and depths of flow are relatively small, the viscosity becomes a
domina.tiQg factor and the flow is laminar. In this case the Newton's
lD.w of viscosity applies. This law expresses the ~~at~.9l1_p~tween the
Example 6-6. Compute the flood discharge through a riv.er reich of 500 it ha.ving
known values of the ws.t"r areas, conveyances, and energy e,oelficient8 of the upstream
and downstream end sections. The fall of water surface in the reach was found to be
0.50 ft.
Solution. ';rI.e. wl).ter areas, conveyances, and energy coefficients for the two elld
sections of the' reach are:
," Au
Ad
11,070
10,990
=
=
K, = 3.034 X 10 6
K. = 3.103 X 10 6
.J,~,~Ofti~T!~
~V~~~,~"
V?",,-~
~ .....
1,;:\.~j
~"
'r-
.~ f' "'"(i"', r
w(Y"l~'Y)S ( _
. t~F"'~~101l
au = 1:13"
O'd
" Ii!
.(
....
.~~
..,
;.:.;
) ' ..,~
,/
'1:':~7
,., AifC
~~1:;;t
. 2(/ ad
.""
2(/
1.354
FHl. fr9. Uniform laminar open-channel flow.
1,42,1
-0.070 .
Since l'u is less than V d, the flo," is contracting, and k = 1.0. Hence, h, = 0.500 0.070 = 0.430, S = 0.430/500 = 0.00086, "IS = 0.0~93, and Q = 3.070 X 10' X
0.0293 = 90,000 ds.
Similarly, other approximations are made, as shown in Table 6-3. The estimated
discharge is found to bl! 91,000 cfs.
TABLE
6-3.
Appr.oximatlon
IAs~umed
1st
2d
3d
4th
5th
Flo. ~
97,000
90,000
91,200
91,000
2g
a~ ~l
2U
6-6
h,
v'S 1Computed
Q
-~----I-!-----
face, so-called S1J.1"!ace flow is produced. The depth of the flow may be so
thin in comparison with the width of flow that the flow becomes a wide-'
open-channel flow, known specifically as sheet flow. In a drainage basin
,.,j
I>
du
dy
(6-41)
dv
g8 (Ym - y) dy
J'
g8 (YYm _
)I
t)
2
fr.
J,;.iii:.
i n7 .... , ~,
.
'\\
' . . . ,
_,(.
.'
1.177
~,'
.'
'flop'li' 't~o
(8-42)
UNIFORM FLOW
150
V = -1
Y
10 ' vdy
11
PROBLEMS
(6-43)
'y
(6-44)
,,-'~
6-1. Detern:ine the normal discharges in channels ha.ving the following sections for
.
:r ." ..... -,
151
:j:
6-2. Prove the following equ1Ltion for the dischuge ina. tl'iangular highway gutter
(Fig. 6-10) having one side vertical, one 3ide sloped at 1 pn z'. MallTlilig's n, depth of
I
I
\
where
. 6~3. Compute the discharge in the triangular highway gutte!' described in the precedmg problem when z = 24, n = 0.017, Y = 0.22.ft, and S = 0.03.
6-~. U~ing the Manning formula, determine the hydmulic exponent N for the [01lowing channel sections: Ca) a very narl'OW rectan(!;ls, Cb) a very wide rectangle, (c) a
Y~ry wide panbola for which the wetted perimeter i~ practically equal to the top
Width, and (d) an equilateral triangle with a vertex at the bottom.
.
6-5. UBing the Ch6zy formula,' ~how that the general equation for the hydmulic
exponent N is
.
1:..
(3 T _ U dP)
dy
(/H9) .
6-6. Solve Prob. G-4 if the determination of the hydl'!l.)llic exponent is based on the
Chezy formula. Compare the results with those obtained in Frob. 614.
I The G. K. formula shows that Chezy's C is a f~nction of the hydrauiic radius and
hence of the depth ii. Thus, the Chezy formula has hot been found very convenient
for determination of the N value. Forcanal~ in earth and gravelly soil the N value
is generally found to havBan increase of 0.30 to 0.50 due to the v!l.ri!l.ti~n in Chezy's
C with respect to the depth. Thi~ i~crease, however, brings the N valuc closlJr to
that bMcd on the Manl1ing formula.
1,
(
-\
;;1
'152 .
UNIFORM FLOW
6-7. Compute the l)ydraulic exponent N of the trapezoidal channel section (Fig.
2-2) having a normaldep.h of 6 ft, using (al Eq.(6-15), (b) Fig. 6-2, and (el the
graphical method based on Eq. (6-16).
6-11. Compute the hydraulic exponent N of a 36-in. circular conduit having a n~r
mal depth of 24 in. above theinvert, usi,ng (al Fig. B-2 and (b) the graphical method
based on. Eq. (6-16).
'
6-9. Using the Manning formula, show that the depths for a maXimum discharge
and velocity in a aircular conduit are, respectively, 0.938d. and O.Sld.. .
6-10_ On the basis of the Chezy formula, determine the respective depths for maximum discharge and maximum velocity in a circular conduit.
.
6-1~_ A~ what depths will the maximl.lin discharge and velocity occur in a square
cOlldUlt laId flat on one side?
. '
. 6-12. Prepare the curves of dis~harge: andveloeity variations with respect to the
depth in ll. square conduit laid on one side.
6-1S. A channel is assumed to have a constant hydraulic radius R for any dep'th of
flow. .Prove that the CrOss section of this channel can be defined by
y = R[ln (x
vx' - R')
II
Q.. The cross section of the stream at 11 typical upstrel1m station in the uniform
Tea<:h is given hy the following cQordinates:
StrLUr:f(1.
Elev. m.s.!.
Sta/ian
Elev. m.s.l.
590.0
580.7
578.2
582.0
581.0
580.0
6. + 00
8 + 00
10 + 00
11 + 00
12 + 00
14 + 00
543.7
540.0
572.2
573.2
568.5
590.0
Left b8lnk: 0
1
1
3
+ 00
+ 00
+ 50
+00
4 +00
5 + 00
b. The value of n for the main channel is estimated as 0.035, for the side channels
0.050..
e. The nl1~ural slope of the stream is about 1 it/mile.
Construct a synthetilJ re..ting curve. It is suggested that th", wl1ter areas of the '.
main channel and the side channels be seplJ.lated by the extended sides of the main
1l.S
ch~.nnBl.
- In R]
(6-50)
= R when y = o. Draw the sketch of this section and discuss its properties.
(~IN;~: Frrm~ the giv~n condition, R = A(P =dA/dP "'" "dY/Vd:!"' + dy'. Solve
where x
thIS mfferentlill equatlOn, and evaluate the mtegration constant by the condition that
x = R when 11 = O. MathematicallY, the section is formed by two catenaries as
sides. For practical purposes, an. artificial bottom should he provided ~iilce the .theo~eti.cal section is bottomless. A.uniform-flow fOl'mulfl, such as the Manning formula,
lIlfiIcates that the hydrfLlrlic radius is the sole shape parameter for the velocity. The
adequacy of this indication can bo verified experimentally hy testin'l' a channel built
of tha section oi constant hydraulic radius. If the indication is tru:, then, once this
cl~annel is designed for a safe velocity; it should I;>e nonscouring and nonsiit,ing bver a
WIde mnge of stages. In .earthen canals, however, the large variation in water surface during the change of stage wouldetode the sides very easily.) .
.
6-14. Verify Eqs. (6-17)to (8-19).
6~16. A rectangular testing channel is 2 ft, wid'e and laid on a slope of 0.1035%.
When the channel bed and walls were made smooth by neat cement, the measured
normal depth of flow was 1.36 ft fOT a discharge of 8.9 cfs. The same' channel was
then roughened by cemented sand grains, and thus the measured norrne,! depth became
1.31 ft for a discharge of 5.2 cfs.
a. Determine tile discharge for a normal d~pth of 1.31 it if the bed were roughened
and the walls were kept smooth.
, . '
.
b. Detel'mine the discharge for a normal depth of 1.31 it if the walIs were rOl.]ghened
and the bed were smooth.
_
.'
c. The discll!ir~es for the conditions described-in a and b were actually measured and
found to be' G.50 'and 6.20 cis, respectively'. Determine the corresponding n; values,
and compare these values with those computed by Eqs. (6-17) to (6-19).
'
.
6-16. A chann~l consists of a main section and two side section.s (Fig. 6-6) .. Compute the total disbhfLrge, assuming that the main section and the two side sections are
separated Ca) by vertical division lines and (bl by extended sides of the main channel
Given: n = 0.025 for the main channel. n = O.OBOfor the side channels, andS i= 0.001:
6-17. The hydrographic survey of a stream jndicates tha.t the hydraulic poperties
of the stream llre relatively unifOl'm for a length of over 2' miles. The data 0btained
by the survey" are:
'
6-18. Compute the discharge in I1n overfliJ~ved highway gutter (Fig. 5-10) having Il.
depth of flow of 3 in. and a longitudinRl slope of 0.03. The gutter is made of conprete \vith n = 0.017 and has a t~iangular section with a vertical curb side, a sloped
side of z = 12, and a top width of T = 2 ft. The overflowed soil-aggregate pD.vement has a crosg slope of z. = 24 and n, = 0.020.
6-19~ For all equal amount of discharge, an ice-covered .channel should have
greater depth of fl(HV than an uncovered channel, lor two reasons: (1) thc wetted
perimeter is greater in an ice-covered channel and thus results in greater resistance or
less velocity. and (2) the thickness of the ice cover is greater than a depth of wa~er oi
equal weight, since the spGCific gra\'ity Qf i~e is about 0.917. Sbow that the increase
. in depth due to resistanee in' an iue-covered wide open channel may be expressed by
. I:>y
[1.31! (nn.)% - 1 ] Y
(6-51 )
where n, is the roughness coefficient of the channel with' ice cover. n is the roughness
coeflicient of the channel without ice cover, and y is the depth of flow in the channel
carrying the same discharge but without ice cover.
.
6-20: Compute the conveyauce and velocity-distribution coefficients of a channel
ser.tion 500 ft downstream f!"om the section described in Example !l-l. The survey
data at the secti.on 'fol' the same flood are:
Subsection
A., ft'
P, ft
'"
(3
5,320
205
408,
0.035
0.040
1.12
1.10
1.05
1.0-1
"':1
5,670
UNIFOll.iVI FL01-V
155
REFERENCES
a.nd that
thi~
= 14.5n>
crt.
(6-53)
lIH
6-27. Deterrnhie the limit slope of the channel described in Exa.mple 6-4,
6-28. Construct the critical-slope curves of the cha.nnel described in Exampie 6-5
for bottom widths b == 1 ft, 4 It, 2() ft, and "'.
6-29. Determine the critical-slope. curvl',s oi the channel desi::ibed in Example 6-4
for side slopes z
I, 0.2, 0.5, 1, 2, 5, and "'.
6-30. A cha.nnel reac)11,OOO ft loug h<L~ a fall of 0.35 ft in wa~er surfa.ce during It
flood. Compute ~he flood discharge through this reach, using the following data:
A, ft'
Subsection
Upstream:
Main channel. , . .......
Side channel .... .. , ....... .
Downstream:
channel ... , . ......
Side channeL. ....... ...
~
..
4,250
25,620
2,050
5,760
25,610
l,gOo
320
11.
Of
0.038
0.038
1.10
1.0'1
1.20
L08
0.042
0.038
1.10
1.18
1 ..04
1.06
I,.
I
!I
I"
.'!
"> ,,/.-:'
1. Yen Te Chow: Integrating the eqltation of gradually varied flol\', paper 838,
ProceediniJs, American ,"!ocietl! of Civii Engineers, voL 81, pp. 1-32, November, .
1955.
2. R. R. Chugaev: NekotorY(l voprosy neravnomemogo dvizheniia vody v otkrytykh
prizmaticheskikh ruslakh (Abollt some questions concerning nonuniform flow of
w!).ter in open channels), /zv6stiia. l'Besoiuznogo Nauchno-Issierlov(Ltel'skoIlO [nstitllta.
aidrotekhniki (Tri.lnsa:clioM, AU-Union Scientific Re6eard~ institl,/e of Hydraulic
Enyincerinp), L~ningl'!ld,vol. 1, pp. 157-227, 1931.
3. Phillip Z. Kirpich: Dimeru;ionless cons~ants for hydr>l.ulic elements of opellcha.nnel cro~s-sections, Civil Engineering, vol. 18, no, 10, p. 47, October, 1948. .
4. N. N. PavlovskiI: "Gidra.vlicheskii Spravochnik" ("Handbook of Hydmulics"),
Onti, Leningra.d and Moscow, 1937, p. 515.
5. A. N. Rakhmanoff: 0 post.roenii krivykh svobodnolpoverlthnosti V. prir-maticheskikh i tsilindriche.~kikh ruslakh pri usta.novivshemsia dyil!henii (On the construction of curves 01 free surfaces in prismatic nnd cylindrical channels with
established flow), hvestiia V sesoiuznoyo Naudmo-Is:sledouaf.eI'.kO'lo bl.lititula
Gidrotekhllim (1'ranS(lctions, All-Union Scienli.fi.c ResC(Lrcfl InstituteDf Hydraulic
Engineering), Leningrad, vol. 3, pp. 75-114, 1ll31.
6. Robert E. Horton; Separate roughness coefficients for channel bottom and sides,
Jj7'dl""'.""di News-ltccord, vol. 111, no. 22, pp. 052-653, Nov. 30, 1933.
7, H. A. Einstein; Del' hydraulische odsr ProJil-Radius (The hydraulic (11' cross section radius), Sch",ei$~rische Ba.uzeilv,1!g, ZUrich, vo!' 103, no. 8, pp. 89-91, Feb. 24,
1934.
8. Ahmed M, Yassin: Mean roughness coefficient in open cho.nnels with different,
roughness of bed and side walls, Eid(jelll!,~.rische technische H oCMchllle Ziirich,
Mitieilungen ruts der Vcnu.chsa.nst(lltflIT Tfr(lsse!'oau una Erd:bau, No. 27, Verlag
Leemann J ZUrich, 1954.
9. N.N . .Pavlovskii: K voprOStl 0 raschetnoI formule dlia ravnomernogo dvizheniia
y yociotoka.hk s llcodnorodllymi stenkami (On a design formula. for uniform move
men~ in channels with nonhomogeneous walls), Izul1siia VsesoiUZllogo Nau.cir.lIoI 8sledovatel' skolJo Instiluta. Gidratekhnikt (T"(LMrtClions, ,111- Unio;~ Scienlifu;
Resear~h Inslill1le of Hyd1"altiic Erl9ineeril~g), Leningrad, vol. 3, pp. 157-164, 1931.
10. L. MUhlhofer: Rauhigiteitsuntersuclmngen in einem Stollen mit betonierter Soble
und unveddeidete~ Wand en (R(Jughne5S investigations in a shaft with concrete
bOUom and unlined wo.lis), Wass8rlcra!t u.na Wasserwir/,sdtafl, Munich, vol. 28,
no. 8, pp. 85-83, 1933.
.
ll. H. A. Einstein and R. B. Banks: Fluid resistance of composite roughness, Trans- .
actions, Am.erican Geophysical Union, vol. 31, no. 4, pp. 603-6jO, August, 1950,
12. G. l{. Lott.er: SOQbrazheniia k gidravlicheokomu raschetu cusel s l'!>tlichnoI
sherokhovatosliiu stenoI!: (Collsidera.tionson hydraulic design of channels wi~h
different roughness of \ir8.1I8), /zu.estiia. V~e~oiu2nogo Nauchno-!ssledvvateJ,'skof/Q
Instiluta. Gid1'otekhniki (Transa.ctiqna, .till-Union Sde1\tijic Reseprch In.stilutc of
Hyd~at;lic En(Jineering), Leningrad, voL 9, pp. 238-241, 1933.
.
13. G. It Lotter: Vliianie uslovii ledoobrazovaniia. i tolshchinY l'da naraschct derivatsionnykh ka.nruov (Influence of condi~ions of ice formation and thickness on the
design of derivation ca-nlLls), IZTleiltiia. Vsesoiu.znogo Nav.chno-I:ssled()va.t~l'slco(jo.
Ins/itu/a Gidrolekhltiki (Tra.nsactions, All-Union Scimtiji.c ReseO;rch 17Ultitute of
H1!d~G7j.lic Engineering), Leningrad, vol. 7, pp. 5&-80, 1932.
14. G. Ie LoLter; Metod akademilca N. N. Pavlovskogo dUo. ojJcedeleniia koeflitsienta
156
. UNIFORM FLOW
CHAPTER
157
./
'~.
UNIFOR\I1 FLOW"
158
159
a.now the pradicable Use of just about any type of lining or lining tl'eatment now or in tho future ant.icipatcd by the Blireau.
7-6. Freeboard. The freeboard of a channel is the vertical distance
fi'Oill the top of the channel to the water surface nt the design condition.
This distance shouid be sufficient to prevent wnves or fluctuations in
water surface from overflowing the sides. This ftwtor becomes imp or. tant particularly in~he design of elevated flumes, for the Rume substructure may be endangered by any overflow.
There is no universally accepted rule for the'determination of freeboard, since wave action or wat~t-8urface fluctuation in a channel ma.y
bi;: created by many uncontrollable causes. Pronounced Waves and
fiu.ctuatioll of water surface are generally expected in channels where the
velocity is so 11igh and the slope so steep th~t the flow becomes very
unstable, or on curves where high velocity and large deflection angle
may cause appreciable superelevuted water surface on the convex side of
a curve, or in chnrmels where the velocity of flow appt'oaches the critical
state at which the water may flow at alternate depths and thus jump from
the low stage to I;hc high stage at the least obstruction. Other natural
causes such as wind movement und tidal adion may also induce high
waves and Tequire special considemtion in design.
Free boards va.rying from less than 5 % to grea,ter tha.n 30 % of the depth
of flow are commonly used in design. For smooth, il1terior, semicircular
metal flumes on tangents, carrying water at velocities not, greater than
. 80 % of the criticnl velocity with a maximum of 8 fp.'l, experience has
indicated that a freeboard of 6 %. of the flume diameter shouid be used,
For flumes on curves with high velocity or deflections, wave Ilction will
be produced; so freeboard must he increll.sed to prevent water from slopping over,
.
water to pick up lining blocks and push them out .~~ ??SlulOn.
movmg
..
h Id be designed against such POSSibilItIes.
Accordingly,
the, Immg
SOli, 'bi e VelO~l'ty
h M'
urn PerrnlSSI
' - . The mim:mum pel'missible
.
7 -3, T e
mlm "t'n
is the lowest velocity that Will not"
velocity, ?L' I'.he ~onS'O" ~d ~ duc~ the growth of aquatic plant and mo~s.
start sedlmentatlOn an IJ1 t ' . d its e"act value cannot be er.sIly
1'1 .
I ' ty is very uncer aJl1 :Ln
l\.
h'
liS ve. OCI
' ' no
ISlt load
d For water carrymg
, or for desiltecl flow,, t IS
.
, t for .its effert
fdetermme
t . h s little
si"'nI'fi canee excep
- onplant growth.h Gen'h
. ac IIOJ a kino'' .a mean
t>
l
'
t
'
f
2
t
'Hps
may
be used safely w en"
ve OCI y 0
0
.
't ef
era y spea
"".
,
t' the channel is small, and a mean veloci y 0
percentage of sIlt preseDillln.
t a growth of vegetation that would
not less than 2.5 fps W pi even,.
'
th
,. i'lg capaCity of the channel.
,
sedously decrease
longitudinal bottom Glope of cha.nnel is
7,.4. Channel Slop ,
ra hv and the energy head required for
generally governed by the topog , P', h slope may ,depend also on the
h f!
f ater In many cases, t e
. ,
t e . ow 0f the
w channe
,
1 F'"
'or exanlple , channels' used for W:Lter-dlstrlh d
r
' pur~ose o.
"
' b se used in irrigation, water supply, y :au I~
butlOn plll poses, such as t o ,
.
h' , level at the pomto!
. . ,
d h droDower proJects reqUIre a Ign
.
mll1Jl1g,
y . a smaL1 b-I'
'~d eSI'rable in order to keep the los3 III
delivery;~J.l1
therefore,
ope I::>
v~locit'lJ
::allihe
elevation to a minimfum' h
1 deIJend mainly on the kind of materiaL,
The side slopes 0 a e anne
. h
.
" 7-1 give:, a genera1 I.d ea o
f the slopes
Table
. suitable for use Wit 'l'fLrJOUS
'
TABLE
7-1.
' sLOPES
SUl'rABLE SID;;;
OF MATElRIALS
. 111 aterial
Rock .... ,.,..
. .. , ........ ', ... , *
Muck and .peaL soils . .. ", .. . .... : .'........ , ... ~ ........ .
Stiff clay or earth with, concrete Immg ... , , . ' , .......... .
"
Etlft.h ,Vlth
Btone I"Wing, or earth
. for large eila,nnels ....... .
Firm clay or earth for small ditches ........... , , .......
Loose sa.ndy earth. , ....... , ... , .... , .... , , .. , ... , : . , .
Sandy I06.m 'or porous clay .. , . : .. : ... , ...... , . , .. , .... ,
~
Side slope
Nearly vertic al
~:i: 1
H':1 to 1:1
'an
1:1
1%:1
2:1
3:1
. .
erodible material, however, a more accurate
kinds of
h'd b e checked
[wainst
the criterion of
. m~LterJal.
,
f th For
lopes SOUl
"
"0
determll1atwn 0. . e S l i t (Art 7-10) or by the principle of tractive
maximum permli5s1bb~~e. f:c~ors to' be considered in determining slopes
,~1.
~dition of seepage loss, climatic change,
force (Art. 7-14).
are method of constru<..; lon, COl
. . 'd 1
hould be made as
.
t
Generally speakmgj SI e s opes s
channel Size, .e c.
d
Idb designed for high hydraulic efficiency
steep as practICable a~ s ou 1 e th US Bureau of Reclamation [4]
and st:Lbility. For 11l1ed can~~,
e'~'5'1 slope for the usual sizes
has been considering standarfdlzhl~lg ~n a .. that it is sufficiently flat to.
of canals. One advantage a t IS S ope IS ,
~.
I
I
1;,
Freeboll.rd in
unlined cann.'l OJ' lateral will nornu;.lly be governed by
considerations of canal size and locatiOJ1, storm-water inflow, and watertable fluctuations eaWled by checks, wiud action, soH characteristics,
percolation gradients, operating road requirements, and availability of
excavated material. According to the U,S. Bureau of Reelamation [4J,
the approximate range of freeboard frequently used extends from 1 ft
for smalliateral"/ with shallow depl;hs to 4 it in canals of 3,000 cfs or more
capacity with reheively large \vater depths. The Bureau recommends
that preliminary estimates of the freeboard required under ordinary
conditions be made according to the following formula:
(7-1)
where F is the freeboard in ft, y is the depth of water in the canal in ft,
and C is ll. coefficient varying from 1.5 for a canal capacity of 20 cfs to
2.5 for a canal capacity of 3,000 cfs or more~ This approximation is
.\
II
160
I,
UNIFORM FLOW
be noted that a best hydraulic section is the section that gives the minimum area for a given discharge but not necessarily the minimum excav!1tion. The section of minimum excavation 'occurs only if 'the water
surface is at. the level of the bank tops. Where the water surface is
below the bank tops, as frequently occurs, channels narrower than those
of the best hydraulic section will give minimum excavation. If the'
water surface overtops the banks and these il.l'e even,with the ground
level, wider channels will provide minimum excavation,
TABI,E 7-2.
AArea
Cross se<;tion
II_~__.
I'
Tmpezoid, half I
of a hexagon I
Rectangle, half
of a square
I
Triangle, half of
a :;quare
Semicircle
T=2V2y
Hydro~tatic
catenary
Capacity. cIs
T_I_D~ _ _Z_ _
il, _ _
.:
4y
,II
y'
I 2 V2 11
Section
. factor
R-I_ _
Hydraulie depth
%'
%,y
~YI
1. 39.58Sy'
2, 9836y
'%y
2YI
%y'"
2:\1',6
Try
y'
v3" 11
2111
20YI
4-y
-4 y'"
%YI % y'3y u
.' The pri~ciple of the best hydraulic section applies only to the desig.n
of nonerodlble channels. For erodible channels, the principle of tractive
force must be used to determine an efficient section (Art, 7-15). .
(U.S: Bureau
of Reclamation.)
~Y' i
I "~YI
72YI
"
I
1% V'2 )%'V2
HYt
!
r
.)
P,arabola,
161
based upon average Bureau practice; it will not, however, serve for all
conditions.
.
For lined canals or h1.teralS, the height of lining abo"';'e the water surface
will depend upon a number of factors: size of canal, velocity of water,
curvature of alignment, condition of st0rm- and drain-water inflow,
fluctuations in water level due to operation of flow-regulating structures, .
and wind action. In a somewhat similar mannel', the height of bank.
above the water surfitce will vary with size and location of canal, type.
of soil, amount of intercepted storm or drai.n water, etc. As a guide for.
lined-canal design, the U.S. Bureau of Replaination [3] has prepared
curves (Fi~. 7-1) for average freeboard and bank heights in .relation to
capacit.ies.
Example 7-1. Show that the best hydraulic trapezoidal 8ecti~n is one-half of a
hexagon,
Solutio ll . Table 2-1 gives the water area. and wetted perimeter of a tl'apezoid as
A = (b
+ zy)y
and
P = b
+ 2 VI + .' y
whe~e y is th~ depth, b is the bottom width, and .: 1 is the side slope,'
.. Flrst, conslder A and z, to b,e constant. : Differentiating the above two equations
with respect to y and solvmg Simultaneously for dP Idy,
.
.
dP
dy = 2( v'f+Zi
b
y
z)
f
t
I
,+ ZZ
2)
Substitl!ting this equation for b/n the previ~us two equations for A ELnd Pand solving
11
simu1t,a.neDusiy for: P,
p
=z
of
z =
163
UNl.F'ORMFLD'W
1,62 .
~3 ~ ta.n30
35
AR'i=~
1.49 vIS,
practicability.
5 .. For the design of irrigation channels; the channel section is some~
times proportioned by empirical rules such as the simple rule given bythe
early U.S. Reolamatiau Service [51 for the full supply depth of water in
-vA
25
'"<>.
..
'0
<:
20
"
of Art. 6-6.
.
4. If the best hydn.ulic section is required directly, substitute in
Eq. (6-8) the expressions for A and R obtained from Table 7-2 and
sol verOr the depth. This best hydraulic section may be modified for
y= 0.5
'-
.=
~
(6-8)
The determination of the depth for the computed value of AR7~ can
be simplified by use of the design chart (Fig. 6-1). Some engineers
prefer a. solution by trial and erro!", similar to Solution 2 for Example 0-2
feet ..
30
== 15
,,,!
"E
2
0
OJ
:>
Q!
10
100
IPOQ
'1,000
'!,pOD
FUIG ' 7B-2. Experience cur:re.sshowing bottom, width and depth of lined channels
( .S . .ureau of RG~aJltatum,,)
. '
ex~~!~I~t~ting
A, = (b
+ zy)y
a,nd R
(t,
+ zV)y!CIJ + 2 Vi + z' y)
-::::---'-":......:..~~;.--'
y)%
(7-2)
== 167.7
in the
~ove
= 2 and simplifying,
7,680
164
i'
UNIFORM FLOW
It should be noted that thia solution is exactly the same as the computation of
the normal depth given in Solution 1 of Example 6-2. Accordingly, the solutions by
trial il.nd error and by the graphical method described in Exo.mple 6-2 can also be
applied to the present prublem.
Similarly, assume other suitable values of band z, and compute the corresponding
depths. The final' decision on dimensions will depend on practical considerations.
If the 'values of band z are decided o.t the 'beginning of the compuio.tion, the depth will
be c,)mputed only once.
Suppose that b =20 ft, 2 = 2, and y = 3.36 ft are the final vo.lues. Assign a freeboard of 2 ft; the total depth of the channel is, therefore, 5.36 ft and the top width of
the channel (not the width of the ,vater smface) is 41.4 ft: The water area is 89.S ft',
and the velocity is 4.46 fps, which is greater than the minimum permissible velocity
'
for inducing silt, if any.
When the best hydraulic section is required, substitute A, = vl3y' and R = 0.5y,
obtained' frem Table 7-2, in AR~~ = Hi7.7 !!p..d simplify; the depth is found to be
y = lUI it. Add 3 ft freeboard; the total depth is 9.1) ft. The corresponding: bottom
width is 7.6ft, the top width of the dUlllnel is 18.7 ft, the water area is 75.2 ft', afld the
velocity is 5.32 fps. Since the, best hydraulic trapezoidal section is the half hexagon,
t,he side slopes are 1 on
7-3.
CONVERTED BY THE
va/a.
Material
I It has been noticed that certain channels are erodible whereas others v:ary similar
in cha,nnel geometry, hydraulics, and soil physical properties are not. As a further
step in investig~tioIl, the chemical properties of the material forming the channel
body should be explored, It may be that an ion exchange between water and soil or
hydration of the, material is providing a binder in some places and thus affecting the
erosion. For a general discussion of the compiexity of this problem, see [81 and [91.
Clear water
_I
'7\
7-8. Methods of Approach. The behavior of flow in an erodibl~ channel is influenced by so many physical factors and by field conditions so
complex and uncertain that precise design of such channels at the present
stage of Imowledge is beyond the realm of theory.! The uniform-flow
formula, which is suitable fOr the design of stable nonerodible channels;
,provides an insufficient condition for the design of erodi.ble channels.
This is because the stability of erodible channels, which governs the
design, is dependent mainly on the properties of the material forming the
channel body, rat.her than only on the hydral..lllcs of the flow in the channel. Only after a stable section of the erodible channel is obt,ained can
the uniform-flow formula be llsed for ,computing the velocity of flow and
discharge.
Two methods of approach to the proper design of erodible channels
are described here: the method oj perrn1;sSible velocity and the method of
tractive f01ce. The method of permissible velocity has been used exteu- .
sively for the design of. earth canals in the United States to ensure freedom
from scour. The method of tractive force hOes sometimes been used in
Europe; it is now under comprehensive investigation by the U.S. Bureau
165
v,
fps
ilb/ft2
V,
fps
rD,
TO,
Ib/ft'
--'1---------
'1
iI
i
0.020 1. 50 O. (127
0,020 1. 75 . 0,037
0.020 2.00 0,048
0.020 /2.00 0,048
0.020 2.50 0.075
0.02012.50 0.075
0.025 I 3.75 0.26
0,025 3.75 0.26
0.025 6.00 0.67
2,50 0.075
2.50 0,075
3.00 0, n
3.50 0.15
3.50 0.15
3.50 0,15
5.00 0.413
5.00 0.46
6.00 10.137
i: ~~
4.00
4.00
5,00
0.43
5.50
0.30 .16,00
0,91
5.50
,.1
0.80
I 0.67
1.10
I
I!
The Fortier and Scobey values were recommended for use in 1926 by the Special
Committee on Irrigation Research of the American Society of Civil Engineers.
f
f
ties tha.n new ones, becau$e the old cha.nnel bed is usually better stabilized, particularly with the deposition of colloidal matter. When other
conditions are the same, a deeper channel will convey water at a higher
mean velocity without erosion than a shallower one. This is probably
because the scouring is caused primarily by the bottom velocities and
for the same mean velocity, the bottom velocities are greater in .the shal~
lower channel.
i'
i
f
J
IL
!~
,
I
UNIFORM FLOW
LG6
Attemp~1 were made early -to define a mean velocity that would cause
neither silting nor Scoill:ing. From the present-day viewpoint] however, it is dOHbtful whether such a velocity Mtually exists. In 1915,
167
J
_Ir
j
!
,g 0.7
:
~,
~
'"
.~
.2
'0
g 0.5f----j--,--+-..---/---';-',,\},c-:-+---f-1--i
.)
0.4
,."
:is
'"
"iii
j o~~,-=l==I=t:I=l=
0,.8
()'5~H-++--t
-{
O.4!-H-++-fflt-f--+
:3
o.:;f--f-+-++-+H-H--t
Q.~
!-,-,-.---l...--'---r",-,--,--W.-..c
Cloy
FIG. 7-4, Curves showing' U.S.S,u, da.ta on permissible velocities for cohesive soils.
FIG. 7-3. U.S. and U.S.S.R. data on ]lermissible veloCities for no,nco,hesive soila,
(7-4)
where V, is the nonsilting and noneroding mean velocity in fps; y is the depth of flow
in ft; C - 0.84, depending primarily or. the firmness of t,he material forming theqhannel body; and z "" 0.64, an exponent which varies only slightly. Ba.sed on later
studies by other engineers, the valu~s of C generally recommended are 0.,55 for
extremely fille soils such a.s tho,se found.in Egypt; 0.84 for fine light sand soils such
as those found in the Punjab, India; 0.92 for coarse light sandy soils; 1.01 for sandY
loamy silts; and 1.09 for coarse silt or hard-soil debris. For clear water, a "alue of
z = 0.,5 has been suggested.
For the design of canals carrying sediinent-laden water, the Kennedy formula is
now practically obsolete and is being replaced by methods based on Lacey's regime
theory [11-16], Einstein's bed-load function [17), a.nd Maddock-Leopold's principle
of channel geometry [18J. There are voluminous writings on these methods. Comprehensive bibliographies can be found in [19] to [25].
,,
l
I~
l~
.
I
.suitable n values for various materials and the converted values for the
corresponding permissible tractive force, which
will be discussed later (Art. 7-13). In 1036, a
.u.~l'''''.<1iH magazine [28] published values of m[l.ximum permissible velocities (Figs. 7-3 and 7-4)
ahove which scour would be produced in noncohesive materiai of a wide range of pa.rticle sizes and
various kinds of cohesiVe soil It also gave the
variation of these velocities with channel dep!;h ~
(Fig. 7-5).
The maximum permissible velocities mentioned
abo Ire are with reference to straight channels.
For sinuous channels, the velocities should be
lowered in order to 'reduce scour. Percentages of
reduction suggested by Lane [29] are 5% for
slightly sinuous canals, 13 % for moderately
sinuous canals, and 22 % for very sinuom; canals.
percentage val tieS, however, are veJ;'Y approximate, since no accurate data are available
Carreclion foctor
at the. present time.
FIG.
7-5.
Curves showillg
7-10. Method of Permissible Velocity. Using
U.S.S.R. corrections of
the maximum permissible velocity as a criterion] permissible velocity fo,r
the design procedure for a channel section, as- depth for both collesive
silmed to be trapezoidal, consists of the following and noncohesive materials.
steps:
r
i
-j
168
I,
l
It should be noted that the unit .tractiv~ force in channels, except for
wide open channels, is not uniformly distributed along the wetted perimet.er .. Many attempts have been made to determine the distribution of
the tractive force in a chann~l. Leighly [31] attempted to determine this
,distribution in many trapezoidal and several rectangular and tl'iangular
. channels from: the published data on the velocity distribution in the
1. Fot' the given kind of material forming the channel body, estimate
the roughness coefficir.nt n (Art. 5-7), side slope z (Table 1-1), and the
maximum permissible velocity V (Table 7-3 and Figs. 7-3 to 7-5).
'
2. Compute the hydraulic radius R by the Manning formula.
3. Compute the water area required by the given discharge and permissible velocity, or A = Q/V.
'
4. Compute the wetted perimeter, or P = A/R.
5. Using the expressions for A and P from Table 2-1, solve simultaneously for band y. The solution may be expedi~ed by
the charts
given in Appendix B.
'
6. Add a pl'0per freeboard, and modify the section for practicability.
i
1.5
'
Example 7-3. ,Compute the bottom width and the 'depth of fiow ,of !l. trapezoidal
cha.nnel laid on a slope of 0.0016 and carrying 8. design dischfl.rge of 4{)O cfs. 'rhe
channel is to be excavated in earth containing noncolioidal coarse i1;rllvels and pebbles.
Solution. For the given conditions, the following rm!estimated: n
0.025, z = 2,
, and nlll:llili1um pe;l'missible velocity = 4.5 fps.
Using the Manning formula, !wlve for R.
4.5
O:97QwyS
!1
1.49
-0.025 R~ ";0.0016
R ." 2.60 ft
or
ThenA = 400/4.5
A = (b
and
169
UNIFORM FLOW
+ zy)y =
(b
AIR
+ l!y)y
+ 2 v'l+ ii Y =
(b
Now
88.8 ft'
+ 2 y'S y)
)\4.2 it
18.7 ft and y
3.46,ft.
'I -11. The Tractive Force. When water 'flows in' a channel, a. force is
developed that acts in the direction of flow on the channal oea-:-"This
fm'cb, which is simply the pull of wilter on the wetted area, is known as the
tract~'ve force. l
In It uniform flow the tra()tive forceis apparently equal to
the effective component of the gravity force acting on the body afwater,
parallel to the channel bottom and equal to wALS, where w is the unit
weight of wa.ter, A is the wetted area,' L is the length of the channel
reach, and S is the slope (Art; 5-4). Thus, the a.verage value of the
tractive fOl'ce pel' unit wetted area, or the so-called tmit tractive f~rce
TO, is equal to wALS/PL = wRS, where P is the wetted perimeter and
R is the hydraulic radius; that is
.
. [TV
w~
Ii,
l- fl--+--!--+---l- -
123'156789
-I
rmm]
9 10
!4-+-1-+--+=1t++i~
:-.-JI:~
o
o
bh
bh
On sid s S 01 channels
On bcttom of cl1annels
channels. U nfoitunately,.
to deficiency of data, the results of his
study were not very conclusive. In the CS. Bureau of Reclamation,
. Olsen and Flol'ey[32] and other engineers ha,ve used the membrane
analogy and analytical and finite-difference methods for determining the
distribution of tractive fOl:ce in trapezoidal, rectangular, and tria!1g ular
a trapezoidal chanchannels. A typical distribution .of tractive force
Itel reSUlting from the m~mbrane-analogy study is shown in Fig. 7-6.
The pattern of distribution varies with the shape pf the section but is
q-5)
In a wide open channel, the hydraulic radius is equal to the depth of flow
y; hence TO '" wyS.
1 'l"his is also known as the shear force or the drab fOlce.
The idea of tractive force
is gellorally believed 1.0 h!we been first :introdue~d intohydraulia litera.ture by du
Boys in 1879 [po 149; of 30]. Howelrer, the principle of balancing this foroe wit~ the
in
channel resistance in 11 uniform flow was sta.ted 1;Iy Brahms earIy in 1754 (see Art. 5-4).
I
.Ii".
170
UNIFORM FLdw
171
,c!
,.
tan
(7-9)
Simplifying, )
PIG. 7-8 .. Analysis of forces ac.ting on a pa.rticle resting on the surface of a chn.nnel bed.
which tends to cause the particle to roll down the side slope. l The
symbols used are a = effective area of the particle, T, = u\lit tractive
force on the side of the channel, W, = submerged weight of the particle,
and", = angle of the side: slope. The resultant of these two fO;'ces, which
are at right angles to each other, is
VlV,2 Si112
'"
+ oh,2
tan 2
= COli '"
J( =
~ 1 - tan 2 0
/1 _
'\;
Si.ll '"
Sill" ~
(7-10)
(7-11)
It can be seen that this ratio is a fUllctioll only ~f the inclination of the
sloping side", and of the angle of repose of the material O. Fot cohesive
and fine noncohesive materials, the cohesi vc fOI'ces,even with comparatively clear .vater, become so great in propprtion to the gravityforce component causing t.he particle to roll down that the gravity force
wn safely be neglected. Therefore, the angle of repose need be considered only for coarse noncohesive materials. According to the U.S.
Bureau of Reclamation's investigation, it was found in general t.hat the'
angle of repose increases with both size and anguiarity df the material.
For use in design, curves (Fig. 7-9) were prepnred by the Bureau, showing
values of the angle of repose for nOli cohesive material above 0.2 in. in
. diameter for various degrees of roughness. The diameter referred to
is the diameter of a particle than which 25% (by weight) of the material
is larger.
1 Equation (7-10) was presented by the U.S. Bureau of Reclamation [35,361 and
Eq. (7-11) by Fan [341. The two equations are mathematically identical.
172
UNIFORM FLOW
0.1
. 0.2
L-----.rL--;
42
~-
, I
3/16 1/4
lIB 112
~j
.....
-T--[I
I I I I
3/4
""
,6
-1--
:3 4
..,'"
I
!,
'"c:
~<~I
;;"'/ /
.If':
~/
~
b'"
Il
I.;;:,~
'0
.!?
t1
$O~
'2 6
f--- ...
24
2. 2
'ti
V
if
J~Ci
V."O- V
~ ~0 .~~
L ,'") r
---V~.'
//
///
V / / V/
V//V
L/
~
.'"...."
3.0 4.0
173
VI/';I/I
Ii
t7-
~o
.;
'"c.e
''"
~~"'/
~'"
."
//
~~i/I
o
..,
V
, V
2.0
I
H/2
I...
40
-;:;
,i .
;VL
.---;----
1
"
.I
"-:;
i---
'b
canals in coarse l
noncohe~ive material~
V:o."':;"
20
I)
7-13.
Permissible
Tractive Force. The pe?'m-issible tract1've'01"ce
is the
J
maXImum Ulllt trnctive force that will not cauge seriOlls erosion of the
material forming the channel bed 011 a level surface: This unit tractive
fol'c~ c"n. be determined by laboratory experiments, and the value thus
obtamed IS known as the critical t1'uctiz,c JOT'ce. However, experience has
shown that actual canals L.'1 coarse nUl1Cohasive material C[l.n st!].nd substantially higher va.lues than the critical tractive forces measured in the
laboratory. This is prohab!y because the water {md soil in l.ctUtll canals
(J?nt~in slight amounts of conoidl1.1 and organic matter whiGh provide a
bmdlllg power and also because slight movement of soil particle" can be
t.?lerated in P:a.~tical designs. without endangering channel stability.
B.mce the permISSIble tractive force is the design criterion fOl' field. condi- '
tlOl1S, the permissible value may be taken less than the critical value.
Tb,e determination of permissible. tractive force is now based upon
50
100
FIG. 7-10. Recommended permissible unit tr.active forces for canals in noncohesive
(U.S. Burea:!' of Rec!amalit.m.)
Ino.terial.
-\
1 The pla~ticity index is 'the difference in .percent of moisture between plas~ic limit
and liq:uid llmit in Atterberg soil tests. This index has been investigo,ted by the U.8.
Bureau of Reclamaiioll as a soil characteristic that can be lL~ed to indics.te resistance
to SCour for cohesive materials. For cs.nal deSign. a plasticity illdex of 7 may be.
taken tentatively as the criticnl vll-Iue, with SCOill' OCCUrring for moder:.te 'tractive
forces below this value. However, acours a.re still observed in many cases where the
index is above 7. Research'shows that determinat.ion of the plasticity index in (:00"
junction with consolidated-shear testa may possibly be necessary.
174
UNIFORM FLOW
\'0
.,
E==E=E~j=R+q:B=t==+=a
r\
,.....
Llilil! I I
175
7 -14. Method of Tractive Force. The first step in the design of erodible channels by the method of tractive force consists in selecting a11
approximate channel section by experiene.e or from design tables,l collecting samples of the material fo;:ming the channel bed, and determining
the'required properties of the samples .. With these data, the designer
investigates the section by applying tractive-force analysis to ascertain
probable stability by reaches and to determine the minimum section that
appears stable. For channels in noncohesive materials the rolling-down
effect shouid be considered in addition to the effect. of the distribution of
tracth-e'forces; for channels in cohesive material the rolling-down effect is
negligible, and the effect of t.he distribution of tractive force alon:e is a
criterion sufficient for design. The final proportioning of the channel
section, ho\-vever, will depend on other nonhydl;aulic practical considera!;iops. The analysis for tractive force is best described by the following
example:
Design a trapezoidal channel laid on a slope of 0.0016 a.nd carrying
The cha.nnel is to be excavated in earth containing noncollnidal coarse gravels and pebb['~s, 25 % of which is 1.25 in. or over in diameter, Manning's n. = 0.025.
.
SoLution. For trapezoidal channels, .the maximum unit tracth'e force OIl the slopillg sides is usually less th!>ll that Oil the (,ottom (Fig. 7-7); hence, the side force i~ the
controUing value .in the analysis. The design of the cha.nnel should therefore include
(It) the proportioning of the section dirnensioIlll for the maximum unit tractive force
on the sides. and (b) c.heckiilg the proportioned dimensions for the maximum unit
tJ:active force on the bottom, .
a. PTopol'lioning the. Section D!:men,~ions. Assuming side slopes of 2: I, or z = 2, 3,nd
a base-depth ratio bly ",. 5, the maximum unit tractive force on the sloping sides
(Fig. 7-7) is 0.775wi;s = 0.775 X 62.4 X 0.0016Y = 0.078y lb/ft'.
Considering a very rounded ma.terial 1.25 in. in diameter, the angle of repose (Fig.
7-9) is 8 = 33.5". 'Nith fJ = 33,5' and z ~ 2, or cJ> = 26.5", t,lle tractive-fol'ce ratio
by Eq. (7-11) is K = 0.587. For a size of 1.25 in., the permissible tractive force on
a level bottom is TL = 0.1 X 1.25 = 0.5 lb/ft' (sam.e from Fig. 7-10), and the permissible tractive force on the sides iST. ~- 0.587 X 0.5 = 0.294 lb 1ft'.
For a state of impending motion of the particles. on side slopes, 0.078y = 0.294, or
y = 3.77 ft. Accordingly; the bottom width is b = 3.77 X 5 = 18.85 ft. For this
trapezoidal section, A = 99.5 ft' and R = 2.79 ft. With n = 0.025 aud S = 0.00113,
the discharge by the Manning formula is 470 crs. Further computation will show
that, for z = 2 and bIy = 4.1, th" ~ection dimensions are y = 3.82 ft and b "" 15.66 ft
and that the discharge is 41"4 cfs, which is close to thO} design discharge ..
Alternative section dimensions may be obtained by as~uming other '(alues oI z or
side slopes.
.
b. Checking the Proportjalled Dimensions. With z = 2 and bly = 4.1, the maximum
unit tra,ctive force on the channel bottom (Fig. 7-7) is 0.97wyS = 0.97 X 62.4 X
3.82 X 0.0016 = 0.370 [b/ft', less than 0.5 Ib/ft', which is the pel'mu;sible tractive
. force on the level boLtom.
Example 7-4.
Voids ratio
FIG. 7-11. Permissible unit tl'l1ctive forces for canals in cohesive material as converted
from the U.S.S.It. data on permissible velocities.
.
are tentatively recommended (1) for canals with high content of fine
sediment in the water, (2) for canals with 10',11 content affine sediment ill
the water, <lnd (3) for canals with clear water.
For cohesive materials, the data based on conveision. of permissible
velocities to UIJit tractive forces and given in Table 7-3 and Fig. 7-11 are
recommended as design references.
'
The pel'missible tTactive forces mentioned above refer to straight channels. For sinuous channels, the values should be lowered in order to
reduce. scour. Approximate percentages of reduction, suggested by
Lane [29J, are 10% for slightly sinuous canals; 25% for moderately sInuous canals, and 40% for very sinuous .canals.
176
UNIFORM: FLOW
transfer (if tractive force has little effect on the results and can safely be
7-15. The Stable Hydraulic Section. The section of an erodible channel in which no erosion will occur, at a minimum water area for a given
dischilrge is called the stable hydraulic 8ection. Empirical pi'oJiles, such
as the ellipse and the parabola, have been suggested as stable hydrauUc
sections by many hydraulicians.. The U.S. Bureau of Reclama.tion [38]
has employed the principle of tractive force t o
a theoretically
stable section for erodible channels carrying clear water in noncohGS'ive
.materials.
.
In designing trapezoidal sections, as described in.the preceding article,
the tractive force is made eqnal to the permissible value over only a part
pf the perimeter of the section, where the forces are close to maximum;
on most of the perimeter forces are less than the permissible value. rn
other words, the impending instability occurs only over a small pr.rt of
the perimeter. In developing a stable hydraulic section for maximum
efficiency, it is necessary to satisfy the condition that impending motion
shall prevail everywhere on the channel bed. For material with a given
angle of repOSE> and for r~ given discharge, this optimal section will provide
noi only the chl1Ilnel of minimum water area, but also the channel of
minimum top width, maximum mean velocity, and minimum excavation.
In the maihematicalderivation of this section by the Bureau, the follow'ing assumptions are made:
..
.
L The soil particle is held against thechanllel bed by the component of
r,he submerged weight of the particle acting normal to the bed.
2. At ,and above the water, surface the side slope is. at the
of
repose of the material under the action of gravity.
3. At the center of the channel the side slope is zero and the,
force alone is sufficient to hold the .particles at the point of incipient
instability.
4. At points between the center and edge of the channel the particles
are kept in a state ,of incipient motion by the resultant of the gravity
component of the particle's submerged weight acting on the side slope and
the tractive force of the flowing water.
5. Thc tractive force act.ing on an area of the channel bed is equal to
the weight component of the water d,irectly above the area acting in the
direction of fiow. This weight component is equal to the weight times
the longitudinal slope of the channel.
If assumption 15 is to hold there. can be no lateral transfer of tra.ctive
force between adjacent currents moving at different velocities in the section-a situation, however, that never actually occurs. Fortunately,
the mathematical analysis made by the Bureau I has shoWn that the actual
I
ignored.
.
. '.
It y
. According to assumption 5, the tractlve force actmg on any e em en ar.
n.rea AB OIl the sloping side (Fig. 7-12a) per unit length o~ the channell~
equal to wyS dx, where w is the unit weight of water, Y is the depth 0
'water above AB, and S is. the
longitudinal slope. , . Since the area
AB is VCdX)2 + (dy)2, the unit
tmctive force is equal to
tal
+ (lL) tan
(dy)t
dx
Yo
Takiflg the effect of lateral tnl.litive force into accoUnt, an a.!ternative assumptioll
was ina.de by the' BUreau, which states that the tractive force acting on a particle is
prop~rtiona.1 to the square of the ll1ean velocity in the channel at 'the point where the
177
I
i
(I
ya' and
:t
tan~
fJ
= O.
pa.rticle is loea~d. This assumption gives ~ solution .that agrees very clqselY w~th
the solution bailed on assumption 5. Therefo.re, neglect of the trB?"fer 0,1 .tractwEl
force in the a.~!l.lysis will give equally sn.tisfactory results, Bad wll:.h cOJ;lslderably
less ,,'ork.
.
.. ~
.178
Yo
o.97wS
l'
A
(7-14)
11
(7-15)
= 0.99
(1 - ~ T
Example 7-6. Determine the profile of the stable hydro.ulic section to replo.ce the
tra.pezoidal section of the channel described in Example 7-1.
SolUlian. For the given conditions, TO = 0.5 lb/ft l , S = 0.0016,. e.= 33.5, a'J.d
n = 0.025. By Eq. (7-13), the center depth is Yo = 0.5/(0.97 X 62.'1 X 0.0016)
5.16 ft. From Eq. (7-12), the shape of the theoretical section is
which
is
II
l
I
I
I
(7-16)
On the other hand, if the channel is to can'y more than the theoretical
section will carry, it is necessary to add a rectangular section at the center.
Suppose the. discharge to be carried is Q", which is greater than Q, and
the top width of the added rectangular area is Til. The value of T))
may be computed by
/I _
n(QI! - Q)
(7-17)
T - 1.49Yo~~SI~
(7 -13)
T'
179
UNIFORM .FLOW
l
[
I!
it by 180/11" or 57.3. The top width mllY be computed by Eq. (7-12) with if = 0, or
cos 0.128::; = O. Thus, 0.128::; = 11"/2 and x = 12.3
Tli.e top width is T = 2::; =
24.6 ft.
By Eq. (7-14), the mean velocity is V= (1.35 - 1.19 tau 33.5') 5.151' X 0.0016}'/
0.025 = 2.69 fps. By Eq. (7-15), the water Mea is A = 2.04 X 5.16'/tan 33.5 =
82.2 fV. Hence the discharge is 220 cfs. Since the design discharge is 400 cfs, it is
necessary to add a rectangular area at the middle (Fig. 7-12b). The width of the
rectangle may be computed by Eq. (7-17) as
n.
. ,,_
0.025{400 - 220)
_
T - 1.49 X5.l6~~ X 0.00W1 - 4.9 ft
TherefCJre the top width is 24.6 + 4.9 = 29.5 ft.
If the channel is 'designed to carry 100 cis, it isnecesso.ry to remove !l. vertical area
ir.:im the middl~ (Fig. 7-12c). The top width of the removed a.rea may be computed
by Eq. (7-1il).
T' = 0.96(1 - V 1 0,%20)X 21.6 = 7.7 ft
Therefore the top width is 24.6 - 7,7
lB.\) ft.
C. GRASSED CHANNELS
be
-1
. . ---'i
180
UNIFORM FLOW
DESIGN OF CHANNELS FOR UNIFORM FLOW
lSI
. Cover
Condit.ion
A Very high
B. High
Reta.rdo.nce
Moderr.te
.l
I.ow
Very low
;.
only the average ctll've is shown, together with the curves for low retardance. The classification of degree' of l'etardance is based on the kind of
vegetation and the condition of growth, as described in Table 7-4. The
term" stand" used'in the tabie refers to the density of grass, or the count
of vegetation, which is sometimes expressed as the 'number of stems per
square foot. The n- VR curves thus developed may also be applied to
other kinds of gmss, provided that their characteristics and degree of
I'etardance can be identlfied. For this purpose, Table 7-5 is provided as a
guide in the selection of the vegetal retu.rdance for diffetent conditions of
stand and average length of the grass.
.
U.~.
Soil Conservation
[41J.
182
UNIFORill FLOW
183
.sl---l--t--........f.--+-++-+-!-l-f---+
~1---4--4--.--+---+--~-~+++----r--+--~~~-+-+~~+---+--+---~
r::
.1 :
.09
.OS
.O~
.00
.O~
...
-.I
VR
VR
.\,
VR
. (b) Curves for B or high vegetal reta.rdllnce.
FIG.
(d) Curves for D or low vegetal reta.rda.nce, a.nd a.n .e.vernge curve for E or very low
vegetal reta.rdll.nce.
'f
I
i
.1
~I
184
UNIFORM FLOW
TABLE
7-5.
Good
Foil' .
Degree of retarciance
>30
11-;-24
8-10
2-6
<2
A . Very high
B High
C Moderete
D Low
E Very low
>30
11-24
6-10
2-8
<2
B
C
D
D
E
High
Moderate
Low
Low
Very low
, I
DESIGN OF
7-19. Sel~ction of Gra~s. The selection of grass for the channel lining
depe.nds mawly on the clImate and soil in which the plant will grow and
surv.I~~ under ,the given conditions. From the hydraulic viewpoint,
stablhu y . and other f~ctors should also. be considered. In general, a
higher dlsch:.l.l'ge requu'cs a stronger or better lining. On steep slopes
?unch grasses; such as alfalfa, lespedeza,and kudzu, will develop channel~
mg of the flow and" hence, are unsatisfactory for lining. For slopes
greater than 5 %, only fine and uniformly distributed sod-forming grasses
sueh as Bermuda grass,. Kentucky bluegrass, and smooth brome ar~
re~om~nended fOl' lining where the main flow occurs. Because' of the
obJec.tlOnable spreading nature of sod-forming grasses, the top portion of
the. Sides and the berm may be planted with gras.:;es that do Ii.ot.spread
easliy, such as weeping love gl'B,SS. 11'01' fust establishment of the lining,
Bermudagl'assand weeping love, grass fLl'e recommended. Sometirne~
an~uals are used as temporary protectibn until permanent coveb by
native gras~e~ are ~stablished. Silt deposition in channels may be con~rolled by Iml?g With bunch grasses, which will develop channeled flow
lnCl'ease velOCity, and thus reduce silting.:
,
'
. 7-20. Proced?re of ~esign., After the !kind of grass for channel ]ining
IS selected, the qegree of retardrmce can be determined from the condition
of the stem length and the density of growth. During the period of
I
!
CHANN~LS
~85
establishment, tht; grass will grow and the channel will be stabilized under
a condition of low degree of retardance. The channel will not reach its
maximilm capacity until' the. grass 'cover is fully q,eveloped and well
established. Therefore, it is suggested that the hydrll:ulic design of a
gras:;ed channel ,consist of two stages. The first stage (A) is to design
the channel for stability, that is, to determine the channel dimensions
under the condition, of a lower degree of retardance. The second stage
, TABLE
7-0.
Slope IMlge,
Cover
----
Bermuda grass
0-5
5-10
>10
GrllSs mixture
I.
I
I
Easily eroded'
soils
6
5
4
7
8
0-5
5-10
>10
7
5
5
4
3
(}-5
5
4
5-10
II Do not
use
all
0-5
,3.5
2.5.
Lesp,edeza sericea, -w-e-ep-i-~~g-lo-v-e
grass, ischaemuffi (yello.v blue- I Do not use on slopes steeper than 5 %, except lOr
stem), kudzu, aHalf!\., crabgrass
side slopes in a combination channel
------
REMARKS.
The values apply to average, uniform stands of each type of covel'.
lIse velocities exceeding 5 fps onlr where good covers and proper maintenll.nce cll,n be
obtairied.'
,
. * If.s. Soil Conse.rv!l.tion Service 1411.
(B) is to review the design for maximum capacity, that is', to determine
the increase in depth of flow necessary to maintain a maximum capacity
under the conditiun of' a higher degree of retardance. For instance, if.
common lespedeza is selected as the grass for lining, the comn-lOn lespedeza
of low vegetal ret!l.rdar\.ce (greell~ average length 4.5 in.) is used for the
first stage in design. Then, in the second stag~, the cummon lespedeza
of moderate vegetal retal'dance' (green, uncut,: average length 11 in.)
shou1.d be used. Finally, a proper freeboard is added to the computed'
186
UNIFORM
depth of the channel, which will further increase the maximum design
capacity. The design procedure I is described as follows:
A. Design Jor Stabiltty. Given the discharge, channel slope, and kind
of grass, the Ilrst stage of design may proceed in the following stiOlPs:
1. Assume a value. of n, and determine the corresponding values of VR
from the n- VR curve (Fig. 7-14).
2. Seler:t the permissible velocity from Table .7-6, and c()mpute the
value of R.
3. Using the Manning formula, compute the value of
L49R%SH
Trial no.
1.49R~5SH
Side slope
z:1
Bottom width
b, ft
Depth
17.0
12.5
0
0.53
0.62
1.00
Top width
T,ft
!I, it
....
3;1
6:1
Triangle ...........
Para.bola ..........
10: 1.
187
---~-
20.18
19 94
20.00
2().00
o.n
11.
'Trial
1
' L35
0.34
4.72
2,50
The correct va-lues for the determination of sections o.re.R = 0.50 It and A = 3% =
10 ft'. Several cha.nnel sections meeting these requirements are proposed as follows
(using charts in App~ridb:: D ) : '
,
I
noQ.
!I'
l'R
1.491mSH
11.
2
3
0.70
0.60
0.65
13,4
11.3
l2.3.
0.G3
0.54
0.58
3.73
4,42
4.07
2.35
2.39
2'.36
0.051
0.050
0.051
4.2.2
3,96
4.07
1
1
The correct depth is 0.55 ft. Adding a freeboa.rd of 0.2 tt,.the total depth is 0.85.ft.
. For the trapezoidal ::;ection with a side slope of 6: 1 and b 12.5 ft, t~12 computat~on
results in a total depth of 0.94 ft. Simila.rly, the tot!).l depth of the trlangula.r sectIOn
with a side slope of 10: 1 is found to be 1.33 ft,.
.1
..;....
-!~
"
188
For the parn.bolic :3ection of y
as follows:
Trial
no.
---
0.80
0.90
0.85
2
3
1
UNIFORM FLOW
0.85
I
I
21.0
22.2
21.6
21. 7
11.2
13.3
12.2
12.4
0.52
0.60
0.57
0.58
4.406
3.76
4.10
4.03
I- - -I- - VR
2.32
2.25
2.34
2.34
0.051
0.051
0.051
0.051
1.49m1S~~
3.78
4.15
4,03
4.07
It should be noted that the computation for the parabolic section is simplifi<od by
the lI,se of t}leequation for R in Table 2-1 and by the foct that the depth is proportional
to the squ,ue of the top width; that' is, T = 20 -Vy/O.73. Allowing!l freeboard of
0.20 ft, the total depth is 1.06 ft n.nd the top width is 24.1 ft.
.The final choice of the cllannel section and its dimensions ...... ill depe-nd on practicability
nnd on the circllmstances under which the problem is proposed..
' .
PROBLEMS
, i
7 -1. Show t~at the most efficient rectangular or triangular section is one-half of a
square.
. 7-2. Explain (a) that any section formed by a p'olygon which can be inscribed by
!\ semicircle with the ccnter in the water surface will have its hydro,ulic radius equal
to one-half the radins of th~ inscribed circle, and (b) that such section will have the
best hydl'll.ulic effiuiency.
.
7 -3: Detenuine the best hydraulic section of the channel in Example 7-2 if the section is (n) rectangular, (b) triangular, (e) cirl!ular, (d) parabolic, and (e) in the form of
It hydraulic catenary.
. .
7-4. Solve Example 7-2 by the empirical rule of Eqs. (7-2) and (7-3).
7 -6. Design a nonerodible channel carrying 200 cfs with n = 0.020 and S = 0,0020.
Use your own judgment and assumptions.
7-6. Based on the p.ractice of t1u~ U.S. Bureau or'Reclamation, det.ermine (0) the
freeboard of the channel designed in Example 'r-2 when the channel is unlined and
(b) the heights of the lining and benk if the channel is lined.
'
7-7. Solve Example 7-3 if the material forming the channel body is fine silt having
an average particle size of 0.006 mm. Estimate the permissible velo,~ity with the aid
of (a) Fortier and Scobey's table, (b) U.S.S.R. data, and (e) the Kennedy formul~.
modified for clear ,vater.
.
.
7-8. SOlVE Example 7-3 if the mn.terial forming the channel body is fairly compact
heavy clayey soil with a voids ratio of l.0.
7 -9. Solve Example 7-3 if the channel has !l. pa.rabolic section.
7 -10. Design the section of a canal to carry a discharge of 200 cfs throllgh a land of
erodible soils with n = 0.020 and S = 0.0020. Assume other necessary data and use
your o~vn jlldgment.
.
. 7-lt. 111e All-American Canal is desigr:ed to divert 15,155 cfs of desilted water from
the Colorado River to irrigate the Imperial Valley in southern Califpmia. This canal
is 80 millls long. . The typical maximum section has a bottom width of 160 ft, width
at wo.tr surface of 232 it, water depth of 20.6 ft, minimum freeboard of 6 ft and bank
width of 27 to 30 ft. The terminal capacity is 2,600 cfs. The canal was 'excavated
189
mostly in n.Uuvi!ll soil, ra~ging from light sand~ or silty [oams to adobe and having
an average particle size of 0.002b in. Review the hydraulic design of the channel
section.
7-12. Review the stability of the section dimensions obtaincd in Example 7-3 by
the method of trn.ctive force, !l.Ssuming that 25 % of the material forming the channel
bed is L2b in. or over in diameter.
.
7-13. Solve Example 7-4 for the following conditions, respectively:
(a) If the side slopes are. assumed as 1 on 1.5. .
(b) If the mateJin.1 forming the channel bed contains fine noncohesive particles,
50% of which are larger than 1 mmin diameter. The water is clear.
(e) If the m~.terial forming the channel bed is cohesive compact clay, having !L voids
ratio eqllal to 0.5 ..
(d) Ii the channel is moderately sinuous.
7-14. The conversion from the maximum permissible velocity to permissible tractive force (Table 7-3) is based on 0. flow depth 0' 3 ft _and an avernge channe.l section
ha.ving a bottom width of 10 it and side slopes of L5: 1. For a.llllvial noncolloidal silts
and. clear-water floW, the m!lximum permissible velocity recommended by Fortier and
Scobey is 2.00 fps r.nd the n value is taken as 0.020. Compute the corresponding
permissible tractive foree.
7-15. Compute the maximum tractive force per unit area on the section of the AllAmerican Canal described in Prob. 7-11.
.
7-16. Determine the cross ser.I,ion and discharge of the st~ble hydraulic section of 0
.channel excavated in a' noncohesive IDr.terial ho.ving To = 0.1 lblft', S = 0.00040,
o "" .31", and n = 0.020.
.
7-17. Determine the modffied profile for the channel section obtained in the preceding problem if the channel is tOCllUY (a) 75 cfs ILnd (b) 300 efg.
7-18. Design a waterway lined with Bermudn. grass on- erosion-resistant soil and
cn.rryillg a discharge of 200 cfs .. The n.vernge slope of the channel is 3 %. Use the
curve for modemte vegetn.l rel,ardance.
7 -19. Determine the total depth for maximnm capacity of the channel section propo,,,ed for the preceding problem. Allow 11 fre,eboard equal to 20% of the computed
depth.
.
REFERENCES
1. Victor L. Streetel': Economical conal cross section~, Transactions, American
Society of Civil Engineers, vol.. 110, pp. 421-430, 1045.
2. I van E. Houle" Irrigation Engineering," vol. 2, "Projects, Condllits, und Structures," John Wiley & Sons, Ill~., New York, 1956.
3. ",Linings for lrrigrttion Canals," U.S. Bureau of Reclamn.tion, Jl1ly, Hl52.
4. Cll-nals !lnd related structures, U.S. Bureau of Reclamation, Design and Constnlclion Manila!, Design Supplement No.3, Apr. 17, 1952; vol. X, 'pt, 2,. ghop. I,
paragraphs 1.15, 1.8, and 1.18. .
.
5. B. A. Etcheverry: "Irrigation Practice and Engineering." vol. II, "Conveyance
of Water," iVIcGmw-HiIl Book Company,Jnc., New York, 1st ed., 1915, p. 122.
6. G\lilford L. Molesworth: "Pocket-book of Engineering Formulae (Useful Formulae
and Memoranda) for Civiland Mechanical Engineers," E. & F. N.Spon, London,
7th ed., 1871, p. 175.
'
7. Isidro D. Carino: A graphical solution for flow in earth channels, paper 1360,
Proceedings, American Society of Civil Engineers, J oumal, Irrigation and Dra.ina.ge
Division, voL 83, no. IR2, pp. 1-9, September, 1957.
.
190
UNIFORM FLOW
8. 'Pete W. Terrell [LIld Whitney M. Borland: Design of stable canals and channels in
erodible ma.terial, Transactions, AnLerican Society of Civa Engineers, vo!' 123, pp.
lOl-U5, 1Di8.
.
9. E. W. Lane: Stable channals in erodible material, Tmnsactirms, American Society
of Cilia Engineel's, vol. lO2, pp. 123-142, 1937; and discussions.
10. Robert G. Kennedy: The prevention oI silting in irrigation canals, Proceedings,.
Inslit:l.tiolL c,f Civil Engineers, London, vol. 119, pp. 281-290, 1895.
11. Edward S. Lindley: Regime channels, Minutes of Proce,edngs, Pun}ao Engtneering
Congress, Lah01'e, India, vol. 7, pp. 63-74, 1919.
12. Gerald Lacey: Stable channels in alluvium, Proceedings, Institution of Civil
Engineers, London, vol. 229; pp. 2.'i9-384, HJ30.'
13. Gerald L"c"y: Regime flow in incoherent alluvium, Cenl,.al Board of h1'1:galion,
.
PlLblication No. 20, Simla, Iudia, 1940.
14. Gerald Lacey: A general theory of flow in alluvium, Journal, Instit'ution of Civil
Engineers, Londan, vol. 27, pp. 16-47, 1946.
'.
15. Thoinas Blench: "Bydrs-ulies of 'Sediment-bearing Cano.ls and Eivers," Evan~
'Industrie$, Ltd., Vancl)uver, B.C., Canada, 1951.
16. Thomas Blench: Regime theory for oclf-formeci sediment bearing'channels, Tmnsactions, American Bociety oj Ciliil En.gineers, vol. 117, pp. 38.1-400, Hi52.
17. H. A. Einstein: The bed-load fUllction for sediment tran~Jlorta~ion in open channet
flows, U.S. Depa"bILent of Agriculture, Technical BuUelin No. 1025, 1950.
18. L. B. Leopold and Thomas Maddock, Jr.: The hy.draulic g~ometry of stream
cha.nnels and some physiographic implictttions, U.B:G'eological BUTVP,Y, Professional
Paper 252, 1953.
19. ,K W. Lane: Sta.ble channels in erodilJle material, Tra1tSdctio,is, American Society
of Civil Engineers, vol. 102,pp. 123~142, 1937.
.'
20. Serge Leliavsky: "An Introductiqn to F1uvial Hydraulics," Constable & Co.,
Ltd., London, 1955.'
21. Instit\ltion Research Committee: Recent developments hydraulics, Proceedings,
In.stilution of Civil Enginp,cr.s, London, pt. III, vol. 4, pp. 990-1049, December,
1955.
22. Ning Chien :'The present status of research on sediment transport., Tl'anoactions,
A lILerican Society of Civil Engineers, va!. 121, pp. 833-868, 1956.
23. Ning Chien: Graphic design of olluvial channels, Transactions, Amer':can Society
of Cilia Enuineers, vol. 121, pp. 1267-1280, 1956. '
24~ Ning Chien: A concept of the regime theory, Transactions, American Society of
Civil Engineers, vol. 122, pp. 785-793, 1957.
25. Eno~ J. Carlson and Carl R. Miller; Research needs'in sediment hydraulics, paper
953, Proccedngs, American Society of Civil Engineers, J oumal, Hydraulics Dvisio1];,
vo!' 82, no. 11 Y2, pp. 1-33, April, 1956.
26. D. A. Etcheverry: "Irrigation Practice and Engineering," va!. II, McGraw-Hill
Book Company, Inc., New York, 1915.
27. S. Fortier and Jf. C. Scobey: Permissible cana.l ;relocities, Tran8~tions, AlILerican
Society of Cillil Engineers, vol. 89, pp. 940-956, 1926. ,
28. The maximum' permissible ,mean 'velocity in open channels, Gidrotekhnicheskoie
Stroitel'stvo, (Hydrotechnical Constmction),I\{oscow, no. 5, pp.5-f, May, 1936.
29. Emory W. Lane: Design of stable channels, Transactions, Am.erican Society of
Civil' Engineers, vol. '120, pp. 1234-1260, 1955.
.
30. P. du Boys: Etudes du regime du B,h6ne et l'action exercee par les eaux su~ un lit
~ fond de graviers indetiniment affouillable '(The Rhone and streams with movable
bedsj, Annales rles ponts et Cha1L8S~eS, ser. 5, vo!. 18, pp. 141-195, 1879.
in
191
36. A. C. ~~rter: Critica~ tractive fones on channel side ~lo]Jes, U.S. Bgrea.l! of
Rcclmrt(lMTc, Hydrav.hc LaboratoJ'Y Report No. Hyd-3G6 (supersedes Hyd-295)
Feb. 18, 1953.
'. .
.
,
37. E. W. Lane ~tl1d E. J. CarIllOn: Some factors affecting the stn.bility of canals coustructed in coarse granular materials, Pr'oceedings of the Minnesota [nlernalional
IIydmulics Convention, Sept. 1-4, 1953, Joint Meeling of Inte1'1wtional As.sociation
for H ydtl}.ulic Research and HydmldiC5 Divi.sion, American Bociety of Civil Eng,;rwers,
pp. 37-48, August, 1953.
.
38. R. E. Gloyer and Q. L. Florey: Sta.ble channel profiles, U.S. But'eau of Re,clama,ti01l, HydTflHl'ic Laboratory Report No. Hyd-325, Sept. 27, 1951. The work was
started by R. G. Conard and revie.wed by E. \Y, Lane.
39. W ..0. Ree: Hy~raul.ic characteristics of veget.ation for vegetated waterways,
Agnculttlral .Enqmeenng, vol. 30, 110. 4, pp. 184-187 and 189, April, 1949.
'
40. ~'. O. Ree and V. J. Palmer: Flow of water ill channels protected by vegetat.ive
lmmg, U.S. Boil Co,~seTVation. Service, Technical Bulletin No. 9lH February 1949.
41. Stillwater Outaool' Hydraulic Laboratory: Handbook of chandel desig;ri' for soil
an~ water conservation, U.S. Soil Conservation Service, SCS-TP-Gl, Marcll., 1947;
reVised, ,June, 1954.
.
.
42. V. D. Fredcnhagen and E. H. Doll: Gmssed waterways, Agricullural Enuineming,
vo!. 35, no. 6, pp. 417-419, June, 1954.
.
.,\
.j
I
1
-\
1,
(
. CHAPTER
I
lI
193
bution; (3) the depth of Row is indefmitely 1111'ge,so the depth of ftow cun
be considered constant as the water enters the channeL 1n th!'i,~ha!!neJl
the effect on the velocity di!J.trib!Jti~t9. boun.1.::IY-XQ.hlghnes~ is
indicatedbythe- ii~~" IlHf,- Outside the surface represented by ABC,
--7'- th~ ;~I~city7 d'i;tribution is practically uniform. Near the channel surface and within the region ABC, v~locjty"'yl.ries l\.ccording_tQ_ 4isww.e
iromthe channel surface. . The region inside ABC, though not distinc'ti;;;,i;k~~~~ boundary layer 1 and its thickness is deSignated by o.
Since the bouhdary layer is not distinctive, jiB. thi<lkn.es.JLh~~_I?een de.i
....fu!ed arbitrn.rily in."yarioufi..~ays .
.~~_......:;.-> A common definition is that the
lrei&-C,;11 _ '
._~
thickne;:;s 6 is the magnitude of the
':((5 w/,bu l( ~n
normal distance from the boundary
;ll~ft1,ceat ,~hi()h th;Vclocity ;; is
'
equitnQ99%Of theli:rrd~~:
-n:y Vo, ''I'hTcli.the velocity:'distribu1ion-:-b~ur:ve1n the boundary Tayer
approaches
asy!llptoticall~(Fig.
. g:<~---~-~
....--_~
The e,1lranCE
condition is
hypolhelicol
r-
FlO. ~L Development of the boundary layer in an open chl1nnel with an idea.l entrance
condltlOn.
.
.
roughness, will vary with the distance over wllich the water t!'avels in
the;. channel. (Fig. 8-1).1 If the flow is uniform and stable and if' the
channel is prismatic and of constant roughness, the veiocity rustdbution
will eventually reach a definite pnttern. For simplicity of discussion the
following m'e assumed; (1) th~ fiow enteX'ing the channel is lamil1al' and of
uniform velocity distribution; (2) no restriction exists at the entrance that
will cause abrupt disturbanc~ of the water surface and the velocity distri. ,1 For the sake of simplicity, the two-dimensiona.l profile of:a. wide open channel
With exagger~ted verticll.l sc.ale isahown.
192
r",-
. '
(8-1)
"-
;
. I-I/:.
.,. .If {J. ,(.r
C>
-=- vI(.')j"~oJ.fCL 1'r"G<:.-' .tId .;)' l.tfol
V{~U_1. .d~ ~/..
_'
~-"
.t:d _ . i
/'/-t . . . ~7;1'.,A;..
< /i
194
195
is
V,k
<5
or
<
51'
(8-2)
)I
where V f
a term known as the friction lJe/ocitu (Art. 8-4).
Using the Gh6zy formula. it can be shQ\rn~ above condition
that, for a surface to be hydraulically smooth, the roughness height must
be less than a critical rotlghne.ss expressed by
k.
'~-----Y-
5C ;
(8-3)
where C is CheZ~I's C, I' is th.e kinem(\tic viscosity, and l' is the mean
velocity. The above conditionisgiveri for r0l1ghi1ess obtained with sand
. having \'alue;s of C greater than 100, prob"bi,r. For the avcri1ge COlldi~ion, Schlichtillggives kc = 100"/1', which corresponds to C '" 113.5.
ail approximD.tion,
(8-3) U1:;ty be applied to channels.
If the roughness height is gl"eater tha~l th'l critical va111e..d~y EJi: .
(8-3) (Fig. 8-3c), the roughness elemeuts will have sufficient magnitude
~ angu!a.ritY..to extend...tht'lis~ffects beJ:::9l!.c!. the !l).mil~[1,r subl~r..i1ncl.
thus to disturb the flow in the chi1nne1. The surface is therefore said to
bC"ro1Lg/i:--in rough ch~i;,nJS, ~
the-~eI;;~ity distribution will depend 011 the
form and ~ize of the roughness projections, .!1l1d a stable laminar subli1yer
can no longer be formed.
, i
lI
I It should be noted that the roughness height is mel'eLy a measure of the linear
dimension of the rQughness elcmeuts but is not necessarily eqoaL to t.he ac~ual, or ev~n
an average, height. For example. two roughness elements ll13.Y have diffctcllt linear
dirnr>,nsions, but, owing to the difference in shape and orientation, they may produce
identical roughness effect and, thlls, their roughncsscs will be clesignated by the same
roughness height.
2 The position from which the rOllghness hcight should he measured is IL disputable
mli\tter. It is a.'<StlInJad here tha.t k is measured from 1\ datUllI that li,,~ at. n distance of
O.5k below the average bottom of the channel.
~I
196
UNIFORM FLOW
r->'I'h~
I,
I
;
I
I
.i
i97
!i
!.~etermiM,
and
TA1l!.E
8-1.
ApPROXIMATE VALUES
OF
ROUG.lNESS HE1(lHT
. Mc.lerial
k,
It
it
0.0001:"'0,0030
0.0002-0.0080
0.0004-0.0070
0.0005-0.0150
0.0008-0.0180
0.0006-0.0030
0.0013-0.0040
0.OU15-Q.0l00
0.0020-0.0100
0.0030-0.0300
0.1000-3.0000
(0)
.-'--;:;,
. , 1"
.
slcimmt'ng) flow.
---------
'
~ solated-roughne~~Jow.
( c1
198
UNIFORM FLOW
mental 'data from many different sources. The concept can also be
extended to surfacs of variable roughness by u5ing average values of the
roughness dimensions that vary or by combining the friction factors for
each flow type to give all over-all apparent friction factor of the flow.
83. Computation of Boundary Layer. For the development of a
turbulent bound~ry layer in wide channels, an-;:pprQximate but practical
meth~2!L h~_~.E!:Q12Q~_Qy_Jl~lJ&t.f.1, This method
waS' developed primarily for flo\v i.,1l; ~.han~els.oI large ~l'?.P.e, but it has
been found applicable also to channels of small slope; provided the flow is
.
-.::- -.... ,-...... -- .--:'
. computing the varied-flow surface is used. In the case of tIle boundarylayer development of a uniform flow, the application of Bau61:'s method is
just the same except that the water surface requires 110 computation, for
it is simply parallel to the ehanneLbottom.
Example 8-1. Aooncret.e overflow spillway of indefinite length has a. surface slope
angle 11 = 538' (Fig. 8-5) and a roughness snch that k
0.005 ft. When the discharge is 3130 cfs per foot of spillway width, compute the length for boundary-layer
dev!:')opment, the profile of the boundary layer, and the water surface,
Solution.. The computation is shown in Table 8-2; the headings are explained as
follows:
"
.j
now
,j
TA.BLE
Dam
8-2.
= 0.005 it
x/k
BOtiNDARY-LAYER COMPUTATION
(J
538'
u. i /2g
li/x
360
cfs/n
v.
glv,
( 4)
F" .
50
Fm. 8-5. Growth of bounda.ry layer over the surfac<:l of an overflow Rpillway.
100
200
400
th.~l~~~idlr
600
The st,udy of boundary-layer de.velopment by Bauer was made on concrete overflow spillways (Ii'ig. 8-5). In this case the transition from the
la~illar to the turbulent boundary layer usually occurs far upstream
from the zone under consideration; hence the laminar boundary layer can
be ignored, fo'r it is all insignificant part of the .problem. From the
results of Bauer's investigation, the following equation may be written:'
l) _
1 X 10;
2 X 10
4 X 104
8 X 10'
L2 X (05
0.0073
0.001313
0.0060
0.0055
0.0052
0.36
40
50.S
0.136
1.20
2.20
3.12
SO
71.8
7.09
5.00
7.13
5.07
160
101.5
3.~
3.65
320,
113.5
175.8
250
2.72
2.35
480
2.04
The computed water surfa.ce a.nd t,he bounda.ry layer are plotted on logarithmic
paper (Fig. 8-6). The point of intersection of the two profiles indica.tes the location
of the section where the bOlmdary layel: reache.s its maximum, or where it is fully
developed. The lengtb of the development is shown to be approximately 460 ft.
It shOUld be noted that the simplified method of 'computing the water-surface
profile of a varied flow is justified in this problem because the flow is relatively thin
and the slope is very steep.. As the frictional loss in this case is :negligibly small, the
potential energy of water is practics.lly entirely converted into the lunetic energy
expressed by the velocity head. For the computation of surfl:\Ce proj\Ies of varied
flow in general, various methods will be described in Part Ill.
To illustrate the effect of change in roughness upon the growth of a boundary layer,
the computation has been repea.ted using ~he8amc da.ta but with a rOl\gbness twice as
large, The line marked with It 11 = 8, k = 0.01 itO (Fig. 8-6) is the result of this
computation. The change in water surface ispractic.!Llly negligible. It can be seen
tha.t the difference between the values of the development lengtll for the two cases is
0.024.
x - (x/k)o.13
where 8 is the thickness of the turbulent boundary layer at distance x
from 0 in the diiectionof flow (Fig, 8-5) and where k is the roughness
height, Bauer has shown the application of this method by the 'following example. It should be noted that the flow on the spillway surface
to be described in this example is varied and that a sirtiplified method of
1
I
!
I
I
I
.. I
<I
200
~I',
UNIFORM FLOW
10
:FEfIi
! I t
,~
! ~u{j ,
i'.r-.,
I,
t-
~>'er.....
~.,
"
"/.bf:
JiV
....
'+-
OJ
U
,,'
i5
"..'"
't>'
"\<-
~I
O.
0.4
v:
/
I
r
[/1/
/" V ,
i-
{j
200
DistallCe x, f j
300
400
600
BOO
P,y./('UViJ.f-
vgRS
(8-8)
Vr
. 8-~., Veloc!ty Distribution in Turbulent Flow. The velooity distributiOn In a ulllform channel flow will become stable when the turbulent
",-_Q.9un~ar~ la!el' 1l3i'!!!.v ,develoE~d. In the turbulent boundary layer,
the, clistnbu:lOn can be shown to be approximately logarithmic. ","
The shearing stress at any point in a turbulent flow moving over a solid
surface has been given by Prandt.l [7] as
T
= pl2
(::)2':
L~~:::) .~
0:.
tVl ~
(8-9)
about 30 ft, or a variation of less than 10%. This shows that an increase in rough,ness has a tendency to speed up the boundary-layer growth or to',educe the developmenL l e n g t h . '
,
(8-5)
where
'1-
-L-J
(8~7)
yo
-----iY' it varies
0.3 c,.
40 50 60 70 80 100
FIG.
I~ In JL
Vp
I
.~
V,
v = 2.5
'/
.,~
//"'~
o.61--:- f-.
0.5 f--.
'++
Lbo..:\) . '
..
I
oP~c:7
epV:l
J.QLth~n near the acEd surfac\l, ,Prandtl introduced two assumptions:' (1) that themixh}g 'length is proportional. to 'Ii, and (2) that the
shearing stress is constant; Since the shearing stress at the, channel
surface is equal to the unit tractive force (Art. 7-11), the second assumption 'gives T = TO.:, From these two assumptions,Rq. (8-5) may be
. .
wrItten
(8-6)
where K is a constant for the proportionality between L~nd }I:, The value
of K has been determined py many experiments [8] to be about 0,40.
11ltegr'ating li<J. (8-6),
'
c.( '1 .=::> I '" ~ "3'
'b'
o. 9
0,7
/'
,\
201.
% for
smooth surfaces. 2
For wavy
1 Von Karman .[9J also proved this law by II similarity hypothesis which assumes a
linear shearing-stress distribution, the mixing length being proportiona.l to (dv /dy) /
(d~/dy).
011
20~
!
i
UNIFORM FLOW
surfaces, the' value of m wHI depend on the shape of the wa.ve profiles.
Substituting Eq. (8-10) for Yo in Eq. (8-9). and simplifying,
v
9yVI
575V
.
r log - v -
if:
= 5.75YI
log
30
-!Co
~Q'
P - 'YY
(8-15)
(8-11)
'k
11
}o
B d.y = Ph - 2: 11,2
2
(8-16)
Substituting in Eq. (8-14) the value of v from Eq. (8-9), that of B from
Eq. (8-15), n.nd that fO!' A from .Eq. (8-16) and then integra ling and
(8-12)
203
.<
Substituting
(8-13)
lli
-.
8-0. Theoretical UnIfOrm-flow Equations. Using' the Prandtl-voll
Karman universal-velocity-distribution law, KeulElgan (13)2 has derived
equations for mean velocity of turbulent flow in open channels. . :I;<"o.r
the sake of clarity l;1.nd simplicity, Keulegan's derivation will be modified
below.
By the cQntiuuit.y equation, the. total discharge through an ordinary
channel section (Fig. 8-7) may be written
~h
FA
t V=k v dA
=.
1.= O .
lch vB dy ,
0
v=
or
-V = VI {5.75 log
(8-17)
(8-18)
This is the general theoretical equation for the mean velocity of uniform
. flow in open channels.
For. smooth channels, Yo mn:y be represented by Eq. (8-10). Also,
[n
lf
...... .
obtl~ined:
[li!!:.
exp (-1 _1'h~)l
Yo R
. 4A J
V Yr (A
where h is the depth of water, A is the vlater area, B is the length of the
curve of equal velocity, and y is the vertical depth measured from the
boundary to the curve of equal velocity. Since the laminar sublayer is
relatively very thin, 00 can be assumed to he zero. It is further assumed
that the maximum velocity is at the free surface and that the length B is
",,,',
2.5Vr In
to illustrate notation.
(8-14)
This' value is derived from Nikuradse's experimental data on Tough pipes [121.
Nikuradse's experjments, sand grains were cemented to the inner walls of pipes to
. simulate roughness. Such roughness is known as artificial T"Ougimess.
2 Other references on this subject fLre [141 to r21J~
~ec~iol)
I
t
.J
-I
204
UNIFORM FLOW
"d'
ber
functions 0 f t h e F. rou e num
.
as follows:
IT = VI
(8-19)
l'
Umng Bazin's data [22] for wavy surfaces, Keulegan obtained a value of
Ao = 1.3 for smallwood channels and -3.0 for large wood channe1s.
For rough channels, yo may be expressed by Eq. (8-12). From Keulegan's study of Bazin's data [22], the valu~ of Ao was found to have a wide
range, varying hom 3.23 to 18.92. Thus; a menu value of 6.25 for
Ao may be used. The theoretical uniform-fiow equation for Tough
channels is, therefore,
v=
V, (6.25
+ 5.75JOg~)
205
I
_ _
RY /)
"A,
+ ~.7u log--;-
(8-21)
(8-25)
t'"
..
, .' f'
ns of the }'roude number ..
In the above eouatlOns, A. and l~ ale .mc 10
.
ft.
The
1 d t'
f" ct~ of other mlllOI' ac OIS.
These constaill.s may al,;o ll,/; u e ne e Ie"
.'
L
(8-20)
7
v
VI
Yo
(8-21)
.;
"c:c
20.8R
C -- 32"61
"
ogc;-
(8-22)
(8-2:3)
4:
-I
-2
-3
0.2
0.3
0.4
F
FIG.
. .
. F' 88 by the
lotted data are found to be much scnttered(as shown III 'Ig" I - I "1
.
. t " l ' ffects "When t 1e c lu,nne
shaded boundaries) probably OWlIlgo t le:;e e .
I
h
slope becomes too large, the flow will be unstab~e (Al:t. 8-8).. l~ suc
cases, the \u,ws of turbulent flow described in tIllS o.rtlCle are no onger
valid."
.. , R
hess Coefficient.
S-S. Theoretical Interpretation of M.an?m g s oug. n ..
he theoIt should be interesting to study Mannmg s n by relo.tl~g l~ tfn\he preretical channel roughness for rough channels, as ~e(v8e ~~e Ma ning's n
ceding article. Eliminating C from Eqs. (5-7) [1.n
-~,
n
-="
206
UNIFORM FLOW
n
where
~ (~)k;6
(8-26)
(R/k)'A
21.9 log (12.2R/k)
(8-27)
The plot. of t.his equation, (Fig. 8-9) indicates that, for a wide range of
R/ k, the variation in </I(R/k) is small. As an approximation, (R/k) may
be assumed constant and equal to an aVel'!l,ge value. On the basis of
actual observations made in S,vit;r.e!land, Strickler [23J ar.rived. at a
formula which, when compared with Eq. (8-26), gives an average value l
of .p(R/Ie) = 0.0342. The roughness height used by Strickler is the
I
...---o~':-
S J feems in
No,lhweslern U.S.
(I
~ ..;,n.. _ _ _ _
II
"
'\.
Sfricklerts
__ _
~
,'"
Similarly, let
UO.s
R/l<.
5.75Yr log
median sieve size of the material. Data [24J collected from sever",]
streams in the l1orth\ycstcl'll United States and from the 'Mississippi River
are also shown in. Fig. 8-9 for purposes of comparison,
If </J(R/k) is assumed constant, Eq.(8-2i3) willler.d to the interesting
conclusion that Manning's n varies with the
power of the
height. In other words, a thousandfold change in the linear
me~~,sure of the roughness height !'esults in about n. threefold change in n.
Thus, the roughness height as a measure of channel roughness is more
sensitive than Manningls n. Consequently, the effect of errors involved
in estimating the roughness height for the determination of Mallning's n
by Eq. (8-26) is comparatively smalL
. Bakhmeteff and Feocloroff [Z5} have made a comparison of the Manning'
formula with the Prandtl-von Kirma,n formula [in a form corresponding
to Eq. (8-26)), the G, K. formula, and the Bazin formula, Their results
indicated that the Manning formula in the form of Eq. (5-7) is the best
suited to all formulas under consideration.
8-7. Methods for Determining Manning's Roughness Coefficient.
Based on the theoretical velocity distribution in rough channels, two
.
....;,.....-~approaehes for
's n have
----
og'k =
(8-28)
f='-~x-l
(8-29)
V
Equating the right-hand siues of
24y
5.75V r logT
Uo.t
.
;;.
207
(8-30)
(8-29) and (8-30) and solving for 1'1.,
(8-31)
This equation gives the va.lne for 1'1. for a wide rough channel with
logarithmIc velocity distribution. When this equation is applied to
a.ctual streams, the value of it may be taken as the mean depth. A pli1t
of this equ:1tion (n/y;f, vs. x) has been made to compare with a similar
curve developed by Boyer [24Jand also with actual observations taken
. from several streams in the northwestern United States and the Mississippi Rivet;. The. comparison leads to the belief that a general relation-
208
, UNIFORM FLOW
RELATED TO UNIFORM FLOW
THEORETICAL .CONC""FTS
'"
2U\:l
'.1,000
eooR~~~~~1!~flE~~~!~!~~f~~~~~~t~~
so
600
400
\.
. . rl'1
20
I'"
~
'0
'"
'"
E
0
10
B
:>
4
0.8 L -_ _ .;.._+0.6
O. 4
-l---l++-i4+t~-+-".ci--f';
L__-.l--~4--+-\-+-H+-~~:---'l--"r'":
,-l~4~'Il-~n:-1
0.2.
(8-32)
~here the value of R'I R depends on the hydraulic radius R, the slope S,
and the grain sizes k65 and ka By computing values of (R/k G.)'/.! and
k361 RS, the value of R' / R Can be obtained from the semiempirical curves
of Fig. 8-10. These Curves are based on data taken from seven typic",l
rivers in the United Sto.tes. 1 When <f>(Rlk) is computed by Eq, (8-32),
the va1ue of n can be determined by Eq. (8-26), with k = k65. When
Values of
x35
'liS
.
there is little
moving sed'lment 'In th ~ .St' 1'eam , or R - R' = 0, it can
.
. be
seen that Eq. (8-32) is reduced t.o Strickler's co~stant .(Art. 8-~),
The method described above doeS not comudel' kmds of !IOu~hness
other than :surface roughness an d t h:e rong hness d ue "to moving-sedIment
1
beds. Therefore, it is applicable on1y to problems 111 whlch the bUS1IY-
':Missouri River at Pierre and Ft. Randa.U, S.D., and Omaha, Nebl'.i Elkhorn
Rhier at Waterloo, Nebr.; Big Sioux River at Akron, Iowa; Nioprara Ri~er at Butte.
Nepl'.; Platte River at Ashl!l.nd, Nebr.; SBlina.~ River at San Lll,c"s and Paso Robles,
Calif.; and Nacimiento River.at Junction,' CElIif.
.
j
I
I\
UNIFORM FLOW
,,
bank friction is negligible and in \vhich the active channel, even though
split by sand ba'rs and possibly by islands) is free of debris and vegetation.
8-8. Instability of Uniform Flow. Uniform floW' will become unstable
'when the velocity of How is very high or the channel ,slope is very steep.
When this happens, the instability of the free surface is characterized by
'the formation of a series of roll waves. This phenomenon was first
reported by Cornish [31J in 1910 when he observed it in open :channels in
the Alps. The roll Wave is a phenomenon of unsteady flow and its nature
will be described ill Art. 19-9,
Many attempts [32-45J ha.ve been made to develop a criterion for
instability of uniform flow. In 1945, Vedernikov [35], employing certain
approximations of Saint-Venaht, developed a criterion which may be
called the Vedemikov number V. Later, Craya (41} and Iwasa [43}
studied the initiation of continuous tilliegrowth or decay of an infinitesi1110.1 di5tUl'bo.nce wave 011 fluid surface, and developed similar criteria.
The Vede1'llikov number may be expressed as
II
210
xyV
V",- V
= xyF
providing the value of X,. For lack of such a formula, the Manning or
Chezy formula is used as an approximation only. Since the channel
slope is us,tally large when instability of [Jaw develops, the Froude number
ill
(8-35) should be computed by Eq. (3-13), which considers the slope
effect.
PROBLEMS
8-1. Show that Eq, (8-0 is deriv~d from the condition that the presence oi'the
boundary Lg,yer raises the bottom of the cha,nnel to a virtual distance equal to the dispLa.cement I;hiclmetls.
8-2. A laboratory [ecta.ngular channel 1.5 ftwide carries a unirorm flow of 0.34 cis
at p, depth of 0.15 ft and ll. slope of 0.0009. The roughness height of the channel surf!>ce is 0,00215 ft. Plot the profiie of the turbulent boundary layer, 1).0(\ est.im!l.te the
length -of channeL required for a fun development of the boundary layer \vhich begins
tobe turbulent u.t the entrance.
8-S. Determine whether the c.hannel'described in the preceding problem is hydrauli'
cally smooth or rough.
8-4, A trapezoidal channel (Fig. 2-2) with b ,.. 20 H, z
2, and S = 0.001 ca.rries
a uniform Row at a depth of 6 ft. Compute the
unit tractive force :Lnd the
fric~icm velocit,y developed in the channel,
Wha.t are
maximum friction velocities on the sides and bottom of the dUlIInel?
8-li. Snow t.1U1.t
(8-34)
<:t.
+ 3.'
- 2.'
(2-6)
(2-7)
(1=1+.-
(8-35)
211
i,Il
(8-36)
Vi
I
r
1>;
where.
2.5V';V. Also show that ( = H.2/C ~ 0.883
= 9.5n/RH .. (v~r/V)
- 1 where C is Ch6zy's resistance factor,:j is Darcy's friction factor,:n is Manning's roughness coefficient, R is the hydraulic radius, VA( is the maximum velocity,
and V is the mea.n velocity.
8-10. Plot a curve showing the relationship between a a.nd {J defined
(2-6)
and (2-7), respectively.
8-11. Comput,e the velocity-distribution coefficients. of the flow in the channel
described in Prob. B-7.
1\
1:
212
UNIFORM FLOW
THEORETICAL CONCEPTS RELATED TO 'UNIFORM j,'LOW
18-~~:
J
' . an
- 2/3y - 797 where R d '
ano, A IS in [('.' Construct tl. e disc lIar
t"
'
an yare m ft
"
.
,
ge-ra mg curve of the strean'
, b-14. Show that the Vedernilcov' number in wide channels is V ~ 2F
'.
Row, V = O,67F fo), turbulent flol'v: 'f th "T " f '
for lammal'
1
e ,',-annmg orm'll- IS us d
dV
0
turbulent flolV if the Che~y formula is used.
,~e ,an
=
,SF for '
8-11i. Using t.he Vedernikov' criterion expla'n tl '
. I""
.
.
f<t0tors Upon 1Iow instability. (a) th
'I'
f I " I" genela e"ects ot the follOWing
I
.
e va I,e D x m the unIform flow f rmll lib) h
Slape of U.tb thannel cross section, (c) the viscosity ,of tho flu'd -(tl) thO
a"
t ,e
clent, and (e) the channel slope.
"
I,
e enel'gy codfiREFERENCES
1. Hermann Schlichting: "Boundary Layer The r . "
,
,
'
tv
J Kestl'n 'f G.
H'II B k C
<JlY, transluted from t.he Germau
J.
,
1. C >IIl.W- I
00
ompany I
K
Y
Ltd:, Lor~don, and Verlag G. Braun, ICarlsrulle,n c . ew ork, Pergamon Press
l 955
2. J. W. ~elieur: The boundary layer development in open channels a '
,
/;e\ ll~:-;,
6.
7
.
" ' ,
William J, Bauer: -Turbulent boundarv layer on 3teep slopes T
A me"iC(nI S C', ( C' '1 E , , ransaC,1.OnS
L
.
. 0 ~ .. y 0,
ngmeeu, vol. 119, pp. 1212-1233, 1954.
'
P udwIJ,:P:tandtl: Uher dIe ausgebildeie Turbulenz(On fully developed turbulen )
TOCte mgs of the 2d Inte?natianal Congress Of Applied Mechan'''cs Z" 'h ~e ,
~
"
Hone, pp.
62-74, 1926.
'
tv:
;h~~d~r
von
I{~rm.!in:
213
'
13. Garbis H. Keulegan: Laws of turbulent flow in open channels, Research Paper RP
1151, Journal of Research, U.S. National Bureau of Standards, vol. 21, pp. 707-741,
Decemher, 1938.
14. Erik'Lindquist: 00 velocity formula.s for open channels and pipes, T"ansacl-ions
of the "World Power Conferer.ce, Sect'ional M feting, SfX!ndinavia; Stockholm, vol. 1,
pp. 177-234, 1933.
15. A. E. Dretting: A set of practical hydraulic formulae based on recent experimental
research; comparison with older formulae, Appendi:r; 24, 2d !Jleeling, International
Association of Hydraulic Str'!J.ctuTes Research, Stockholm, Apr. 7-9, 1948, 20 pp.
16. Ralph W.,Powell: Flow in a channel of definite roughness, T,ansaclions, American
Society of Ci'j'il Engineers, vol. 111, pp. 531-566, 1946. '
17. Ralph "'. Powell: Resistance to !low in smooth channels, Transactions, American
Geophysical Union, voL 3D, no. 6, pp. 875-878, December, 1949.
18. Ralph L, Powell: 'Resistance to flow in rough chs.nnels, Transact~ons, American
Geophysical Union, vol. 31, no. 4, pp. ,575-58:2, August, 1950.
19. A . .fi.. Robinson and M. L. Albertson: Artificial roughness standard fOr OPll
~hannels, Transa,ctions, A.merica" Geophysical Union, vol. 33, no. 6, pp. 881-888,
December, 1952.
"
20. Yuichi 11vagaki: OIl the laws of resistance' to turbuleut !'low in open smooth
channels, lI1emoil's oj the Faculty of Engineering, Kyoto University, Japan, vol. 15,
no. 1, pp. 27-40, January, 1953.
21. Yuichi Iwagaki: On the law of resistance to turbulent flow in open rough channeia,
Proceedings of the,4lh Japan National Congress for Applied M echanies, pp. 229-233,
1(l54.,
22. H. Darcy and H. Bazin: Recherches hydrauliques:I, Recherches cxperimeotales
sur l"ecoulement de I'eau dans les canaux decouverts (Hydraulic research: I,.
Experimentftl research on',the flow of water in open channels;, ],f emoires present~s
par divers savants a l'Acud~mie d~s Sciences, vol. 19, no. 1, Dunod, Paris, 1865,
23. A. Si,rickler: Beitriige zurFrn.ge der Geschwindigk'eitsformel und der Rallhigkeits~!l.hlen fiir Strome, Kanale und geschlossene Leitungen (Some contributions
to the problem of the velocity formula a,nd roughness factors for rivers, canals, !l.nd
closed conduits), Mitteilungen des ddgenossisehen Amtes 1iir Wasserwirtiieha1t,
:Bern, Switzel'land, no. 16, 1923.
24, M,. C. Doyer: Estima.ting the Manning coefficient from ~_n average bed roughness
in o[3en channels, Transactions, American Geophysical Union, voL 35, 00. 13, pp.
957-9&1, December, 1954 .
25. Boris A. Bakhmeteff and Nicholas V. Feodoroff: Discussion on opeu cha.nnel flow,
Transactio1'1s, Americcm Society of Civil Engineers, vol. 1'08, pp. 492-502, 1943.
213. M. P. OiBrien: The vertical distribution of velocity in wide rivers, Tra.nsactions,
American G~ophysical Union, voL 18, pt. 2, pp. 4137-470, 1937.
27. F, C. Scobey: The flow of water in il'rigation channels, U.S. Department of Ag,'iculture, Professional Paper, BI"Uetin No. 194, 1915.
28, W. R Langbein: Determination qf Manning'S n from vertical-velocity curve,
Transactions, American Geophysical Union, pt. II, pp. 618-620, July, 1940.
29. Hans A. Einstein !lnd H. L. BarbuTossa: River channel roughness, Transactions,
Amen:can SocieLy of Civil Engineel's, vol. 117, pp. 1121-1132,1952'30, James J. Doland and Ven Te Chow: Discussion of River channel roughness, by
Hans A. Einstein and H. L. Ba.rbaross3., Trr;msaclions, American Sociely 01 Oivil
Engineers, yol. 117, I'p. 1134-1139, 1952.
31. Vaughan Cornish: "iVaves 'of the Sea and Other Water Waves," The Open Court
Publishing Comp!l.ny, La Salle, Ill., and T. Fisher Unwin, Loudon, 1910.
214
l,
UNIFORM FLOW
32: Harold Jeffreys: The flow of water in an inclined channel of rectangular section,
London, Edinburgh and Dublin Philosophical i1!agazine and Journal oj Science,
ser. G, YfJL 49, pp. 79:3-807, May, 1925.
.
33. G. H. Keulegan and G. W. Patterson: A oriterion for instability of flow in steep
channels! T,'o,nsactions, American Geophysical .llnion, vol. 21, pt. II, pp. 594-&96,
July, 1940.
34_ Harold A. Thoms,s: The propagation 'of waves in steep prismatic conduits, Proceedings oj Hydraulics ConJere1J.ce, University oj Iowa St'~dies in Engineering,
Bulletin 20, pp. 21!l--229, 1940.
35. V. V. Vedernikov: Conditions at the front of a translation wave distributing a
steady motioll of a real fluid, Camptes renclHS (Doklady) de I' Acad~mie des Sciences
de t' U.l?.S.S., vol. 48, no. 4, pp. 239~242, 19i5.
35. V. V. Vedernikoy: Characteristic featur!)~ of a liquid flow in an Open channel,
Comptes rendus (Dok.lady) de l'Accidemie desScience~ de l'U.R.S.S., vol. 52, pp.
207-210, 1946.
37. RalphW.Powell: Vedernikov's criterion for ultnHapicl flow, Transactions,
Ame,icanGeopilysical Union, vol. 29, no. 6, pp. 882-886, December,1948. Discussions by V. V. Vedernilwv a.nd Ralph W. Powell, voL 32, no. 4, pp. 503-507,
August, 195138. Robert F. Dressler: Ma.thematical solutions of the problems of roll waves in
inclined open <lhanneis, COlnmttnication.~ on PW'e a;nd Applied Mathomat'ics, vol.
2, no. 2--3, pp. 149-194, 1949.
39. Robert F. Dressler: Stability of uniform flow and roll-wave formation, Proceedings
oj the Gmuily Waves SVm.posium, U.S. Na;tiona.l Burea;u of Standards, Circu.lar 521,
1952, pp, 237-241.
.
40. Francis F, Esr.offier: A graphical method for inVestigating the stability 9f flow in
opim channeJ.o:i or in closed conduits flowing partly full, Tramactions, American
Gcophy.~ica.! Uni011;, vol. 31, no. 4, pp. 583-58G,.1950.
41. A. Ontya: The criterion for the possibility of roll wave formation, Proceedings oj
the G.ravity 'Waves Symposiu.m, U.S. National Bureau of Stan.dards, Circular 521,
1952, pp. 294-332.
42. Tojiro Ishihara, Yuichi Iwagaki, and Yasuo Ishihara: On the rollwave-trains
appearing in the water flolV on a steep slope surface, llfemoirs oj the FawUy oj
Engneering, Kyoto University, Japan, vo:' XIV, no .. 2, pp. 83-91, March, 1952.
43. Yoshiaki Iwusa: The criterion for. instability of steady uniform Hows in open
channels, 111 c1Iloirs oj the Faculty oj Engineering, Kyolo University, Japan, vol-.
XVI, no. 6, pp. 264-275, March, 1954.
44. Chia-Shun Yih: Stability of parallel laminar flow with a free surface, Proceedings
oj th~ 2d U.S. National C01t(JreSJ oJ Applied lif echanic8, American Society of
Mechanical Engineers, 1954, pp. 623-528.
45. C. C. Lin: "The Theory of Hydrodynamic Stability," Cambridge University
.Press, Oambridge, England, 19.55 ..
PART III
GRADUALLYVARffin FLOW.
f
Ii
~k'~
I.
}.
2.
.Y!r~~ IJ<!Lf..
/'''' e-c..
~J:7 !'Wut.(.f.'q'
.(
r
\
L
i
I
1
[
!.
",
.)
J
CHAPTER
~l'il1i;;~~ &~"Rtff
lV1301HOd qn'gt.:;~
218
.
H
I
Ii
T,
V2
= z + d cos e.+ a 2g
(3-2)
,4
dx = cos ()
.=
I .
'I~
,
".,
So - Sf
(9-2)
+ Ctd(V~j2g)jdd
'(
This is the general differential equation for gradually varied flow, referred
to hereafter as the'dY1iamicequa:t~Q'n;of gradually variedflow,or simply as
, the gradual/y-va7'ied:'Jtow eqttation. It represents'the slope of the water)
surface with "espect /0 the bottom of the channel. The depth d is measured
from the bottom of the channel, lJ.rld the channel bottom is taken as the
x axis. Thus, the slope of the water surfcLCe is equftl to the bottom slope
r
So if ddjdx = 0, less than .so if dd/d.'C is positive, and greater than So
1
if ddjdx is negative. In other words, the water surface is paranel to
l It should be notEld that the frictional loss dE is always a negative quantity itl the
'direction of flow (unless outside energy is added to the course of the flow) and that
the change ill the bottom elevation dz is a negative quantity when the slope descends.
. :";
Substituting
sl~le
-'!""'--~
.. -....;;:.;...=~~---....:---'~
. "
t
(V2)
:)
.f
""
:.5
<1>
54J<.~~
Y'~>3'"
i,
~ .
.:; i .
220
the cbannel bottom when. ddjdx. = 0, rising when dd/ dx is positive, and
. loweling when del/dx is negative.
..
In the abovH equation, the slope angle 0 has been assumed constant or
independent of x. Otherwise, a term -d sin (J(d8/dd), where (J is a function of X, wonh:l have been added to the denominator. For small .OJ
cos 8"" 1, d "" y, and cld/dx ~ dy/dx. Thus Eq: (9-2) becomes
So - Sf
.;:rJ,x - 1 + ad(Tn/2Q)/dy
(9-3)
In most problems, the channel slope is smaH; accordingly, Eq. (9-3) will
be used in subsequent discussions.
The term ex d(P/2g}/dy ill the varied-fiow equation represents the
change in velocity head. The coefficient a has been [;,ssumed to be constant from section to section of the channel reach under consicielp.tioll.
Ot,hel'wise, the change in velopity head would have been expressed as
(I(Cl V2/2g)/dy, where IX is a function of x. Since V =Q/ il, Q is constant,
and dA/dy
'1', the velo~ity-hf.)ad term may be developed as follows:
1
d (V~)
a dy
\
Since
21}
Z = VJP/T,
aQ2 dfi- 2
2g--dy
aQ2 dA
=
gA3 dy'= -
aQ2T
gfP
(9-4)
d (V2)
a
ely \ 2y
aQl
= -
gZ2
(9-5)
'\
dy _
221
In: a
Vi
=
GZR
(9-9)
Kn~
So =
](2
(9-12)
(9-7)
where r
So/ S,n, or the ratio of the channel slope to the critical slope at
the normal depth of discharge Q (Art. 6-7) ;
The term Sf inEq. (9-3) represents the energy slope. According to..the
:first assumption in Art. 9-1; this slope at a channel section of thegr!ldually
~al'ied flow is equal to the' energy slope of the uniform fiow that has t),e
(9-15)
dy
dx
. 1 - (Kn/KF
So 1 - r(K,,/J()~
(9-13)
(9-14)
222
where Q is the given discharge of the gradually varied flow at the actual
depth y; Qn is I;he nQrma.1discharge at ~. depth equal to y; and Q. is the
cntic!1r discharg;;:t a depth equal to y; [\nd
'
dy
(IX
;'
So - Q~/C2A2R
- aQ2/ gA2D
= ,1
(9-16)
1. 1 - (K n /K)2
2. 1 .- (Kn/K)"
(9-17)
<:: 1
'-0
(Yn/V) 3
1 _ (yjy)3
(9-18)
2.1- (Kn/KP
1-
(Z./Z)~
(Zc/Z)2
<0
<0
(Z./Z)"
>
and 1
Considering Un
1. Yn
>
>
Yo
00,
1 The term" backwater curve 11 i$ used primarily to indicate the longitudinal surface curve of the water backed up above .0. dam or into a tributary by flood in the mll-in
stream. Ma.ny authors ha.ve extended its meaning to include all types of flow profiles.
dy/ilx is s
(9-19)
~,draWs1oWl~\::::e~ :~~:e~o_yac._."
flow with
ve.
e second case represents a supercritical
~.- ...
a ackwater curve, ~s shown as positive.
0\
0:-----.
~I
T:~ ::s:c:;e:e:rese~L~b~~t~cal
<0
.!
dy _
-(Q/K)~
dx - 1 - (Zo/Z)2
< 0 and
> 0 and 1
The first case indicates that Yo > Y > Yn. and, thus, that the flow is superCi:itical in a steep channeL 'Slffiiiarly, the second case indici,tes that
~ V > Yo, 9E.. that the flow is subcritical in a rild channel.
vVhen the water surface is parallel to the bottom of the channel,
dy/dx = 0, and Eq. (9-13) gives 1 - (1(',/ K)2 = 0, or Y == yn, which indicates a uniform now. Tbe flow is a unjform critical flow if y = Yn = Yo,
8, Ulliform subcritjcal flow if y =
y ... > Yet and a uniform sllpercritieaIflow
if Yo > Yn = y.
70-r purposes of discussion, cha.nnel slope may be classified as sv.stair.ing
and nonsusta.ining. A sustaining slope is a chai1l1el slope that falls in
the direction of flow. ,Hence, a sustaining slope is always positive and'
may also be ,called a positive slope.' A sustaining or positive slope may be
critical, mild (sllbcritical), or steep (supel'critical). A nonsustaining
slope may be either horizontal or adverse. A horizontal slope is a zero
slope. . An adverse slope is a negative slope that rises in the direction of
flow.
.
In a channel of horizontal slope, Qt" So = ~ (9-11) gives Kn = 00
. or ~ince K ... ~ = Q, Eq. (9-13) fo~ hOrizo\ltal channels may
be written
223
224
Ij
In the first case, dy/dx is negative, and .the flow profile is 3. drawdown
curve. In the second case, dy/dx is positiYH, ami the flow profile is a
backwater curve. Since Ie. is imaginary, the value of Yn cannot be
easily expressed. 1 In fact, a finite positive .value of Yn is physically
. impossible because a uniform flow call never occur in an adverse channel.:
(Why'?)
The above discussions are summal'ized ill Table 9-1 and in Figs. 9-2
and 0-4. Some special features of the theoretical flow profiles are' as'
follo",'s:
.
A. Discontinuit1{ in Flo'/j) ProjjJ.e. When Y. y., Eq, (9-13) indicates
tMt~-= co, that
that the flow pl'(jfile will be vertical in crossing
the critical-depth line. If the depth of flow is ch!mged suddenly from a
low stage to ll. high stage in crossing the critical-depth line, a hydraulic
jump will occur, l'epresentinga discontinuity iIi the flow Profile. If the
depth changes from a high to low stage, then a hydraulic drop wi.ll occur.
It should be noted that, at or nep.l' the critical-depth line, the flow profile
is bent to produce such great curvature that the parallel-flow assumption'
for ~he definition of gradually varied flow will. introduce large errors.
. In fact, the flow may become so curvilinear or rapidly varied that the
theory and equations deve~oped in the preceding artiCle become inapplicable. Therefore, Eq. (\)-13) cannot be used to describe or compute accu!'ai-ely the flow profile near the critical depth.
.
B. Behavior of Flat/) P1'Ofile at Specific Depths. For the subsequent
discussion, it is important to recognize the theoretical behavior of the fiow
profile at several specific depths. When y = <0; Eq. (9-13) shows that
dy/dx
SD, that is, that .the flow surface is horizontal. When y = Yn;
Eq. (9-13) shows that dy/dx
0, that is, that the flow surface is paranel
to the bottom of the channel; th,is signifies unifonn flow, When Y YG,
~ hydraulic jump or drop in flow profile may occur, as noted. When
Y = Yn = Y" the flow is uniform and criticaL
.
C. Points of Inflection on Flow Profile. When y
0, Eq. (9-13)
seems to produce. an indeterminate form co /00 for dyjdx. However,
1 Some auth.ors l'egard Un as ~ega.tive [9].
Then the Ch.ellY formula. may be used for
explallation. By the Ch.':zy iormlila, K = CAR~' and K' ,= C'A.'R. ThUll, for So'
ilegative va.lue of K', Rand llJnce 1/n must be negative. The value Un hw; also been
\lssumed to be positive [10J. In this case, the Chezy C is tak~n 1M! an imaginary Y!l.lue
in order t.o keep values of A aud.R inY!l.riably positiye.
I
'
225
I Rela.t.io~
I
Deaign:,tioil
Oha.nnel.
P ..QFILES. IN
~.I<lJ'e
PRISMATIC CHANNlliLS
Zone 2
G.n~raJ typO I
of c,u{"ve
Zone 3
TYP8 of fiow
'i
:
Non~
None
Horizontal
S, 0
H2
fill
'1>
Mild
l)
-"12
<1.. < B.
NJ
1.'1
i
Critical
11>
C2
8,-8,>0
8.
81
182
> 8, > 0
'11>
133 .
J.!
None
Adver
A2
!In.
>
y~
yn
>
y.
lin
>u>
y"
">
y,
1i>
Dra.wdown
>y
'I Subcritieal
i SupercrhJcal
1:1
I B ...kwnt;,r
! SubcrlU",,1
Drnwdo\itn
Suberitir.t'l
);. >!I
Bllckwn.ter
Supererltjoal
I/n
Ba.ckwa~er
Sub.ducal
Parallel to
Uniform~
11-
'lJ~
II,
y. > 11
Bll.ckwJlter
Super-critical
1I.
>
Yn
Backwater
Subcrit.icul
y,
>!I
>y.
11
>
'lin> !I
>
II'
1),:;
C3
Steep
> Y>,'J
1'.
i None
:=
(lIA)*
channel
bQttom
I"
.-
en tlcnl
Sup.relitlcal
8 -lpexcri,ti.en.l
1
None
8ubcriticnl
S, <0
A3
11,,~)
'* >
lie
>11
'D,
Supercritical
it can be shown that the theoretical behayior of the flow profile at or near
y = 0 depends on the tYIJe of uniform-flow formula used in the computa,tion. For a wide rectangular channel, Eq. (9-17) represents the slope of
flow profile ii the Manning formula is used. By this equation, it can be
shown that dy/dx becomes infinite. when y = O. This means that the
curve is vertical at the channel bottom. If the Chezy formula is used, it
can be sh~wn that dy/dx == So(y.,,/y.)3 for y = O. This means that the
curve \vill make a certain angle with the bottom. It is apparent that
there shou;ld be a point of inflection pn the flow profile when y: < y. < Vii'
Mathematical investigations by Gunder [11] and Mouret. [12] have
;revealed that this point of inflection is at [l, depth very close to the channel
226
227
Downstream
pooll&>e i
lD
None
Yn
! r---------------,
,'I,
.l 1
-"
---
'
~'0;;W7_';w;wj&a0
Ml
_
2--- -----
FIG.
-----
,r.. ~.:....------_._.
<
II
i
'~
"0.o
E 1c
~
-----_ .. --
r-------
\
_______ "-1..--
Yo
;!11~
l.:-~
.
.
0-W//////////''''''#~
f-----~-------i~--
I~ 7/%!
L~T/lP/7/.i/~
Yo
_ , _ _ _ ..sL-
Vi
bottom.
9-a).1 Similarly, Moui-et found that another point of
inflection exjsts on the flow profile when V"> y" > y, (Fig. 9-3).~
9-4. Classification' of Flow Profiles. 3 For the given discharge and
channel conditions t.he normal-depth and criticahiepth lines divide the
space in aehannel into three Llones:
.
Yh='ic~---
None
"
of
!
1
228
ally varied flow, they are shown with short dashed or dotted lines,
Various flow profiles are discussed below.
A, M P1'ofiles. SQ < Se and Yll >Y.
The -"11 pt'ofile
the most well.. known backwater curve; it is
the most important of all .How profiles from the practical point of view.
ThiS profile oeem's when the downstream end of a long mild channel is
submerged in a l'eseryoir to a; greater depth than the normal depth of the
flow in the channeL This flow profile lies ill zone L The upstream end
of the curve is tangent to the normal-depth line, since dyldx = 0 as
y = y,,; and the downstream end is tangent to the horizontal pool surface,
Rince dy/dx = Sn as y
c:Q.
Typieal examplea of the 1111 profile are the
profile behind a dam in a natural river (Fig, 9-4a) and the profile ill a
canal joining two reservoirs (Fig, 9-4b).
.
An MZ profile occms when the bottom of the channel at the downstream und is submerged in a reservoir toa dept.h less than the normal
depth. The upstream end of the flow profile is tangent to the normaldepth line, sinae dyld:r: = 0 as Y
Yn. If the amount of submergence
at the downstream end is less than the critical depth, the flow profile will
terminate abruptly, with its end tangent to a: vertical line at a depth equaJ
to the criticald~pth, .since dyJdx = <0 for y = y.. This means the.
creation of a hydraulic drop. If the depth of submergence at the down.stream end is
than the critical depth, then as much of the profile
will form as Jies above the water 'surface in the reservoir. Examples are
the profile at the upstream side of a sudden enlargement of a callul cross
Stctioll (Fig. 9-4c) and the profile in a cana'! leading to a reservoir, where
the pool level is shown both above and below the critical-depth line
(F'ig. 9-4d).
The 113 profile starts theoretically from the upstream channel bottom,
at either a vertical-angle slope or all acute angle, depending on the t.ype
of uniform-flow formula used (Art. 9-3), fl,lld terminates with a hydraulic
jump ut the downstream end. This type of profile usually occurs when a
supercritical flow enters a mild channel. The beginning of the profile,
although it cannot be defined precisely by the theory, depends on the
initial velocity of the
water. The higher the velocity, the farther
downstream the profile will begin. 'The theoretical upstream end of the
profile will intersect the channel bottom. At this end y = 0, thus the
velocity would become infinite. Therefore, the theoretical upstream end
of an M3 profile can never exist physically. Examples of the M3 profil.e
are the profile in a stream below a sluice (Fig. 9-4(3) and the profile after
the change in bottom slope from steep to mild (Fig. 9-4f) .
.B. S Profiles. So > S. and y" < Yo_
The Sl profilebegins with a jump' at the upstream and becomes tangent
to the horizontal pool level at the downstream end. Examples are the
~l
229
/"
dy!d
--...._~~
C'_
---
--:~
230
,;1.
THEORY
profiles of flow behind a damin a steep channel (Fig. 9:-4g) and in a steep
canal emptying into a pool of high elevation (Fig. 9-4h).
The 82 profile is a drawdown curve. It is usually very short and rather
like a transition between a hydraulic drop and uniform flow, since it
starts upstieo.m with a vertical slope at the critical depth and is tangent
to the normal-depth line at the downstream end. Examples are the
profiles formed on the downstream side of an enlargement of channel
section. (Fig. 9-41.) and on the steep-slope side as the channel slope
changes from steep to
(Fig.
The S3profile is also of the tra!Jsitional type, formed between an
issuing supercriticoJ flow .and the normal-depth line to which the profile is
tangent. Examples are the profife on the steep-slope side as the channel
slope changes from steep to milder steep (Fig. 9-4k), and that below a
sluic~ with the depth of the entering flow less than the normal depth on a
steep slope
9-4l).
C. C P1ofilcs. So
and y ... = y.
These profiles
the transition conditioi15 between Ai and S
profiles. Assuming a wide rectangular channel,Eq. (9-17) shows that
C1 and C3 pl'ofiles Iwe curved and that the Clprofile is asymptotic to a
horizon tal line (Fig. 9-4m and n). When hha Ghezy fot'mula i~ used J
Eq. (9-18) will show that the two profiles are horizontal lines. The C2
p7'ofile
the case of uniform critical flow.
D. H P1ofili3s. 8 0
0 andy" = 00
. These are the limit.ing cases of ]..f profiles when the channel bottom
becomes horizontal. The H2 o,nd H3 profile3 correspond to the 1lf2.and
11{3 profiles, but no H1 profile can actually be est,ablished, since y ... is
infinite. Examples of H profiles are shown i!l Fig. 9-40' and p.
E. A Profiles. S~ < o.
.'
The A 1 profile is impossible, since the value of y. is not reaL The /12
and A3 profiles are similar to the H2 and H3 profiles, respectively. In
general, A profiles occur infrequently. Examples are shown in Fig.
9-4q and r.
F. Profiles in Conduits with a Gradually Closing Tflp. .For nny conduit
with a gradually closing top, the normal discharge will increase as the
depth of flow increases. It will increase first to the value of full discharge
at a depth yo' less' than the full depth yo. Thereafter, the discharO'e
will reach a m!)'ximum value Q,m at a depth Yn *. Further increase
depth of flow will decrease the discharge eventually to the full discharge
at the moment when the flow surface touches the top of the conduit.
9-5a .shows the variation of normal discharge in such a conduit.
fll the case of a circular conduit, the depth Yo'"" O.82yo and y" * = O.938yo
where yo is the diumeter of the conduit (Art. 6-4). Within the region of
IJ
ya' and y = Yo, there a.re two possible normal depths for a given dis- .
charge, namely, the lower normal depth y" and the upper or conjugate
A~D
ANALYSm
231
in
these figures .. It shuuld be noted that the critical depth in Fig. 9-5d is .
greater than .the normal. depths y,.' and Y.., but that its corresponding
lower normal depth is less than Yfo' and Yn. Consequently, the corre, sponding critical slope should be less th!l.l1 the normal slope, and the
channel slope is considered mild.
232
~l
I
(a)
~l
~--_.'_.::::=:-
____ _
M;ld slOn
.
~e
~----~-'--'-'--"
,
233
,
em)
--~
///~~~~-~~~-:.
-"....
CI
;
... ~------....
Cr'
JlicG/
Milder slope
~
(q} ...~
-"-.
Slo,oe
'!
..
- - -.
-~.--~- .:':'.:::.:..-::
.... -- -~ ..--~"-.----~
~'
'::""~~\\':~
Mild. slope
(n)
(s)
-.:....
(P)
~.
___ ."_-/:.~_.~~~~~~i.~~.
Critic
Depending
MI
,._:~mp
Dependinc;; on
downstream
~
(f) ---'---:::::.-_ _----:.----
(tJ
t2.
'lIl
downstrecm
con~?:
~
P ./oPe
FIG. 9-6. Profiles of gradually varied flow in a long prismatic channel with
bottom slope.
C"\
'\:\
..
Adverse. slope
LEGEND
'0
1171777Tmn,.,!._
I " 1/
FIG. 96 (Continued).
:3
,~"
,se s\O?e
(>.a~e
01 s~
/.
OPe
",,~y
~".--"';'1:.~.:;:- .. ---- Jump
Sfs e '
Depending on
aownst ream
Cl'
11
hreak in'
234
235
4-5). If the downstream water surface is very high, it may raise the flow
surface at the upstream control. When several st.eep reaches occur in
succession, the control section is at the upstream end of the uppermost
reach. Upstream control also occurs in long mild reaches, because the
lVIl or lIlZ curves will approach the normal depth at the upstream end.
b. DOWNSTREAM CONTROL SECTION. This occurs at the downstream
end in any long steep reach, because the flow will approach th.e normal
dep~h at the downstream end. If the downstream end of a mild channel
terminates at a free overfall, the control Section 'may be assumed at the
brink where the depth is crit.icaL r
C. ARTIFICIAL CONTROL SECTION.
This occms at a control structure,
sur.h as a v;eil', dam, or sluice gate, at which the control depth either is
known or can be determined.
5. Starting D,t the control depth at each control section, trace in each
reach a continuous profile. The position of the profile in each reach can
be corre~tly located with respect to the normal~ and critical-depth lines.
For this purpose, typical profiles described previously (Art. 94) should
be found useful.
B. When flo'''' is super critical in the upsi:.ream portion of a reach but
subcriticaI in the c1ownst,ream portion, the flow profile has to pass the
critical depth soniewhere in the reach. In crossing the critical-depth line,
a hydraulic jump is usually crsl:'.ted in raising the wat,et surfa<;e from a low
depth to its
depth. The exact location of the jump will be discussed later (Art. 15-7).
,I
"""""ntu,,, 9-1. Th":l normal and critical depths of the flow in a channel have been
W'"'!,,'""_'" !Lnd ~,re shown in Fig. 9-7,
1 Also depending on
relative roughness and shape of the two connecting chan-'
nels. In this discussion
factors arc assumed constant.
The term "control section" used here has a. broad meaning. It refers to any'
section at which the depth of flow is known or can be controlled to !l. required sta.ge.
1 Actua.lIy
-s.:~
(Ar~.
3-4).
236
a.
.a
M2
--- ...
a.
E
_-
.:!.
Mild slope
Critical
control sedion
U.C.
upstream control;
Critlcol
control section
I [;ubCritical flow
I'.
~
~t
Critical'
!S b control section
[s Upercrj~.
flow
1------- ------l
,
critlca/
flow
f',O,
CrJ t i co I
'C;:~~~".-L
conirol section Critlcol
1ICoi
.~~
Critical
control section
Steep
"
dy
dx
leD'.~_:~_
l Superr:rir'
flo", cal
I
237
---- .......
or
(9~20)
~,
(9-21)
(a)
(b)
FI~. 9-8. Location of critica.l c'ontrot section. (a) Flow in a prismatic channel;
(b1 flow ill a nOllprismatic channel or spatially varied flow.
and
The solution of Eq, (9-21) Fi = 0 will giveQ = Q" or y = y".
(9-22)
Hence,
1 At this ~oment, the normal depth coincides with the critical depth, bec.oming a
sO-C!tlled transitional depth (Art. 9-6).
2 See another method of anillysis by Merten [20,211.
nonprif:lmatic channels, Fl
0 I'esults in a fictitious normal-flow profile.
It is fictitious because U1~iform flow in nonprismatic channels is unrealistic.
Similarly, Eq. (9-22), P 2
0, represents the critical-flow profile. Since,
Q< is independent of the channel slope, the concept of the critical-flow
profile is valid in channels, of variable slope.
Yl
,C
'j
I:
---~--l-...
_______
'"
"
. ~I
. ~~.-1- -~--;:::;:.::;:::;_.~ .:__:,.'
..
@.,-'-- -----"-'1'-""
'/ j /
;'I.L
,
tLLZ
'
'
\'\
,*""".". ' /
~;,'
If'
If"
' "
"
1/,'
/,
l ,
"
' I 't
I
F,O
)
(d
dx ,
profile
---Transitional profile
------otherf!ow profiles
3:,
r
II
8.
(9-24)
(9-23)
239
qtw,si-norma! flow.
to uniform ch!l.nnel~,
I
I
(9-25)
,1
\
~I
240
108 0 (d Y)
dx ,
1J: (~o)
3
dx
= 0
I
I
(9-26)
This equation has two roots opposite in sign. The negative root will
define a flow profile
through the singular point P, wi.th depth
decreasing in the direction of flow. Since the r_ow in the present probJeCljl
changes frorn. 8ubcl'itical tosupercl'itical; this flow profile is evidently the
profile of the flow under consideration. At x =
this profile is
ttsymptotic to the profile of the quasi-normal flow, or FI = O. The other
root of Eq, (9-26) "till define a. flow profile also passing through P but not
asymptoti'c to 1\ = O. Other profiles, which do not pass through
will intersect the transitlona.! profile with a horizontal slope, pass through
Pi = 0 with a hydraulic jump, and become asymptotic to the two profiles
that pass through P.
The flow profile
through P but not asymptotic to PI =0 is
not the profile of the flow under consideration. However, such a
profile may become real under some other circumstances. In
9-10,
C() ,
241
Iu3 these operations are carried out, the jump GH is forced upst.ream until
THEORY AND ANALYSIS
. Saddle
Spiroi
: I This exa.mple wiw:! suggested .by Mr, F. F. Escoffier of the U.S. Army Corps of
Engineers. For simplicity, the.curv6 for Fl = 0 is not shown)n Fig. 9-10.
Nodal
Vort
FIG .. 9-11. Flow profiles around El. .~ingula.r point, P =0 singular point; F
flow
profile passing through P and asymptotic, to F, == 0; F' = flow profile passing throl.\gh
P but not asymptotic to F ... 0; f ... o~her flow profiles; F, = quasi-norma.l flow
profile; F.= oritical-flow profile; and T ... llrallsitiomu profile.
,
I
I
~
where the profiles are plotted with the ordinate Y representing the distance
above a horizontal datum and the abscissa X representing the distance
along the "datum.
.
. The p0s4ble flow profiles pre$ent~d in the example of Fig, 9-9 are
of the saddle type. In this case, the flow profile that passe~ through
P and is a~ymptotic to FI 0 indicat 9ll eo continuous How: changing
I
242
,
l
Ii
243
1
,
J
.J
II
I
I
-' (
(oj
_---"1 -=-..--T,'- __ _
..
"--
1
-..
- __ a
(9-27)
~
= n~gD
Q/VU, and
"
A,fIT.
'
//,
UncriticO/.S/o pe
(9-28)
, (d)
fi---'
PAl
Here .the 'upstream normal flow changes to tbe downstream normal flow at an
abrupt tr'ansition formed by a hydraul~{l jump.
, See [f2], [II)], [28] to [30]. and [32] .
, I
1:
.1,
\
)
0,,)
244
\"j
I"
t'
i.
(Fig. 9-12b), tIlen So < Se and Ye < y" < Ya < Vb. Since
So < Sc, the flow is sub critical, and the profile should be of the 1VI1 typ~ .
However, the profile will contain two points Ta. and Tb at which the slope
is horizont.aL Bet-';veen these two points a p(')int of inflection apparently
exists. The depths at the two poihts are transitional depths Ya. and y~.
If Qa < Q < Qb (Fig. 9-12c), then So > S,and y. < Yn -< y. < Yb.
Since So > Sc, the flow profiles are of the S type. However, there will
be a point Tb where the slope is horizontal on the Sl profile and a point
T a where the slope is horizontal on the 83 profile.
If Q. < Qb '< Q (Fig. 9-12d), then So < S, and Y. <Yb < Yo < Yn.
The transitional depths will intersect only the M3 profile, and the backwater profile will be of the ordinary M1 type.
B. When the slope is fo.r r;.way from the limit slope. In ihis case, Q.
and Ya are relatively very smail, and their existence is practically insignif]cant. If Q > Qb, the flow will be sub critical and the profile will be
of the M 1 type. If Q < Qb, the flow will be supercritical and the profile
.will be of the S1 type.
C. When Ihe. slope is very large. In this case, the large transitional
discharge Qb is considered to exceed the maximum expected discharge
(see Fig. 6-8). Thus, the flow is supercritical and the profile is of the
S1 type. The highest point of theS1 profile is very close to the downstream end.
The above discussion was developed for the case in which the point L of
the limit slope is below the curve of maximum expected discharge (Fig.
6-8a) and in which the channel sections !1re rectangulal' or trapezoidal or
similar to such forms. If the point L is above tbe curve of maximum
expected discharge (Fig. 6-8b), the larger l;ransitional depth of the flow
will be greater than the maximum expected depth, or Yb > Ym, and the
larger transi~ional discharge will be greater than the maximum expeeted
discharge, or Qb > Qm. The foregoing discussion, however, remains
valid as long as the actual discharge Q does not exceed Qm. If Q exceeds
Qm, the discussion has no practical meaning. Similariy, the flow profiles
remain t,he same, but the useful part of the profiles will be where the
depths al'e less than y",.
If Q
."
.1
'",,
< Q.
LEGEND:
-------~-
EIJ272
:s2
EI.f270
--x-jEU266
..
(""I'/;":~:;::(
.-+----j.500'----~.t<I.---
[=
I
''''11'//''7777,,---,1
I
)
EI.1274
9-1. Show that the wnter-surface slope S" of a gradually varied flow is equaJ to the
slim of the energy slope S and the slope d~e to velocity change d(", V'/2g)/dx.
9 c 2. Show that the gradually-varied-fiow equation is'reduced to a uniform-flow
formula if du/dx ... 0,
9-3. Prove Eq.' (914).
9-4. Prove Eq.: (9-15),
9-5. Prove Eq. (9-16).
9-6. Prove Eqs. (9-17) and (9-18).
Criticol-c\eplh line
- - - -Noqnal-deplh line
PROBLEMS
245
9-7. f?ketch the possible flow profiles in the channels sbown in Fig. 9-13.
,:1
"
246
GRADUALLY VARIED
~'LO"'Y
S. - Sf
+ (",Q'ylgA~) (dbldz)
1 - ",Q'b/gAJ
...._........
(9-29)
REFERENCES
1. J. B. Belanger:
sur Ill. solutionnumiirique de quelques problemes relatifs
au mouvement. permanent des eallli: courantes" ("Essay on the Numerical Solution of Some Problems Reletiv~ to the Steady Flow of Water"), Cn.rilian-Goe\lry,
Pa.ris, 1828.
.
2. J. A. Ch. Bresse: "'Cours de mecJl.nique appliquee," 2e po.rtie, Hydraulique
("Course in Applied l\'Iechanics," pt. 2, Hydraulics), Ma.1let-Bachelier, Paris,
1860.
3. Boris A. Bikhmeteff: "Hydraulics 'of Open Channels," appendix I, Historicnl and
bibliographical notes, McGraw-Hill Book Cr)mpany, Inc., 'New York, 1932,
pp.299-301.
4~ Charles Jaeger: Steady flow ill open channels: The problem of Boussinesq,
J Ot,rnal, lnetitution of Civil Engineers,. London, vol. 29-S0,.pp. 338-348, November,
.
1947-0ctober, 1948.
5. Charles Jaeger: "Engineering Fluid M.echanics," translated from ~he Germa.n lly
P. O. Wolf, Blackie & Son, Ltd., Glasgow, H,56, pp. 93-97.
6. F. Bettes: Non-uniform flow in channels, Civil Engineering and Public Works
Review, London, vol. 52, no. 609, pp. 323-324, March, 1957; no. 610, pp. 434-436,
April, 1957.
.
7.(-Allen: Streamline and turbulent flow in open channels, The.Lond(}tl, Edinburgh
and Dublin Philosophical Magazine and Journal
Science, ser. 7, vol. 17, pp.
1081-1112, June, 1934.
S. Hunter Rouse and Merit P. White: Discussion on Varied flow in open channels
of adverse slope, by Arthur E. Ma.tzke,. Trallsadion.s, American Society of Civil
Engineers, voL 102, pp. 671-676, 1937.
9. ShermanM. Woodwa.rd and Chesley J. Posey: "Hydraulics of Steady Flow in
Open Channels," John Wiley & Sons, Inc., New York, 1941, p. 70.
10. Ivan M. Nelidov; Discussion on S~face curves for steady nonuniform flow, by
Robert B. Jansen, Tra.nsa.ctions, American Society of Civil Engineers, vol. 117, pp.
1098~1l02, 1952.
.
0;
',~-
247
11. Dwight F. Gunder: Profile curves for open-chruloel flow, Transadions, American
Socieey IJj Ci.-il Engineer", vol: lOB, pp. 481-488, 1943.
12. G. Mouret: "Hydraulique: Cours de rneca.nique appliquee" (It HydJ:aulies:
Course tn Applied Mechanics "), L'Ecole Nationalc des Ponts et Chaussees, Paris,
1922-1923, pp. 447-458; revised lIB "Hydraulique giinerale" ("General Hydraulics"), cours de 1'Ecole NatiQnii.le des Ponts et CtwLusseesi Paris, 1927-1928.
13. A. Merten: Recherches 'sur Ill. forme des axes hydra.uHques dans un !it prisma.tique
(Studies on the form of fiow profiles ill. a. prismatic channel), Anna!e.~ de jl.:l18socic:Ucm de. In(16nieuTS Bortis des ~cole8 8pec.iales de Gand, Ghent, Belgium, '101.5, ser. 3,
1906.
14. M. 'Boudin: De I'axe hydraulique des coms d'eau contenus dans un lit prisma.tique
et des dispositifs reaiisll.nt, en p~!l.tique, Bes formes diverses (The flow profiles of
w.ater ill a prisma.tic ch!1nnel' and actuI'1.1 "dispositions ill various forms), A nnales
des trcwaux pu.bliqucs de Belgique, Bru~s!'<l!;, vol. 20, pp. 397-555, 1861-1802.
15. Pierre Koch: JuStification de l'etucle rationnelle du remous dans les aqueducs de
forme circulaire, ovoide <iu IIna.logue (Justification of the ra.tional study of backwater in oinmln.r conduits of ovoid or similar shape), Annall13 d.s P(}tl/s e/ cJlIluGs,les,
pp. 153-202, Sep~ember-October, 19~3.
.
16. L. Gherardelli: SuU'equaziOlle del mota permanente ill alvei prismatici (On the
cqua.tion for steady flow in prismatic channels), L' Enerllia eleUrioo, Milano, vol
28, no. 4, pp. 185-189,April, 19tH.
17. Gianni Formica: Nota sui profili di dgurgito delle correnti permanent! gradualmente variate defluerrti in galleria cilindriche (Not.e on ba,ckwater curves of
gradually v1l.riedsteady flow in cylilldrico.! closed conduits), L'Energia eleltrica,
Milano, vol. 29, no. 8, pp. 480-491; August, 1952; repdnted as Istiluto di Idraulica
e CostrU~ioni idrauliche, ilfi!ano, Mp'lrIorie II studi No. 97, 14 pp.
18. R. Silber: Sur Ill. forme deJl courbes des remous en gal erie COl~verte (On backwater
curves in closed condui.t), cxtrait des Comples rendu.s des s~ance$ de I' Aoo.demie
.des Sciences, vol. 235,pp. 2377-2379, June 22, 1953.
19. R. Silb~r: ":fl]tude et truce des ecoulements permanents en canaux et rivieres"
("Study and Outline of Steady Flow in Open Chnnne~s"), Dunod, Paris, 1954.
20. A. Merten: Thoof<lmes fondamentaux d'hydrnulique fiuvia.1e (Fundamenta.l
theorems of fluvial hYdraulics), Annales ,de I' A8sociation des Inglnieurs sOTtis
des Ecofes Speciales de Gand, Ghent, Belgium, ser. 5, vol. 5, no. 2, pp. 200-210,
1912.
21. L. J. Tison: "Cou:s d'hydrauliquc" ("Hydraulics"), Uriversite de Gand,
Ghent, Belgium, 1953, pt, n, pp. 170-182;
22. H. Poincare: Memoire sur les courbes definies par uno equation differentielle
(Memoir on the curves defined by a differentia.l equation), Journal de malMmatiqul!3 pures et appliq1J.~es, Paris, vol. 7, pp. 375-422, 188L
23. Pierre Masse: Ressaut et ligne d'eau dans les cours d'eau apente.variable (Hydraulic jump and flow profile in channels (If variable slope), Revue gencrale de l'hydrau!ique, Paris, vol. 4, no. 19, pp. 1-11, January, and no. 20, pp. 61-64, April, 1938.
24. 'Charles J!1.eger: Erweiterung der Boussinesqschen Theorie del! Ahlliisses iu offenen
Gerinnen und der Abflilsse ilber abgerundete Wehre (Extension of the Boussinesq
theory of flow in open channels and over a round-crested weir), Wasserkraft und
W~serwirischafl, Munich, vol. 35, no. 4, pp. 83-86, Apr. 15, 1940.
25. Giulio De Marchi: Sui cambiamento di regime di una correnti Iineare a pelo
libero in Ull alvea. di sezione costn.nte (On the transition beLweell 5upercritical and
sub critical conditions in free-surface gradually varied flpw in a cylindrical che.nnel),
L'Enerflia e!ettrica, Milano, vol. 27, no. 0, pp.125-132, March, 1950; reprinted as
Istiluto di Idrau!ica 8 COllt1"urioni ldrrJuliche, Milano, Memorie es!udi No. 82, 1950.
......
'
.~
248
!
I
r
\
CHAPTER
METHODS OF COMPUTATION
, 1
10
['". dx =.lY'dX-dy dy
-,'"
y.
(10-1)
o~~-~~~~~~~~~----~Y
of
249
.'-'"
250
1
in
T<l.BLE
10-1.
10-1
251
MlllTaODS OF COMPUTATION
,I
290.2.
1 - (74.0/290.2)1 . = 760
1 - (10,000/20,800)'
\,
"OOT
\ .
"ot
6,000r
Br-
GJiAPRICAL INTEGRATION
400 cfs
5.00
40.00
ag.20
38.40
37.60 \
36.80'
36.00 i
35.20 i
34.801
34.40
34.20
! 34.00
1 33 ,88
33.76
4.80
4.60
4.40
4.20
4.00
3.80
3.70
3.60
3.55
3.50
3.47
3.4\1,
3.42
3.40
3,36
So = 0.0016
= 0.025
2.22 ft
l(
AA
d:t:/dy
0/
= 1.10
150.00
142.08
134.32
125.72
119.28
112.00
104.88
101.38
S7.92
96..21
\14.50
93.48
92.45
33.68. gl.aO
91.12
33.60
33.44
89.78
3.54
3.43
3.31
3.19
3.08
2.96
2.84
2.77
2.71
2.68
2.65
2.63
2.61
2.60
2.59
2.56
T
40.00 ft
A "" 150.00 fP
R '" 3.54'ft
2.323
760
79:! 155
155
836 1(;3
318
913 ' 175
493
1,000 191
684
1,140 214
8913
1,430 257 1,155
1,750 l!,\} 1,314
2,260 201 1,515
2,770 126 1,64.1
3.480 /156 1,797
4,520 120. 1,917
5,990 i 158 2,075
7,930 139 2,214
10,760 187 2,401
"
5.000[-
rm
R~;,
y,
I
I
d,
d)'
4.000~
i
'MO~
I
200i
I
1 I
I II
',OOor
. 151011
IINI::!I
11 II
I I I
(\J
t<i
<.0
r'l
",I
t:::1 ~I ~!
4
I
I
..
5 Y
r<i
\
1
;:;.,
252
10-2.
T,uiLE
Ye6.r of
puhll-
Type 01
cba"nel
InvOBtigo.tor
catioo
1848
Dupui.
1850
Bresse
1875
Gr... hof
l!;il!(!
R1iblm~nn
1896
1900
1912
(1932)
1914
1S
Broad'tee.
tangle
Broad rso .
tangle
Br?ad ree-
Effect or
change iu
kinetic
energy
1928
1930
1938
1947
1~51}
1954
Veloeity
formula
N - 3, M = 3
JV"", 3. fyf = 3
i2J
Con.id,..."d
ebhy
N = 3, ftf - 3
[31
N-3,M=3
l~]
22'.1~'1I:uSil>lJ
Ignored
CMzy
tiJ
Ohby
eM.y
'"ngl.
Approximate
trApezoid
KotallY
Refer-'cnce
Ignored
Cb~.y
Ignored
Brond ,...
tan;:le
Tolkrult~
Considered Cilhy
I Broad para
bola
Masoni
Common Tee"' C"".idered CM.y
'"ng)e
Bakhmetefl
All eh"pea
Con.idero.d CMz:!,
by.teps
C1l{)nBo.~
Senafl'nnak. Broad reo19:nored
23.7~'1IFJ.TlIS~~HI!J
Ehrenberger
~angle
BIltic16
Oon.idered
1921
It should be noted tllll.t, when the depth approaches the no~mal depth, the incremental area varies so greatly with the change in y value that it becomes difficult to
pirmimeter. In such Il. c.tl.Se, the area. may be computed by assuming it is a trapezoid.
For instnnce, the incremental area between y '" 3.42 a.nd 3.40 is t,,1 - (7,930 +
10,7(0)(3.'12
3.40)/2
187.
.
253 ..
METHODS OF COMPUTATION
N = 4, M - 4
lSI
N=a.1I1=3
(5]
.It:; c: UH
Ii,S)
N ~ 3.5. M = II
N = 3.552. M n= 3
N = 3. HI, M: = 3
[{:t .- ol"
A,.1R. where
- is a varilll}le and
dll/dz = eonet
N - 3.;1. 1>f -.8
I~J
1111
N=2+2m.~1-3
(12]
0:::
UI;OD"\ A.!c
:rv~
K:
tt
K.:
ct lI'~1
None
tt
tfM:t
AI
cc lI CO I.l,at
Z2
1%
1JJl
[IOJ
lIS1
[14.15]
[16]
(17)
the integration is carried out by short-range steps and with the aid of a
varied-flow function.
In an attempt to improve Bakhmeteff's method, Mononobe [13] introduced two asllumptions for hydraulic exponents. By these assumptions
the effects 9f velocity change and friction head are taken into account'
integrally without the necessity of dividing the channel length 'into short
reaches. Thus, the Mononobe method affords a more direct and accurate
computation procedure wherebyresuH.s, can be obtained without recourse.
to
in kinetic energy
the friction slope, or r in Eq. (9-14), i3 a.ssUI;ned cOMbnt in each
rea.ch. Since an increa.se or decrease in depth will cha.nge both these fa.ctors in the
sa.me direction, their ratio is relatively stable a.nd can-be IloBsumed constant for pralltical purpose!!.
254
~ ~: [ 1 -
ltN
(10-3)
This equation can be integrated for the length ,; of the fio,v profile. Since
the change in depth of a. gradually varied flow is generaIJy small, the
hydraulic exponents may be assumed constant within the range of the
limits of integration. II:_. ?-~I1~e ..U!.e _hY'dra~lic eXQonentL1l<re
not}ceably deE~ld}nt O}l ~jn the limits _Q,f ~l.vell rea:.c::h, the reach
shoul<i Qe....l!.b,dividlill. foy )ntegrf),ti::m; tl:!.el} .th~ J.IYcir al1lic e~p.o12~!lls Ln_
e~gl:j~e~J~~9_h may be assu~~consta.nt. IntegratingEq. (10-3),
(10-4)
The first integral on the right side of the abo,re equation is designated by
F(u,N), or
'
F(u,N) =
fu ~
1 -:- UN
(10-5)
)0
fv.
)0
U N- M
1 -
.'
UN du
J f'
N )0
dv
1 - vJ =
.J
N F,v,J)
( 10-6)
)0
(10-7)
This is a varied-flow function like F(u,N), except that the variables u and
N are ;'eplaced by v and J, respectively.l
. Using the notation for va,ried-flow functions, Eq. (10-'1) may be writt.en
or
where
=~:,
[ 11 -
F(u,N)
A[u - F(u,N)
_ (y,)M
J
-,
B -
y"N
,
'M J
1
+ G:)
N Fev,J). + canst
+ BF(u,J)] +
1t =
't,
= x~
I.
(l0-8)
( 10-9)
const
J
Yn.
. .-
N - M+ 1
where
Let u
255
METHODS OF COMPUTATION
. 'j
Xl
+ B[F(U2,J)
- FCli"J)] I
(10-10)
-I
._.
./
.. ''1'.--
L
,
~.,,-..:~ In computing, a flow profile, f,!rst the flow in the channel is analyzed
J
I
,
l
\
Example 10-3.
(Art. 9-5), and tho channel ,is divided into a numherofreaches. Then,
the length of each reach if! computed, by Eq. (10-10) from knmV'll or
assumed \lepths 8.tthe ends of the reach. Tile procedure of computation
is as follows:
'1. Compute the normal depth y" and critical ,depth Yo from the given
data Q and So (see Arts. 6-6 and 4-4).
'
'
2. Determine the hydraulic exponents Nand M for an estimated
average depth of flow in' th~ reach uuder considerl1tion (see Arts. ~3 and
4-3). It is assumed that the channel sectivn under consideration has
approximately constant hydraulic exponents.
3. Compute J by J = N/(N - M + 1).
4. Computc values of U = Y/Vn and ,1> = uN/J at two end sections of
the reach.
5. From the varied-flow-function table in Appendix D, find values of
F(u,N) a.nd F(1),J),
,6. Compute the length of the reach by Eq. (10-10);
The above procedure is illustrated by the following examples:
.
'U
l.488
5.00
3.40
DiL ...
},012 :
0.476
I
I
v :
1 . 0,1.5
}i'(''',N)
1 025
-0.B77
1'-
Q =,400 cis n
FI..OW
O.O~- 0.0036
1_
y
-'-~!
:
10-3 BY TIiE
'" - 1.1Q
y;
2.22
2.14
1.B7
1.60
1.33
1.07
0.80
0.53
0,27
0.00
11
F(1_,N)
, F(v,J)
L
~--
0.831
0.800
0,700
0.6')0
0,500
0.400
0.300
0.200
0.100
0.000
0.792
0.755
0.638
0.525
0.420
0.;315
0,219
0.132
0.055
0.000
0.979
0.917
' 0.756
0.627
0.511
U.404
O.3D1
0.200
0.100
0.000
0.962
0.888
0.699
0.552
0.431
0.319
0.219
0.132
0.055
0.000
20:6
20'40
188
161
134
102
71
IB
45
72
10i
135
i63
1BB
200
43
18
0
to dischllrge 400 cfs with a depth equal to 0.55 It at the vena contr~ta: C~mpute the
flow profile. If a hydraulic i ump occurs at the downstream end, suartmg Wlt~ a depth
of 1.6 ft, determine the di,stancl; 'from the vena contro.cta to the foot, of the Jump,
SQlulion. From the given data, li. = 2.67 ft and y. = 2.22 ft .. Smce y.. > y" the
channelsiope is mild. As the depth of flow issUing from the slUIce !!;ate ls.1ess than
the critical depth, the flow profile is of the M3 type. .
T '
Considering aflilverage depth of 1.61 Ct, the hydrauhc exponents are I\ 0= 3.43 and
M = 3.17. Thus, J .. 2.72, N IJ = 1.26, lind (Ycly~)MJ IN = 0,442,. ' : .
Table 10-3 'shows t.he computation of the flow profile. For convenUl[lce m mter-
- ... - .... -
F(v,J)
[.293
-1,080
+ 0.Hl7
TliE
----
5, The .varied-flow funct.ions k(lI,N) und F(v,J) are obtainbd from the table in
Ap~(lndix D ancl given in the fourth and fifth colmnns of the o.~ove table.
6, In Eq. (l0-9), A = y,./So =: 2,100 and B = (voly,,)MJ IN F 0.197. The length
of the backwater profile is" thereCbre,'
i
.
L = 2,100[0.476 - (-+0.877)
DIRECT-INTEGRATION MET,BOD
Example'1O-2. ,i,rith reference to the chs,1l11EJI described ill Example 10-1, compute
the length oC the backwater profile extending frorn the dam site to an upstream .'3BCtion where the depth of flow is 1 % grea.ter than the normal depth.
Sohaion. The given data are Q = 400 cis, b = 20 ft, Z = 2, So = 0.0016, a = LID,
and. n = 0.025.
,
1. FOlIo'wing Example 6-2, y,.= 3.36 to Following Exa.mple 4-2 with a = 1.10,
y, = 2.22 ft.
i
'2. The depth at the downstream end of the backwater profile is Y2 =,5 ft: At the
upstream end, the depth is Yl = 1.01 X 3.36 = 3.40 ft. The average depth may be
to.ken as 4.20 ft'3.nd ylb = 0.21. From Figs. 6-2 and 4-2, the corresponding hydraulic
exponents are fonnd to be N = 3.65 and 'llf = 3.43,
3. The vnJue of J = 3,65/(3~65 - 3.43' + 1) = 2.90.
4. For each section, values of H and v are computed, as given in the second and third
columns of the following table:
'Y
257
METHODS OF COMPUTATiON
256
258
x =
(1 "-:,::)
tI -
+ const
1"(u,3) ]
F(u,3) = )0
_ 1-:1'
du
-113 -
7,
11'
+ g +1
(u - 1)'
_ _1_
.H,2l~
V' cOu
As the depth of flow varies from 5.00 to 3,4,0 ft, an average value of'y = 4.20 ft
may be assumed for the evaluation of Chezy's C. For a wide' rectnngular channel,
Eq. (5-7) gives C = 75.13. Since y. = 3.36 ft a.nd So = 0.00113, A = 2,100 and B =
0.715, Eq. (10-17) gives
.
x == 2,100[u - 0.7151"(n,3)] + const
(10-19)
Assuming a constant of zero in Eq. (10-19), the computation of x is as follows:
+1
-Y3
I F(t':3~1
-------I---~---I
5.00
1.488 ) 0.260 \2,720
3.40'
1.012 t 1. 360 .
80
---'----
The length of thB backwater curve is, therefore, equal to 2,720 - 80 = 2,640 ft,
about 10 % larger than the values determined by the previoLls methods.
.
. (10-11)
(u
259
METHODS OF COMPUTATION
Mathe-
dy
dx = 1 +
(10-12)
-S~
0'.
- s
(10-20)
d(V2j2g)/dy
'\J-g
F(u,N)_s,
and
Thus,
G'So
or
~0 T
u
L
x
,.f[u - B1"(tt,3)]
UN
+ const
(10-16)
(1U-17)
Example 1.0-6.
(10-21)
F(v,J)_s,
j.~
a u lf - N
}o 1 + uN du
d1t
+ UN
J
(10-22)
r
v
d!,'
N }o 1 -+ vJ
(10-23)
where v
u,Y!f and J = N/ (N'- 111 + 1). --For evaluating these functions a t.able [18jl has. been prepared which is Table D-2 in Appendix D.
Accordingly, the length of flow profile betwe~n two sectiqns 1 and 2 may
beexpressed by Eq, (10-10), where A = -'y,,/So, B = -(y,jYn)MJjN,
and the varied-flow functions are replaced by those for' adverseslo pes.
(10-15)
= -g-
+ 1tN
ra
(10-14)
Yn = "C'So
Ye
(l
N-MduJ + const
(y,)M
r"u
Yr.,}o 1 +
(10-1:3)
.317
[~ - j(U ~ _
31qi
Ye =
and
x = - y"
S~
(10-18)
c7t/.
'- 1
dy
d2
-(Q/K)'
(Z,/Z)'
= 1 -
is
(9-19)
Since the critical slope S,.is defined as the 'slope that will produce a discharge Q :at
Il.
.!
I Tables of varieq.-flow functions for adverse slopes. with limited range af the
hydraulic exponent N hiwe also been prepai'ed by Matzke [20J and others [2IJ.
!.,
\
\
260
for the range of Yl to Y2; let Me be the M vaille at the critical depth Ye;
and let M be the average M value for tpe range Yl to Y7. Thus, Eq.
normal
depth equal ' to the critical. depth !I' (Art 6-7) , the d'ISC1~arge may be
' expressed
,
!IS
VB,
K,
' (10-24)
Substituting Eq. (10-24) i(l Eq. (9-19) and letting (K.jK)' = ( /)1'1 (Z,/Z)'
(yJy)Dl, and p = y/y"
. " . y, y ,
~:!! '" S
dx
pM-N ,
(10-25)
'1 - pM
berBduced to
pN-.lf+1
N _ M
'
+1 -
pH+-\)
+ 1 + ~onst
(10-26)
dx
This equation can be llsed for the computation of the length of fJ/)v
fil'
h'
zontal cha.nneL
' .
~ , pro e In a Orl-
It sh ou~ld, b e llotec1 t h at
, the assumptIOn of COllstnnt
hvdraulic
exponent<:
0'
d'ISCUS.
J
.
... ~ l'n I-he
v
legomg
N
y N"IN [ 1. -
x = ~
'"
log K
N"2tonli
18 ,.,/
------~:-~
I I I
I
I
I
!
I
I
I
,.,"
".
2
!I
I I I
I
I
"'''''''
0>".
0>
!2E
log Y
1
1t~r-,I!
1 ___
_,
-;- (y/1'IM)M
-- - du
'
1 - UN
y"N .. !N
1 - 'UN j
(10-28)
F(u,N)
J
'] ,
+ (y,M'IM)H
~>.fN N F(v,J) + c.onst
(10-29)
(108).
The length of the flow profile between two consecutive sections of
depths Yl and Yz can be computed by Eq. (10-10), except where A
YnN.fN ISo and B = (YeM,IM IYnN.IN)!':J IN.
ChanrLeis with Gradually Closing Crown. For channels with gradually
closing ero\vn, the hydraulic exponents are variable near the crown, and
the m'ethod proposed r.bove may be used. For more accurat_e results,
however, the integration of the dynamic equation maybe performed by
a procedure of numericp.l integration. Such a procedure has been
applied to circular c,onduits hy Keifer and Chu [17].
Let Qo be the discharge of a circular conduit, flowing full at a depth
equal to the diameter do of the conduit and having the energy gradient
equal to the bottom slope So, and let Ko be the corresponding conveyance.
I I I
log Kn
I I I
,8
el
y,/J,IN duo
r I I
log Ze
__ y"N"IN
__
So
Channels with
l'a'I'iablc Hydrauli-vETponen'"
.
".
261
METHODS OF COMPUTATION
'-----.l.:-c--J_--LL"'>.
.>..>.'
01
. ..2
log y
2. .2 52
Thus,
(10-30)
i
1
263
METHODS OF COMPUTATION
where (ICo/KF is evidently a function of y/d o and, hence, can be represented by fl(u/do).
Ji'r"omEqs. (\)-4) and (9-7), the following may be written:
262
0:(.221'
ZC)2
= 'qA3
(Z
T /d o "
aQ2
aQ! (Y)
-([;Sf. C4
(10-3.'3)
where (1'/ d o)/ g(A/d 0 2 ) Sis appa,rently afunction of y/d o and, hence, can be
represented by hey/do).
Substituting Eqs. (10-32) and (10-33) in Eq. (9-13) and simplifying,
(10-40)
(10-34)
(10-41)
dey/do)
1 - (Q/QoFh(y/d o)
x _ do [ (v/do
- So}o
aQ2 (v/e.,
-~)o
or
Ia(y/do) d(u/d o) ]
1 - (Q/QoFfl(y/d o)
.(x _aQ:~
y.) +"const
do'
(y Q) .(Ii/d. -dey/do)
X
FI do' Qo =}o
1 - (Q/QoFh(y/d o)
Y = F? (X, R) =
-f.(u/do) dey/do}
- do Qo
)0 1 - (Q/Qo)2!t(y/uo)
x = - do
So
where
and
+ const
."
(v/d
(10-35)
V2
E=y+cr.2q
(10-36)
(10-42)
fOl"
the deri-
. (10-37)
(10-38)
(9-8)
These are the varied-flow functions for circular conduits, depending all
y/d o and Q/Qo. They can be evaluated by a procedure of numerical
integration, say Simpson's rule. A table of these functions for positive
slopes, I prepared by Keifer and Chu, is given in Appelldix E.
The length of flow profile between two consecutive section3. of depth "Vl
and Ya, respectively, in a circular conduit may be expressed as
(10-39)
where A = -do/So and B = cr.Q2/d o
10-3. The Direct Step Method. In general, a step method is chal"acterized by dividing the channel into short reaches:and carrying the computation step by step from one end of the reach to the other. There is a
great variety of step methods. Som.e methods o..ppear superior to others
in certain n~spects, but no one method has been found to be the best in all
5
,
t
;
If So = 0, then Q, = 0, Q/Q, = "", and the varied-flow fund-ions hecome meaningless. If S, is negative, Eq. (10-30) shows .that Qo> is negat,ive. Since the actual
discharge Q must be positive, (Q/Q.)2 bee"omes" negative. Thus, the integration
procedure must be done for negative values of (Q/Qo)l in the two varied-flow functions.
"
Example 10-7. Compute the flow profile required in Example 10-1 by the direct
step method.
.
Solution. With the data given in Example 10-1, the step computations are carried
out as shown in Table 10-4. The ,,&lues in each column of the table are expln.ined
as foll<JW5:
, Col. 1. Depth of flow in ft, arbitrarily assigned from 5.00 to 3.40 ft
Col. 2. Water o.rea in ft' corresponding to the depth 11 in col. 1
Col. 3. Hydrn.ulic radius in ft corresponding to 11 in col. 1 .
Col. 4. Four-thirds power of the hydrauli~ rn.dius
Col. 5. Mean Velocity in fps obtained by dividing 400. cis by the water area in
col. 2
CoL 6. Velocity head in it
Col. 7. Specific energy in ft obt,ained by n.dding the velocity head in col. 6 to the
depth of flow iI) col. 1 ~. :
'
1 First suggested by the Polish engineer Charnomskil (22] in 1914 and then by
Husted [23J in 1924.
f", >,
,
,~-
.: -,
.. --/
METHODS OF COMPUTATION
1 (
21
;g ~ ;g
I.(')
-..:tt 0
e,p 00
r-...
...-I
C"I .D C'l
NMl!"Jt-...~;:S~~~~8~g~c::.
~ r--: u:cr: ~ c ~ 00 00 00 J:-..l"-..;! ~
'""""!
~~~~~~~~~~~~~~~J
<01 !:::;:E~~~~g;$~::g~...-IN(QC'-) t
~!:::, ~~
~ ~ MO'~
~~ ~
~ ~, .
C'I') M::'t;l
'~r ~~
I 000000000000000 _
I ~ ~ ~ ~ ~ ~ ~ ~ ~.~ ~ ~ g;; ~
Jl:'""'i
3/
rt;
;;01
~I
! ./1
_
".
Note that, in either tpe direct step method or thE'l standard step method
which w:ill be described in the next article, the step computation should
be carried upstream if the flow is suhcritical and downstrea.m if the flow is
supercritical. Step computations carried in the .wrong direction 'tend
inevitably to make the result diverge from the correct flow profile .
10-4. The Standard S!~Jl_~et~9d. Thill method is applicable also to
nonprismatic channels. In nonpril'lmatic channels, the hydraulic elements are no longer independent of the distance along the channel. In
natural channels, it is generally necessary to conduet a field survey to
conect the data required at all sections considered in the computation.
The computation is carried on by steps from station to station where the
hydraulic charactel'i3t~cs have berm determined.. In such c'ases the distance between stations is given, and the procedure is to determine the
depth of flow at the stations. Such a procedure is usually carried out by
trifl.l and error.
In explaining t.his niethod it is convenient to' refer the position of the
water s\:rface to a hQrizontal datum. In Fig. 10-6, the water-surface
, elevations above the, datum at the two end sections are
gj "/
'"
C'\tNC'lMC"t')C"t')MM~~~''';~j;'';~
I
J
'<!
J
)
{,
--.1
~'
~I
I
."
--.1I
.~
~I
265
Col. 8. Change of sp.ecifi~ energy in ft, equal to the difference between the E value
in col. 7 and that of the previous step
CoL 9. Friction slope computed by Eq. (9-8) with n "" 0.025 and with V as given
in c'o1. 5 and R~i.in col. 4
Cbl. 10. Aver~ge friction slope between the steps, equal to t.he arithmetic mean of
the friction slope just computed ill col. 9 and tha.t of the previous step
Col. 11. Differe,nce between the b0ttom slope 0.0016 and the average friction slope
CoL 12, Length of the reach in ft between the consecutive steps, computed by
Eq. (10-41) or by dividing the value of b.E ic col. 8 by the vahle in col. 11
.
Col. 13. Distance from the section under consideration to the dam site. This is
equal to the cUlmllativ~ sum of th~ values in coL 12 computed for pr\lvious steps.
Th~ fiow profile thus computed is practically identical with that obtained by graphi~l integration (Fig. 10-3).
.
.,.,... 'Example 10-8. A 72'-in. reinforced-concrete pipe culvert, 250 ft long, is raid. on a
slope of 0.02 with a free outlet. Comput,e the flow profile if. the. culv;;l't dischs.rges
232 ds, n = 0.012, and" = LO.
Sol'utian . . ,Fronl tl:e data, V' = 4.35 it and Yn = 2.60 ft .. Since y, > y", the chr.nnel slope i~ steep. As shown in Fig. 10-7, the control section is at the e.ntrance; water
will enter the culvert at the critical depth !lnd .thereafter flow: at a depth less than 'Y,
but grer.ter than y . The flow profile is of the 82 type.
Table 10-5 shows the computat,ion of. the fio\v profile, which is self-explanatory.
The computed profile is plotted Il.S shown in Fig. 10-7. Plotted also in the figure is
the energy line indicating the variation of energ:' along the culvert. The comp\,tll.tion has be~.n car~ied to exceed the length of the eulvert, so. that the depth of flow at
the outlet can be interpolated. This depth is fOHnd to be 2.81 ft, and the corresponding outlet velocity is 19,4 fp3. It should be noted that, if the. pipe were fiowing full
at the outl'et, the outlet veloCity would be only 10 fps.
264
Zl
..
So t.x
+ Yl + ~2
(10-4:3)
METHODS OF COMPUTATION
IJ) C'J 0
1""""1
1""""1
1""""1
(10.,.44)
and
cno:.
267
NM
(10-45)
CO"'lflCON~I""""ICOO
<1
where the friction siope S/ is taken as the average of the slopes at the two
end sections, or as Sf.
Substituting the above expressions in Eq. (10-40), the following may
be writt.en:
.
V 12
)12 2
(10-46)
Zl
al -2-- = Z2 + 0:2 -2 + 11,1 + h.
g
g .
O(O<Ot-t-<:OlQO
.-tcr:,t-l,Q"'d'I<!:l
O':IM"'lflt-O"lCCflr-
o:JcqO~LO-tK:.Otn
Lf') "'IfI C"I C"J
to ~
~""""'I""""II""""IOOOO
r-
tn
00000000
00000000
where h. is added for the eddy loss. which ma,Lbe appreciable III nOI1..NiBma~ic~. No rational met,hod of evaluatin edd' loss is available. r~~epends mainly on the velocit,y head change a.n . lllay
be expressed as 11 part Of It, or k(~aV2!2!l) where k is a coefficient. For
-graduzHy COl1vel'gmg and-dive-rgll1g reaches,!, = 0 to 0.1 andO.2, respectively. For abrupt expansions and con}.r.actions, k is about 0.5. For
prismatic ~uTai'Chailiiels) the eddYloss is practically ~'Ie = O.
For convenience of computation, h. may sometimes be cOll03idered part of
the friction loss and Manning's 11. may be properly increased.in computing
_!}l> I~.~PI h~.i~9 in the cor,nputatiol1.
The total heads at two eUd sections are
000000000
0')00<1:><0<0000'"
1""""I0l~U"J"'IfIC'-I~t
OON~t-M"l~
00000000
OO,.......(cnl.O-.;.t-tOtO~
~ ~ ~ ~,~ ~ ~ ~ ~ ~
<0 t..OcO
jC'~ 00
;"'..... "'1'1
(0
.qt
t-t"-o 0000 00
1""""1
t-' to to
~
f.O
CO
Q
...-I
(10-47)
('>
..
and
1100 <'> <0 "" ..., \I') <0 0 ....,
.... ~O)OC"lC:OCf.lt-C*:loC()
,r- ,....;....; N ~
~ 00 ~ c;)
to
1""""11""""1.1""""1'_. 1""""1
I
I""~,
A:
.... ....
""
t-COCOO
.....
.....
0'>
t-
~I
t:'-
t;-
C'~
t-
00
HI = H 2
....
'"
..... .... .... .....
::g~g~~
<'l
\OMtnHMC'O~t:OO
OlI""""l~t-O':I....-l~Ot-
~ ~
. . .;
jN N
r--:
~ ~ M~
C-l
'1"""0000000<:'1'"'
M cq, en co CI'J 0
00 c.c l'-o
,"'IfI "dt r:v'l cr:,
I'"
C1:I.~ C\l N
C'l
QC'lOmo\l')oaor-<b
........
"r-r-<1:>,","'\t)~"""<j'
~
"
-,1""""11""""1
io,-,tI"",",,,,r-<o<o<o1
COtn t-t-tO.-t loON 0')
) ....... Oo>oor-,.:><O\I')
.....
C'1 C'l 0:-1 ,.... .....
A:
(10-48)
000000000
Z56
I
"
..f
+ hI + h.
(10-49)
This is the basic equation that defines the procedure of the standard step
method.
The standard step method is best suited to computations for natural
...l1annels. 4Jl.is!!latic channel, howevelJ.~ill be used in trw following
'example in order to ~ the iI1ustration and to' aUows_oml?J!.risoll with
reSiiItSobtained by the other methods that have b.een: described. For
apphcatiOilto natural channels, .an example will be shown later (Art.
lO-G}.
Example 10-9. Compute the flow profile required in Example 10-1 by the standard
str.p method. As~ume that statiollS along the channel are j1x~d at the distances
determined in the solution of Example 10-7. The elevation at, the dmnsite is 600
m.B.!.
--I
268
Solution. The step computations are arrauged in tabular form, as shown in Table,
10-6. Values in each column of the table are explained as follows:
CoL 1. Sectionidcnbiiied by stiJ,tion n'lmber such as "station 1
55:" The
location of the stations is fixed at the disto:nces det~rmi!led in Exampie 10-7 in order
to compa.re t.he procedure with that of th~, direct step method.
Col. 2. ,Vfl.ter-surface elevation at the station. A trial value is first entered in
this column; this will be verified or rejected on the basis of the computations made 'in
the remnining columns of the table. For the first step, this elevation .must be given
'or assunled. Since the elevation of the darn site is 600 m.s.1. nnd the height of the
dam is 5 ft, the first entry is 605.00 m.sJ. When the trial value in the second step
has been verified, it becomes the basis for the verifieation of the trial valu~ in the
1I1lxt step,and so on.
CoL 3, Depth of flow in ft, conesponding to the water-s'll'fll.ce elevation in col. 2.
For instunce, the depth of flow at station 1
55 is equal to \Vater-surface elevatioh
minus ehvation at the dam site'minus (distance.from the clam site times bed slope),
or 605.048 - GOO.OOO - 155 X 0.0016 = 4.80 ft.
'
CoL 4, Water area corresponding to y in coL 3
CoL 5. Mean velocity equ!l.l to the given discharge 400 cf" divided by the wa.ttoi'
area'in col. 4
CoL Ii. Velocity head in ft, corresponding to the velocity in col. 5
Cnl. 7, Total head computed by Eq. (10-47), equal to the sum of Z in col. 2 and
the velocity head in col. 6
.
Col. 8. Hydraulic mdius in ft, corresponding to 1J in coL 3
Col. 9. Four-thirds power of the hydraulic radius
Col. 10. FYiction slop~ computed by Eq. (9-8), with n = O.O~5, l' from col. 5,
and R~~ ]rom coL 9
CoL 11. Average friction slope through the re~.ch between the sections in each'
step, approximately equal to the arithmetic mean of the fdctloll slope just comp';lted
i.n col. 10 and that of the'previous.step
.
Col. 12. Length of the reach between'the sectioll.S, equal to the difference ill station numbers between the st-ations
CoL 13. Friction loss in the reach, equal to t,he product of the values in cols, ] j
and 12,
Col. l~. Eddy loss in the reach, equal to zero
Col. 15. Ele\ation of the total head in ft, This is computed by Eq. (10-49), that
is, by adding the values of hI and h, in cols. 13 and 14 to the eleva~ion at the lower end
of the reach, which is found in col, 15 of tli __ previous reach. If the vallie so ohtained
does not ng'ree closely with that entered in col. 7, Il. new trial value of the water-surbee
elevation is assumed, Ilnd so on, until agreement is obtained. The value that le.1ds '
to agreement is the c!!rrect 'water-surfuce elevation, The computatiDn may then
proceed to the next step. The computed now profiie is practically identical with that
obtained by the gl'aphical-integmtion method shown in Fig. 10-3.
i
)
,I
. 0
09.0
000
:. f6
~ '[3;.~'~ ~ ~ ~ ~ ~ ~ g ~ ~
0 0- 0
N
C'l
C'1.1""""1 C'1 N N
I""'"'f
-I
1""""1
:00000000000000
1
!
r
~!::;~~~~g~f2~~~~t;::;~
u-; L~ ~ ~ ~ ~ ~ M M ~ M ~ M M M
C'f':l 1""""1' m co u:; ~ t- ..... IX) Lr.J C'o':: ..... 0
0')
It)~M'''''OOl;:Qt-t-~t.D~tO~U'::I
-.::tt
M~MMe..)~~NNC'1~C'i"l~~
. . ;tt
---~----------------------t-cot-c.oV":l~O-e::tl~CO[-......-I~~C"'-\
."
:;;""
i:>-
is
II
----
~~;;;~~~~~t2~~;3~~~
......-I
-f
......-I ~
.......-4
C~
C"";)
C"";)
C"";)
000000000000000
C") a.: ~ ....,.. C"t ~ co Lt':l O"J M 00 ~ r- co
~ ~ t- U') L"':) t- !"""'4 ~ co lfJ M t- C"l U") co
~ co O"l ....... , eo"J: lfJ co OJ a
......-I ,,1 ~ ~ CC (r"
l-
~C'1NM~e...JMer)~~~.qt....;tt~~
-0 co C'l N ctJ 0
o 0 C'f':l t- C"l 0
00 00 C'-1 .....-I 0
00 11";10 C'l
00 M C1 ~ H;l ~ ~ co ~
oc-i...;~~~~.....;r...:cr;~MN~"";
t....,~.;1JC'I.....-I....,oIoocn~O":IC":IO":ImCJ)
~......-I......-I_"""-f~......-I
8~g~~g~f;~!2g~;:~~
u-)~~";~~:MMMM~~MMMM
lSI'
'';;
,,$
.[
ell
~I
~ to 00'::: ~ ~ ~ ~ ~ ~
269
;i
271
METHODS OF COMPUTATION
and the friction slope vary with the square of the discharge. Therefore,
after the curves for one value of Q, say 400 cIs, in Figs. 10-8 and 10-9 have
been computed, CUI'ves for othol' discharges, say 200 and 800 cfs, may be
obtained t'ithel' by shifti!lg the curves horizontally (Fig. 10-8) or by shiftthe abscissas (Fig. 10-g). The distance by which either the curve or
the abscissa should be shifted is determined by the square of the ratio of
the !lew to the original discharge: .
270
-l
-j
0.05
0.1
Veloci~y
FIG.
0.5
need, ft
1.0
,/"
l<
272
the cur~e fOl'. renc? 1...2 as shown in the figure is obtained. For simplifythe IdentdieatlOu of the curves, elevations at sections of oddnurnber
are . represente~ ,by abs~i8sas and those at sections of even number by
ordinates: . ':Vltn the ,dlagram thus prepo.red, the flow profile for any
a,ssumed 1mbal elevatlOll can be determined, Taking the initial elevation as 605.00 and fonowing the dotted line in the direction of the 'arrows
the corresponding water-surfaue elevations Il(; subsequent sections can b~
obtained easily, The diagram is const,ructed for a discharge of 400 cfs.
For any other discharge under investigation a diffe;'ent diagmffi must be
prepared.
GQar
~
.i 607~
J
:;
C
,2
606 ,-
.'"
Q
"
"
~
,"
...
505
'"
'0
if:
606404~----c*~----~!~____~I~____~
606
607
608
FIG. 10-10. The Le!Lc~ diagram for flOW-pl'ofile computatlon when a large numbel' of
flow profiles are reqUIred for the same discharge.
C. The E'l:m lIfet!.wd. The Ler;.ch diagm,m is recommended if flow profiles Iorthe same dlScharge are required fOl' val'ious initial st!Lgcs. When
flow profiles for various discharges are desired for different initial stages, a
method developed by Ezra; (25) may be used. Similar method':! have also
'
been developed by others (26-281.
~ul:lstituting Eq. (10-45) in Eq. (10-46), the following equation may be
wntten:
','
I.
(10-50)
'o/here
!l.nd
(10-51)
(10-52)
MlllTHODS OF COl\:fPUTATION
!r
n V1
'8 1 = 2.22Rl~~
2 V 22
~
~2 = 2.22R2~i
2
and
81
,82 =
a.ud
"
273
n
( Q)2
1(1
(10-53)
(10-54)
(1O~55)
(!l)2.
1(2
It can be seen that functions F(Zl) and F(Z2) are proportional directly
to the squares of.the veIQc1ti~s or of the discharge Q, For any other disQs, the corresponding values of these functions can be obtained by,
muit.iplying the functions by a factor (Q./ QF.
The Ezra method gives a graphical solution of Eq. (10-50). There are
two major parts of this solution:
L COMPUTATION ,o!..ND CONST1H)'C'l'ION OF Z + F(Z) CURVES. For each
chosen section in a charulel under consideration, several values of \Vi:J.tersurface elevation are selected, and necessary geometric elements A and R
for each of these water-surface elevatiofL'> are determined ,and tabula~ed,
'This procedure is the same as in the standard step method.
For given values of Q and n, values of C/ P /2g and SI are determined at
each section for each of the selected elevations. In artificial prismatir.
channels, the friCtion slope Sf 1l10,y be computed by Eqs, (10-53) and
(10-54). In irregular r..hannels, the v(,lull of 1C is determined first and the
value of Sf is then computed by Eqs. (1()"'55) and (10-56). This will be
shown in Example 10-12.
'
For each section, t,he corresponding vl1llleS of F(Z) are then computBd
by
(la-51) and (10-52). It should b! noted that two values, F(Zl) ,
and F(Z~), should be computed for each section. In computing F(Zl) of
the section, the value of ..6.x is the value ..6.Xd measured downstream from
that section, In computing F(Z2), the value of..6.x is the value ..6.x,. measured upstream from that section. In the computation, this rule ensures
that values, of ax will
common to sections at both ends of any reach,
since it will be shown that the same value of ..6.x is used on both sides of
(10-50), In other words, at each section Zl = Z7 = Z.,
Now, fol." each section' with Z = ZI = Z21 values of ZI + F(Zl) and
Z2 + FCZ2) are computed, and curves of Z1 + F(Z'l) and Z2 + FCZ2) l1re
plotted against Z for each section (Fig. 10-14).
, 2. DE'rEP.MINATlON OF WATER-SVRllACE, ELEVATIONS. The effect of
eddy losses may be included in the value of the roughness coelficient n, as
described in Art. 10-4; thus, the term h. in Eq. (10-50) isi zero. The
resulting ~alue of Z2 + F(Z2) for a \:l.ownstream section is the:refore equal
to Z 1 + F(Zl) for the next upstream section of the same reach, and vice
versa..
be
- 1
"
274'
METHODS
, For subcritical flow, the flow profile is determined in an upstream direction. Star!.hlg with a given initial water-surface elevation at a down:;;tream sectioll, the value of Zz
F(Z2) ClLn be obtained from the appropriate Z2 + F(Z3) curve. Entering the curve of ZI
F(Zl) for the next
upstream section with this value, the corresponding, water-surface elevation is determined directly, This procedure is J'Elpeated from section to
section, tracing the desired flow profile.
For supercritical flow, the flow profile is traced in a downstream direction. Starting with the value of .?I
Ji'(Zl) at the initial section and
taking this value to the Zz
F:(Zlz) cw've for the next downstream section, the corresponding water-surface elevation is determined.
If eddy losses are not included in the roughhess, the term h. may be
expressed as k{O'V2/2g), where k is a coe.fficient described in Art. 10-4.
Accordingly, the procedure of computation should be modified.
First of all, it necessary to plot curves of k(a V1j2g)
the elevation Z for eil,ch section.
.
For subcritical .flow, water-surface elevations for two consecutive sections are determined in an upstream direction by the procedure described
O. Values of k(aV2/2g) may, therefore, be obtained
previously for h..
from the plotted k(a V2/2!7) curves for these elevations .. The difference
ka:{VN2g. T'~2/2!l),or he'in the reach, is then added to the value of
Z2
for the lower one of the two sections. The resulting value is
taken to the
+ F (Z 1) cun'e for the hight'll' of the t.wo sections, and the
corresponding corrected water-surface elevation is determined.
For supercritical .flow, the procedure is similar. However, the flow
profile should be traced in a dowllstreant direction, and the correction for
eddy losses should be deducted from the value of ZI
F(ZI) before this
value is taken to the Z2 + F(Zz) curve.
The application of the Ezra method will be illustrated by examples ill
Art. 10-8.
10-6. The Standard Step Method for Natural Chll1lnels. For flow in
most natural channels at a normal stage, the flow profile in short reaches
is very close to that of a uniform flow, but slightly modified by local
channel irregularities .. For apparently gradually varied flow, an approximate solution may be obtained by either the di!'ect-integration method or
the direct step method, assuming a prismatic chp.,nnel having the average
geometric and hydraulic characteristics of the natural channel. For a
practical and precise Bolu tiol1, however, the standard 'step method is
recommended.
.
The standard step method has many obvious advantages in application
to natural cfU'.llllels. When the velocity head is small, ~he step method
can be carried even in t e wrong direction wH~ resulting in serio~
~rr~rs, al~ is always advisa Ie 0 ca;ry the computation upstream
is
~~---.-:--
COMPUTATION
275
OF
, !
)
, I
--1
l
..
I
.. i
-, !
tj
277
METHODS OF COMPUTATION
TABLE
10-7.
MlSsounr
Mo.-"
KANSAS CITY,
<0
'-
'Q
'Q
.:;""
::
"0
"
I'il
.E
'o"
""t::
"
'0.,>
">
:>
-"
.0
0
\:.)
"
~
~c:
..
"<l
cor
r:Q
~,~
",,,
o~
,,0 '""'~~
:0; :::
~ c:
- "
]~ .E:>:
l-
I-
~t
...'"""
e ...
"E
.I
J
o~
Cl
92
Ch.a.nnel
1 + SO Hi t ' 2 1,520
1.550
H:'L9 Channel
2 +IH 23 + 72 2.0S0
~'.
ltO
720
7l!.D
3,020 Levee 1 on 4
2,040
2.080
47,SOO Bank 1 on 1
720
720
0+61
DVc.r ..
1 +W
bnr.. k
t:
a
~
l,il00
<Il
>::
il
:<
...,
7.16.2
37S.U5"
Lef~
over-
Channel
I, S8il
>-
378,95
Ch3une~
880
1.320
01.
'-
:;:
"
OJ
'"
...J
i.,
'"
'"
...
ovet-
lD
0-
ol
b.nk
Channel
Dike
E
X
OJ
Lovee IonS
bo~h sides
:12 + 80 4,4 +
10n3
44 + 50 5.7'+40 1.29()f
460,
57 + 40 62 -I- 29
1,310
490
34,600
5,700 Levce 1 on-4
640
2,430
<=:
>-<
Fllirfax 384.00
Bridge
,...;
,....
0
,....
5GO
b~nk'
Chllnn~1
"
r;
RigMoverbank
11
3SS.57
216
840
It,!iS50utL RivQr
0
'-
"t:l
Le\c. 1 0:1 3
but.h slue3
;:;
"
3~.
54()
";::
0
<1;
Riv~r
ol
,~
,~
I{ansfls
"0
>::
c:
;.
'C
3,020 Le"'eg-l on
~
~
""E
c:
"
"'
bank
..,
<:>
B9
745.5 Left
:3
f;
W
..J
1.200
Chfll\nel
84.5
.:I
178
b~nk
<5
Channel
1,060
.". 1 745 . 3 Left ovor
b
'-'
I
f
Ri\~3r
'"
.!!~
Missnuri
E-
.1
1
1,400
...... 1
125
278
:!:
~
.,
.,
.,;
."
....
- - . --
.,
..-!::-
c:0
;e
;;
,,;
,,;
c:;M
"":.
co
....
--~-
'"
;:;
!;
'"0
0
0
c:0
'"
--Oil
---;;;-
.....
0
'"8'"
Example 10-10. Compute the flow profil;; in the Missouri River near Kansas City,
Mo., for a. design discharge of 431,000 ds. The plan indicating the locations of the
cross sections is shown in Fig. 10-11. The geometric elements of these cross sections
a.re shown in Table 10-7 for given water-surf!t.Ce elevations .. For other elevaLions, the
geometric elements are to be estimated from such data. l
.
Solution. The computat.ions are tabulated as sho',vn in Table lO~p. The column
headings in the table are explained below:
Col. 1. . Section number in conformity with the plan of Fig. 1O~1l
Col. 2. Subsections, for which M.C. designates the main channel section and L.O.
designates the left overbank section
Col. 3. River mileage above the mouth of Missoud River
Col. 4. Water-surface elevation. The initial elevatiol1.of 752.25 at. seotion 1 was
estimated from the rating curve of the Kansas City gaging station on the Hannibal
Bridge at Missouri River mile 3.77.58.
. Col. 5. Water area. For instance, section 1 is subdivide'd into le;:t-overbank and
main-channel areas. At elevation 752.25, the area is determined for each subdivided
area from Table 10-7 or from a curve prepared as shown in Fig. 10-12.
0
0
'"
'"
-_....'"'"-
,,;
0
--0-
-'
...
0
'"t";... ..'"
~
'"o.
.!!
00
"
'"M
Ii
X
...'"
~
h'
xx X
">
<1l0"O:;
0>
p;j
.;::
'"
~~
"
0
gj
.~
:;;"
.."'""
<OM '"
'"
iii
--",-
II
0>
;:!
:6
:l
.
"
p..
~"'I~
-0-
'"
'":
000
""'"
'"
.,;
::::
I
<0
""
......
~I
--;0-
",
",.
'"
...!::-
",.
A in 1,000 fi2
",.
FIG. 10-12. A and R vs. \'liater-surface elevation for channel section No.1 of Missouri
Riyel" at Kansas Cit:!;, o. (U.S. Army Corps oj Engineer. )
1 The
plan and data used in this example were obtaine.d from [291.
values, however, have been modified to suit the present purpose.
Some numerical
o
~l
d
~
;:
.0
~..:l
I'"
I""
1.5.
0
~,j
-I
280
Col. '6. Wetted pcdmeter. :For elevation 752~2p a.t section 1, the wetted perimeter
is determined from' Table 10-7 or from Fig. 10-12.
Col. 7,' Hydmulic radius, obtained by dividing the area in col. 5 by the wetted
perimeter in col. 6.
,
'
Col. 8. Two-thirds power. of the hydraulic radius in col. 7.
Col. 9. The value of Manning's n. It is assumed ~hat the general rosses due to
contraction, expansion, find bend are included in the friction losses coniputed from
the selected n values.
'
Col. 10. The conveynnce K = 1.49 A R* /n
Col. lL The value of I(} / A'
Col. 12. Energy coefficient for nonuniform velo~ity distribution. From col. 5 to
col. 12, the procedure of computation is the same as that described in Art. 6-5. Tne
coefficients for the .subdiviLh,d channel sections are assumed equal to unity.
CoL 13. Mean velocity, which is e~luar to the section discharge of 431,000 ds
divided by the water area in col. 5. Above section 5, the dischaJ;'ge is divided between'
the Missouri River (350,000 cfs) and the 1{ansas River (81,000 ds). The velocities
shoold be computed for the divided discharge.s accordingly. The division of discharg;e
is balled on a hydl'ologic study of the drainage basins of the two rivers.
Col. 14. Velocity head
'
Col. 15. Total head, whi~h is equal t.o the sum of thc elevation in col. 4 and the
velocity head in col. 14
Col. lB. Friction slope, which is equal to (Q/ K)'. The 1C v!Llue is the total
value for the section under consideration. Thus, for section 1, S, = [431,000/
(290.1 X 100)]' = 0.000220.
Col. i7. Average friction slope through the reach between the two sections, that'
is, the arithmetic mf!fLn of the friction slope just computed in coL to and that for the
previous step
Col. IS. Length of the reach between the sections, that is, the difference in river
miles between the sections as converted into ft
Col. 19. Friction loss in the reach, th!lt is, the product of the slope in col. 17 and
the reach !ength in col. 18
'
Col. 20. Eddy. loss in the reach. The general losses are included in the friction
loss cOlIlputed in ,col. 19. However, at the entrance of the KansaB River to the Mi,~.
souri River, an additional eddy loss at the C1mfillence is expected. This is estimated
at, 10% of the increase in velocity head,or 0.10 X (1).63 - 0.13) = 0.05 ft.
CoL 21. Total head, which is obtained by t1.dding the SUIll of the losses hl in col.
19 e.nd h, in col. 20 to the total head in the same column for the previous section. If
the value so obtained does not agree closely with that entered in col. 15, the watersurf!lce elevation in col. 4 should be assumed again until a fair agreement is reached.
For the computation ot' water-surface elevations at sccHons1K, 21(, 6, and 7, see
Art. 11-10.
METHODS OF COMPUTATION
281
+ h~2L -
hvI
(10-57)
'where F is the fall in water surface and h'J2 -' h~1 is the change in velocity.
head. If hvi - hVI is zero~r negligible" then S, = F /L, and the normal
disch[J,rge of a uniform flow by the Manning formi.tla is
= l.~~
_4RH
(J)'Il
(10-58)
For a gradually vlLl:ied :flow with backwater effect having a' discharge Qz
and a corresponding F r in the same re!wh, it may be assumed ,that ft. form
similar to Bq. (10-58) may be written, i.e., that
.Q~
1;9
AR>1
(~ys
(10-59)
(~)Z
= Q/v'F
(10-60)
, ~vhere Q/ y'F is called the discharge for 1-ft fal!.1 !his equationC8,l: be
used ill the flow-profile eomputntion if the stage-fall-discharge relatlOnshtp for uniform flow in the reach is known.
The stLlge-fall-discharge relationship f.or 11 selected reach may be determined froni. records of observed stages and discharges Cfable 10-9). The
stages or water-surface elevations at the beginning section _of the reach
are plotted as ol'dinates, and corresponding values of Q/y'F are plotted
as ~bscissas, resulting in a sta.(je-ve1"!:;us~Q/....IF C'urve (Fig. 10-13). When
any water-surface elevation ~\t the_beginning section of the reach fS given,
the corresPQnding value of Q/ v'F can be read from the curve, and the
fall for a dischrLrge Q. be computed by Eq. (10-60). The computed fall,
when added to the water-surface elevation at the beginning section of the
-reach gives the water-surface elevation at the end section of the reac.h,
whicl~ is also the water-surfaee elevation at the beginning section of the
next reach. The procedure is repeated for each reach until the complete
,required flow profile is ob,tained.
: The stage-versus-Q/v'F curve is generally cOl~structed as an average
'curve for varying river conditions; such as rising and falling of stage,
,lIn a similar method devel~ped by Rakhmanoff [341 a term F /Q' is. us 7d in lieu of
Q/ ~F. This term has the nature of /J. resistance factor and therdore IS given !I IIame
of resistance'modulus by PavlovskiI [21, p. 1151.
282
fluctuating stream bed, and effects of wind, aquatic growth, ice, .and overbank flow .. Owing to these varying conditions, the plotted points 20re
often scattered; and a smooth line, giving consideration to the varying
conditions, should be drawn through the points, representing the average
condition of the-channel. Where sufficient measurements are available,
data of doubtful accuracy should be rejected. In general, the more
recent meaSurements should be given greate: weight, as reflecting recent
channel changes. Other factors that should be considered in constructing the curve are the relative accuracy of individual discharge measurements; the ,flow condition during the measurements, whether rising,
falling, or stationary; conditions affecting the stage-fall-discharge relationship, such as the changes in channel roughness, levee bteaks, and shifts of
cha,nnel controls; and the existence of substantial local inflow between the
stations.
The stage-yersus-Q/VF curve may qe extrapolated !].bove or below t.he
range of the observed data by extending the curve at its ends in accordance with the general trend of the curvature. Howe-.:er, any abrupt
change in hydraulic elements of the channel section will produce a, eorresponding change in the curvature of the curve. In this case, a correc.tion for the change, if known, should be made in extrapolating the curve.
This method is used most advantageously when a number of discharges
corresponding to known stages, or vice versa, a.re desired in a stream.
By making proper allowance for variable conditions, satisfactory results
ca.n be obtained for reaches of large rivers 50 to 100 miles from the measuring station. The data required by the methOd are often .les;:! expensive
the,n those required by the standard step method. However, this advantage.is usually.offset by the inaccuracy of the-results, because the effecl of
the change in velocity head 1s ignored in the present method. For this'
reason, the stage-fall discharge method is more satisfactory for problems
. iriwhich the velocity is well below critical end decreases in the downstream direction.
E:x:amplelO-ll. Compute the water-sudace elevation at section 1 of the Missouri
River problem in Example 10-10 by the sto.ge-fall-discharga method. The reach from
section 1 to section 5 istakea as the first reach. The water-surface elevations are
available from stage records for gages located at sections 1 and 5. The discharges
have been observed at the A.S.B. Bridge located about 3,000 ft downstream from section 1. These data are tabulated in Table 10-9.' .
Solution. The data arid cornputa tions for the sta.ge versus discharge for a I-it-fall!
curve are given in Table 10-9, which contains the following headings:
Col. 1. Recorded water-surface elevations at section 1
Col. 2. Recorded wo.ter-surface elevations at section 5
Col. 3-. Fa.ll in ft, which is equal to the differ~nce between elevations entered in
cols. 2 and 1
1
e.;.-,
283
METHODS OF COMPUTATION
in
755 752,25
~ 750
crt
---I-~-lf--,-~-
745
!
~;
1- '
I
740r--r-
..,
I~
II
~i----~
'0
'0 735
L~
.12
0;
>
-[- w-
'" 730
."'g"
. 725
"":
~
0
3;
720
715
0
-r~!
50
100
ISO
200
250
I
1
h
_ _I
300
,I
I
.,,350
".,
".,
Values OfQ/IFinl.o00
lJ~its
Eq.
"
284
285
METHODS OF COMPUTATION
e eva,.lons
e III, erme la/,,", sectlOns may be obtained by in'erpel-t'
" Th.e stage-versus, 1,
}
Ion.
Q/
. v F curves can be drawn for each section and the cbmputat'o
b
. d
fol' the subdivided reaches.
'
.
I n can e carne out
TA.BLE
Section 1
Section 5
(1)
(2)
724.8
725',3
729.13
727.4
727.8
730.2
730.8
731.3
734.6
735.8
736,6
745,0
722.2
724.6
725.0
725.3
725,7
726.2
730.2
728.3
728.8
731.2
731.7
732.3
735.6
736.7
737.7
746.6
723.1
725 6
726.0
726.4
"
10-10. ConiPUTA'I'I~N
(Missouri River
Sea.
no.
River
mile
D'ISCh al'ge,
cfs
ft
(3)
0.9
0.0
O.G
0.9
1.0
1.0
0.9
1.0
1 0
0,9
1.1
1.6
0.8
1.0
1.0
1.1
THE
(4)
.'
Length of
reach
377.58
.....
378.65
5,655
. ....
VF
!~5)
33,600
36,100
66,100
69,500
76,000
97,200
105,000
113,000
141,000
157,000
104,000
326,000
22,900
45,400
49,900
52,300
35,40',)
38,000
85,30U
7'6,200
76,UOO
97,200
111,000
113,000
141,000
165,000
156,000
' 258,000
25,600
45,100
49,9UO
'19,800
......
OF
FII,11 ,
".'
~rater-silrface
elevation
, '., =
10_11
BY
431,000 ds)
Q/-./F
F. =~-2
(Q/ -./F)
752.25
1.65
r 335,000
1.65
753,90
.... ' ..
To be continued if desired' ...
10.,8. The ~zra Method for Natural Channels.' If flow profiles for a
numbel' of discharges or stages are desired, the stage-fall.discharge
metho~ can be use.d most advantageously for a simple and economical but
approXlrnate solutlOn. However) if a precise computation, including the
10-11.
TA.BLE
R I R%
".
Ie'
A'
+F(Z.)
0.12
.... 11,&3
753.93
0.11
.... 11.80
754.80
I'
co
0>
!;;;:~ . 3:~c-.-_-.t
I~i
I.
0.10
...
0.13
1.46 1.70
753.46
753.7il
(l.l~
1.39 1.65
154..39
754.66
1.3111.55
755.31
755.M
1.71
7.55.71
J
-0.151
-~'.~,--
;;<
TAllLE
10-11.
COMPUTATION OF
Zl
+ F(Z,)
AND
z. + jt'(Z.)
FOR l:i:::CHtI'LR
10-12 (conli'n"lJ.ed)
8ee.
+ F(Z.l
no,
0.24
0.21
.t
""
co
1.111 1,45\
753.11
753."'5
l.oJ 1.341
754.04
754.34
-0.08
0.. ~g
1.00 1.27
75S.00
755: 27
-O.Hi
O.1.~
0.5J
0;90
M2.61
752 ..90'
-0.15
0.12
0.5, 0.83
753.5u
753.g3
-0.13
0.11
0,~4
751.54
754.78
-l
0.78
752,62
0.62
H'OUI,,u;onr'" x
In~.LJII)<.orovl:t'U!;~I"".1
7,900 7411,10.6 4.B3,O.Q50 11
0.58
0.54
,.....---.
... 1
753.58
288
TABLE
10-12.
FO~
EXAMPLE
10-12
DY
Sec.
no.
3
4
River
mile
Z,
377 58
377.78
377.9,1
378.33
378.65
+ f(Z.) z, + F(Z,)
. . . .. .
754.14
754.41
754.G8
754.93
754.. [4
754.41
754.68
754.93 .
= 431,000
cfs)
-~
'ffater-surface
elevation Z, It
no.
752.25
. 752 72
753.38
7154.15
764.43
F(Z,)
(2)
(3)
752
;::
.:o
:;::
.~
0;
754.43
'-'
75~5~=====
.,
~ 754
"'",
F(Z.)
z. + F(Z,) z. + F(Z.)
. (4)
752
753
754
75:1
754
OJ
289
METHODS OF COMPUTATION
(5)
754.59
755.41
75G.20
2.59
2.41
2.29
1. 96
1. 86
1. 78
(6)
2.36
2.22
2.08
753.{16
754.8G
755.78
75 1L36
755.22
756.08
1.94
753.49
754.40
755.31,
753: 94
754.80
755.70
752.82
753.75
754.73
753 21
754.11
755.05
752
753
754.
1,49
lAO
1.34
1.80
1. 70
752
753
754
0.82
0.75
0.73
1.11
1.05
752
753
754
0.83
0.78
0.73
1. 21
OJ
753
75ZL-____~_____L_ _~~L-~
756
755
FIG.
754
Zl+F(Zll,ft
10-14. 'Curves of Z
753
+ F(Z)
.754
755
Z2+F{Z21.ft
756
--,-----~----~--756,,----_.----_..------._--~
752.83
7fi3.78
75<1.73
The Z + F(Z) curves a.re then plotted (Fig. 10-15). These can o,lso be'obto.ined by
a simple graphical metbod. In Fig. 10-14, straight lines Z = Z may be drawn along
with the Z + b'(Z) c;urves. For example, referring to the curve for section 2, the
intercept AB between the line Z "" Z and the curve for Z+ F(Z) is "qual t{) F[Z) for
Q = 431,000 cfs at the water-surface elevation of 754.00. This intercept A. B multiplied by 1.34 gives the wrresponding intercept A'B' in Fig. 10-15 for the same elevation but different discharge. In this manner, the point Bf Bud, hence, the Z
F(Z)
curves for Q = 560,000 cfs can be plotted. From these curves, the required watersurface' elevations are determined, as shown in Table 10-11.
TADLE
10-14.
10-13 BY
THE EZRA
METHOD,
River
mileage
li77 .58.
377.78
377.94
378.33
378.65
755
756.
'2
3
4
5
Z.
+ F(Zd z. + F(Z,)
= 500,000 ds)
I Water-surface
elevation Z; ft
~
754.83
755.20
755.56
755.83
754.83
755.20
755.56
755.83
752.30
752.97
753.84
754.80
755.18
,290
Example 10-14. Solve th(, problem in Example 10-12 by treating the eddy losses
separately. It is assumed that the values of the roughness coeffident n do not
include the effect of eddy losses and that the latter is to be computed ILS 50 % of the
change in velocity head when the velocity head is decreasing downstream or as 10 %
of the increase in velocity head when it is increlLsing downstream.
Solution.. Curves of Z + F(Z) constructed for Example 10-12 (Fig. 10-H) are
also applicable to this solution. The computation for the determination of t.he wD,tersurface elevation is tabulated in Table 10-15, in which the numbers in parentheses
755
Nt-~~~
t-t---t.... t-t-
\fJ~OOOM
C"Ir-~..-I-::t'
NNM-,jI ....
to l..() L.r:! lO
l- t- t- t- t4"J
*;::;r::-~G~
i-f;)r O.f.,Vz/2g
r~
'-""-"~
__
l-
11"
(/)
(/)
'"
VJ
~ ;8
.8
a.,,.
,,<
+
N'
-"-------.,
754 I
.;;
c
.9-
u.,
"
rJl
tOt-lQ"ItI
..-I'd'tt-G
-,j<",-,j<-,jI
It;!1.()1J)'U')
t- t- t-.t-
N'
- - ; - - .-;:N~.n.--::-,;<..--:......
::;-I
-00'00
0000
~\,I
0;
:g'"
~
.'.,
'"
Il'l
'0
C1
'0
753
vl~ NI~
~ c 5 c0
:r;
:fi~~
'" ",,,
(J)(/)Ul
"
U)
I-
752
"
"
0.2
0.4
~
0.6
<,0"
]~
o.a
ka VZ /2 9
......
tL.,l
o~
. C"')
.. ..
~t-O>
~c.:.
IJ)
If)
It)
t-t-t---r--
g+
+ F(Z.).
t'- LC
.-----,...........
C'l
OJ
......-t
:8~3~
--
indicate the order of the computation steps. The curves ~f 0.5,. V'/2g and 0.1,. Y'/2g
are then plotted (Fig. 10-16). The:;e 'curves are used to correct for eddy losses, as
'
,
explained previously (Art. 10-5).
Starting at section 1 with a water-surface elevation of 752.25 (see step "1 in Table
10-15), the ,value of Z. + F(Z,) is found from the Z, + F(Z,) curve for section 1 as
754.14 (see step'2). The corresponding values of 0.5,. Y'/2g and 0.10: Y'/2g are found
from Fig. 10-18 to be 0.89 (see step 3) and 0.18 (see step 4), respectively. Taking the
value of 754.14 to the Z, + F(ZI) curVEl for the next upstream section 2 (see step 6),
the corresp6nding water-surface elevation is 752.72 (sEle' step 7), and: the value of "
0.5",Y'/217 is 0.78 (see step 9) and that of 0.laY2/217 is 0.16 (see step 10). Since the
.,.;
......,
-.!'..,;t-,jl-,jl
r..
~:I,CJlQtClJ)
..-~ ~~
t- t-.. t- t-
": ~;;a ~
I----~~--=-=-=--I ~
~
e:5~~~:g
t-~l-or-cooo
~~~~f.;
JS
I----~-----------l~
.,,; 0
lJ.la
291
~.
METHODS OF COMPUTATION
292
)
velocity is in~ren.sing downstream (or decreasing upstrenm), the eddy- los8 is 0.18 0.16 = 0.02 (see sLep 11). In order to include the effect of this eddy los5, add 0.02 to
754.14 (see step 2) which was previously found to be the value of .2. + F(Z.) a.t section 1. The resultIng v!l.lue is 754.16 (see step 13). Referring to the Z, + F(Z,)
Curve for section 2 with this value, the "'ater-surface elevation is found to be 752.74
(see step 14).
.
Repeating thll !l.bove procedure a.nd correcting for the eddy losses accordingly, the
,vater-surfllce elevations at all sections over the reach nre obtained.
a.
PROBLEMS
10-1. Show that the flow 'pro file in a wide horizon tal channel. may be axpressed ns
\.
:c =
"'uC' (v - '!y"
~) + const
(10-61)
(l0-G2)
10-3. Complete the computation of the flow profile in Example 10-2. ,
Complete the computation of the flow profile in Example 10-5, nnd compare
the result with that of Prob. 10-3.
10-a. Show Lhat the .flow profile in a frictionless rectangular channel ma.y be
expressed' as
10-~.
:c =
_V,
2,1).
r_ 2 -
3M:
+
Y
=;
(it:)3]
v
graphical-integration method
direc t-in tegration method
direct step method
standard step method
10-f)' A free fail instead of the dam controls the depth at the downstream end of the
channel described in Example 10-1. Compute the flow profile from the control to an
upstream section where the depth of flow is 1 % less. than the normal depth, using:
.10-7. Compute the flQw profile in the chl\unel described in Example 10-1 if S. =
0.0159 .. Assume that the profile starts Ilt the dam site at a depth of 5 ft and terminates
upstre~m at a .critical depth. Use:
'
a. The
b. ,The
c. The
d. The
a. The,graphical-integration method
b. The direct-integration method
c. The direct step method.
d. The standard step method
10-11. Solve Example 10-3 if S, = O.OOlo,using:
,(10-63)
+ const
293
\
i,
The
The
c. The
d. The
(t.
o.
grE1phical-integration method
direct-integration method
direct step method
standard Ilt,~.p method
0.OlG9,using:
a,
ustng:
The
The
The
The
gmphical-integr!l.tion method
direct-integratioll method
direct step method
stand~rd step method
" ,
,
. ' P b 10-8 is horizontal, compute the flow profil"
'h'~ the depth of ~0\V is 10 n. U~c:
10-14. If the cha.nnel cescnbed III rO. t'
from the control section to an upstreu.m sec lon'~ e o l ,
,
294
METHODS OF COMPUTATION
a. The flow profile between the reservoir a.nd the control section
b. The friction loss through this spillway measured in it
c. The elevatio11 of the pool level in the reservoir
10-19. A prismatic tntpe20idal earth spillway with '3: 1 side slopes and a bottom
width of 75 Jt is discharging 1,500 cfs. The spillway h!ls Il horizontal bottom upstream
from a. critical control section for a distance of 90 ft Itnd an advene slope of 10: 1 foro.n
upstream distance of 100 n. from the reservoir. Manning's 'II. is ,0.035. Using any
. method given in this chapter, determine:
,\\
., ,
295
10-20. Solve the problem in Ex::o.mple 10-10 by the standard step method, io~ a
design discharge of 500,000 ds.
10-21. Carry out 'the computation in Example 10-11 by brep.king up reach 1-5 into
four short reaches at tho intermediate sections.
10-22. SOlve the problem in Example 10-11 by the stage-fall~discharge method for
a design discharge of SOO,OOD cis.
10-23. Solve Example 10-1 by the stage-fall-discharge method.
10-24. Solve Example 10-1 by the Ezra method.
10-25. Solve Example lO-i'by the Ezra method if Q = 500 cIs.
REFERENCES
1. A. J. E. J. Dupuit: "Etudes tht\oriques at pratiques sur Ie mouvement des eaux"
("Theoretical and Pmctical Studies on Flow of Wo.ter)J), Paris, 2d ed., 1863'.
2. J. A. Ch. Dresse: "Cours de mecat:.ique appliquee," 2e j;:artie, Hydraulique,
(tr Course in Applied Mechanics," pt. 2, Hydraulics}, Mallet-Bachelier, Paris, 1860.
3. F. Grashof, "Theoretische Maschinenlehre" ("TheoreGic~,l Course on MD.'
chines"), Leipzig, 1875, '(01. 1.
4. M. Ruhlmann: "Hydromechanik" (" Hydromechanics "), LeipZig, 1st ed., 1857 ;
Hahnsche Buchlumdlung, Hanover, 2<1 ed., 1880.
5. G, Tolkmitt: "Grundlagen der Waaserbaukunst" ("F'llndamentals of HydrauHc Engineering"), Ernst & Sohn, Berlin, 189S.
6. U. Masoni; "Corso .di Idraulica Teoriea e Pratica" (" Course of Theoretical o.nd
pro.ctica.\ Hydraulics"), 2d .;d., Naples, 1900.
7. Boris A. Bakhmeteff: "0 Nern.vnomernom Dvizhenii Zhidkosti v' Otkrytom
Rusle" ("Vo.ried F10w in Open Channels "l, St. Petersburg, Russia, 11112.
8. Boris A. Bakhmeteff: "Hydraulics of. Open Channels," McGraw-Hill Book Company, Inc., New York, 1932.
9. Philipp Forchheimer: "Hydmulik" (" Hydraulics"), 1st ed.} Teubner Vcrlags.
gesellschaft., Leiptig and Berlin, 1914.
10. E. Baticle: Nouvelle methode pour la d6termination des combes deremous (New
method for the determination of. backwater curves), Le Genie civil, vol,71l; no. 23,
Dec. 3, pp. 4.88-492, 1921; no 24, Dec. 10, pp. 515-516,1921.
1,
~.
.1
-j
.j
i
I
296
;;Y:
0; R
~t
Ii
I'
CHAPTER
rI
I
11
PRACTICAL PROBLEMS
!
t
11-1. Delivery of a Canal for Subcritical Flow. When a cana.l connects two reservoirs .. having varyill;g levels, the discharge of the can'al
under variou~ conditions of reservoir level is called the del1'uery of the
canal; this problem was discussed by Bakhmeteff [lJ. Bakhmeteff has
treated this subject for prism:3,tic canals with subcritical flow under thr,ee
general cases. The Cl1SI3S are clMsified according to the condition of
three v2.l'iables, namely,the depth of flow Yl at the upst~eam end of the
canal, the depth of flow Y2 at the dovvnstream end, and the discharge Q
of the canal.
A. Case of Constant Yl. This is the caSE in which th~ water level at the
upstream end oUhe canal does not change (li'ig. 1l~1). The depth Yl is
FIG. 11-1. Delivery of a canal with subcriticl!.l flow IJ.nd C0l15tIJ.Ilt YI,
. .
I
I
298
299
PRACTICAL PROBLEMS
l
\
The value of thi~ discharge is Qn = /(" v' So, where I{,. is the conveyance
of the cross section with the depth Yz equal to YI and where So is the
bottom slope.
2. FLOW OF MAXIMUM DISCHARGE. When Y2 is equal, to the critical
depth Yo of.section 2, the discharge will reach its ma.ximum possible value,
since Y2 cannot be le~s than Yo and tlw head is at its mfLximum. If t.he
downstream pool level B falls below the depth Ye, a free overfall at that
depth will occur. This discharge, as indica.ted on the delivery curve, is
equal to the critical discharge at section 2, or Q, = z .. vg, where Zo is the
section factor of cross section 2.
For the determination of the maximum discharge, a trial computation
is required. The procedure is first to take a series of discharges, beginning
from Q" and going upward. Then, making Yz = Ya in each case, determine the corresponding YI. The discharge that makes YI equal to the
given d'epth at the upstream end is the Qm ... required ..
3. FLOW OF .l1.fl PROFILE. When Y2 > 11", the flow profile is of the 1];I1
type. The upper limit of this curve is a horizontal level, indicated by
az; at this condition the discharge is zero, since the head, or difference
between the pool levels, is zero. For Y2 > Y., the fiow will l:everse its
direction .. The lower limit of the 1111 profile is, apparently, the uniformflow surface an. For any intermediate flow between these two limits, the
depth Yz and thp. corresponding discharge can be determined by a trial
flow-profile computation. The procedure is first to assume a. discharge
less than Q,. and then to compnte the corresponding depth Yf. Consequently, the delivery clirve can be plotted.
:
4. FLOW OF 1112 PROFILE. When Y2 < y", the flow profile belongs tQ
the .1112 type. The lower limit is, evidently, the critical flo\y surface ac.
The Y2-Q relationship can he determined by the procedui'e described
above for the lIfl profile.
From the delivery curve, it becomes evident that the portiori NC of the
curve' is very steep, so that Qm.., exceeds Qn by only a very small amount.
It will be seen that this is true in most practical cases except for very Ehort
or very fiat canals. Referring to Fig. 11-2, it is assumed that the end
point d of the limiting 11{2 curve is loeated at a depth O.99Yn. If the
length L of the canal is greater than the length L' of the limiting J1'2 profile,' then no change in Yz between Yeand Vn will affect the condition
upstream from the point dj that is, the discharge will rema'in the same.
As long as L > L', the maximum discharge QJD.x will be pra.cticaUy equal
to Qn, that. is, the portion NCoi the corresponding delivery cllrvewil! be
practically vertical. The flow-profile equation indicates that the length
of the flow profile is :inversely proportiOlial to the bottom slope; the
smaller the slope,the longer the profile, and vice versa. For this reason,
reducing the slope will have an effect similar to that of making the canal
~
,
r
I
-1
,.
1,
Deli,ery
! ---'-------:--------,---1----- .--~-
cu"~:
~,=
I.
. conslonll
0""':'"
MZ"p'orile
\
-. -----..---
~
~--T-i--
~,
----------I c
I ~%~
-----.:....
-.__ l ,/_ _ ,_UJ__01
-.-- l
---__ ' - ,
r-- 0
---..,.
[1,
!
:
k'
11-2. Delivery of
'0"'< -~
df
tbe falling
ID~~
3.
1. UNiFORM FLOW. When :111 = 'yz = Yr., the flow is u.niform, the flow
profile. nb is parallel to the channel bottom, and the dlschar~e Qn corresponds to point N on the delivery curve. The value of Qn IS equal to
)
'!'
r,
I
~I'
"1
'I
~-
J _ -t
FIG.
o;/(yzl
.1
...
300
'-
K"
.where Ie, is the conveyance of cross section 1 with dept.h Y1 = Y2
and wher~ Su is the bottom slope.
2. }I'.fow of Maximum DischaTge. When YI reaches a depth Yn< that corresponds to a clitic111. discharge at section 2, the discharge becomes
maximaL Any depth Yi > Yrn is out of consideration, because it would
sill1ply raise the flow profile mb to the position shown by the dashed line
m'b' and, consequently, require an increase in the downstr~am depth Y2.
The~o.lue of (J . ~ is equal to the critical discharge at section 2, or Zc -iii,
where LIe is the section factor at section 2 for a depth
to Y2.
3. FLOW OF M"l PROFILE. For Rny depth 1:/1 < Y", the flow profile
belongs to the Ml type, and the
is less than Qn. The lmve"t
possible limit for YI is y.; at this condition the flow profile is horizontal and
the dL':!charge is zero.
4. FLOW OF 1\-12 PROFILE. For any depth Yl varying between Ymand
Yn, the flow profile. belongs to the 1112 type, and the discharge is less than
Q,,,"x b:llt greater than Q~.
C. Calle of Constant Q. This is the case in which the delivery of the
canal is constant,while the pool levels at the two extremities of the canal
fluctuate.
1. UNIFORM :now.. Referring to Fig. 11-4, various possible flow pro- .
files are sketched.. "Vhen YI = y., = y", the flow is uniform, and the
surface is a straight line ab pal-aIlel to the char~nel bottom. This normal
depth 11", cu.n be determined by the Manning formula for a
con~tallt
discharge Q.
2: .FLOW OF 1111 PROF'ILE. For positions above ab, the flow profile
belon'gs to the 1\{1 type. The upper limit of the 1111 profile is a horizontal
line; at this. position 112 approaches Yl + SoL as a l~mit. As this c0nditiol1
ilS approached, the difference between pool levels and hence the head, or
velocity of flow; decrease, However, the water area increases as the
depths increase; whereas the discharge, as a product of them'e[j, and the
velocity, cnn still be kept consts.nt and equal toa given value.
.
3. FLOW .OF .312 PROFILE. For positions below ab,the flow profile is of
the JlII2 type. The 10we.'3t possible position. of the JJ12 profile is a'b'; at
this position y~ is equal to the critical depth corresponding to the given
discharge Q.
.
The Q-constant Curve. The relationship behr~en the depths Y1 and 1'2
for const~nt Q can be plotted (J:"ig. 11-4). The resulting curve CN P is
known as the Q-constant curve. Several auxiliary e:urves have also been
constructed to make clearer cert,ain chamcteristic features of the Q-constant curve.
.
The N line is a straight line drawn from the Qrigin of the coordinates
and inclh1ed at an angle of 45 with the coordinate axes. This line is the
locus of the normal depth for all discharges. For any point on this line,
Q
301
PRACTICAL .PROBL'EMS
_ ,_
at-
2 for
cdrosshse~ IOn
L
tion
ept . a se,c ..
Q.
rr.
The Z line is a straight line drawn parallel to the lh~e fl;om a point on
.
.
d' t
S T from the orilrin O. ThIS hne represents the
the Y~ aXIs at a is ance I)J.J.
c
.
fil
H
+
SJ." or the upper limit of the lJtil pro e.
enee,
' .
d
con ItlOn Y2 - Y l ,
t f 11 f m the left
, the Qcconstant curve apPl'oaches this line asymp 0 1(\a y 1'0
when both YI arid Y2 become vel'Y large.
. ,
The coordinates Vl and Y2 of any point P on the Q-collstant curve 101 [l.
-. I
302
303
PRACTICAL PROBLEMs
3. As the tailwatel' level rises further, the jump will move upstream
maintailling its height a~~d form in the uniform-flow zone nb, until'
reaches point n. From there on, the jump will move upstrecun on the en
CUl've, gradually ciiminishing in height. The height, of the jump becomes
zero when it reaches the critical depth at e. In the meantime the flow
profile reaches its theoretical limit, cb l l of the SI profile.. Be~ond this
limit the incOiriing flow will be directly affected by t.he tailwater, and the
entrance acts as a submerged weir. III practical calculations, the'hol'izontallil1e Cb'" may be +,aken as the practical limit of the tail water stage.
This wi\] avoid the compntation of bUlb" and will also provide a margin
of safety;
11-3. Problems Re.1ated to Canal Design. Knowledge abollt the
delivery vf a c<onal, as described in the preceding articles, has useful
-I
it
t ______! __ ~----'~z
o
-
i=.:::::.:::::::~ ~
,
/'
t:.y
.
-=.--=
_I
--,.ily'
Conal bof!om of
jncrec!;ed slope
.[
Omin Ome.
11>-
Service discharge = Q
.. [
is
,i
I
,t
I
II
II
I
::.pplications in the hydraulic design of cana.ls. Several important problems related to such applications are described below.
A. Change n Depth Due to Changing Delivery. In designing 11. canal,
it is often necessary for the engineer to anticipate t.he fluctuation ill depth
of flow due to any possible change in delivery. This fluc(,uation in depth
can be estimated easily from the delivery curve of the canal for the given
range of fluctuation in discharge.
In most cases, except where the canal is very short or where the bo~tom
of the canal is unusually fiat, the canal can be designed for a uniform-flow
condition, because 'the maximum discharge will be practically equal to
the normal discharge. The procedure of canal design for uniform flow
h::),s been discussed in Chap. 7. The relationship between the depth and
discharge can be obtained easily on the basis of any uniform-flow formula.
such as the Manning formula.
When a canal is designed to deliver water from a reservoir of constant
pool level to a service chanilel at the downstream end, the discharge in
the canal should meet the variable demand, when a,nd as required by the
service channel (Fig. 11-6), This; wi~h a subcritical fiow, falls into the
.[
I
.)
-\
.o:.~'
.[
I
'304
P~ACTICAL
case of constiUlt Yl, A. in' Art. 11-1, in which the l~pstream depth is kept
constant while the dOWllstream depth fluctuates. The delivery curves in
Fig. 11-6 will explain how the fiuetuation in the downstream end depth,
caused by the variation in the service discharge demand, can be redlacec1.
The method is 'simply to increase the bottom slope of the canal. ,This
, increases th~ normal diseharge, and the delivery curve will change in position from ZNC to ZN'C'. It is apparent that, for the same range of Q",.~
to Qmi., the fiuclu:1tion in depth fJ.y' is reduced and becomes lestJ tha~ fJ.y.
PROBLEMS
305
,VAl
YA '= Yt
+ h. + ayg
(11-1)
Y...
Yl
+ h.
(11-2)
For subcritical flow, the term h, is the head loss due to friction and may
be expressed in terms of thE) velocity head at section 1, that is,
TI i ~
(al
(11-3)
= C'-2
g
h,
where. C. is a coefficient which has an average value of 1.25 for a wellrounded entrance. l Solvili.g for TIl from the above equation,
VI
, -vc: v'2Yh.
=, ~-
(11-4)
Q = V 1 .4 1 =
.
(b]
Jc:
,~
Al ;;2(Jh.
v~
(11-5)
In most pract.ical problems, the depth YA, instead of YI, is given. For
example, in the C[l.se of the constant upstream depth, A iJ:l Art. 11-1, the
depth YA would he given as constant. For any given condition of YA,
the relation between Q and Yl can be established by Eq. lI1-5). A curve
representing this relRtion can therefore be constructed. By means of
this so-called inflow-discharge-rating c'Urt1e, the reiations.among YA,YI, o.nd
Q can pe detei-mined.
For supercriticalflow, the flow, at section 1 is criticaL, The problem is
simplified by the fact that the rf',lation between Yl and y.{ is practically
fixed, irrespective of the entrance friction loss.
C. Elimination of Hydraulic Jump in a Steep Canal . .. As pointed out
in Art. 11-2, when the tail water-pool level is higher .tho.n th~ critical depth
in ~ st~ep canal, a hydraulic jump will develop in the callal (Fig. 11-5).
Such a;hydraulic jump is objectionable and dangerous, par,t.icularly
when
,
,
Dat~
the losses in entrance structures may be found in some hydraulics textbooks, handbooks, or other literature [21.
1
OIl
306
s'
Tailwaler
II
PRACTICAL PROBLEMS
307
to this design, the channel starts with a 140-ft-wide rectangular concrete section.
The 140-ft ,base width extends from statio'n 10 -/- 00 to the sill at station 10 + 45.
From station 10 + 45 to station II + 45 the walls converge on circular curves to a
base width of 60 ft. From station 11 + 45 to st.ation 15 + 82.50 the walls convei'ge
in a straigh t-line transition to a. base wid th of 25 ft. The channel below the transition
consists of a,'rect.angula,r concrete section with a constant width of 25 ft. Compute the
liow profile in, the spillway channel for' a design discharge of 6,200 ds. The control
for this discharge was designed to occur at theinlet sill station 10 + 45. Use ex ,= 1
and n = 0'.014 in the computation.
Solution. In t.his probiem the flow is supercritical, for the COIl trol sectib,n is placed
at the upstream stat.ion 10 + '15., The critical depth RI, this section is equal to Yo ;",
,'l' (6,2007140)'/32.2 = 3.93 ft. The critical depths at other stations may also be
computed. If the' computed depth happcns to be greater than the corresponding
critical depth, the flow will be subcritical.
The coinputation of the flow profile is given in Tnble 11-1, with the fo!lcwing
'
headings:
Col. 1. Station number
Col. 2. Length of reach in ft, w11ich is equal to the difference between the two
station llumbel's at the two ends of the reach
CoL 3. 'Width of the channel inft
Cols. 4 to 16. Same ns the st.eps from cols. 1 to 12 in Table 10-4, except t.hat an
extra column for So (Col. 14) is provided since So is not constant throughout the whole
channel length under consideration.
The computation is arranged in a forin similar to that used for' the direct step
method, but it is performed by trial and error. This procedure is introduced because
an additional variable for the channel width is involved. In this computation,the
depth of flow y is assumed and entered in col. 4 at each step. The assumed depth is
considered correct when -,he resulting value of AX entm:ed in col. 16 agrees with tbe
length of reach in coL 2. It. shouid be noted that the depth of flow computed in this
eXD:mple has been car.ded'to more decimal places than would be necessary for practical
purposes.
The flow profile thus computed should be corrected for air entrainment for highvelocity flow occurring at the downstream end of the channel.! 'When a high-velocity
supercritical flow occurs in a noriprismatic channel, it is likely that standing waves
would appear as a result of the lateral-boundary changes (Arts: 17-3 and 17-4).
-I
-I
-I
-I
g.j?;~~=
~'-
0,1>'"
'
SPILLWAY
.~
C1>
'~"
>
~I;
~~
~I;'
.~.'"
~I~
~
..ci
PROFl~
FIG.
\. SPILLWAY
,-.
TA,llLE
II-I.
~~
)j
l/}i
a:V'/2g
3.72
3.36
3.29
3.36
3.45
3 . .')0
3.70
3.87
5.78
5.04
3.88
(UO
11.27
13.48
114.85
15.75
10.71
17.80
19.00
20.35
21.62
201.14
1.97
2.82
3.42
3.S5
4.33
4.92
5.60
6:44
7.26
So ., Sf i AX
,(2!_~L
10 + 45
550.0
14U.0 :l.9~
10 + 51
{50.0
6.0 I I:~O.O 3.54
10 + 5S
417.5
7.0 120.0 3.48
10 -I- 65
7.0 1l0,0 :1.58
393.9
10+73
8.0 100.0 3.71
371.0
10.0
CJO 10 + 83
1)0.0 3.87
348.2
~ 10 + 95
12.0
320.2
80.0 4.0S
11+10
70.0 4.35
304.5
15.0
11 + 22
12.0 . M.O 4.41 I 2~6.7
11 + 45
2:J.0
60.0 4.28 1200.8
12 + 01.5 62:5
55.0 .3 123 J71.1)
12 +70
62.5
50.0 2.855 142.8
40.0
13
10
46.8 2.817 1:1l.8
13 +'32.5 22 5
45.0 2.()58 133,1
13 + 95
62.S
40.0 3.399 135.!l
14 + 57.5 62.5
35.0 3.948 138.2
15 + 20
62.5 [ 30.0 4.050 lag 5
15 + 82.5 1 62.5
25.0 .5 05Z 141.3
1
11-1
6,200 cfs)
3.75
2.81
2.57
2.51
2.61
4.90
5.04
,).22
5.44
5.72
6.08
5.83
3.97
3.52
3AI
3.60
2.!)1 . 4. 1G
3.22[4.75
3.55 5.42
3.89 I 6.11
6.. 36
0.90
7.43
8.04
8.n)
9.68
10.79
11.67
9.05 13.33
3(l.09 20.21 23.33
43.41 29.27 32.13
47.04 :34.38 37.20
46.61 33 .74 . 36.70
45.62 32.32 35.72
44..89 31.28 35.23
44.45 30 09 3&.34
43.87 I 2~.8B 35.53
0.4610.0032
0.54 0.0042
0.5:1 0.0043
.0.61 0.004.6
0.75 v.D052
0.89 0.0056
1.11 0.0060
0.88 0.00G8
1. GO 0.0088
10.00 0.0290
8.80 0.0473
5.07 0.0572
-O.GO 0.0532
-0.98 0.0442
-0.'19 0,0374
0.11 0.0322
019 0.0278
5.9
7:1
7.0
8.1
10.0
11.9
15.0
12.0
23.0
62.5
62.5
40.0
22.5
62.4
62.8
61 2
G3.3
- :1
:no
PRACTICAL PROBLEMS
eliminate cross wn.ves and other turbulence, and to provide safety for the
structure and waterway. "Yhen the transition is designed to keep streamlines smooth lind nearly 'parallel and to minimize standing waves, the
theory of gradually varied fiow may be used in the design. The essence
of such a design has been di~cu8sed earlier (Arts. 3-5 and 3-8) in connec-
in terms of the change in velocity head between the entrance and exit
seetions of the structure.
For inle t; structures, the entrance velocity is less than the exit velocity;
hence, the water surface must always drop at least a full difference
inlet
311
betvveen the velocity heads, plus a small oonversion loss known as the
In
'~his
lOllS.
The drop Lly' in water surface for inlet structures may there-
fore be expl'essed as
(11-6)
wher~ 'h.
loss.
For outlet struGtures, the velocity is reduced, at lea.st in part,in order
to lift the water surface. This rise in water surface, known as the
recovery of velocity head, is usnaily s.ccompanied by a conversion loss known
as the ouae/loss. The rise t:.y' inwater surfnce for outlet structures may.
be I?xpressed as
(1l-7)
Co
Type of Transition
Wal'ped typ'e, ......... .
Cylinder-quadrant type ........ _.. .
Simplified straight-liM type ....... .
Straight. line type .. _.............. .
Square-ended type. . .... _....... .
,
i
'"
0.10
0.15
0.20
0,30
0.30+
. 0.20
0.25
0.30
0.50
0.75
--1
....... !
-I
I
,r,
312
313
PRACTICAL l'ROBL:fj)M!;
'~ 51.40~~~~-~--~~~-.~-.,~~~~--~f~~~~~-,~_Je
"
'" 51.30
'"
'"
."
:- -,
:+o
o
'"
on
"
+
r'1
ll'l
"
.n
+,
0oJ
'"
oJ
cO
'"
1'1
'"+
HYDRAULIC PROPERTIES
Hume
Canol
-L-J"
'"
U"l
"'6.0
0.315
V'3,24
Ci'315,
5'0.0009 ,,-0.014
V'5.98
Q'314.5
PLAN
PLAN
I-
\
"
CD
~~~~~~~~~~~~
iii<
N
'"
SECllON ON CENTER
line
Top 01 bonk
SECTION D-O
,Rip-fop if r~~ui,ed
315
PRACTICAL PROBLEiHS
. head from l' - 2.715 fpl! to 5.98 fps is equal to tlk. = 0.553 - 0,117 = 0.436 ft.
Neglecting the chnpllel frietjoll fol' the time being, the total drop in water surface is,
thet'Afol'e, LlO X 0,436
0 ..480 ft.
.
For a. smooth Itnd continuous .flpw, the theoretical flow profile'may be u.sswlled 11.[ ,
two equal pll.rn.bolas, !.IJ.ngent to ea.ch other o.t point B ,\nd hori~ollt,al, respectively.
a,!; A and C, Strictly spellking, the parabolas should be tangent to the water surfaces'
in the canal and the flume, but !l. small deviation of these water surfaces from the
horiwnt.at is not important in ~he present example.
A number of sections are then aele(;t~d a!.:mg the transition, where the /lOll' prolile
will be computed and the stnlctund dimensions determined.
3~ Cc.mplilatio'll of the Flow Pmfilc Inch,diug Friction. The computation ]" shown in
Table J 1-2 with the following heo.dings:
Cd. 1. Number for stD-tioll1; eqtHI.l!y spaced every 5 ft aud lneiLsured in the
direct.ion of flow,
'
Col. 2. Drop in water surface. The tolal drop in \I'D-tel' surface from A to C,
neglecting friction, is OA80 ft. The drop from it to t.hemid-point B of the antisymmckicill reversed p~\,l'abola is made equal to half th-e drop, or 0.240 ft.
CoL 3. Change in velocity heo.d. Assuming thnt the conversion loss i~ distributed
over th" entire l~ngth of the transitiQl). in proportion'{'o the change in velocit,y head,
vo.l:Jes of tll~, are obtaine.cr by dividing value.<; of t!.lf' in col. 1 by 1.10.
Col. 4. Tot.[11 velocJ;,y heD-el, equal to the cumulative v?Jue of t!.h, ent~ring the
preceding coiumn
.
,
,
'
Col. 5. Veloci(" in ips corresponding to .the velocity head in the preceding column
Col. 6. Water area in ft~, equa.l to th~ d~"charg~ 314.5 cfs divided by tbe velocity
in the
column
CoL 7. Half the' top width in ft, obtained from the cross sections or the sketched
plan
11-10). The piau m~}' be chos@ either ~rbitmrily or by trial untilsatisfactory reslll!;s !l.re obtained., The choice of a proper shape for the plan is a matttlr
of judgment ..
Col. 8. Half the bottom width in It, obtained f"om the, ~ketched plan
tV.)!. 9. The dt'pth of flow ill it, equal to A/CO.aT + 0.5/;)
Col. 10. Hydraulic nlrliu& in ft
CoL 11. Til'.! friction slope, computed by Eq. (9-8) with n = 0.014 for all sections
the tmnsition
If
Col. 12. The friction head, equal to 'the dist.a.nce between stationli, or 5 ft, multipliedby the avernge of the friction slope of the section and that of the preceding section
Col. 13. Cumulative frir:t.ion head in ft
Col. J 4. The water-surface elevation, including the Ilffect of the ch~,l1ncl friction,
- 'Zhf. The flow profile thus obtained should be free from
equal to 57.41
objectiona.ble irrllgularitics;' if not, the plan ma.y ')B t\:tered. It shOuld bG lloted,
hQwever, that, a slight change in th<! elevation of waier .;;udace at a. given POlot may
cause an r.ppreciable chltr.ge in the dimeMions of the skucture,
COl. 15. The elevr.tion of the channel bottom, equa.l to Z - y
4, De(enlll:rmiion of Situct.Ul'd Dimerl$ions. After t,he arbitl'lirilr 6kf:tched plan
for'the tral;sition.is fOl~nd to be satisfactory, th~ s,tructnral dinlensions may be dlJf.ermined ns given under the following headings of Tnble 11-2;
CoL 16, The side slope z =' (Q.ST - 0.5b)/y
Col. Ii. The elevation of tile top of lining. The l'ecommend<3d heigl'lt of lining
above the witter surface for a discharge of 314.5 cfs is o.bout 1 ft (Fig. 7-]). Draw a
strci.ight line above the computed water sui-face at an
distance of approxima.tely 1 ft. From this line the elevation ZL IS obtained.
314
~Qer9Y
energy
line
rne
energy
line
d water .st.,foce
Tunnel
for tramitions and friction losees, and a safe coefficient of roughness was
used to determine the depth of flow in the tunnel. After construction,
however, the transition losses were found to be practically negligible; so
the actual normal depth of the floY!
the tunnel entrance was
cOl1sirler::.bly less than the assumed vahle. As a result, an objectionable
hydraulic jump was observed inside the tunnel. Actually, the transition'
should have been proportioned to avoid the jump on the basis of negligible
t.ransition losses. Since the structure had already been constructed, the
hydraulic jump was iinally eliminated by bolting cross timbers to the
channel bottom, thus increasing the friction and bringing the normal
depth up to above the critical depth.
Example 11-2.' It is required to design an inle:t st,ructure connecting an earth ca.na.l
huving 8, bottom width of 18 it and side slopes of 2: 1, to It rectangular concret'l flume
12 it {} in. wide, The hydrauli9 propert.ies of the cana.l and the flume arc given (Fig.
11-10). The design discharge is 314.5 cis.
Solulion. The design procedure~ involves thll following steps: '
L Delermimr!ian of the LmgU; of the Tra:ll..'l1:tion. The length of the transition is
detllrmined so that B, straight line joining the. flow line !It the two ends of the transition
will ma.ke an angle of a.bout 12.5" with the 'axis of the strnetllle. This length in the
design is found to be 50
1 .
2. DeterminatiDn. of fhe Flow Profile Neglecting Frict.icn. For the type of stl'uctul'e
under contemplation, the inlet loss
be safely assumed to be 10% of the change in
velocity head, or 0.1 ail-". ,The total
in water surface is, thel'efOl'e, equul to L1
ah. pillS the drop necessa~y to overcome channel fridiOn. The change in velocity
ft:
in
-I
I
I
PRAC'l.'ICAL PROBLEMS
317
+ 0.50
Col. 20.
OI.l?CIt:lO\,/')OI.t)0U10
OO
__
(\l<:,-~c-;tf;l"';l'l""1t:I
+++++++++++
00000000'000
316
are
PRACTICAL PROBLEMS
0.'
... h~
c,
'"(;
J
t;:;
;:'
::>
0
a.
,-,
Z
0
:.:: 9!.
c..
U;
W
I-
UI
tr
"
r.I 0Z
'I:
t)
'..J
65
i=5:
+0
<.l
>-
(0
'
WI-
~-;
t;
::1,
.;
C
.9
.<;:l
Ii)
C!
;:!
..,
+'
'::l"
:;:;
0
"0
..,m
,2;)
318
319
- I
--I
-\
"\
320
(a)
,
,,
,,
Wo!er-surface elevation at
id
(bj
fOI'
321
PROELE:&!S
proceed upstream from the d~wnst,eam point B Where the divided flows
unite again. The initinl water-surface elevation at puint B may be
determined from the
curve at this statiori for a total discharge Q.
The computed water-snrh1ce elevation at point A tor channel 1 is then
plotted against the sa,me for channel 2. A ct,U've may, therefore, be
dl'a\vn (Fig. 11-14b) fo1' several assumed sets of :dischal'ges Ql and Q2 of
different proportions. :Since the flow is divided ~t point A, the two computed water-surface elevations at this point fOl~ the channels should be
\;!qual if the assumed qivisiol1 of flow is correct.: Thus, the elevation at
point A corresponding to this correct division of flow may be obtained
from the plotted c.urve at which the curve is Intersected by a.dashed line
bisecting the coordinate axes. The dashed line represents the condition
that the two computed elevations are equaL In the meantime) the computed elevation at point Afar chfml1el 1 is plotted against the discharge
Q1 (Fig. l1-l4e). From this curve the correct discharge Ql may be
obtained for the correct elevation. The corresponding discharge in
channel 2 is, therefore, equal to Q2
Q Q1.
If the dividfld flows are supel'critie~\l, the control point will be at A;
hence, division of flow will depend on the entranee condition of the divided
channels. During the normllJ-flow eonditioll, it may be assumed that nil
flows ~\re uniform, and division of flow may be roughly determined from
the following I'elations: Q1
1(1 vs;., Q2 = K2 -VS;, and Q = Q1 + Q2.
11-10. River Confluence. SVhen the flow-profile computation is
carried upstream through the confluence of r. river and its bributary, it
is neeessary to determiQ.e the water-surface elevations immedIately
uostream hom the confluence. The procedure for solving this problem
is' illustrated in
10-10 fOl'the confluence of the Missouri and
Kans&s Rivers. In this example, a dischi.rgeof 81,000 cfs from the
Kansas River is combined with 350,000 cfs from the Missouri River to
give'a total discharge of 431,000 cfs immediately below the confluence.
Cross sections lK and {) are located immediatelyupstl'eam fromthe confluence of the two streams (Fig. 1O-1l). In Tr.ble 10-8, the hydraulic
elements, ;'elocity heads, ancl total heads are computed separately at
the two sections. The velocit)~ head at cross section 5, is, compnted for 11
total discharge of 431,000 cfs. This value is divided between the cross
sections 6 and lK corresponding to th9 dischal'ges of 350,000 cfs f.l1d
81,000 cis, respectively. The friction slope at each cross section is computed fol' the discharge of 81,000 cfs at cross section IK and fOl' 350,000 ers
at cross section ().The friction head loss hJ is then computed, using
thA average of the friction
from cross sections 5 to lK in the
Kansas River and from cross sections 5 to 6 in the Missouri River.
At the confluence, eddy loss is usually high, In Example 10-10, this
is estimated as 10% of the increase in velocity head from cros~ section
II{ to 5, or 0.10 ({l.G3
0.13)
0.05 ft. The total energy in cross section
.5 is, therefore, equal tohJ
h~ + H = 0.10
0.05
755.0,1: = 755.HJ.
Subtracting the velocity head' from this value gives 755.19 - 0.13 =
755:06, which should be equal to the assumed watel'-sul'faceelevation ab
cross section lIe.
,
Th~ above method should bli:flpplied to sub critical flow of relatively
low velocities l1Qt exceeding about 10 fps. At high velo'cities, the eddy
10sses'al'e high, and the error il1~olved in the estimation may become quite
appreciable.
The problem of riYer confluence may be further illustrated by an exam~
322
,323
PRACTICAL PRDBLE.IS
pIe given by Sto~;:m' [10], using the Junction of the Ohio and Mississippi
Rivel's (Fig. 11-15). The initial conditions of the uniform flows in the
Upper l\::Ii.~ssippi, the 'Ohio, and the Lower Mississippi Rivers are,
respectively, as follows: the normal depths, Ynr, Yn2, and tina; the channel
slo'pes, SOl, S02, and Soa; and the roughness coefficients, nI, n2, and n"
Fwm these da,ta the normal
Qnl, Qn2. and Q,,3 call be computed.
Now, ,a flood wave is assumed to initiate in the Ohio River at a place L
flow condition is expected to develop; at thifi time the depth at the pla.ce
of flood initiation becomes Y2, but ~he depth far upstream in the MissisRiver remains Ynl. As a general rule, the backwater effect in a long
sr,ream resulting from even fairly large discharges of its tribntaries aoes
vel"y far ,upstream. Therefore, the backwater curve in the
not
Upper Mississippi Riyer should be l'elatively short. At the final steadyflow condition, steady b:l(',kwater curves will be formed in the Ohio and
Upper Mississippi Rivers. Downstream in the Mississippi River, the
depth will change from Y1l3. to Ya, but the flow may beasslimed to
remain uniform.
ThE:; flow involved ill the [I.bove-mentioned problem is considered to be
subcritical, which is the usual COl.'le and can be verified easily by a camput,~tiol1 of the Froude numbers. At the final steady-flow condition, the
following conditions are evident: (1) the discha.rge in the Upper Mississippi River remains the same, or Ql = Qnl; (2) the depths of the three
channels at the junction n.re all equal to Yi at x = 0; and (3) the Slim of
the
from the Ohio and Upper Mississippi Rivers is equal to
the discharge in the Lower Mississippi River, or QI + Q2
Q3. By
assuming a value of Vi> Q3 may be comp:lted un1t.hell
initiation of flood
I
.. I
Point of
floDd
Since YiJ Q~, and Y2 ~are ilow known,. the leng!;h L 01 the backwater curve'
can be computed. If the computed L agrees with the given L, the
assumed Yi is the correct value. Otherwise, new vulues of, Yi should, be
assumed unt.il a correct valLle is obtained. This is a trial-and-error
solution:. Ii solution may 0,180 be obtnin<3d:byassurning Q2 at the beginning l1l1d ~heclr.ing finally for either Vi or L.
PROBLEMS
11-1. A rectangular channelS ft w'ide and 500 'ft long connects two reservoirs of
va,rylng surface levels (Fig. 11-16). Assuming 3. frictionless channel, construct the
Q-constant curves of Yl
J(y~) for Q = 10, 50, 100, 150, a.nd 200 cfs, The loss at
t,he channel ent.rance is negligible,
~T
y.
1
FIG. 11-16. Profile of
~T Ya
11
It CMlld
')\
324
II,
.yzg
n. The flow p::-ofiles for discharges of 600, 800, 1,000, 1,200, 1,500, and 1,800 crs in
the spillway
,
/), Curve showing dis(.\harge \'5. elevation of pool level in the res,ervoir
c. Friction loss in spillway between the reservoir and the control sectioll ror I.l. discharge of 1,500 cfs
.
11-1"{. Revle\v the design of nil O\lt.let tl'll.l)sition from Ilume to canal (Fig. ll-ll).
The flume r..nd canal are the same as those described in Exnmple 11-2.
11-18. Review the design of ihe siphon inlet and outlet transitions (Fig. 11-13).
GiYen th.e hydraulic properties: A = 01.52 it', R = 2.89 ft., SQ = O.OOOS, nnd n =
0.0225 fa! the canals; and A = 38.48 ft2, R = 1.75 ft, So = 0.0012, and 11 "" 0.014 for
the circular siphon ba:reL
11-19. Compute and construct li bnckwater envelope curve for the backwater'
caused by a 5-ft dam in the channel described in Example 10-1,
...,.,,---l(.-8'
? ~----'-:-~
FIG. ll-17. A raft chute for PIC'b. 11.13.
Determine the following items by the Manning formula and by any method of flowprolile'computation as described in this book:
a, Slope of the steep reach
325
PRACTICAL PROBLEMS
11-2. Construct the Q-constuut curves of YA = f(vB) for Prob. 11-1, The entrance
discharge ,nt\y he eomputed by Q = 3bf]1.l where b is the channel width a.nd H is the
tot.a.l hettd.
11-3. A tmpe;;oidal channel connecting two reservoirs 2 miles Ilpart has b = 50 ft,
z = 2, fr = 1~ n = 0.025, and So =; 0.004. The upsireamdepth V, is mninto-ined
consLant !).;ld equal to 6 ft (Fig. 11-1). Construct the d'olivery curve Q = f(y.),
11-4. COllstruct the delivery curve Q = f(ytl for Prob. 11-3 if the downstream
depth:v~ i.; lTIaint(Lined constant and equal to 6 ft (Fig. 11-3).
11-0. Assuming a variable y, in Prob: 11-3, construct !l. curve of Y, against Q", .. for
0. range of diEcharge varyiag from 0 to 2,000 cfs.
il-6. Assuming v1l.riable depths !/t and y, in Prob. 11-3, construct the Q-constant
curves fordischalges having tile normal depth equp.l to 2, 4, G, and 8 ft, respectively.
11-7. Sulve Prob. 11-3 if tht: reservoirs l\re 2,500 it ap:.i,rt.
11-8. Solve Prob. lJ-4 if the reservoirs Me 2,500 ft c.part.
11-9. SolvePIob. 11-5 if the reservoirs !lre 2,500 ftapart.
11-10. Solve .Frob. 11-6 if the reservoirs are 2,500 it llpRI't,
11-11. Construct the delivery curn Q = f(YB) for Prob. 11-3 if the upsLream reservoil' depth Y II ia Dlaintained con~tant and equal tc B ft. The fzee entranoe is wtlllrounded, or C, - 1.25.
11-12. Construct tbe delivery curve Q = f(Y,J) for Prob. 11-4 if the downstream
resp.rvc-ir depth YB is maintained constant and equal to 6 ft. C, =1.25.
11-13. A rectangu~ar ra.ft. chute 20 ft widc, as descril::ed by Bllkhmeteff [lj, is to be
built between levels A. and B(Fig. 11-17). The design conditions ,,,'e:
L The'ta.ilwatm fluctuates by 8 ft.
2. The depth ill Lhe steep reach is kept to a minimum nav'igl1ble depth ~f 2.5 ft.
:l. The (lvCl'nge velocity should not exceed 14.7 p8.
4. The dischal'ge should be kept below 750 ers,
5. The roughness coefficient n = 0,03.
6. The entrance discho.rge is computed by L\ weir formula Q _ 0.4
by", l.~.
frcl~board
of 5 in.
11-20. A dil!charge of 1,000 cfs is divided between two rectangular channels exoavated in rock. The chnnneis are later joined !J,gain (Fig. 11-14). Channel 1 is 10 ft
wide and 200 ft long. Channel 2 is 15 ft wide lend 150 ft long. Thebottom of channel2 is on the llverage 'about ::l ft lower than that of cho-llnel 1. Assuming n = 0.035
aml a tot'll drop of water surfa.ce between the dividing !lnd joinip.g poillts of the chan7
nels equal to 5 in., compute the divided uniform flows ..
11-21. Wi~h reference to the problem shown in Fig. 11-15, the following data are
assumed: Yr.' = Yn~ = 1/ .. 3 = 20 ft, SOL = So. = 0.5 ft/mile, S03 = 0.49 ft/mite,
n, = n . = no
o.oa, the widths of the Ohio and Upper Mississippi Rivers = 1,000
ft, the width of the I,ower l\{ississippi Riv~r = 2,000 fti II> = 40 ft, and L = 50 miles.
All rivers are assumed to have rectallgular channels. Determine the junction deptl\
and the flow profiles ill the rivers after the flood flow approaches a .stead); condition.
REFERENCES
1, Boris A. Bakhliletefi: "Hydraulics of Open Channels," McGraw-Hill Book
326
2. Julhtrl Hinds: The hydraulic design of flume and siphon ~ransitions, TrcT.nsadions,
America" Society of 'Civil Engineers, vol. 92, pp. 1423-1459, .1928.
'
'3. "C;vil Works: Flood Con~rol in the Los Angeles Area," TIl[) Engineer School,
For~ Beh'oir, Virginia, 1950, E206.00 (4-50) ML, pp. 22-28 and plate 10.
4. L. S~alldi6h Hall: Open channel flow at high velocities, in Ent.rainment of air ill
flowing \\'flter: a symposium, Transactions, Ameriwn Sodetv oj C;~il Engillec",3,
vol. lOS, pp. 1394-1447, 1943.
5. Fred C. Scobey: The flow of water in flumes, U.S. DepC!1'lmenl oj Ag"ic7l!t1l7'e,
Techniml B1t/lelin No. 303, December, 1933.
G. Hrdraulic design d!\1;a, appendix I of CanalS and related structures, U.S. Bure(1.(L
oj Reclamalion, De~ign and Cons/rllction .liJ(l.n1Lal, Design Su.pplwtent No.3, 1952,
vol. X, pt,. 2, paragro.ph I-13.
'
7. Julian Hinds: The hydraulic jump and erit,ical depth iu the design of hydraulic
S~r\)ctllres, Engineering Ncws-RsClJrd, vol. 85, no. 22, pp. 1034-1040, Nov. 25,
1920.
8. WallD.ce M. La.nsford and WilliD.1n D. Mitchell: An investigD.tion of t.he backwater
pro(ile for steady flow in pi'isnl:t~ic channels, University oj Illinois, En!1ineering
EXpcTl:lI1cnl Stalion, Bulletin Series No. 381, vol. 46, no. 51, lViD.1'ch, 1949.
1.1. Willi"'ffi D. Mitchell: Sf,age-fall-clisoharge reiatioils for steady flow in prisma~i~
ehfl.Tln<:.ls, U.S. Geolo!!ica.18urvey, Wa.ler Su.pply Paper 1164,1954.
10. J. J. St.oker: "Witter Waves," vol. IV of "Pure and Applied Mathemat.ics,"
Interseience Publishers, Inc., New York, 1957, pp. 45G-461.
CHAPTEIt
12
.j
-I
'.
328
329
from the study of flow in roof gutters [lOJ to the design of wash-water
troughs in water-treatment plf1nts [11,12] and of side-channe1 spillways
on dams (Fig . .12-1).
B. Flow with Decreasing Discharge.. Fundamentally,
type of
spatially varied flow may be treated as a flow diversion where the
diverted. watel; does not Ilffect the energy head. This concept has been
verified by both theory and experiment,. Therefore, the use of the energy
equation will be found more convenient in solving this problem.
The theory of spacially varied flow with decreasing
was
probably employed first in the design of lateral spillways or side-spillway
wehs. This type of structure is usually a long notch installed along the
side of a cha.nnel for the purpose of diverting or spilling excess flow.
Laboratory tests on such structures w~re first made by Engels tI3] and
by Coleman :and Smith [14J.I Forcl~heimer [15J has approac.\1ed th~
1 Engels's ex~eriments indi~f.lted a rising and those by Coiemnn and Smith, on the
contra.ry, a dropping flow profile along the spUlwa.y crest. This confusion \vas lEIter
expla.ined by De :!vI.arahi [16] as resulting from the f8.Ct thu.t flow was sllbcritical iil
Engels's but supercritical in Coleman and Smith's experiments.
QV
where w is the unit weight of water, Qis the discha.rge, and V is the veloc- ,
330
ity.
Similarly, the momentum "' ..,""',' ... section 2 per unit time is
~{] (Q
+ dQ)(V + dV)
where dQ is the added discharge bet\veen sections 1 and 2. The momentum change of the body of water between sect.iolls 1 and 2 is, therefore,
equal to
'
~ (Q
(J
+ dQ)(V + dV)
:! QV
f!
~o [Q elV
g
+ (V + dV) dQJ
,
W sin
u;So(A
f) =
::
+ J1: dA) dx
331
P2
w(z
+ dy).t1 + ~ dA dy
The
w8 QA d,o;
where dy is the difference between the depths of the two sections 1 and 2.
Neglecting the term containing diffel',mtials of higher order,
w(z +dy)A
-~.
where slope So is equal to sin IJ and th.., term containing the product of
is dropped.
p. - P z
w A dy
Equating the momentllln change of the water body to' all t;he external
forces acting on the body,
w
'
(V
g '
- [Q dV
+ dY) dQ]
"
=
P1
P2
.
+W
sin
f)
FJ
(12-1)
The friction head between the two l3ect.ions is equal to the friction
slope SI mUltiplied by the length dx, or
.
hf
Sf
ax
The frictional force along the channel wall is equivalent to the pressure
due to friction head multiplied by the average q,rea (see Art. 5-4) J or
Ff
w(A
+ Y2 dA,)Si dx = wASf dx
.Jo
If.-
S/ne'
1(
"f! . V dV
+ dV
(12-2)
+. AV dQ ) + (Se - S,)dx
(Q + dQ)/(A + dA), the above flqua-
ely -
{J
(So - Sf) dx
( 12-3)
(
I
.1
'
333
This is the, dynamic equation for spatially varied flow with increasing
discharge. Theoretically spealdng, a momentum coefficient should be
used in the equation. However, the energy coefficient is' used because
the friction slope Sf is evaluated by a formula for energy loss, such as the
MaIming formula. l When q. = 0, this equation becomes the dYnlimic
equation for' gradually varied flow of constant discharge.
B. Flow with Decreasing Discharge. ,For t.hc analysis of this type of
spatially varied flow, the energy principle is directly applicable. Let z
.be the distance of toe bottom of the channel section above a horizontal
datum (not shown ill I~i.g. 12-2); tIH') total energy at a channel sectiou is
Adding this term to the right side of Eq. (12-6) and differentiating) the
resulting equation will be identical wit.h Eq. (12-5).
12-3. Analysis of Flow Profile.. In discussing the analysis of the flow
profile, a simple example will be given first.
H = z
CiQ2
+ y + 2g112
(12-6)
+ dy +
dx
a (2Q dQ
2Q2 dA)
2g A2 dx - AJ dx
-Sf, dz/dx
-So) dQ/dx
(dA) (ddxY) -_
dx,dy
~,! =
(12-7)
q., and
dy _ So - Sf - aQq.luA2
dx- - 1 - CtQ2/gA~
?, '
(w dQ dt)(a V2)
IJw(A d:c)
b'y.3)
1(LZ + lJ2'1.
(12-8)
which is' the dynamic equation jorspat'ially varied flow with deer'easinu
_ _-;,.. discharge. It should be noted that this equation differs from
(12-5)
only in the coefficient of the third term of the numerator.
Now it is interesting to know [33] that the momentum principle can
also be used for the derivation 'of Eq. (12~8). In 11. spatially varied flow,
with decreasing discharge, no momentum is added to the water. Followa ,procedure similar to the derivation of. Eq. (12-5), the term containing dQ may be dropped from Eq. (12-1); the resulting equation will
be identical with Eq. (12-8).
_~ Likewise, the energy principle can also be used for the derivation of
Eq, (12-5). In applying this principle to spatially varied flow with
incre~sing discharge, the energy due to the added discharge dQ per elementmy lellgth d:r: should be added to the total energy
the course
of the flow during the time interval dt. This kinetic energy per pound of
water is equal 'to
mass X velocity2
g X unit weight of water X volume
(12-10)
where I: is an integration con.taIlt which may be determined by the bOUl,da,y conditions of the flow profile,
At the outlet, x = L lIud 11
Yo, Thus, Eq. (12-10) gives
TdiY
dx
(12-11)
11.
where
1 +_1
2F.1
).JL_
y.
('J!)3
y.
(12-12)
(12-13) '.
F,' =
It is apparelit that F, is in the form of n Frqude number oi the flow at the outlet.
When free overfali occurs at the outlet, the :!low is critic",l; therefore, F, = L The
equation of the flow pre file becomes
11
(11)~
%v. -H -y.
(12-14)
The above example illustrates spatially varied flow in horizontal channels of rectangular -cross section. Similar analyses can be made for
channels with parallel side walls having irregular bottoms and for channels with sloping side walls. For channels with sloping bed, hbwevel', a
334
335
(
I
Value of G
FIG. lZ4:-SO"luiioDs for channels with sloping bed and par:>ilel side walliJ.
W. H, Li [5].)'
,
.
FIG. 12-3. Types of spatia.lly varied flow as determined by F, nnd G.' The diagram
is shown for [\, channel of an arbitntry sec~ion. Fvr rectangular channels, the line
dividing regions Band Cis G "" 1 + F.; for trio.ngular channels, it is G '" 2. The
dashed line in each channel is the critical-depch line.
(After
I~
336
,(
\
of the \,rater surface at the outlet will not affect the entire flow profile.
'The flow profile upstream from the
. jump cannot be determin!ld froni the
: value of YO) but it can be determined
from the critical depth yaand the
position of the critical section Xc. A
dimensionless flQ~"
in the
sllpercritical reach ('[;~ig. 12-5, ill
which A" and A .!we water areas,
respectively, of the critical section
and the section a distance x from
the upstream end of the ch~nnel) has
been computed by numei'ica[ inte~
gratioll. This curve can be used to
FlO. 1,?-5. Dimensionless flow profile in
compute the flow profile below the
Btipercriti!\a~ reaches in a sp .. til1!lyvaried-liow cht,nnel of sloping bed and critical section and. above the hyparallel walls, (After W. H. Li [5),) , draulic jump.
.
The positi'on of a critical section
in the lateral-spillway channel can be determined by the method of singular point (Art. 9-6),
Regio-n D. This region represents the flow,condition Fo.t which there is
supercritical flow throughout the downstream portion of the channel but
where the depth of submergence at the ouMet is not
enough to
create a hydraulic jump in the channeL Thus, the value of Fa is not
determined by the depth of su'bmergence. The dividing line between
regions C and D (Figs. 12-3. and 12-4) was obtained by numerical integration on the condit,ion of a millimum depth of
required
to prOdtlCe a hydraulic jump at the outlet. This Jnllllmum depth 'is
. equruto the depth neeeSS1J.ryto set the downstream pool level at a seq).lent
depth. A depth of submerge'nee greater than this. minimutn depth wlll
force the jump to move upstream into the channel, and the oondition of
flow Ifill be represented by region C.
When the slope of the channel is extremely steep or when the value of
G is very large, the flow ,vill become unsteady. The limiting value of G
that will keep the flow in a steady condition has not
determined.
In the above a.nalysis, the' effect of friction has
This
has been verified as justifiable for the design of wash-)vater troughs and
side-channel spillways. For effluent channels around 6ewnge-treatment
tanks. however, the effect of friction may increase t~e upskeam depth
Yu ~s :much as 10%.*
'
),
relmitof friction in
337
Example 12-2. Analyze the flow in iii rectangular ch:wuel .of .small slope with a
bottom mck (Fig. 12-6) and derive the equation of the ,flc\v profile.'
SoluiiOn. The flow in a channel with bottom rack is a case of spatially vaded flow
.with decreasing discharge. The ra.ck is usually made of parll.lJel bars 01' perforated
screen. There arC va.rious application;s of such 0. device. For exo.mple, the channel
may be' an "intake" to withdraw'w{l;ter
from, say, a mountain torrent, or a "skim, mer." to reduce the ,volume of wa~er required to transport, say, fish.
A~s;lming '" = 1 !l.nd 8 "" 0, the speoific
Bn"rg'y at any section of the channel (Fig.
12-6) is
E
V'
=. '11 +"
~g
11
Q'
+ 2gb'11
(12-15)
For a spatially vaxicd flow with decreasing discharge, the specific energy
can be considflrcd constant aiong the
channeL' Thua, dE/dx = OJ or, from
Eq. (12-15),
(0'
(12-16)
338
hea.d of the flow over the rack. It may, therefore, be Msumed that the effective
head on the racle is [lqual to the !!btic hee.d, or the depth of flow. over the rack.
A. }I'or VeNieal ,"'low lhrough Ih~ Rack. In this case the discharge through a length
ax of the rack may be expressed by
(12.17)
where ~ is the ratio of the opening area to the total area of the rack surface and c is
the coefficient of discharge through tb:e
. From Eq. (12-15), the (Iischarge is
(12-18)
by
Substituting Eq. (12-17) for -dQldx and Eq. (12-18) for Q ill Eq. (13-16)~nd
simplifying,
'1
SUbstituting'Eq, (12-24) for ~dQldxand Eq, (12-18) for Q in Eq. (12-16) and
simpliiying,
dy
2ee ',/iJ[1?=-Tj)
(12-25)
;h "'"
3y - 2E
Integra.l.ion of this aqUlltion gives the eq1!ation of the flow profile as l
~
[~i sin-I
EO
~ }~
(1 - 2lt)
- %~ "VE
-/!J. (1 1!_)]
+ 01
E
E
-J~ - % ~~ ( 1 .. ~) ] + ~2
(12-2B)
(12-27)
or
The integration eon.stants in the above equ:1.tion may be evaluated by the condition
that 11 - !II and x = O. Then, when 11 = 0, Eq. (12-26) wi!! give the length of the
rack required for!!. complete withdraw!l.1 of the main flow througll the I'l'l.Ck, or
x =
(12-19)
-E
.c
tntegratioll of this equation gives tlle equation of the flow profile o.s
C05-
[v ~1I1E
/1!
'Ei!l) - ..,',.sm- (. 1
( 1 - -
'~
2Y ,)
(12-28)
TM'J>Trjm"
J1
x=
For !I
y, and x
(12-20)
<c .E
= 0, the integra.tion
a.ll
Q.
flow
~
(!L!
.,; .E
(12-21)
Wilen y = 0, Eq. (12-21) gives the length of the rack I'equire,d for e. complete with~
(\iawal of the main flow through the rack. or
.
E
EC
(!L!E
(12-22)
L,
(12-23)
. where QI is the discharge through the ent.ranceto the reach of the rack and is also
equal to the wit.hdrawal discharge Q" through the rack.
B. For Incl:ined Flaw IArough Ill.e Rack. In this case the discharge through a. length
dx of the mck may be expressed by
_dQ_.eb
ax
(12-24) .
and (i2-18). By
.cbz V'iiE, By
Eq'tating these
For a complete withdrawal of the main flow through the rRck, it is evident that
Q, = c'bE,6, from which .E = (Q,/e'b),', Thu8, E may be computed if the
incoming discharge Q" b. and c' are given.
,
'The v~lue c of the coefficient of dischll.rge throug;h tho mcl. cpenillgs actually varies
considerably along the rack. For fJx"mple, typical values dcf,ennined experimentally
were found Lo vary from 0.435, for a grade of 1 on 5, to 0.497, for a horizonkl.l slope of
the racks of parallel bars; and from 0.750, for Il. grade of 1 on 5, to 0.800. for a horizontal slope of the racks of perforated scrsnns [271. III general, the value is higher for
racks of
screens than 'for racks of parallel bars. The value is higher for
horizontal racks than for inclined racks. The local value increases as the; ftow depth
on the rack increases if the bars are parallel to the direction of the main flow bilt
decreases with the depth if the bars are in transverse direction,
'
The n.nalysis of this problem may be further extended by considering t,he effects of
the streamline curvature, the nonuniform velocity distribution, and the bottom slope,
and by classifying various types of the flow protils thus created. In general, there
are five types of flow profiles, which are similar to those to be described in the neXt
exampie,
It may
to note that, when a critical state of Row exists OIl th" upstream
side of the rack, the critical depth will occur u.t a. section somewhere upstrea11l frOl:n
Q",
~ Equat~on (12--26) is given by Mostkow [27J and Eq, (J 2-27) by Noseda [2Il. THe
tw.,' equatIOns aremc.tbematically idellticll.l. The relatitlnship between the integra- .
tiol! constants ill 0 1 = Ot
,..EIS.c,
'.
340
the entrance section. This phenomenon is, therefore, similar to that of a free overfall (Art. 3-4). The ratioYliV" of the entrance depth to the critical depth -decreases
\I'itll increase ill the value of and of the rack slope. Typical average :atios vary
from 0.70 to 0.90, which correspond approximately to the valuGs of VilE frGm 0.47
to 0,60.
'
Example 12-3. Analyze the ftow through a side weir in a prismatic horizontal
rectangular channel.l
Solution. The tlow th;:cugh a side weir is a case of spatially varied flow wit.ll
decr~asing dischD.rg~ (Fig. 12-7). According to Frazer [32j, the foHowing five
typm; of flow profile can be produced:
_--'-'T~
{OJ
===tl
~----:'. YIJYJ=:=::::J~J2 .
Qy( -dQldx)
dx = .Ilb'y' ~ Q'
--,-'--
12-7e)
(12-16)
The discharge over ailY given length of tho weir can be computed by a weir formula,
~..!!!.
dx
(12-30)
where c is the discharge coefficient and s is the height. of the weir sillll.bove the bottom
of-th~ channel.
The sill of the weir is pa,'allel to the bottom -of the channel. EquatiQn (12-18) also is applicable to the present problem. Tl1us, the discharge at -any
section is
(12-18)
Q = by -/2U(B - 0
where b is (he width of _the channel and E b the specific energy.
With Eqs. (12-30) and (12-18), Eq. {12-16} becomes
2c-V(E - y)(y - s)3
ay
ax
II
(12-31)
311 - 2E
12"7d)
7'ype d. Depth of flow less than critical at the entmnce with supercritical flow
in the \veir section, the depth of flow
decrell.Sing along t-he weir section (Fig.
Thus, assuming
dy
12-7c)
(e)
341
where
(1)
\E
~c F (1)
+ canst
E
(12-32)
2E - 3~ ..JE - y _ 3 sin-1 - IE - Ji
E-s
V-s'V y - s
(l2-33)
WiLS
-w
lo-"Y
Ady
+ wSo !o~%
A d:1;
-wA 6.y
+ wSoA 6.x
}o
- wB,A 6.x
(12-34)
,,,here A is the average area. Since the discharge varies. with the finite
increment of the channel length, the average area may be taken as
A = (Ql + Q%)/(V 1 + V 2). Also taking Q = QI and V + 6.V = Vz
and simplifying,
342
dy'
-dy
+ Sodx
TABLE
+ So c.x
II
:;
(~;
4
48
6 I 78
8
112
10
150
12
192
14
238
16
288
18 "342
20
400
22
462
24
528
26
598
28
572
(12-37)
(12-38)
I.'
1. 7 t
2 44
18
'3.11
20
3.75.
224.36
24
4.95
25
5.541
28
5.11 II
30
6 67
32
7.22
34
7.76
36
8.31
38
8.84
1-l..
16
II
10.49
12.52
14.15
1.5.53
16.75
17.86
18'.88
19.82
20 .71
21. 55
22.34
23.12
23.84
504
978
1,585
2,330
3,216
4,252
5,440
5,780
8,284
9, 960
11,800
13,820
16,020
----~------~--~~----~------~-----
TADLE
12-2.
I'Q+Q,ly, I
v,'!V,+v,
(4)
400
l,pOO
2,000
4,000
8,000
8,000
10,000
12.0001
14.000
10.000
1
I
I
--I
8.77
9.25
--~-------
(5)
(8)
400
3.4 '10.01
1,4GO
5.2 12.5
3,000
9.2 ' 14.91
5,000 13.5 1 17.6
10,000 16.9119.31
14,000 1 19.7 20.61
18,000 '122 I 21.B:
22 ,000 242/22.4'[
26.000 25.2 23.2
30,000 28.0 23.8
(7)
10.'0
22.5
27.4
32.5
36.9
39,\1
42.2
44.0
45.6
47.0
12-4
AQ
(3)
I
----.--------j-.----
,0
(2)
-:-:~~I-
0+ 10' 10
0+25 1.5
o T liO 20
1 + 00 50
1 + 50 50
2 + 00 50
2+50 50
3+00150
3' + 50 50
4+00 50
1.52
2.52
3.33
4.01
4.63
5.22
5.76
6.29
6.82
7.31
7.81
8.29
" ""I
12-4
(7)
- - 2 - ---;;-)1--'-1-'-2-jl--0'-.,..9-2- ---'-7-'-6-8- 1 - - -- - 1 69
Substituting Eq, (12-35) for l1y in Eq. (12-37) and introducing !l,n
energy coefficient a for nonuniform velocity distributbn,' the drop in
water surface is
&~'=O:~(I~~r~+Q~2)(6V+ ~:6Q)+Sf~.1;
12-1.
(12-36)
343
(8)
11:1 R, hi
--.-------oV
(9)
A'l'l:lly'
(10) 1 (11)
(12)
(13)
Q.03
0.05
0.08
0.16
0:16
0.15
0.15
0.15
0.15
0.15
4.30 4.30
4.99 9.29
6.93 16.22
5.37 21.59
4.48 26.07
3.90 29.97
3.45 33.43
3.1.9 3fi.52
2.88]39.50
(14)
Co\. 5. Critical depth in it, interpolated from Table 12-1 corresponding to the discharge in col. 3
Col. 6. Critical velocity in fps, interpolated from Table 12-1 corresponding to the
discharge in col. 3
Col. 7, Sum of the velocities
the previous sta.tion and in the station under con
sidera.tion
Col. 8. Increment. of discharge tlQ = Q, - Ql
Col. 9. Increment of velocity t. V = V, - V,
Col. 10. Drop in water suriace due to impact loss, or
.1
in
(12-39)
,i
J'
344
Col. 1 L Criticl\\ hydraulic radius in ft, interpolated from Table 12-1 corrc;:;pond'ing to the di.schmge in col. 3
Col. 12. Friction loss, be.sed on :&t. (9-8) with n. 0.015, V from col. 6, and R
from coL (ll), Since this is ll. minor item. compared with tlie impact loss, it may be
ignored if desired.
Col. 13.. 'The I,ctal drup in water surface 6y'
Ay,: + h,
CoL 14. Cumulative drop in water llurface
0""U'.l"''''' ....
000::>""0> ....
t-r-t-l.Ce'to
~<6u:ice>~
~C'lC'40000
C"l'~
CJ)
:::g:s~~g
000000
oeoco
..
..
COOMC~OO
~<.OtOtOu?-.i
c,ot-t-t-C()
C'3 .......
Thecumulati:"e drop in water surfaee is plot~ed as the heavy dashod line in Fig. 12-8,
starting from !l.U a.rbitrary eleva.tion 120' ft from some station at x = 10 ft. The
critical depths from col. 5 of Table 12-2 are then ptotted from the dashed line?s ShO'Uil
by the dotted line.. It is a.pparent that this dotted line represents the bcttom of II
rl~O{,O
~~~~c:
Ci)C:~"""'~(D
~~t"CeoOCJ
NQ ....... OOCO
1_ t""'
r-
~~ ....... .-;~
\1:1 C'l 0.
"'t'tlCO<oCOID
....... 1.C"OO\Qot'-
~~~L.)~c..i H
N'l~N-C-'1
co r-
~' ct)
r..:::;
en
r:Q
(0
O'J ':"')
....
lOC";It.OC'fllO
Oischor~e In cfs.
'
00000
~_8~~~g_
....... N
100
C"1 'C'1
c,~,
OM'tt-(H
90
en
~'0l'Q)
l"':'c-.i
00
~ C')
.... "'lOU'.lCO
t!l
00000
00000
L..":)... 0,..
0_ 0_
"'00,""
c:.
,$
...... _N
jt\-~--~------~
O~LOC"lt'-OO...-i
. . . .
l",",OOCi)t"--('I':>~r-
:;i'>o\;l'j"'I':TI
....... - .......
~
'1""""1
lJ")
000
-0
. N
.......
-0
<oo.ncQU")04?
C"?X>I-l""""'.MtOrI~ : 'ZO t'- ..-! t- '-di
DislonCi '.1:0ng
c~onnel
~C)C'1'C'-,C'll"""'i,....c
length In It
t:;:;
1--'--1
:3
'
....c
J3
'.;s
t.Ol"--o:c:r;C'!~~
W.
CO-..:ttM""t'O'l,....tC'1~
- ' t-
OiIct~r-
,....~NN
I--<o-~-.-"'-,_-----1
~-
g8g~~~~~
r...: ~ L..j ~. c-i ~ ~ cO
oo(>?.;f:1
l'_,
e'8"
] 1--------------------, ~
. fictitiolls channel in which the flow at the givl'..rt dis~ht\rge condition is critical at e~ery
section throughout the full length of the channe!. The c!l.I.Shedline is. the corresponding water surface. A tangent parallel to the bottom of the actual channel can be
drawn to the dotte~ bottom line of the fictitious' critical-ftow channel. The point
of tangency, at Iyhich the two bottoms have the same slope, gh'es the location of,the
critical'section, which is found at Ste.. 1 + 64. It is evident that the slope required
to ma.inta.in critica.l flow to the left of this section i~ greater than the actulll slope ~and
that to 'the right it is)esa, which is the con<;lition ne'cessary for the formlLtion of.!l. control. If more than qne point of tangency is possible, the one giving th'e lowest position of the ta.ngent ,viii be likely to control., It is:also p08~ible to ha.ve two or more
control sections with'hydraulic jumps 'between.
:
Having located th~ control section, the tlow-profife comput.ntion can be carried: ou,t
as shown in Table 12-3. The computation proceeds upstream from the control section for the subcritical flow in the upper part Qf the channel and -downstrenm for, the
'"
g
',B
.,.,'"
>.
,.0
.....-
;....
~8:;;;;;
M~M~
1--------------:
Ol.OCf')"I:t'
1..... (,0 CD
1;'-
CO ~
....-i
::l.
'"
<:or-r-.~t-C).jooco
OOt-OOI'..Ot-~
~~~~c:~~~
"'t-t-r-LOC'<laO
000000
....... 1;'-'.1r-c-,1LC";l
O~~OO~M
~,""cQ'lN1"""f
'"
to
'<"0
'" 0
++
OlQCO
tOC'l"""'O'
++++
0000
345
<f'
0000
L..":l 0
(Q
tQ
++++
C'lMM~
346
347
ally, this method can be applied equally well to any type of varied-flow
equation for flow in prismatic as well asnonprismatic channels.
12":6. Spatially Varied Surface Flow. An important type of surface
flow encountered frequently in engineering, problems dealsvrith runoff
from a plane surface as the result of rainhdL Apparently" this is a
problem in spatially varied flow with illcreasing discharge and can be
treated as such; it may, however, be very complicated, becoming a threedimensional problem if the surface is CUl'vedin space, as in the case of
a road pavement that has a cambered transverse profile and a longitudinal slope. The theory of SlHl,tially varied flow was first used in
sudace flow by Keulegan[40], and the equation thus derived wa.'3 applied
supel'l~ritical flow in the lower part of the channel. ' The procedure of computation
is the same.as thnt explained for Table 12-2, except that the wllter-ourface drop /J.y'
in col. 4 is finally (Obtained when it agrees with the computed Ay' in col. 17. This is
shown for the computation at SLa. 1 + 00. In cols. 3 and 5 are elevations of the channel bottom and the water surface, respectivelJ. 'The value of Ay~ between :z: = 1{l
and x = 0 cannot be computed, but it is arbitrarily fLssumed to be twice the velocity
head at X = 10 ft. The final flow profile is constructed as shown in Fig. 12-8. The
accuracy of the computatior.. will depend on the length and' number o(subdivisions
assum<ld.
-1
1
II
(12-40)
Owing to the variable velocity distribution in the channel cross section,
the value of the energy coefficient may be very high. According to
Schmidt [251, values up t.o 1.30 have been observed at the beginning of
the spillway, and even higher values were found at the end of the spillway
crest. By experimental study, Schmidt Was able to develop an adjustment procedure to correct fOI' the effect of the nonuniform velocity
distribution.
The value,of 6.Q in Eq. (12-4:0) is the discharge over the spilhvayper
6.xof the crest length. :Nfany formula.'3 have been proposed for its determination. For practical. purposes, the formuh\ for. the regular weir of
similar crest shape may be used if the corresponding discharge coefficient
is reduced by 5 %.
12-5,' The Isoclinal Method, For a simple but approximate computation of a flow profile, a graphical method suggested by Werner [39J may
be used. By this method the spatiaIly-varied-fiowequation iIi any form
is plotted with y against z for different values of dy/dx aspararneters,
i'esulting in a numb'el' of isoclinal curves (Fig. 12-9). Starting from the
depth at the control sectioll C, aline is drawn with a slope (= 0.03) equal
to the average value of dy/d.'1; (= 0.05) indicated by the isoclinal curve
passing through the control depth and dy/dx (= 0.01) of the next isoclinal
curve, which the line interdects at P. Starting at P,' repeat the procedure
to determine P', and determine similarly other P9ints of intersection.
,The flow profile is the curve joining all the points of; intersection. Actu-
.'.
.~
..
')
i
x, II
to experimental data by Izzard [41]. For flow on a Toad surface, a comprehensive analysis was performed by Iwagaki [42].
For practical purposes, all approximate equation for discharge of
surface flow is generally assumed, s11ch as
leY'"
(12-41)
whereq is the discharge per unit width of the flow, y is the depth of flow at
the point of outflow, aild k and, m are constants. At equilibrium conditioll, the discharge q at a point x distance below the drainage divide is
q = xq*
(12-42)
where q* is the constant inflow dne to rainfall excess, or s,upply rate, per
unit area. The rainfall excess is equal to rainfall minus infiltration and
other losses that will not become surface runoff. Combining the above
i
Iy.
348
(12-43)
This is the eguation for the flow profile, which is generally applicable when
x is not too iarge. The value of k has to be determined experimentally
since it depends 011: the surfaoe charaoteristics, slope, type of flow, and
viscosity (in the.case of laminar flow). The value m depends on the type
of flow; it is approximately % fqr turbulent and 3 for laminar.
For tmbulent flow, the differential equation for the surface flow, from
Eq. (12-4), may be written
.
dy
dx
(12-44)
'I"here F2
V 2 /gY. If the raindrop momentum is ignored it can be
shown that the coefficient 2 in the numeraotor will becom~ 1 [of. Eq. (12-8)].
For analytical studies, the profile of the surface flow can be computed by
the method of numerical integration. The control section of the flow
profile can be uetermined by the method of singular point or by a criterion
developed by Keulegan [43].
For laminar flow on a road surface, I wagaki [<12] has performed a.n
elaborate matherna~ical analysis, in which the continuity and momentum
equations are applied to n, three-dimensional element of the flow .. By
consielering a, general case in w.Uich the velocity D,nd depth of the flow do
not change in the longituuinal direction of the road surface. he was able
to derive a differential equation as follows:
'
ely
where
and
1
I
I
j
~.~.-------
L -----
Fl
ilx - F~
349
(12-45)
(r)"-l
6q.2X
- --g;;2
!.
12-1. A rectangular wash-water trough 20 it long a.nd 1.32 ft wid~ l)<1rries a discharge at a slope of 0.065 to a. free-fall outlet. If the measured upstream depth is
0.34 ft, I)ompute the discharge by means of the chart in Fig. 12-4.
12-2. An
formula for calculating the discharge capacity of rapid sandfilter wash-water tronghs,has been developed by Miller [Ill by asSuming a parabolic
flow profile at a maximum discharge. The formula is
Q "'"' 1.91b(y..
-+ L tan 8)
(12-46)
where b is lhe width of the rectangular channel in ft, y .. is the upstream depth in H,
L is the channel length in ft, and IJ is the angle that the channel bottom makes with
the horizontal. Using this formula, compute the discharge required in Prob. 12-1.
12-3. A rectangular wash-water trough 30 ft long is required to cal'l'ya discharge
of 8 cfs, having a free fall at the outlet. Design the trough for the least o.mount of
material required fOT theconstl'uction(neglecting the end wall and making the total
wall and bottom wall of the channel a minimum). Assume:
a. A horizontal channel
a. That the channel has a slope
-I
GRA.DUALLY VARIED FLOW
350
SP_~TrALLY
Assume
a. A horizontal channel
b. A channel with slope equal to 0.0(35
12-5. Derive the equation of the flow profile in Example 12-2 if a is not equal to
unity.
12-6. A horizontal bottf)In rack made of perfor:ttcd screr.r, is. designed to divert
water from a channel . . Determine the length of the rack req1:lired to withdraw the
total main .fioll' of 26 ds from the channel. Given:. = 0.5, e = 0.8, c' = 2.80,
b = 3 ft, and y,/E = 0.60.
12-7. Solve tb.e preceding problem if the rack i~ Illade of pam!lel bars.
12-8. A side weir is ased to ciivert the excess of a storm flow of 75 cfs from >\ 'IS-in.diameter sewer. The sel'!er has a grade of 1 in 400,a iullflow eapacity of is cis, and
an unrestricted outlet. The dry-weather flow is 5 cis. Detimnine (a) the height of
the weir sill, and (b) the lengt.h of the weir, assuming a = 1. It is ,.Iso assumed that
tile top width of the water s.re:t is const2.nt and equal to the diameter of the sewer; so
the equations derived for recta.ngular channals can be applied.
12-9. Solve the preceding problem if '" = 1.20.
12-10. By converting increments to differentials, show that Eq. (12-35) is ident.ical
with Eq. (12-4).
.
12-11. Determine the control section in EX!l.mple 12-4 by the method of .~ingular
point.
.
.
12-12. Demonstrate analytically tha.t Rinds's method for the determmatlOn of
control section is identical with the method of singular point.
12-13. C~mpllee the flow profile in the channel described in Example 12-4 carrying
a varying discharge of 50 cfs per foot of ch!l.nnellength.
.
12-14. Compute the flow profile in Example 12-4 by the isoc!inalmethod.
12-15. Verify Eq. (12-40).
12-16. Shol~ that the flow in a rectangular prismatic channel with a lateral spillway may be expressed by
Q
,----zy
-= b \) (Hy' --
1/3) ~
(12-47)
where b is the channel width, y is the depth, H is the constant energy head in th.e spillway section measured above tile channel bcittom, and a is the energy coefficient:
12-17. Artificial rainfall of a constant intensity "qual to 3.6 in./hr is applied on a
concrete pavement ha.ving a slope of 0.0'.1 and a roughness coeffkient n = 0.025.
Compute the flow profile, ignoring the raindrop momentum. Assume that:
1. Julia.n Hinds: Side channel spillways: Hydraulic theory, economic factors, and
e)cperiment.al determination' of losses, TTansactions, American Society of Civil
Enr;ineers, vol. 89, pp. 881-927, 1926.
2. H. Favre: "Contribution o.l'etude des courants liquides" ("Contribution to the
Study of Flow of Liquid "), Dunod, Paris, 1933. .
.
3. E. Meyer-Peter and Henry Favre: Analysis of Boulder Dam spillways made by
Swiss laboratory, Engineering Ne1J)s7Record, vol. 113, no. 17, pp. 520-522, Oct.
25, 1934.
VARIED FLOW
351
1I.
1
I
I!
~-
3;)2
19. H. Favre: Sur ip.s lois regissant Ie nlouvement des flu ides dans les conduites en
, charge avec }l.<lducLioll Io.terale (On the laws governiilg the flow in :CDnduits with
iaterai disciwrgej, Rev1,e '((niver-selle des mines, Liege, vol. 13, sel'. 8, no. 12, ,pp.
502-512, December, 1937.
20. W. H. R. Nimmo: Side spillwo.ys for regdating diversion canals, Tmnsactions,
Amel'ica.n Society oj Ci..'il Engineers, vol. 92, pp. 156r-1584, 1928.
21. Giorgio Nosedo.: Operution and design of bottom int.ake Tl!.cks, P"oceed-ings of the
6!h Genei'al 'Meet'iny, Intwno.tional A.ssociatio'n of Hydrl.LlI/ic Hesea'rch, The Harjue
i955, vol. 3, pp. Cl7-1 to Cl7-ll, 1955; repTinttd flS Ist'Unto di ldm'alica e Cosln!.ioni IdmuliciJ.e, Milario, 11femorie e stud'i No. 130, 1956.
'
22. Giorg-io Noseda: Correnti perinanenti con port,.to. progressi'lamente dec\'escente,
defluenti su gl'iglie eli fondo (Steady flow with graduallydecl'easing discharges on
bottom intake mcks), L' En~r(riu. eleU'ricu. r l"'filatl,(J, vol, 33, no. 1, pp. 41-51, January,
10.56; reprilltecl as Istituto di Idmul'ica e Costr,,~'ion'i Idra1l/i"h6, 1I1'iiano, Memorie e
8t1l(/;i No. 132, Hl56.
23. Giorgio, Noscda: Corrt!!'.ti permanenti eon portat<t prllgressivament,e decrescente,
defiuellti su griglie di fondo: Riee:-c!1, sperimentale (Steady flow with gr!l.duo.Uy
decl'eMing discharge on bottDm intake racks: Experims:nt!l.1 results), L'E?wyia
eieUriCll., !lIilano, vol. 33, no. (1, pp.565-588, June, 1956; Teprinted as ]:;!ituio di
Id"a'ltb!ca e Cosln,~ioni IcZrauliche, Alilano, il-lemo1'ie e St'lldi !'lo. 13,J, 1950.
24. Martin Schmidt: Zur Frage des Abflusses liber Streichwehre (Dlseharge \lver side
weirs), Tecfmische Un'ilJel'si~{;U Be1'l'in-CharlaUenb1Lrg, Ins/itut jiil', Wasse"bau,
Milteihmg 41, 19M.
25. 1\'h\1'tin Schmidt: Die Bel'echnung von Streichwehren (Computation of side
weirs), Die Wasserwir!schaft,Stuttgart, vol. 45, no. 4, pp. 06-10,0, January, 1955.
26. Martin Schmidt: "Gerinnehydraulik" ("Open-channel Hydraulics"), VEB
Verlag Technik, Berlin, and Banverlag GMBH, Wiesbaden, 1957, pp. 188-19B.
27. M. A. Mostkow: "Handblch der Hnlraulik" (" Handbook of Hydraulics "), VEE
Verlag' Technik, Berlin, 1956, pp. 21)4-208 and 213-221.
28. Michel A. NIostkow: Sur Ie calcul dcs grilles de prise d'eaa (Theoretical study of
bottom type water in~ake), La Houille blanche, Grenoble, 12th yr., no. 4, pp. 570580, September, 1957.
29. Peter Ackers: A theoretical consideration of'aide weirs as storm-water overflows,
Proceedings, Instil.'l1tion of Civil Engineers, London, vol. 6, pp. 250-269, February,
1957.
.
30. John WilEam Allen: The dis~harge of w~,ter over side weirs iIl cireular pipes, Proceedings, Inslitu'ion.of Civil Engineers, London, vol. 6, pp. 270-287, February, 1957.
31. Vincerit Knight Collinge,: The discharge c!l.pacity of side weirs] Proceedings, Institlltion oj Civil Engineers, London, vol. G, pp. 288-304, February: 1957.
32. Willil1Ln Frazer: Thc behavibur of side weirs in prismatic reCltangular chdnnels,
hoceedin(ls, Instit'lliion of G-ivil Engineers, London, vol. 6, pp. 305-328, February,
1957. ,
33. Ven Te Chow: Discussion of Flood ,protect-ion of canals by lateril spillways by
Harald Tults, paper 1077, Proceedings, American Society of C~vil' Engineel's,
Journal, HydTa11Zics Div'ision, vol. 83, no. H.Y2, pp. 47-49, April, 1957.
34. l!'. Gl1~ot: De Watervang met liggelld rooster (Channel with bottom grid), De
Inveniimr in Nederll;ndsch-Indie, no.:7, 1939.
35. M. B~uvard: Debit d'uIle grille pat en .dessous (Discharge passing through a
bottom griq), La Houille blanche, Grenoble, 8th yr., no. 2, pp. 290-291, May,1953.
,36. J. Kurjtzma~ll and M. Bouvard: Jt~ude theorique des grilles de pri,ses d'eau du
type ':'en dessous" (Theoretical study of bottom-type water-intake grids),
La H ottilie blanche, GreIloble, 9th yr., 'no, 5, pp. 569-574, September-Oc:tober, 1954.
J. Orth, E. Chardonnet, and G. 1'I~ynareli: Etude des gri1~es paul' pris~s d'eau cl.u
37. type "en dessous" (Study of bottonl'type W[I,ter-intake gmls), La Houdle blanche,
G'renoble 9th Yf .. 110. 3, pp. 343-351, June, 1954.
,
.
38. Josef Fr~nk: Hydraulisch~ Untersucbungen. fUr das TiTOler Wehr (Hyc\J'aulic
B.nalysis for the Tirol weir), DeT Bm!~llgenieuT, BerliIl, vol. 31, no. 3,pp. gO-lOl]
1956.'
i
b' I' I ' .
39. P. Wilh. Werner: Wasserspiegelberechullg von Kana~en el g elC lmagSlg?l'
Bewegung lind veranderlichcl' Wassennenge (ComputatIOn of .water surface Il1
channels wi.th steady flow aad variable discharge), Dte Bautechmk, Berilll, vol. 1[),
110.23, pp. 251-252, May 30, 1941.
' ,
.
"
.
40. G. H. l{euleg[l,u: Spatially variiLble discharge over a slopmg plane, 'lransacltons,
Arnorican Geophysical Union, pt. VI, pp. 955-959, 1944.
.
'.
41. C. }'. Izzard: The surface-pn)file of overllJ.nd flow, Transaclwns, Amenca.n Geo. '
.,.
physical Union, pt. VI, pp. (J59-968, 1944.
42. Yuichi Iwagaki; Theory of flow on road Eurface, Jl;ie7ll0lrS of the Faculty OJ EngmcBring, Kyoto Ur..il,ersity, Japan, vol. 1~, no. 3, pp. 139-14:, JulY~ 1951. . ,
~,3. Garbis H. '1Celliegan; Determination of critical depth III sp[l,tl.all y varw,b~e flow,
Proceedings of the !lid Midwestern. Conference of Fluid Mechamcs, The Ohto State
University, ETl.gineeTing Experiment Station, Bulletin 149, Septem.her, 1952, pp.
57-80.
>
"d
S
l:"'C
~
~
~
>-I
t1
."
)~
>-:l
l.""l
._ _ L
. ,
----
).
',.
CHAPTER
13
INTRODUCTION
. _. ___..::;;-
I'
O.
'Eieparation zones, eddies, a.nd rollers that may occur in rapidly
varied flow tend to complicate the flow pattern and to distort. the actual
velocity distribution in the stream. In such cases, the flow is actually
confined by one or more se.e..aration zones I:ather than by solid boundari!"""
13-2. Approach--tontheProbiern. Th~th~ that assumes a parallel
flow wit.h hydrostatic preSSUl'e distribution is known in classic hydraulics
as the Breslie theory;1 such .a theory is used for uniform flow and gradun.lly
varied flow. This theory, of course, does not n.pply to rapidly varied
floW', even wit.h continuous flow profile, For rapidly varied flow of continuous fiow' profile, cln.ssic hydraulics has shown that a mathematica.l
equation of the flow can be established on the hasis of an inviscid- (i.e.,
frictionless or nODviscous) n.nd potential-flow condition [1-4]. A direct
1 Thill is 50 called because of. the early contribution by Bresse to the solution of the
d,rnamic equutioll for gradually varied flow (see Exa.mple lO-4).
357
I
358
INTRODUCTION
7.
8.
9.
10.
11.
359
.i
,i
J,
REI<'ERENCES
-j
j.
I
CHAPTER
14
(
,i
k = A (liY+ B ;
of
voi cos ( J . .
(14-1)
Y = -voi sin !J
+ Yzgt + C'
360
(14-2)
+C
(14-3)
where it = gH /21)0 2 C05 2 8, B = - tan 8, and C = c' / H. 8illce the horizontal velocity component is constant, the vertical
of the nappe
l' m~.y be assumed constant. Adding a term D
T / H to the,above equation, the general equation for the upper surface of the nappe is
(14-4)
The above nappe equations are quadratic; hence, the nappe surfaces are
theoretically parabQlic.
.
, Numerous tests on nappe over a vertical sharp-crested weir have been
made, On the b'asis of data of the U.S. Bureau of Reclamation [1], of
Hinds, Creager, and Justin [2,3], and of Ippen [4], Blaisdell [5] has developed the following equations for the constants ill the general nappe
equations:
A.
FIG. 14-1. Derivation Of nappp. profiles over S!larp-,jr~stcd weir by the' principle
projectile.
351
-0425 + (l25
0.411
1.603
0.150
0.45
(14-[)
~
(14-8)
where m
h;,/H - 0.208, and h. is the velocity head of the apPl'o,\ch
Bow, For high w.;;irs, the velocity of appro[\,ch is relatively small and
call be ignored. Thus, the constants become .4.
-0.425, B
0.055 ,
C 0.150, and D = 0.559. Experimental data hays, indicated that
th~se equations ~re not valid when x/H is leS8 than ab011t 0,5 and that, for
h.IH > 0,2, additional data for verification are needed, For xl}! < 0.5,
the preSS1ll'e within the nappe in the vicinity of the weir crest is actually
ab9ve atmospheric because qf the convergence of the streamlines. Consequently, forces other tha~ gravity a.t'e acting on ,the nappe, which
makes the principle of the Pfojectile invalid.
.
It should be noted that tJie above theory anq. equations apply ollly if
the approach flow is sub critical. For supercritical flow, or F > I, t.he
1
363
L = L' -
(14-10) .
where L' is the mea~ured length of the crest and N is the numbet of contractions. For two end contractions, N = 2. For one end contraction,
N == 1. When 110 contractions are present nt the two enus, N = O.
According ~o a well-known weir formuia of Rehbock [10], the coefficient
C in Eq. (14-9) is approximatety
.
H
3.27
0.4011:
(14-13)
(14-11)
5.68 1
+ hH)!.6
",here h is the height of weir. IVIeasurements by Rouse [4, p: 532] indicate that this equation holds up to Hlh = 5 but can be extended to
H Ih = 10 with f~Lil' approximation. For If Ih greater than about 15, the
weir becomes a sill, and the discharge is controlled by a critical section
immediately upstream from the sill. The critical depth of the section is
approxim.ately equal to H + h. By the critical depth-discharge r,~la
tiol1snip, it oan be shown that the coeffir.ient C is
(14-14)
where y is the height of the opening below the gate in it and H" is the head
on t.he center of t.he gate in ft. for 1.1ngated weir or spillway, (f = O.
The va.lues of C are as follows:
.
(14-12)
The transition between weir and sill (betw~en Hlh = 10 and 15), however, has not yet been clearly defined.
.
Experiments have shown that the coefficient C' in Bq. (14-9) remains
approximately constant for sharp-Grested well'S under varying heads if
the nappe is' aerated.
14-2. Aeration of the Nappe. In the preceding article the ovedalling
nappe is considered aerated; that, is, the tIpper and lower napp'e surfaces
are subject to full atmospheric pressure. Insufficient aeliationbelow the
"
.-.-c-/
0.077
0.5
0.135
1 0
0.175.[0.202
I 0.220
0.225
..
.I
.\
364
(
\
1
365
where X and Yare coordinatei.l of the crest profile with the origin ::It t,he
highest point of the crest, He is the design head e:cch~din(J the velocity head
of the [\,pproach flow, and IC and n nre parmneters depending 011 the slope
of the upstream face. The values of J( [,nd n fl.re given as follows:
Slope of ups/rem" fave
n
1.850
1. 836
1.810
1.776
~ EKamples of othel' well-known pmfile;; a;:8 (1) the CreG.ge, profile [13) developed
from a mathematica.l extension of Ba.iH's data, (2) the modified CnilJ.gcr profile [21
based on t"e U.S. Bmeau of Reclaniatioll dfJ,tn from Di!m-er tests [II, (3) the LClnBDavill projile [l4J Q!lsed on the U.S. Bure!L\\ of Reciall1aLicn data from Fort Collins
tesLs'[l] and the datn of Bazin [12] ,l,nd Scimel11i [15], t4) theSci,.i;emi profile l151, (5)
the SmetmtCl
Ill)], (!l) the J) Marchi profile [17J, and (7) ~he Escande profile. [18] ..
For" good
of various well-known profiles, see \19].
See. [11. The main project WIlS sta.rted in 1936, nnd the test.s were conducted in
the Bureau's Hydlll.ulic Laboratory in Denver, Colo. Earliee experiments conducted
by the Burea.u were performed in H)32 at the Colorado Agricultural Experbllent Station,' Fort Collins, Colo. The former are usually referred to as the Den.ver tests, the
btter as the Port Collins tests.
'From [20], Hydraulic Design Charts: Ill-I, WES 4-1-52; and 111-7 to 111-9,
WES 2-54. The development of the WEB standard shapes is described. in [21].
Experiment Slatioll.)
366
whel:e H. is the total energy head on the crest in ft, 'inclUding the velCl()ity
head in the "pproach chani1el. Model tests of the spillways have shown
that the effect of the approach velocity is negligible whim the height II of
tho spillway is greater than 1.33H,!, where Hd is the design head excluding
the approach velocity head. Undel' this condition and with the design
.......
0
1.04
1.3
1.03
1.2
1.02
lot
367
H.!~ld and C/Cd for spi.lIways d;si~l1ed for WES shapes having vertical
upstrea,m face. For spIllways havmg sloping upstream face, the value
of a can be c.arre.cted approximately for the effect of the upstream-face
slope ~y m?-itlplymg C ~y a correction factor obtained from the attached
chart 111 !lg. 14~4. ThIS correction was developed from the Bureau of
ReclamatIOn da.ta flJ.
Hd " 17.7'
EJ.
.E
983.:.3i~~,.;..;....."._
c: 1.01
.o:!
of-
3.54
8.95'
::'
'.6S
15 1.00
23
r.... Axis of
crest
0.99
/
0.98
0.2
0.4
O.S
1.0
....
Point of to;;;e.1CY
1.2
X=35.9' Y=:3-2.7'
0.6
Ha~veloclty
head
0.5
Ha
1 --
HdI_
0.4
E1.
I
0.\ 1---1--:7""'1
0.70
0,85
0.90
0.95
1.00
a~x:am~le 14-1: Detcr~ine the crest elevat.ion and the shape of an overJlow-spill.
l~ .sect!O? havmg a. vertloal upstren.m face and a cres~ length of 250 ft The d ;
discharge 15 75 000 cfs Th
t
. ,
. as.gn
1 000 0 d h'
.
e ups ream water surface at design discharge is at El .
. .an t 0 avera~e channel floor is at EI. 880.0 (Fig. 14-~)
.
,
Sol,l!MIl.
h' I
fl'l1
0 .
r 1'bl8 ;\sSltffiing
d
a 19l Over ow BpI way, the effect of approach vclocity is
n(:goalgx 2"Oa)n ~ d = 4,03. By the discharge equation, H.u = QICL = 75 0001
.
0
= /4,1 and H, = 17.8 f t . '
)
The approach velocity is V = 750001(250
' 20)
.
i
I ' t h d' H'
a.
I
.,'
X ~
= 2.5 fps, and the COrrespondng :e OOlJ as. IS a ':" 2.5'/217 = 0.1 ft. Thus, .the design head is Hd _ 17.8
0.1 - 17. ( it] and the height of the dam is h "" 120
17.7 = 1023 ft Th' h' ht'
IS greater thai 1 33H
d'
.
.
IS eIg ,
'
' I I . . d! an , ll~nce, tIle. effect of approach velocity is negligible.
'
Th e ares!; e_evatlO n ls.at 1000.0 .~ 17.7 "= 982.3.
. B
.
.
y Eq. (14 15), tbe C,Est shape IS expressed by}'
X 1 s/23 C d' .~ f tl
shupe
t d b hi
.
.
oor Ina""s 0 1e
compu e y T, S equatIon tHe plotted as shown in Fig. 14-5.. The crest shupe
, .
upstrCftm from the orir~in of ~!le coordinates is constructed accordin~ to the d.imensiorus
recommended in the chart of Fig. 14-3. The design of the straight portion ,0 th~
spillway sudacebdow the crest section depends ~n the stability requireme;ut and 011
the features of the sti;Iing basin at the toe of the spl11w!l.Y. Aslope of 0.0: lis assumed
for the straight pOltioll of the downstream spillway surface.
head is greater thp.ll about twice the height oLthe spillway above the
upstream bed.. Such spillways are often found in earth dams.
In using the Br:dley curve, it is necesilary to' know the coofficient of
discharge for the design head. If this coefficient is unknowl1, but the
spillway shape 18 given, 11 method 1 suggested by Buehlel' (23] may be used.
B'T this method,' based on an equation derived by BrudeneU
the
c~e:ffici6nt of discharge is computed by the ~q'.lation
-t
O,5,~
~/
OAI--~:':-:I~I 73'
I\
f
'
369
368
. (He)V'12
(H-17)
C = 3.97 JJ;,;
where H, is all operating head and HDis the theoretical design head,
including tho approach velocity hea-d, for a standard profile havh1g a
.S
Vi
:;:
u
a;
-!'.
3<
ru
.0
.,
'"co
-;;
'6
40~---4-~-4-----i-~4-~-4----1
cr~st
in U
370
Using the chart and Eq. (14-17), the rating curve of. a spillway of given
profile can be cumputed; ,
According to comparisons with actu,al model tests, both the Bradley
and Buehler methods have been shown ~,o give highly'accurate
at all but very low heads. At low heads of 5 ft or so, the computed discharge may be as low as 8% below the actual value. For pra.ctical purpOSGS, however, both methods fire sufficiently accurate.
14-13. Upper Nappe Profile of Flow over Spillways. The shape of
upper nappe profile of flow over 8 spillway crest is significant in the desigu
of spillway abutment Ivalls and for the selection of pivot elevation ot,
tainter
The WES shapes for high overflow spilhlrays with vertical
by the U.S.
upstream face have been investigated, using model
Army Engineers Waterways Experimant Sta.tiOll. Figure 14-8 shows the
shapes and coordinates X r;,nd Y of the upper nappe profile obtained from
such tests for negligible approach velocity, fo], conditions with and without piers, and for three different head ratios. The tenn Hd is the design
he!'.d excluding t.he velooity head, and H is the operating head other tha:n
the design head, also excluding the velocity head. Profiles for intermediate head mtios may be interpolated .. Owing to thecontrac~ion
effect of piers, !J. pronounced hump between X/Hu.
-0.5 and 0 occurs
all the upper.nappe profile along piers when the discharge is high .. Upper
nappe profiles fill' three gate bays adjacent to abutments are
in Fig.
, 14-9, showing the abutment effects on the nappe profiles.
It should be noted that the upper nappe surface is exposed to the
atmosphere and,hence, subject to alteration due to wind and air CUlTents
and the absorption of surroupding air. As s. result, the flow is aerated
and the surface becomes wavy and unstable. The upper-nappe-shape
only the ideal cases,
coordinates given in Figs. 14-8 and 14-9
where air plays little or no part. The upper nappe surface for slt'ping
upstream face should have a lower elevation than that for vertic.'1.1
.upstream face. Hence, the given coordinates may also be 1.lsed safely for
spillways with sloping upstream face for which the actual data are not yet
available.
14-7. Effect of Piers in Gated Spillways. Piers are needed to form
the sides of the gates in gated spillways. The effect of the piers is to contract the flow and, hence, to alter the effective crest length of the lSpillways. The effective length of OM bay ,of a ga.ted spillway may be
expressed as '
L = Lo - J(NH.
(14-18)
1
COO.ft'O!NATE'~ Fan. U1>?ER
H IH;
X Ill.
--=!.O
N A.PPln
W'J,:Tn
No Pume .
1- .4!lol-=t:or::. o.
'
1:210
0
-1. 0
-0.484:,1-0.81-0.9115 -0.8 -1.185
i-OA75 -0.6 -Q.893 -O.G -1.151
'1-0.460 -0.4 -0.865 -0.4 -1.UO
:-0.425: -O.21-0.S21i -0.2, ,1,,-'1.01;0
:-0. 3n
0.,-0.7551 0.0:.,1-1.000
'-0.300, 0.2, -0.681. 0.2 '-0.019
-0.2001 0.'1,-0,580, 0.41-0.821
1-0.075
0,61-0,465; 0.6 -0.71)5
0.015, 0.81-0.320[ 0.81",-0.569
1.0
0.258,
LO -0.145
l.O'-OAlI
1.2
0.470
1.2
0.055, 1. 2 0 >220
1.4"
0.7051
0.29{j 1.4 :"'0,002
1.6, 0.972.
1.6
0 ..;!l3, 1.0
0.240
1.8 i 1. 269[
L~~!~l:~ 0.5$1
-O.ll
.-0,0
-0.4
-0.2
0.0
0.2
0.4
0.6
o,s,'
'I
1.-
1"1
of a-pprnne h.
-1.0 -1.230
,-O,g -.. 1.215 '
-0,6 -1.194
-0.4 -1.15'<;
-0.2 i-I. 122
0.0. -1.071
0.2 i-1.015
0.4 ;-0.944
0,6[-0.847
-1
~:~ I,
3Sfj
o
1.4 -0.102
1.2'r.
1.6
0.172
_ _.......:.._ _...L..._---'-_ _-L_I~.8
0.465
of a.PJ;n"'ou.ch.
-1
-0.6 -I 209
-0.4
1.218
-0.2 -1.244
0.0 -1.103
0g5O
0.4 -0.821
0.6 -0.689
0.8 -0549
LO -0.SS9
1.2 -0.215
1.4
0.011
0.208
0.21-
>
where Lo is the clear span of the gate bay between piers; K is the pier contraction coefficient; N is the number of side contractions l equal to .2 for
1 These tests are d~signa.ted as II General Spillway Tests-OW 801."
The information given here is from [201, Hydraulic Design Charts 111-11 to 111-15 WES 9-54.
371
FIG. 14..8. Upper. n!\ppe profiles of flow over WEB spillways with a.nd without piers.
(U.S. Army En[f<neer:; Waterway/! Expel'iment Slation.)
.J
-i.5
-10
.., -0.5
"x
)::-
a
0.51------+--++..
1.0
I Left abutment
LEGEND
~===:::J~0.205 Hd
Boy No.2 l.078 rid
,"
J.
i Soy No.5
-i2=~3--t---- axis
OVERFLOW "SPILLWAY CREST
1.5
2.0
FIG. ]4-9. Upper nappe profiles of flow over WES spillways showing appro!l.ch channel and abutment effects.
(U.S. Army
x'C ..
o.5
'-
1 .
"
0.5
BAY" NO. 2
-1.0
-0.5
0 " 0.5
X/lid
LO
1.5
2.0
LEGEND"
Left side of bonk
0.5
f---+---~f-
oxis
OVERFLOWSPI"LLWAY CREST
-LO
-0.5
0.5
X/Hd
i.0
1.5
2.0
374
each gate bay; and H. is the total head on the crest including the velocity
head. In comrl1lting the discharge through gated spillways,the effective
length determined by the above equation should be l.lsed. The discharge
cpefficient, however, is assumed the same in both gated and ungated
spillways.
The pier contraction coefficient varies mainly with the shape and position of the pier nose, the head condition, the approa.ch depth of flow, and
the operation of the adja.cent gates. The approxim~,te J( value given by
Creager and Justin [3, p. 120] r1lnges from 0.1 for thick, blunt noses to
0.04 for thin or pointed noses and is 0.035 for round noses. These values
apply to piers having a thickness equal to about one-third th8 head on,
the crest when all gates n,re open, When one gate is open n.nd the adjacent gates are closed, these values become roughly 2.5 times larger.
The U.S. ,Army Engineers Wat.erways Experiment Station has conducted tests (General Spillway Test cW 801 [20]) on different forms of,
pier nose for spill ways of the liVES sha.pe. On the basis of these tests, a
round-nose pier is recommer.ded for general use with high heads. The
[(value for the round-nose pier plotted against the ratio of He/ Hd with
variable distances upstream from the crest is shown in Fig. 14-10. The
effect of other nose shapes 011 the contraction coefficient is shown in Fig.
14-11. The he.ight of the test spillways was 6.67Hd, which had negligible
velocity of !l.pproach. Under the testing conditions, these da.ta are
applicable to high spillways and to the condition that ,the adjacent gates
are open. For low spillways with appreciable approach velocity, the
pier'contr[Lction coefficient fOl; the round-nose pier with various approach
depths is shown in Fig. 14-12. III .. he abs~nce of adequate data, pier contraction coefficients for other nose shapes for low spillways may be
obtained by proportioning from the data for high SpillW8,yS (i.e., from
Fig. 14-11).
14-8. Pressure on .Overflow Spillways. If the spillway profile is
designed exactly in the shape of the lower nappe of a free overflow, the
pressure on the spillway crest under the design head should be theoretically nil. For practical reasons, however, such an ideal profile is
generally modified so that low pressures will develop under the design
head. As the spillway must be operated under heads other than the
design head, the pressure will increase under the lower heads and decrease
under the higher heads. Assuming a two-dimensional irrotational flow,
the pressure on the spillway crest may be accurately determinedanalyticnny by a numerical method, 'graphically by flow-net analysis, or instrumentally by an electronic analogy.l More exact determination of the
pressure, however, will depend upon model tests.
See references given in Art. 13c2, A pra.ctical procedure for the r!l\axation method
appFed to the problem under cDnsideration is given in [25].
'
I
(/)tJ>
f-f-
(/)z
wl!!
a:u
(.J
I~
u:: z
3"-
g~
:r:
--1
"-u a::
a:z W
wo Q.
>ofQ
LI..
,0
we::
f-f-
I-
z u
e>O
U
LI..
LI..
<\l
'E.,..
<\l
Il:l
'"
...
"
<!
S::
:;"
tl
~
...."'"
E5 w i!'.
;r-
n.
...~
'
:::;
~
I,
t
I
,i
,I
~~.~~~~J-J-J-~~~~~~.-L-L-L~O,,-L-L~ ~o~~-L~-L-L-L,~o3
i
!I
,\ .
375
"0
Z
:r:.~
-0
.,
Type I
Type"
.c:
C
.!?
.,'"
1:1
E
~
~
<>
~I
""
"0
''""
.J:.
'0
0
"15
tr
ups~ream
Hydt(~u1ic
----.---..~
1,4
1.2-
'5<II
Hd ~ design head
.J:.
.ffi'
~
"
:L
~ 1.0
:L
.t:f
0
"'c:
.J:.
'"
"iii
'"
O.S
.2
1ii
<.<>
-1
c:
0,6:
-1.5 r.'.
Il:.
-2,0,
'"
.J:.
0.4
r",-"
,
-a,5L-LL
.0,5 Hd rod.
0,2 Hd rod
~
,g
'0
'0
l' "~''"T'~'
\1\
i5 0.5
u
'~
-1.0
-0
-:j
0.0
\-1
Axis of
'reO
Ij~
qt=:r
Type 2 pier
I I ! \ I I I !
O. 0,5 1.0
\.5
2..0
Ratio of horizontal distonce to design head
-'
a::
0.2
LOW GATED QGEE CRESTS
PIER CONTRACTION COE.FFICIENTS
0
0.05.
0.10
Coefficient of pier contractIon, K
[201,
H!ldra1~lic
.. "-.-,-.-
j
378
0.4
0.3
1\
r\V
.",1r 0.2 t
'" 0.1
~ 1~
j!i -o.~
"- ~
r<-
-0.3
,.0-..
r V
-0.4
I
!.I
I
');.
hJ
-0.2
-",
-K
~:~
--+I
--
I
,."..
i". ,./
c-:- -I~
t-
-\
~----------1~T-YP-e-2--Pi-er-{-F-i9-.1-4-_-ln-----1--------J+.-
.\
-, I
-0.2
0.2
0.4
0.6
o.s
(-1S....)
1.0
1.2
Horlzonto! distonce
Dasign head
\Hd
'-1-- 1---
~ -.
i'.,
1.4
....;-'"
H!Hd"1.33
KIKa" 0.50
"./
-0.2
l--
r----., >-...<.;
H/Hd ;'.00
iO
/v
""'"
1'8
'"
~Ii~
w .",
01
-b~50
.... r-'<>H!Hdi.QO
-0.2
-0.5
-t
\ \''''
'\
0.2
0.4
--
0.2
_-t--__
0.6
o.S
Horizontal dislance (X)
Design head
Hd
0.4
.\
0.6
379
1.0
12
l.4
R~0.2 Hd
R 0.5 Hd
., .
,
i
1'"\ 1
-\
I."
l' ~
~
I~
-\
IV
I
Net.: Dota tJosed on CW 60! tests
i'~
FIG. 1'1-13. Cres~ pressur9l! on WES high overflow spillways (continu.ed). (b) Along
center line of pier bay. (U.S. Army 1!l11.glneers Waterways Rxperiment Station [20],
HydrClulic Desiqn Chart 111-16/1; WES 3-55.)
.
.
)5
wiD result in ovel'design of the spilhvay, particularly for high dams ..'
Suc.h overdesign is an acceptcd procedure, providing an additional factor
of safety.!
.
.
The pr'essme teduction on the upstream face of a vertical weir has been
determined both theoretically and experimentally by Harris [26). On
the basis of OW 801. tests for ungatad WES-shape crests of vertical
upstream face and of U.S. Bureau of Reclamation tests of pressure on a
14-13. Crest pressures on WES high 'Jverflow spillways. Ca) No piers. (U.S.
Army Engineers H'ale)'u'ays Expe"imenl Slation [20], Hyd1'rJ.ulic Design Chart 111-16,
WES 9-54.)
FIG.
is
.-.~.'
.,'
.',
..
-.I
I
380
381
shul'p-crestecl wdI' under the design head, the resultant of the reduced
pr~~sure is f?~lnd t~ be approximately 12,9Hd 2 lb per unit length of the
spIlLway, actmg hOrIzontally at a distance of O.161Hd it below the top of
the crest.
. . ,
Uj
1:)
'"
Ol
'" 0
~12
.0.
'"a:i'"
" c
li:./JJ '"
v;
-0.'2
104
I
,f
Rotlo, Hr"
3.2
,L~
3.4
3.5
Dischorge coefficient, C
F.!G: 14-13. Crest pressur.es on WES high ovel'fi~\\' spiHwo.ys (co/!t"'~lled). (c) Along
pletS. (U.S. Al'Il'Y Engmeers lVate1'7l1aYs Expel'tmellt Sin/ion {20] HI'dl'ClL:iic Dcsif!1l
Chm"t 111-16/2, WES 3-55.) .
' U
.14-9. Drum Gates. The drum gate is a hinged g!1.te which floats in a
chamber and is buoyed into position by regulating the water level in the
chamber. Primarily, it is llsed to control the surface eiev!ltibn of the
water upstream. As a measllring device, the drum
resembles a weir
with a curved upstream fa(;e over the major portion of its triLvel. The
rmgle e U;'ig. 14-14) is formed between. the horizontal and a .line drawn
tangent t.o the downstream lip of the. g!\te. This angle is considered posiwhen meastive when it is measured above the horizontnl ilnd
ured below the horiz.ontaL For positive values of 8, the gate acts as a
sharp-crested weir, the control poil'l.t i.~ the downstrea.m lip of the gate,
and the head is measured o.bove this point. For negative values of 9,
the gate !lcts as ~ curved-crested weir, the control point is the highest
point of the gate surface, and the head is measured above this po}nt.
382
Since the d!'um gate acts as a weir, the discharge through the gate may
be expressed as
Q = CLH.!.6
(14-16)
where C is the coefficient of discharge, L is the length of the gate, and H. is
the total head. Laboratory investigations have shown that the flow' ever
this type of gate call be completely defined by H., &, C, the radius r of the
gate, and the depth of approach. The dept,h of approach. however, has
very little influence on the flow behavior when the appro!',ch depth, measured below the highest point of the gate, is equal to or greater than twice
the head on the gate.' This condition is well satisfied by most drum.:gate
installations, especially when the gate is in a raised llositiou, Therefore,
the coefficient C may be considered to be a function of He, fJ, and 1'. .
Bradley [27J has made a comprehensive stu.dy of the drum gate, using
data obtained from 40 hydraulic models of existing drum-gate structures
, of various sizes and scales. . The results of this study are showl1 by a
family of curves (Fig; 14-14) where C is plotted against fJ with the r,atio
H.lr as a parameter. When H.lr .0, t~e ~ate becon:es a straIg~t
inclined weil~, and the cOI'l'espollding dashed Ime III the family of curves IS
based on Bazin's data [12J. The curves extend dowmvard to f) =;0 -15.
The discharge coefficients in the regi011 between.6 = -150 and the gate
cOl1lpl~tely down clln be obtained by graphical i.ntetpolation of theratiI~g
curves of the gate, The comput.ation of the ratmg curve when the gate IS
completely down is the snme as that for a spillway with an ungated crest
(Art: 14-5).
.
14-10. Flow at the Toe of Overflow Spillways. The theoretIcal
velocity of flow at the toe of an overflow spillway (Fig. 14-15) may be
compllted by
.
=--=:---:---=--
V2g(Z,
+ H" -
Yl)
600~
560
II
440
,II I
.----:r-
I'
,I
II
360
i'
-",,,l
'l
240
--I--
II
/1
tnlr--4.-.-17
I !
if
"
11 -01
:I:
;;;
I If I/,
!I / Y//
XI
12C :---
II
80
40
..
I-10
20
30
~
j' I
40 50
I---
-~.-
--
I
~
1
i
---1
60
..
I.
II I
I 17
/ ; W'
II J V
IV I
N!...:......c.tn!+
~'--.>L~.
~r-
8f
I 1/
I /1
I , / '/ /
IiI/ I /
lit
1."1,
:.;,,-,:1:
1
.~,
1 '/
NL1/-;
.;::
II
~I:q
..: J.---.
:>n()
0
0
I
I
# :'7.1/!i
280'
Ii
i
I
. o,r
d- if
JI
i ,I
il
i!
ow..fJ
II
II I
I I
. I! ~ l-+-.~ r--i Tr
I ;tL ~R,r I--I ~I x
.:cr:- i -
T;----it-
i!
.a()n
J
I
-""
'i
520'
(14-19)
383
70
80
90
100
1/0
I
I
f
150 160 170
FIG. 14-15. Curves for detennination l)f velocity. at the toe of spillways with slope5
1 on 0.6 to 0.8,
.
pUrpi)SeS, although it can be refined by addition~l experimental information which may become available in t.he future. .
Experiments by Bauer [30J indicate that friction losses in accelerating
the fiow down .the face of a spillway may be considerably less than the
. normal friction loss in fl0W with well-developed turbulence. Therefore,
the friction loss is not significant on steep slopes; but it woulcJ. become
impcrta.nt if the slope were small, For this reason, the chart in Fig.
I
- \
384
14-15 i,~ not !1pplicable to slopes finttcl'than 1 on 0.6. For fb,tter slopes,
the problem nmy be analyzed by the method described in Art. 11-4.
At the end of the sloping spill W[\y surf&ce, the flow cho.nges its direction
. rather abruptly and therebjr produces appreciable centrifugal pressures.
In order to create a smooth transition of the flow [Lnd to prevent the
impact of the falling watar from scouring the foundation, the surface at
the spillway toe is usm'l.lly designed as a cUl'ved bucket [31]. To be
thoroughly effective the bucket should be t[~ngel1t to the foundation or
neady so. The radius R of the bucket, measured in feet, may be estimated approximately by the following empirical formula:
385
10(VH.4H+!&,;{Z.6H+EiJ
(14-2u)
where V is the velocity in fps of the flow :1t the toe fl.l1d 11 is the head in ft,
exclucHng approach velocity head, on the spillway crest. The design
.bucket pressure and the maximum side-wall pressure sh,mld be equal to
the centrifugal pressure plus the hydrostatic pressure corresponding to
the tai!water depth. The centrifugal pressure may be compnted by
Eq. (2-8) for known radius of the bucket and velocity of the flow .at the
toe. Determination of the bucket pressure, however, can be made more
~.ccurately by flow-net analysis or by the relaxation method, or, still more
precisely, by rnodel tests. The results of such determination
ve indicated that theclIect of the bucket curve on pressure actually extends
even beyond the ends of the curve.
.
The flow leaving the toe of a high overflow spillway is a high-velocity
jet, containing a Hnge amount of energy capable of causillg heavy scouring. Unless the downstream approach is resistant to such scouring,
measures to D.void the danger of scouring should be takeli. in the design .
. Popular methods :11'C design of a ski-jump spillway, utilization of submerged taiIwater as a bl'ake, and use of a hydraulic jurr.:p as an energy
..
dissipator.
14-11. The Ski-jump Spillway.. In this type of spillway the toe is
"'''';;1.",.'''''' in the form of a large specially shaped lip or bucket 'which
throws the whole jet flow into the air. P,nt of the energy in the jet is
dissipated in the air, but in any case the
in.lIs back into the river
channel a.t a. safe distance from the dam. This
was fin,t proposed
by Coyne [32,33]. Figure 14-15 shows a typical ski-jump spillway. : In
this design, the spillway-bottom slab is also the roof of the power house
which is built against. the downstream side of the dam. The slab is
heavily reinforced in order to ta\l:e the
lon.ds due largely to
the centrifugal pI'essure of the jet acting ori the spillway toe. l
un
of
NE YRPI C.)
'"
.,,
VI
("
...
*"b
\I
\~
II _ I
no
:::00
\Jll
rrln1
<II
'"i":::
~
~
Ul"l1
-I:;)
Ul~
n1
2.
-1
. @-o,
---..
--
-~
I
\
I,
I'
..
--~
(J)
-T.
,1
--
~.-~~*--""~"<"l'<"""']~""",,,='"
- ,
-",~----"""-
I---,'+<~-~"""-"'''i''--~'"""",.....
~~:d
..
::-;:.~
+d
Y-'>-i---I~-+---+-.."..l1.5~
1.1;> 0.562
l.:i2.
hd
.He
0.201
0.796
1.53
2.14 1.75
:3
X/He
(0)
0.049
0.OS3
0.142
2.14
0.$55
-1.0
.tv)
FIG.
1~-18.
(d)
Typical pressure (dashed lines) and surface (solid lines) profiles for flow over submerged overflow dams. (Selected
(a) Supercriticn.[ flow; (b) flow involving hydraulic jump; (e) ftow with!l. drowned
hydraulic jump; (d) flow a.pproaching complete submergence.
388
389
(1
pressure reduction of
2 it? Compute the load 0'.1 the weir after this aeratiol1.
14-3. If the channel floor is at El. 975.0 instead of E1. 880.0, and other data remain
the s3-me, determine the spillway section required ill Example 14-1. A trial-anderro, procedure is required for the determination of the value of C.
14-4. If the upstream fa.ce has a slope of 3 on 2 instead of veriic!J.I and other data
remain the s",ne, determine the spillway section required in, Example 14-l.
14-5. Determine the rating curve of tlHil spillwl!.y developed ill Exarnple 14-1.
14-6, AD o,-erflow spilhvay of unknown profile designed br a total hC!l.d of 14.5 ft
has a crest length of 64 ft and a coefficient of discharge equal to 3,48. Determine the
rl>ting curve' by the Beadle.)', curVe. '
14-7. If t,he spillway in the preceding problem has a profile shown by the dashed
line in Fig. 1,*-7, detennJne the rating curve.
14-B. CompuU! ~he upperm.ppe profile of the fiow over the spillway designed ill
EXl1mple 11-1, assuming"'!):!:! piers.
'
14-9. Determine'the wall lIcigh',s for the overflow-dam seetion designed in Example 14-1, assuming six b!l.Ys formed by WES round-nose piers and a maximum operating head 35% higher than the design head.
'
,
14-10. Determine the discharge over the spillway section designed in Example 1"-1
if the spillway has six bays formed by WES round-nose piers.
,
14-11. Detecmine the pleasure on the crest of the spillwn.y designed in Example
14-1, fOJ: operating head equal to 0.5u, 1.00, and 1.33 times the design head; assume
(a) no piers and (b) six bays fOl'med by WES round-nose piers.
H~12., A drum gate (F:ig, 14-11)) 50 ftlong alld 20ft in.radius is installed on top of!l.
high overflow spillway., Determine the rating curve of the gate: i.e., discharge VB.
is
PROBLEMS
H-1. A vertic[L1 sharp-crested weir, 20 ft high and 60 ft long, is built as a.n overflow
spillway without eud contractions. When the head is 20 ft above the crest of the
weir and the nappe i$ completely aerated, compute (a) the nappe profiles, and (/ the
hydrostatic load acting on the weir.
l!!-2. During te~ts on the o.er!ltioJl of the spillway described in the preceding problem, negative presSure~ under tlie nappe of 9.7 it of wCltel' were observed. Compute
the increase ill the t,otalload on the weir. How much air, measured in cubic feet per
1
2
RE;FERENCES
1. Studies of crests for overfall da!llll, BO'l,lder Canyon Praject Final Reparl$, pt: VI,
U.S. B~Lreat' of ReclamatiDn, Hydraulic Inve.tigalions, Bulletin 3, 1948.
2. Juliall Rin.ds, William P. Creager, alld Joel D. Justin: "Engineedng for Dams,"
John Wiley & Sons, Il1c., New 'York, 1945, voL 2, pp. 358-361.
Ji
i
390
,f.
20.
21.
22.
23.
25.
26.
27.
28.
29.
30.
31.
32.
33.
34.
35.
3S.
391
tions of the 4th Congress on Lalgs Dams, vol. 2, Pl'. 105-124, International Commission on Large Darns of the World Power Cbnfcrence, New Delhi, India., January,
1951.
"Corps of Engineers Hydraulic D'"sign Criteria," prepared for OlIiGe of the. Chief
of Engineers, U.S. Army Corps or Engineers, Wat.erways Experiment Station,
Yicksburg, Miss., 1952; revised in subsequellt years.
John C. Harrold: Discuss-ion on Equn.&i.on of the free-falling nappe, by Freel W.
Blaisdell, P:-oceeding~, American Socie(~ of CivaEngineB1's, vol. 80, sp.para&e no.
624, pp. 16-19, August, 1955.
J. N. BradlEY: Discharge cQefficients for irregular overfall spillwaYB, U.S. BIL1'CaU
oj Recla.mation, Engincering Monograph No.9, March, 1952.
Boo Buehler: Discussio'l on Rating curvp.3 for flow over drum ga&es, by Joseph N.
Bl'adley, Tran.~ac1.ions, American Society oj Civil Engineers, vol. 119, pp. 421-428,
1954.
Ross N. Brudenell: Flow o.er rouurled crests, Engincering Ne1ll8-Rec01'd, "oJ. 115,
no. 3, p .. 95, July 18, 1935.
M. K, GanguE and S. K. Roy: On the st.1.ndardization of the relaxation tl'eatment
of systematic pressure comput[l.tions for overflow spillway discharges, liTigation
and Power, '1'h~ Journal. of ihe Central Board'of Irrigation and Power, vol. 9, 110. 2,
pp. 187'--209, New Delhi,India, April; 1952.
.
Charies 'V. Harris; An analysis of the weiI' coefficient for suppressed weirs, Univen":t!! of W(Jahington, Engineering Expm:"!enl Station, Bulletin 22, ]:.923.
Joseph: N. Bradley; Rating curves for 110w over drum gates, Tmnsa;;Cilms, American SoCiety of Ciua Engineel's, vol. 119, pp. 403-420, 1954.
H.obert B. JansGn: Flow characteriRtics on the ogee spillway, paper 1452, P"oceed- .
ings, American Society of Civil EnOl:neers, JOIL1'1lal, Hydmulics Diviiion, vol. 83,
no. HYS, pt. 1, pp. 1-11, December, 19.57.
Research study Qn soilling basins, energy dissipators, and associated appurtp.no.nce~, U.S. Bu.refJ.u of Reclama.tion, Hydraldic Report, No. Hyd-3fl9, June 1, 1955,
pp.41-43.
William J. Bauer.: Turbulent boundery la.ye\' on steep slopes, Trcnsaction.l, American Society of. Civil Engineels, vol. 1U), pp, 1212-1233, 1954.
J. H. Douma: Discussion on De5ign of side walls ia chutes and spillways, by D. n.
Gumensky, Tranlluclio1lS, American Society of Civil Engineil1s. vol. 119, pp. 364368, 1954.
A. Coyne: Latest development of do.ms and hydro-electric power stations in
France, paper read before a joint meeting of the Insti&ution of Civil Engineers and
the British Section of the Societe des Ingeuieurs Civils de France, Loudon, Nov.
25,194i. Editorial review entitled Development of dams in France, Engineering,
vol. 164, no. 4274, pp. 613-614, Dec. 26, 1947.
.
A. COYM; Barrages-usines de I' Aigle at de Ss.in t-:f;tienne-Can tales (Dams and
hydr.oplants of Aigle and Saint-Etienne-Cantales), Trava1~x, Paris, vol. 34, no. 185,
pp. 194-215, March, 1950.
.
R..Mo.itre and S. Obolensky: Etude de quelque~ caracteri,;tiques de l'ecoulement
dans Ill. partie aval des evacu!l.teurs de surface (Study of some flow oharacteristics
in the downstream part of spillways), La Houille blanche, Grenoble, 9th year,
no. 4, pp. 481-511, July-August, 1954.
F. Auroy: Les eV2.cuateurs de crue.s Uti barrage de Chastang ,(The spillways. of
Chastang Dam), Transactions of the 4th Conqresson Large Dams, vol. II, pp. 661f8G, International Commission on Large Dams of the World Power Conference,
New Delhi, January, 1951.
.
E. A. Elevatorski: Trajectory bucke~-type energy dissipators, paper. 1553, Pro-
.-i
I
\
J,
392
ceedings, 'Arnerica.n Society of Civil Enrr.ineers, J oul'nal, POlljf!;'l' DiVi8jon. voL 84"
no. POl, pp. 1,,17, Febru:::ry, 1\)5B.
37. J. N. Bradley] Studies of flo'\v characteristics, discharge a.nd pressure relo.tive to
submerged dams, U.S, Bm'ec.u of Redamalion, Hydm,tiic LabonJ./.lyry Repor! 1-82
HH5.
"
'
38. Report of the Board' of Engineers on Deep Waterways, U.S. Board of Engineers,
pt. 1, p. 291, 1900,
"
~~\). R. E. Horton: Weir experiments, coefficients and formnl!\S, U.S, Geologica.! S1:11!8y.
W(ltel"S'llP1I1V CLnd 111''i~'Q.lion Paper 200, 1907.
'
'
40. Glen Nelson COltl The ~ubmerged weir as a measuring device, Univers'ity of WisEn!/ineeri"g
Blation, Bulietin 67, pp. 48-75, 1928,
41. II. J.
characteristics of I'lllbn:terged spillways, Master's thesis,
Colorado Agl'ic.tll~ul'a1 a.nd Mechanical College, Fort. OolliIlS, December, 1951. ,
42. i\I. Bar Shan;)'] Pressure distribution on downstream face of a. submerged weir,
Master's thesis, Sta.te University of Iowa, Iowa City, June, 1950.
43. Spillwa.y and lock a.pproach, Jim Woodruff
Apalachicola River, Flol'idp":
Model investigation, U.S, Army Engineers lVaierway;s Experiment S'(.!lio~ :i'cch, nual Memera1Ull/lI. 2-340, Mal', 1952.
,
. I
44, Morgantown spillWay, Speciaf tests; lIl).published report, U.S. Army Engineers
Wliterwaye Experiment Station, Vicksburg, lVIis"., 19:'*9.
CHAPTER
15
15-1. The Hydra1!1ic Jump. The hydraulic jump was first ;,o"D"f'i",~
experimentaU}' by Bidone [1,2], an Italian, in 1818. 1 This led
[3]
(1828) to distinguish between mild (subcritical) ::md
(supercl'itical) .
slopes, since he had observed that in steep channels hydraulic jump is fre~
quently produced by a. b!l.lTier in. originally uniform flow.
abundant studies 'were made and the. results were quoted by many
writers. Outstanding contributors to OUI' present knowledge about the
hydraulic jump a,reBresse (1860) [4], Darcy.a,nd Bazin (18fi5) [5], Ferri':'
day and Merriman (1894) [5], Gibson (1913) [7], Kennison (1916) [8].
Woodward and Riegel-Beebe (191'7) [9], 1(och and Carstanjen (1926) (lOJ,
Llndquist(1927) [I1J, Safranez (1927) [12], Einwachter '(1933) [13,14],
Smetana (Hl34) [15,16J, Bakhmeteff and Mat;1ke (1936) [17), Escande
(H)38) [18], Citrini (1939) [19], Nebbia (1940) [20], lCindsvatel: (1944)
(21], Blaisdell (1948) (22J, Forster and Skrinde (1950) [23], Rotlse, 8iM,
and Nagaratl'l.am (1958) [2'1J, and mallY others. 2
The theory of jump developed in early days is for hOl'izontalor slightly
inclined channels in which the weight of water 1."1 the jump has little effect
upon the jump beha:/ior and hence is ignored in ~he analysis. The results
thus obtained, however,
applied to most cha.l1uels encountered in
engineering problems. For channels
large slope, the weight effect of
water il). the jump may become so pronounced that. it must be i.nc,luclecl in
the analysis.
.
Practical applications of the' hydraulic jump are many; it is used (1) to
dissipate enel'gy in water flowing over: dams,
and o,ther hydraulic
structures mid thus prevent sCQuriug: downstream from the struotures
(Art. 15~8); ,(2) to recover head Dr raisethe water level all the dovvnstream
side of a measuring flume El.nd thus maintain high wate,l' level in the
~ The experit'nellt was made ill Paris in 181;8 and re'pul'ted the following year in [11,
The hydraulic jump is also known as a, sta:n.di1lg wave. In French, it is called Ie ressm,1
hlldra.tdique . . In German, ,it is der Wass6rsprung. III honor of Bidone, the; hycil'Alllic
.;jump in Italian is named il sail a di Bidone (the jtimp of Bidone).
=For a, compreltsnsive review of the studies 011 hydraulic jump, see [,25], For a
mathematical trea.tment of the sllbject, see (251.'
,
M3
394
395
(3-21)
il
. ~ 18 r---t-~--r-
~ 16~-4--+-~--+--~74']
14
I/
t /
I
1--1----I'----+---+---1I<--t--r-t--r-
r---
, -I
I-
r---t---t---"-T--l-V+-+I---I-+-I---1---1.
I
I 0 r-~I---r-" t-I/--fj+----+-+-t-~I_--I---f--l
12
I--+--+-j-Y---+--r
6 f--+-,--!-t'--,f--+----l
.2=
~ (.j1+8F'-1\
Y, z
,
F{ V/-/9Y,
4~~-+7L~--l--+--r-'--I---'-''--+--1
2 17't- +----I--I---+--+--+---l---r---t-~"i
L-.--'-_-'--_-'i' _-,-,"_"
'---"--_~~"_'_~_L __~
10
12
14
16
18
20
22
15-1. Relation between FJ and y,jy,for a hydraulic jump in a hQrizontal ~ectangular channer.
. .
FIG.
,J
Il
I
"'"--1-
396
of
Yl
El
397
ydEJ, is .the relative seqtwnt depth. All these ratios can be ,shcnvn to be
dimensionless functions of F 1. For example,
hi
El =
,II
+ 8F 12
+2
F12
(15-2)
Since the relative loss, efficiency, rel2_tive height, and relative initial.
and sequent depths of a hydraulic jump in a horizontal rectangular
F 1, resulting in a
channel are functions of F 1, thcy c[',n be plotted,
Veiues of ,Ft
398
I I i I I I I I ! I
! I I I I I I ! I I
I-
I-6
1/
L
Y;
--
I-I--
II
, r--L=-i
ROII.r~
yll
,..;'P~::- __-- Y2
VI~'''''''''-'''''''''''''
?:?r"
1/
~tld\llor
li-'eolOS(;illtlliJiCj
lUI'fIP4fTlP
f.'
s,,'rlacp.7
jlimp
Sleady lump
f..-Vlo.,,)'
----- f.. --- ---- -- --- --8esl perto~rnOl1ce
lurbulenr.e only
/.
/.
,
b--
"I
I--
-+-
.'1'.""
,
F:-:'
!-.
r-
=i
-r
I I I 1/ I" I I I !
-~~
I,I-
Sirang- jump
--A~''-P'Ob''--T -- ;-p-.n;';';~~i~19b~i;.-'-;;J
per lotm:Jnt;;e
T
2
/~/.'I'I//
rough SudClce
c:ofloilior.s
I I I i I I l-C
IS
16
11. IS
IS
20
F"V.!m,
FIG. 15-4. Length in terms of sequent depth y, of jumps in l,lorizontal channels.
(Bt;!,sed all dala and rer.ommendations of u.s. Bl!rerwof BeclallwUon [34].)
Bureau of Reclamation data [34,35] obtained from six test flumes. Perfect agreement was found between the yJ/E 1 curve and the data. The
agreement between the E2/El or t.E/E l curve and the data was f!lirly
good except for PI < 2. The experimental curve for t.E/E l recommended by the Bureau is shown by the dashed line (Fig. 15-3). The
agreement between the yziE l and h;/E1 curves and the data was good for
high Fl values, but the scattered data failed to define the curve8 B.CCUmtely forF l <:; :3.
15-5. Length of Jump. The length of a j-ump may be defined as the
distance measured from the front face of the jump to a point on the surface
immediately dowIlstream from the roller. This length cannot be deter':
mined easily by theory, but it has been investigated experimentally by,
many hydrauliciaus. 2
'
I 'It is possible that at least part of this discrepancy is due to, the scale effect of the
testing model (see next article).
, Among them Safranez at the Technical University of Berlin [36,37J, \Y6ycicki at
the Federal Institute of Technology in Ziirich, Swit~erland [381, Araviri in Russia [391,
Bakhmeteff and Matzke at Columbia University [17), Moore at the California Institute of Technology 140}, a~d the engineers of the U.S. Bureau of Reclama.tion [34,35J.
3[}9
,(
I
1
!
I
'.
'-"l
400
--c
ever, the length of the jump should be considered. The following will
illustrate the locaGion of a hydraulic jump in three typical cases (Fig.
15-6) :
Case A shows the jump below a regulating sluice in a mild channel.
The profiles A Band CD c.an easily be identified as of J\lI3 and M2 type,
respeet.ively. The methods of computing these profiles are discussed in
Chap. 10. The curve A'B is a plot of the depth sequent to AB. By
"~
1'Ol~~-n
I
-J
'i
.............
Fl.
CIJse A
F,o 1.98~
0.9
0.8
..,
0.,
401
A'
F,=2.92"
I F,=4.09, -,~:-...."-+nl)Y--1r---j
F , ' 5 .53 ~-A,"--:;,,-A"l'---+--+
F l ' 8.63
0.6
~------.
.. 0.1
f.-...I"IA--+-~t-
4
x
/h j
{
J.,
the position of pI, the length of the jump can be estimated. By trial
and error, a horizontal intercept bet,ween the curves A'B and CD can be
found equal to the length of jump. For instance, the horizontal distance
EF is equal to the length of the j'Jmp corresponding to the dept,hyz at F.
It becomes apparent thHt the jump will fOl'm between G r.nd F, Slllce .the ;
depth at F is sequent to the depth at G und the distll,nce EF measures the
length of the jump. It may be noted th~t, if the length of the jump were
not taken into account in the analysis,the jump would have been considered to form at the upstream ]Joint F', resulting in an error represented
by F'F. In case A, it can be sen that, by. increasing the downstream
water depth or raising the curve CD, t~e jump can be moved up~tream.
The downstream depth may be raised to such a height that the ju~p will
eventually be ;drowned out.in front of t~e sluice. Decreasing the' downstream depth PI' lowering CD will move the jump downstream.'
.
The above discussion applies also to the lo!!ation of a jump fOI'med at
the foot of a weir 01' overflow spillway,
.
Case C
F1G.
When there is a hydraulic jump below a sluice, the issuing flow from
the sluice will form 0. jet that possesses a vena. conti"acta, '1'he distance
L, from the vena contract,!L to the sluice opening is usually short. Hegitrd. iug this distance, there is.a generally adopted rule which stc'l.tes that the
vena contracta is located approximately at a distance h from the sluice
, opening. l
1
~ Row from a sluice correspond,S to one-half the flow from a. drclliar orifice and that the
'lena contracta in the flow fro'm lJ. cil'cular orifice is located approximately at a, distance
I)f half an orifice diameter from the orifice, The rule of half orifice diameter was
originated by Weisbach [421.
.,
1
402
Example 1.6-1. Lo;)ate the hydraulic jump in Examp.le 10-3 if the flow downstream
.
~
,
from the jlllnp is uniform.
Solution. From the given data, the ,pecific-energy curve E = lJ + C!I. V~/2g and
the
curve F = 13Q;1gA + fA. of the channel may be constructed as sbown
in
'15-7, InComputing the spenific-force curve, the va.tue of fJ may be ~tima.ted
as 1.04 tor IX = l.10 [using Eqs. (2-6) and (2-7)).
Below the sluice, the M3 profile has been computed in Example 10-3, as ShOWll by
AGB in Fig. 15-8. Using the curves in Fig. 15-7 and following the method described
in Art. 3-7, the curve of sequent' depth A'F'B co~respondirig to tile curve of initial
depth AGE ClJ,n be determined,l The curve A'F'B a.nd the downstream flow p::-ofile
CF D (equa.! to the nOl'ma.l-depth line in this example) interSect at P'. The initial
depth of flow at F' is then found from the M3 profile to be 1.70 ft. The (l/ll'responding F = 1.52 and, from Fig. 15-4, Lly. = 3.6, The length of the jump is
therefore equlJ,l to L = 3.6 X 2,67 = 9.0 ft. At this point there is an approximation involved, becs.use the length of jump should be based on F a.t E (instead of F/l,
which, however, is as yet unknown ..
In this example ,I> is equal to. the OOrmal depth of flow in ,the channel since the flow
downstream.is uniform: If the How downstream is not unif,Qrm but gra.dually varied,
then the depth a.t the intersection F' of the downstream profile with the curve A'F'B
should be taken as 'iI" This i.;;. also an approxima.tion, because the aqtua.l depth lJ1
should be at F, the position of which is as yet unknown.
.
.\
JUMP AND ITS 'USE AS ENERGY DISSIPATOR.
II
is
bR~
~~Cc'
5f-.
<-
/
I
2~
if
\/
1/'
!I
100
1"10.
II
/ I ,
i
I
--
I
I
~E'~_
~ye
"""-....
r-- 1---".- . - .
2
200
I
I
'-,
/,
\1
oV
o
/:/'
403
::;
;)00
6
400
~..;
B
500
I
9
10 Scale (Dr E
600 Scole (," F
157, The E curve and F curve for the determination of jump location.
i
'.1
.\
!
. 1
----:'!.I:.._
~-- -----~ p-~-- - --
M3-proI1l.e
. I
. lOwing to the difference between," and 13, the computed critica.l depths indicated
;by the specific-energy and specific-force CU.J:VBs are not precisely identica.l. However,
the discrepancy is so small that it cnn be ignored.
.
FIG. 15-8. Location of a hydraulic jump.
404
,
"
405
t.han Y2. This means that the tailwateI' depth in' ?l1se '1 is decreased .
As a result, the jump will recede downstream to a pomt ",-here ~q, (3~21)
is again satisfied. This case must, if possible, be aVOlde~ lD deSign,
because the jump, repelled from the
. apr~n, will take place
either on the ;oose rubble bed or, still
m entIrely unprotected
15-8. Jump as
Dissipator. From iL practical viewpoint,
hydl'iLulic jump is a useful means of dissipating excess energy in l3upercritical flow. Its merit is in preventing possible erosion below overflow
spillway:?, chutes, and
for it quickly reduces the velocity of the
flow on a paved apl'on to a point. where the flow becomes incfl.pable of
scouring the downstream channel bed,
The hydraulic ju~p used for energy dissipation is usually confined
pa.rtly or entirely to a chunuel reach that is known as the stilling basin.
The bottom of the basin is p~wed to resist scouring, In practice, the
stilling basin is seldom designed to confine the entire length ofa free
hydraulic jump on the paved apron, because such a basin would be too
expensive. Consequently, necessoria!: to control the jump ure usually
installed in the basin. The
purpose of surh control. is to shorten
the range .within which the jump will take place and thus to reduce the
size and cost of the
basin. The control has additional advantages,
for it improves the dissipation fLU1ction of the basin, stabilizes the jump
action, and in some C!lSeS increases the factor of safety. In designing a
stil:ing basin using hydraulio jump as energy dissipator, the following
practical features should
COllflidered. 1
A. Jump Position. There are three alternative patterns (Fig. 15-9)
that allow a hydrnulic jump to form dr}wnstream from the source (such
source as an overflow ilpillway, a chute, or a sluice):
Case 1 represents the pattern iil which the tailwatel' depth Y2' is equal
. to the depth Y2 sequent to YI. In this case, the values of F\, y), a.nd
Y2' (=: Yl) will satisfy
(3-21), and the jump will occur on a solid a.pron
immediately a.he~Ld of the depth y 1. For ;;eour-protection PlU'POS~s, this is
an ideal case, One big objection to this pattern, however, is tha.t a little
difference between theactual and assumed values of the relevanthydrauliD coefficients may cause the jump to move downstreani from its esti:r;llltted position. Consequently, some device to control the position of
the jump is alw!1Ys necessary.
Case 2 represents the
in which the tailwater depth y{ is less
l For sinlplicity, the lellgth of the' hydraulic jump will not be considered in the
present discussion. See [251, [341, (35), [4S), and" Hydraulic. Energy Dissipators," by
E, A. Elevatorski, Mc'Graw-Hill Book Company, Inc., New York, 1959, for further
iniormatioll 011 the design of stilling basins.
CjJSi! I:
Yz
ccse 3: yi ,. Yz
FIG. 15-9. Effect of tauwllter depth on the formntion of a hydraulic j lInlp below a w"ir
. or sluice.
chm1llel, resulting in severe erosion. .The remedy for the design is to use
a certain co~trol in the channel bottom, which will increase the tailwater
depth and thus ensure a jump within t.he protected apron.
Case 3 represents the pattern in which the tail \V\.ter depth Y2' is greater
than Ya. This means that the tailwater depth in case 1 is increased. As
a result, the jump will be forced upstream and may finally be drowned out
at the source, becoming a submerged jump. This is possibly the safest
case in desi~w, because the position of the
jump can be most
readily fixed .. Unfortunately, the design is no~ efficient, for little energy
will be dissIpated.'
., .
.r
406
yi'toilwa:er depth
.~<>ilL
Coso. I
toilwcler rating
-'~rJU~"p
1
Clfin g
: , /;
~ I
~ ~
Discharge Q
Cose :3
jump will occur on the protected apron is to use sills to create a stilling
basin (Art. 15-9).
Class 3 represents the cot'ditions in which the jump rating curve is
always at a lower stage than the tail water rating curve. This means that
case 3 in Fig. 1.5-9 occurs at all times (i.e., the tail water is higher than the
sequent depth) and that the jump will move upstream and will probably
be drowned out at the Source. Consequently, little energy wilt be dissipatr:!d. An effective method to ensure a jump is to build a sloping apron
above the channerbed level (Art. 15-16). The slope of the apron can be
such that proper conditions for a jump will be developed on the apron at
aJl discharges. Another method is to provide a drop ill the channel floor.
alJd thus to lower the tailwater depth (Art. 15-10).
Class 4 represents the conditions in 'which the jump rating curve is at a
higher stage than the tailwD,ter rating curve at low discharges but nt a
lower stage at high discharges, An effective method to ensure n jump is
to provide a stilling basin for forming a jump at low discharges and to
combine with the basin a sloping apron for developing a satisfactory jump
at high discharges.
.
Class 5 represents the conditiona in which the jump rating curve is at a
lower stage than the tailwater rating curve fit low discharges but at a higher
stage at high discharges. An effective method to ensure a' jump is to
increase the tail water depth sufficiently high by providing a stilling pool,
thUd forming a jump at high dischatges.
C. J'urnp Types. In~, view of the various types of hydraulic jump
df;lf!cribed in Art. 15-3, the U.S. Bureau of Reclamation [34,35J gives the
following practical considerations:
1. All types of jump are encountered in the design of stilling ba5ins.
. 2. The weak jump requires no baffles or special consideration. The
Ollly requirement necessal'Y is to provide the propel' length of a pool,
which is relatively short.. This can be determined from Fig. 15-4.
3. The oscillating jump,.frequently encountered in the design of canal
structnres, diversion dams, ::mcl eVen outlet 'Yorlcs, i.s difficult to handle.
If possible, jumps in the 2.5-to-4.5 Froude-number range shollid be
avoided. In many cases use of this jump cannot be avoided but, in other
cases, aitering of dimensiot's ma'y bring the jump into the desirable range.
Baffle blocks 01' appurtenances are of little va:lue. Waves are the main
source of difficulty; hence specially designed wave suppressors may be
used to cope with them.
4. No particular difficulty is encountered in steady jumps. Arrange"
ments of baffles and sills to be discussed later will be found valuable as a
means of shortening the length of the stilling basin.
5. As the Fronde number inereases, the jump becomes more sensitive
to taihvater depth, For Froude numbel;s as low as 8, a tailwater depth
Cose 2
"'0
\'
Tailwoter
roting
. .
Discharge Q
Cose 4
Cose 5
->-,"
'C
<:
":::'
>.
'"'"co
"
'"
'"
Disc horg e 0
'"
~
'"
407
'C
<:
":::'
'"
"'"
"
(;,
Discharge Q
to show the relation between the sequent depthY2 and the discharge Q;
Because of the difference in the relative positions of the two rating curves,
Leliavsky (43] has suggested that the design may be considered according
to five different classes of condi tions (shown respectively by fLVe cases in
.
Fig. 15-10):
Class 1 represents all ideal conditions in which the two rating curves
always coincide. This means that case 1 in Fig. 15-9 exists at all times
and that a jump will form at the desired place on a protective apron at
all discharges. . Conditions of thisolass are mrely encountered naturally.
Class 2 represents the conditions in whic.h the jump rating curve is
always at a higher stagetha,n the taiIwater rating curve. ThiS means
that case 2 in Fig. 15-9 exists at all times (i.e., the tail water depth is
lower than the jump sequent depth) and that the jump will form at a
certain place far downstream. An effective method of ensuring that the
i
~.
"
SENERGY DISSIPATOR
408
greater than the sequent depth is advisable to make certain that the jump
will stay on Hm apron.
6. When the Froude number is
than 10, a stilling basin using
the jump a dissipator may no longer be the most economical dissipatiun
device. In this case, the difference between the initial and sequent
depths is great, and) generally spe&king, 8, very dj'lep basin with high
basin may .not be
retaining wall is required. The cost of the
commensurate with the results obtained. A bucket
of dissipator 1
may give comparable results at lower cost.
15-9. Control of Jump by Sills. The hydraulic jump can be controlled
or effected by sills of various designs, such as sharp-crested weir, brcadcrested weir, and ~,brupt rise and drop in channel floor. The function of
the sill is to ensure the formation of a jump and to control its position
under all probable operating conditions.
.
Interesting experiments [46J have shown that the forces acting on the
sill in a jump decrease rapidly to a minimum as the downstream end of
the jump is moved upstream to a position approximately over the silL
The force then increases slowly to a constunt ,;alue f.s the jump is moved
farther upstream. This change in force on the sill is probably due to a
change in the velocity distribut.ion from one end of the jump to the other,
since llonuniform distribution of velocity is a characteristic of such
rapidly varied flow. As a result, the momentum in the nonuniformdistribution section is greatly increased. Theoretically: speaking, the
control of hydraulic jump by s~lls can be analyzed by t.he momentum
theory; Because of lack of accurate
of the velocity distribu-"
tim!., however, the theoretical analysis cannot predIct the quantitative
result very closely. For useful design information one has to rely upon
e..~perimental studies.
. Dimensional a::mlysis shows that the relations
the Froude.
number Fl or F of the approachingflo\~) the height 12 of the
approaching depth Vl, the depth Y2 upstream from the aill, the distance X
from the toe aLthe jump to the sill, and the downstream depth 1/3 may be
expressed as
as
:l
=,
~ ( F, ~, ~:)
an
The
1 This is
\Ipturned bucket provided a.t the toe of !I. sp'illway to deflect the overflow up :through the tailw!l.ter. If tlle ta.i1water is high enough t:o submerge the
bl~cket, :Ii. roller wiU form downstrelj.m from the bucket and tend to mo~e bed materials
toward the dam. thus preventing Eel'io~!s scour at the toe of the dam:. Onthe other
h!lD.d,if ~he ta.ilwater is low, the overflow will be thrown up and alit so th!lt it will
strike a solid-rock channel, if any, at a safe distance below the dam; See [44), [45]1
p.nd Arts; 1<1-10 and 1 4 - 1 1 ,
. .
409
,
.
..,.
.
scont~olled by the sill, however, cannot be
exact pOSItIOn of the Jump. a
d l t d this positio.n can be repre. d
1 f cally In the rno e s u YI
.
t
. cletermme ana Y .1
,.
X .d
The ratio is taken as constan
sented by the ra:10 between. an "uY2~iellt to ~nsure a complete jump.
in each tes~1 hav!llg a m~gn~tud~till~ basin should be made at lea~t
In the deSIgn, the lengt .0 a ~
gh' er the length of the basm
.' X.'. }i'<lr economiC reasons, 0',1, ev I
.
equa" 1.w
.
,
F hIli !lnd Xly~ for
FIG. 15-11. Experimental relal.lons among,
t,
(A.jler Fm'Bte1' "'Id Bhinc/e [23J.1
II.
".",,_~. re"C~Ll
weir.
mtatyh' be
a ee
stream channel condit.ion.
. . of ex )el'imental data
o
,
b tween the well' heIght h an t e
the appr~achln~ ~ow, (2~ ~;) ~~:l~ati~ between the distance X fram
approachmg dept Yl, an.
d th dep! tIl y upstream from the Well',
h'
to the well: an
e
2
"
toe 0 f t e Jump.
1 . f the effec. of a (TivBJ;l weir for known
This
pe~mrts an ana !~IS 0
,."h t the norm.:'l.l tailwater
anproach and tallwater condItions, pIovlded t a.
'st.t that iil pro
.
ff t th discharge over t he, Well' Cle}
<,
~~X:~ ~o;: +O~.;5~~ Or~inarilY, ul1su'bmerged conditions pteya~l ,over
glf.l.Dl
(15-3)
"
t?e
:33
.
1 This condition was first observed by Bazin (see reference [12] of Chap. 14) find
. later demon:;trated theoreticn.l1.y by Bakhmetcfi [27].
,
.
)
410
411
RAPIDlJYVARIED FLOW
a wide l;allge of
Thus, v.lvl in Eq. (15-3) can be dropped.
If submergence occurs, the jump is forced upstream with possible drowning as a .result,
In the diagram, any point is represented by a pair of coordinat~s
(F,h/Vl)' If the point lies within t,he curves, hydniulic jump will occur,
with its relative position indicated by the corresponding interpolated
value of X/Y2. Points lying above and to the left of an interpolated curve
represent the conditions under which the weir is t,oo nigh, so that the
jump will be .forced upstream and possibly drowned at the sonrce.
Points lying to the right,of the curve repre'sent the c011ditiol1s under which
the weir is too low, so that the jump will be forced dO'wllstreal1l and
possibly washed out. Y{hen the weir is as low as this, it ma~r be crossed
only by a single undular surface rise, forming a standing swell (Fig. 15-6).
For design purposes, it, is proposed that the curve X/Y2 = 5' in the
diagram be used. Laboratory expedence has shown that the highest
required weir does not necessarily occur under conditiollSOf maximum
discharge [20]. Consequent.!}', the highest required weir should have the
largest required value of h within the expected l'3.l1ge of
as
determined from the diagram,
B. Control by Broad-crested Weir. For a br'Jad-crested weir, ift11e
downstream depth is lower than the critical depth OIl wp of the weir, that
is, if Ys < (2Y2 + h)/3, the tailwater iyillnot affect appreci~~bly the rela-
tion between the head water elevation and the
Thus, .the
discharge over a unit width of the weir can be written'
= V IY1, H
can be reduced to
Since q
0.433
= Yz
2.667:F 2 (1
.
+ )
+ ya!
}tIy2.)
YI
.F
.2
.S1
1:;
~ 2r---~+-----1-----~~~~--~-~---r----~-----1
(15-4)
"-hit
(A/tel'
(15-5)
(15-6)
This equation giv.es the relation between h/Yl and F and can be plotted
. as the curve shown in Fig. 15-12. Forster :and Skrinde (23J have found
that this curve coinddes with the: experimental data for an abrupt
rise with y~ = Yo for X = 5(h + Ya)., Despite the lack of furtherexperiThis is
Values of F,
, i
~.
~ 'H:;'
. Yz
h
II, and F
(l.
.oj
..
412
tH3
jump will be forced upstream and may finally be drowned. If the point
is at the right of the curve, the rise is too low, and the jump will be forced
downsti'eam toward the' abrupt rise and mayfinnJly be wil.shed out.
If a point (F,Y3/Yl) lies in the diagram below the line Y3 = 1'" then the
normal downstream flow is sl.lpercritical. A jump is followed by a critical
section cre..'Lted over the crest of the abrupt rise, so that the rise acts as a
weir, and the design diagram (Fig. 15-11) for the sharp-Cl'ested weir may be
. used.
V1-
v,
Y, ----+-
,
~.
l
\
16~
Volue s of F
FIG. 15-13. Experimental relations among F, YJ/V I, and 7t/y, for an abrupt rise.
Forster and Shhule [23].)
(AJter
For design purposes, the diagram (Fig. 15-13) can be used to determine
the necessary length and depth of stilling basin when V 1, Yr, and Ya are
known. It is proposed that a point (F,1/3/Yl) be first defii,led for conditions at or near maximum discharge and tha,t the corresponding value of
hjyl be determined by interpolation .. By repeating this procedure for
other discharges within the expected range of discharge, a largest required
value of h enD be obtained. This value should be used for the highest
required rise. A minimum height. of rise necessary to prevent the jump
from being washed out can also be thus determined.
15-10. Control of Jump by Abrupt Drop. The control of hydraulic
jump by sills is useful if the downstream depth is smaller than the sequent
depth for a normal jump. If the downstremu depth is btrger than the
sequent depth for a normal jump, a drop in the channel flooi' must be
used in order to ensme a jump. This condition occurs generaUy at the
end of the expansion of R supel'critical flow.
For Il. given appro<lching Froude number, the do\vnstream depth of a
drop may fali in any of the five regions as sh9Wl1 in Fig. 15-14tt. The
lower limit of region 1 is the depth at which the jump will begin to travel
upstream. The uppe.r limit of region 5 is the 'depth at which the jump
i4n~~-l
12
~"'--+---+-----t-.--
,;::- 10 f-.----+---j--.-
;,.
'"
i :1--~qL___;~..w~-~
-f
----+----t-------1
~"i-3-~.:--
.'
OO~~'~"--~--~2~-~~3
4L----::---;:s
5
F=vl/fol
,
Ie
10
.
I
f
I f ns among F Y';YL, and lI'Yl or an
a10'
.
I
i
414
will begin to tr:'L\'ci do "'llS cream , Evidently, the drop does not control
the jump in these two regions. The jum.p is stable and the drop is effective for its desired purpose only in regions 2 and 4. The intermediate
region 3 represents an undular state of flow without a breaking front.
By applying the continuity and momentum equations in an analysis
similar totlutt made for the broad-crested weir (Example 3-2), HSll [47J
has shown that, for region 2,
(15-7)
and, for region 4,
.
,.
i
,
i
(15-8)
I\
F2 = liYa/Y1
[( ~ + 1)2 _ (J0.)2l
7't 1 - yJivl \ Y l 1 l 1 J
A simplified analysis and further experiments were alsomade later by Moore and
Morgan [481.
, The dentnted sill or serrated baffle is also known as the Rehbock sm; because it
was first patented by Prof. Theodor Rehbock after it was developed from experiments
made during 1924 to 1927 at the Technical University of KarLsruhe, Germany.
415
ing velocities, the sill is usually dent::Ltcd to perform the additional function of diffusing the residual portion of the high-velocity jet that may
reach the end of the basin.
Baffle piers are blocks pla,ced in intermediate positions across the basin
floor. Their function is to dissipate energy mostly by impact action.
The baffle piers are very useful in small structures with low incoming
velocities. They are unsuitable, however, where high velocities make
c!witation possibie. In certaiIl circumstances, they must be designed
to withstand impact from ice or floating debris.
There are many generalized designs of stilling basin that use a hydraulic
jump as the means of energy dissipation.! Three typical designs will be
described in the subsequent artides:
1. The SAF basin. This is reCl)mmended for use on snutll scructUl'es
such as small spillways, outlet works and small canal structures whBre
FJ = 1.7 to 17. The reduction in ba~in length achieved through the use
of appurtenances designed for the basin is about 80% (70 to 90%).
2. The USBR basin II. This is recommended for use on large structures such a,s large spillways, la.rge canal stnlCtures, etc., for FJ > 4.5.
The jumwand-basin length is reduced about 33 % with the us~ of appurtenances.
3. The USBR basin IV. This is recommended for use with jumps of
FI = 2.5 to 4.5, which usually OCCll!' on canal structures and diversion
dams. This design reduces excessive waves created in imperfect jumps.
It should be noted that these designs are only typical examples and
applying them to stilling basins under
t.hat caution should be used
entirely different design conditions.
The principle of the stilling basin applies also t.o the design of a canal
drop (.01' canal fall), which is a sti'ucture built to secure the lowering of the
water surface of a canal and the safe destruction of the energy so liberated.
The canal drop is sometimes designed with a contracted width like the
Parshall flume. Such a drop is known as a fiumed drop, which can economically be built together witha crossing bridge and used as a meter or a
regulator as well [50,60-63].
15-12. The SAF Stilling Basin. This basin (Fig. 15-15; SAF denol;es
"Saint Anthony FalLs") W2.S developed at the St. Anthony Falls Hydraulic Laboratory, University of l\iinnesota, for use on sma.lJ ,drainage structures such as those built by the U.S, Soil Conservation Service. The
design rules summ[Lrized by the investigator Blaisdell [22,64] are as
follows:
L The length LB of the stilling basin for Froude nuinbers between
FI = 1.7 and FI = 17 is determined by LB' = 4.5ydF1G.76.
in
1 For mOl'e info!'.mation see [9], [25J, [34], [35], [43J, and [49J to [,58J.
developed by the U.S. Army Corps of Engineers, see 1591.
For designs
.)
j
I
I
(
416
j,
I
l.
CENTERLINe:
TRAPEZOIDA;"
SECTION
it
DOWNSTREAM ELEVATION
(U.S. Soil
'
417
2. The height of the chute blocks ::md floor blocks is y 11 and the width
. and spachlg are approximately 0.731/1.
3. The distance from the upstream end of the stilling basin to the floor
blocks is L11/3.
.
4. No floor block should be placed closer to the side v;all than 3y/.8.
5. The floor blocks should, be pbced downstream from the VjJ"~""'''50
between the chute blocks.
6: The floor blocks should occupy between 4{J and 55$ of the stillinghasin width.
.
7; The widths and
of the floor blocks fOl'divel'ging stilling
basins should be increased in proportiOll to the increase in stilling-basin
width at the floor-block location.
8. The height of end sill is
by c.
O.07y 2, 'tV here y 2 is the theoretical sequent depth corresponding t,Q 1f1.
g~ The depth of tailwater abov6 the stilling-basin floor is given by
yl = (LlO - F 1 2/120)Y2, for
1.7 to 5.5; by YI' = O.85Y2, for
F1 = 5.5 to 11; and by 'Y2' = (LOO
FN800)Y2, for F1 = 11 t'o 17.
10. The height of the side wall 1lbove the maximum tailwater depth
to be expected during the life of the structure is given by z = Y2/3.
II. Wing walls should be equfll in height to the stilling-basin side walls,
The top of the wing wall should have i1 slope of 1 on 1.
12. The. wing wall should be placed at an angle of 45 to the outlet
center li.~1e.
i3. 'l1.e stilling-basin side wail~ may be parallel {as in 1} rectangular
stilling basin) 01' they may diverge as an extension of the transition sidl;)
walls (as in a trapezoid~l stilling basin).
14. A cutoff wall of nominal depth should be used at the end of the
stilling basin.
.
15. The effect Qf entirained air should be neglected in the design of the
stilling basin.
15-13. USER Stilling Basin II. From intensive studies of many
existing structures and laboratory illvestigations, various tYPElS of generalized design of stilling basin have been developed by theD.S, Bureau of
Reclamation [34,35}. L"SBR basin I is the basin created by a jump
occurring on a flat floor with no appurtenances. This can be designed
easily by following the principles descdbed in the' early D,l'ticles of this
chapter. However; such a basin is usually not very pructic!).! because of
its expeI)sive length and its lack of control. USER basin IIi is designed
for a pu~pose similar to that of the SAF basill, but it has a higher factor
of safety, which is neceSS!1ry for Bureau use. The pel'forrhance of this
basin in~icates that the jump-and..:basilllength can be reduced about 60%
. with the appurtenances, as compared .vith 80 % for .the SAF basin.
Therefol'e, the SAF basin is shorter and more economical but, in cons&-
II
<II.
'118
-I
.5
.b..D2
-\
VI
F "--
I~
(c
1
(a]
28
26
24
2220
I
~.
--
' -:;c-=--
I /
18
."'-:;;'
14
1 OJ/
'0/ .
0'/
I?~
1'/ /[
.j
/,/
1/
17
. /Sj/
;} /
!
I I
ii-r-
1<9;.
I
\// i///
~6
I'
j; ",i
A.7Q~
1--..
///
I
:'/.1
'<
Mi~fml,lm
-W
depth
/
11# .
12
, 1/;, ~?~
'/
10
'S
F,'
10
:0
8
6
(dl
/_M
\~
,-
..
..
1
i
f-~
..
;.........-
"/
2r-#"
16
/,1
~'
/
14
10/'/
II
12
Vj
./gD,
.l
t----.
i
6
iO
(3
-F
1
12
14
16
18
20
V,
.,.tgi),
~~
(b)
FIG. 15-16. Design curves and proportions of USBR ba.sin II. (U.S. Bureau of
Reda.;nation [34].) fa} Definition of 5jmbols; (b) minimum. tailwatet depths; (c)
length .of hydraulic jump; (c) n.pprotimate water surface and pressure profiles (conjugate depth = sequent depth); (e) recommended proportions.
I'
I
420
quene!!',. has ~ lower safety factor. USER basin V is us6d where struc~
tU~'al er,onOlll,te8 dl~ti1.~e th6 lise .of a sloping a.Pl'Oll, usually on high~dam
s~dlways. ~ he ?nnclple of deSign for hydraulic
on sloping apron
:Vlll ~e descl'lbed 111 Art.. ,lS-16. USBR basins II and IV will be described
III tillS and the next articles; respectively.
. USBR basin II was developed foJ' stilling basins in common use for
hlg~~dam a,~1(t eUl'th-dam spillways and for large canal structures. The
. ba,sln contall1S {:hute blocks at the upstream end and a dentated sillnel1r
the do:vl1stl'eo.m end. No baffle piers are used because the relatively high
velo?ltles
the jUl,:\lP might canSe cavitation on piel;". The
detailed
and the data for COn1FJUtutiollS al'e shown in Fig. 1.5-16.
The rules recommended for the design are as follows:
1. .s~t apron eleva.tio~l to utilize full sequent tailwater depth, plus ~n
acl~led mctol' of sdety 1f needed. The dashed lines in Fig. 15-16b are
gmdes dn1wl: for vtl,dous ratios of actUttl tailwaterdepth .to sequent
depth. ,::itud1es of exis~il1g
in:dicate that most of the b,usins wer<,
de~lgn~(l for sequent. tallwatel' depth or 1,,8s. HowBver, there is a. lilI:cit,
WhlCh.l~ ~overned by the curve labeled "Minin~um TW depth." TillS
curve IndICates the point at which the front oithe jump mo\'cS [l:\~Tay from
the chute blocks. In other words, any additional lowering of the tail. water depth would cause the ju:np to leave the 'basin; that
would
pro~uce u "sweep-out." For design purposes, the b!l.Sin should not be
deSIgned for less than sequentclepth. For additional safety, in fact, the
Bureau recommends that f, minimum safety margin of 5 % of
be added
to f,he sequent depth.
.
v3.8v
1 The slope of the chute has little effect on the jump as long as the velocity distribution and depth of tiowa.re reasonably uniform on entering the jump.. If the chute is
long and fiat, the velocity maybe concentrated in one part of the flow section, resulting in an asymmetric8.1 jump with strollg side eddies. The operation of suell (l. .jump
is expensive and should be a.voided in ail stilling basins. Also, when the angle of
divergence of the chute is too la.rge for water tp follow properly, the jump will become
rough and uncertain its position.
of
422
The length of basin can be obtained by entering the curve in Fig. 15-16c with
F, = 7.13. Tllll~, LID, = 4.16, or L = Lll = 4.Ui X 36.9 = 154 ft.
The height" wid th, and spacing of the chute blocks as recommended are D I; thus the
dimension can be 3 t 10 ill. The Might of the dentated sill is O.2D" or 7 ft 5 in., and
th~ wicl~h anc1 spacing of the dent,ates can be O.ISD" or 5 ft 6 in.
423
15-14. USER Stilling Basin IV. When F 1 =2.5 to 4.5, an, oscillating
jump WIll be produced in ,the stilling basin, generating a wave that is
difficult to dampen (Art. 15-3). USBR. basin IV (Fig. i5-H) is designed
to combat. this problem by eliminating the w~ve at its SOUl'ce. 1 This is
Imlmllllll//TIIlI//IlIII/I?/dIlIII/llwm
,..j---;---- L
-..j
and those of Moore [4.0J and Bakhmeteff and Feodoroff [65J, Rand [6l1J
found t.hat the flow geometry at straight drop spillways can, be described
by functions of the drop numL'61', which is defined as
o
Fractional
D=L
gh
space'
VI = mox.tooth 'Hid Ih OJ
; - 5 p oce,: 2.5w
Top
~d
(15-9)
where q is the discharge per unit width of the crest of overfall, 9 i's the
acceleration of gravity, and h is the height of the drop. Tho functions
are
'surface on 5 slope
4.30Do.27
(15-10)
Vv
h = 1
_. OODc.22
(15-11)
~!l
(15-12)
0.54Do.m
~2 = 1.66Do.27
!
\
I
I
I
I
."
1.,
(15-13)
where La is the drop length, that is, the distance from the drop waU to
the position of 'the depth YI; Y'P is the pool depth under the napp'e; VI
is the depth at the toe of the nappe or the beginning of the hydraulic
jump; and Y2 is the tailwater depth sequent to Yl. The position of the
depth Vl can be approximately determined by the straight line ABC which
joins th.e point A on the apron at the position of Yl, the point B on the
axis of the nappe at the height of pool depth, and the point C on the axis
of the nappe lIt'the crest of the fall. The fact that these three points lie
on a straight line was also verified by experiment.
For a given height h and discharge q per unit width of the fall crest, '
the sequent depth Y2 and the drop length La can be computed by Eqs.
(15-10) and (15-13). On the one hand, if the tailwater depth is less than
-1
,1
424
I)
r)
(J
\
)
;
!)
r '
If
I)
Y2, the hydraulic jump will recede downstream. On the other hand, if
the tailwater depth is greater thanY2, the jump will be submerged. As
the trtilwater level rises, the spillway crest may be finally submerged.
The spillway will still be effective if the submergence does not reach the
control depth Oil the spillway crest. The upper surface of the submerged
nappe lllay be assumed as a straight line tangent ito the, upper surfn.ce of
the free nappe at the point w~lere the nappe plunges into the tailwater.
The upper surface of the free nappe may be represented by the general
equation given in Art. 14-l.
4.1Q
WN V2gYl
(15-14)
~NERGY
D1SSIPATOR .
~f
425
a space !n ft,
i
I
-I
I
. ,
~se" '
2.
to ,
4 kn. own as drowned-out Jumps,
pI.ac t'IC,al pI'oblems
, " C'"
. . t are
common forms anci usually appear simply as jets of w~ter plunglllg m. ,o.a
d'ownstream pool below the steep slop~. For pra;,tI~al PU~'poses, It~:
believed that the solutions for the typIcal form 01 c_ase 1 a.nd fo: t
drownecl-ollt jumps nre mlltually, applicable. Case b shows the Jump
on' an adverse slope. This is a rare type of jump, and no adequate
experimental data are ava.ilable !!.t the present moment. '
.
For the analysis of the jlimp of case 1, a rectangulm~ channel of umt
'wi.dth is assuH),ed. Considering all effective forces paraLel to the channel
bottom, the mOll'lentulll equation lllay be written
9!!!. ce 2Y2
g
- (3llll) = P l
P2
+ lYsin e -
Fj'
(3-14)
wh~re
Y,/vgch; and.
426
simplifying,
where
(~y
427
Since d 1 = 111 cosO and d~ .= Yz cos fJ, Eq, (15-18) may also be written
(ZGZ
+ 1) ~ + 2G2= 0
(15-10)
Yl
(15-17)
!
(
HC vi-+ 8G2
- 1)
.'
(15-19)
The above two equations ai'e analogous to Eq. (3-21). Since G = f(F1,f)),
these equations indicate that ddd~ l:md
are functions of FI I,nd fJ
'-I
TI 'e is a general belief [70J th,l,t J( and L/(dz - d 1 ) vary primarily with
leI d h
tl at G is a function of F I and 0, or. G = f(F 1,8). .
F I o.n l ellee, h.
'Nith Eq. (3-20) for level-floor Jump" a
If Eq. (1.5-16) IS
.'
r'd t Following the solution
. '1 'ty between the two equatlollS :8 e\ 1 en .
,
SImI an
'
( - 16) .
. .. ltlv
for Eq. (3-20), the solution of
10- ) IS appa.eI
H(V1 + 8CP -
(15-18)
1)
Co se I
\6
17
Ie
'9, 20
FIG. 15-20. Experiment!!.l relll.tions I::ietween F, lind lIz/Y, or d,/d J for jumps in sloping
cha,nnels.
'
case :}
Case 2.
Cose 4
or
co
"1
'-
.J
I
\
I
i)
428
)
429
\
I
'
2l-+-I-;!'+-I-1)
Section A'A
'0 .
;:
:3
or
10
Ij
17
18
19
20
F, 1r,/:I9d,
seq'.le~t
chllnll~ls.
(U.S,
2. Position the apron 80 that the front of the jump will form at the
end of the slope for the maximum discharge a.nd tnihva~el'
condition.
3. The portion of the jump to be confined on the stilliilg basin is a
uecision for the uesigner,. depending on the qUlllit.y of the rnatel'in.l in the
river bed and other conditions. The average ovel'-o.ll apron is about 130 %
of the length of jump for the U1:1xi::num discharge condition.
4. With the apron designed properly for the maximum discharge COll: dition, the next 8t~p is to be certain that the tailwater depth [1.ud basin
: length available for energy dissipation are sufficient for, say, one-quarter,
one-half, and three-quarters of capacity.
.
5. A hCll'izontnl apron will perform on par with the sloping apron for
high Proude numbers, if propel' tailwater is provided.
upstre~llll
\,
.,
I
I"
I
II
!
I
White .[SDI.
t.
'-1
431
F nl
V"l
--=
~gYl
Y 1 sin (3
~gYI
. (3
-113111
(15-20)
Considering a section A-A nOl'lnal tb the wavefront, it is seen that a normal hydraulic jump occurs in this section and that Eq. (321) call be
applied. Substituting Eq, (15-20) for Fl in Eq. (3-21), the ratio of the
sequent to initial depth is
1:!~ =
Y,
1)
( 15-21)
This is the eqllation tlmt represents the condition for ~\:l oblique
hydraulic jump to take place.
Iteferring tQ Fig. 15-22, the ta,ngential velocit.ies before and after the
jump are VII = TT "i/t::m (3 and V 12 = V,,2/tan (3 - 0). Since no Jl1ornentum chfl,l1ge takes place patallel to the wavefront, these two velocities
shou1d be equal, or
V"I
tan-f3
(15-22)
'V~; = tan ({3 - 0)
By the condition of continuity 'VI V"I = Y! Y,,2, 'the above equation 08,n be
written _
Y2
tan f3
(15-23)
Y; = tB.n (3 - 8)
I
i
I
I
,I
II
-\
'-
~.
Elimino,tingudUl from Eqs. (15-21) and (15-2:3), a relationship involving F 1; 0, and .13 is obtained:
I
\
This equation should produce the value of t3 if F 1 and. e are g-iven. However, a direct solution of t.his equation for {3 in terms of Fl and e is practically impossible. Ippen [78J has prepared a four-quadrant graph _(Fig.
1:'5-23) showing all relationships expressed by Eqs. (15-21), (15-23), and
(15-24). This gl'aph is 8elf-explanatory and can be used for the solut,ion
of an oblique hydraulic jump.
Since the oblique jump is a normal jump across the section A-A, the
energy 10:0:0 in the oblique j'ump can be computed by Eq. (3-24). In
practical problems involving an oblique jump, Y2!YI' is usually small.
Thus, the head loss may often be neglected in design. Like~. normal
jump, if vdvi < 2 (or F"I < 1.7, according to Art. 15-3), the oblique
jump becomes undular. This' fact has been confirmed both experimentally and theoretically by It1pen [79J.
It should be noted that an oblique jump will rarely oectll' alone in an
1
I
FIG. 15-23. Genera.! relations among Fl"e, /3, y,iy" CLnd F, for oblique hydraulic jumps.
(Allel' A. T. lppen ~781.)
_
_
15-1. Show that the {oilowing equation lIl[Lybe written for a hydraulic jump taking
place in a. horizontal trapezoidilJ channel (some of the notation is given in Table 2-1):
il
r'
3F l ')(1
+ I)]r
+ 1)2
- 3F,'(t
where ,.
L
...
.- i
1
PROBLEMS
"':'/:":;'"
.-1
y,/y"
1 = b/ZYl, and Fl
=V,;VOy,.
= 0
(15-25)
432
,,
I
I
/.
I!
fOl"
.
.
,
HYDRAULIC JUMP AND ITS USE AS ENERGY DISSIPATOR
now
Ydv'ilih.
where r = Vo/lll and F',
16-3. Verify Eqs. (15-0 and (15-2).
. 16-4. In Bidon~'s experiment it was found that, for V, = 5.59 fps and !II = 0.208 ft
m a rectnngularchannel J the sequent depth y, was 0.613 fi" Determine' (a:) ~he alternate .depth, (b) the tbeoretic~.l sequent depthJ (I;) the energy ioss in the jump, .(d) t'he
relative loss, '(e) the efficiency of the jump, (f) the relr,tive height. of the i limp, (g) the
lengt.h of the jump, and (il) the type of jUlnp;
.
' .
lIi~5. In a !'ec~augular channel with b = 20 'ft, n '= 0.03, B,nd So = 0.04, the depth
?I un.lform flow IS 3 ft. A low dam which keeps 7 it of water immediately behiud it
IS built at thee downstream end of the chanuel.
If thtc backwater surface behind the
do.m is assumed horizontal, find the pussibility of developing a hydraulic jump in the
channel. ,If ~he jump will OCCllr, determine (,:) the discharge over the spillway of the
dam, (b) ,lie !ump heIght, (.G) the energy loss m the jlllnp, Cd) the efficiency of jump,
and (e) the dlstali~.e of the Jump from the dam.
16-6. Locate the hY,d.raulic jump in Example 15-1 if the ch9.nnel has a free-overfall
01itlet.50Q it do\vnstream from the Y~na contracta.
15-7.'\Vater flowing over a low spillway of bl'oael width passes on toa level concrete apron at 12 t below the spillw"y crest, The head over the SpillW!e}, crest is
8 ft, and the tailwn.ter depth is 10 ft. Determine the location of thc jump. Use'
Eq. 0'1--9) with C = 3.61 for computing the discharge n,nd Eq. (14-19) for computing
tbe depth of Ho\\' at the spillway toe with a velocity coeffici<mt of 0.9 (i.e., the actunJ
velocity is equal to 0.9V 1).
.
.
16-8. A wide rectar,gular channel with n = 0.025 is laid with a change in slope from
steep ."I, = 0.01 to llliid Se = 0.002. The depth oLuniform flow in the mild channel
is 5 ft. Determine the location of the hydraulic jump.
.
lli-9. SolVe) the precflrling problem if the steep slope is 0.03.
15-10. Determine the iniLial and seqilent depths oJ a'hydraulic jump in" horizontal
channel 30' ft wide lmd carrying 30'0 cfs. The loss of energy in the jump is 5 ft .
. (HIN'!': The solntion requires. a trial-and-er'ror proeedlJl'p. slwh as.: First e...'5Ullle F 1 ;
~hen comput.e y';Yl by Eq, (3-21) altd Y, by Eq. (3-2c1); finally c!leck for Flo A gl'aph-'
leal solutIOn using specifia-ellergi ancl specific-fo;,ce curves is also suitable.)
. 15-11. The canal fall (Fig. 15-24) ear!'ying 300 cis is Ie recklllgulal' channel having a
WIdth of 30' ft., slope oj' 1 on 5, all';. sufficient length. It. is designed to dissipate a head
Er.ergy line
--::;;:------:-------------r
'" ,
777m7J7.7777~.
.
5'
1~
1
I
I
!
RAPIDLY VARIED
l;
. .
l'
-----~
~oss of 5 ~t by hydraulic jump. Determine the position of the hydraulic jump assuromg negligible channel friction and ignoring the effect of cha-nnei sLope on the jump.
(HINT: Use the result obtained ill the preceding problem.)
433
16-12. A canal fall 40 ft long and c3:i'rying 240 cis is designed to dissipate a head loss
of 4 ft .. The channei of the fall has a.slope ofl on 5 and diverges fr.om a width of 8 ft
at the upstream end to 24 ft at th~ downstream end. Determine the loeation of the
hydraulic jump, assuming a frictionless channel and ignoring the slope effect on the
jump. (HIN'l': Divide the channel into, say, four reaches, n.nd COlll'pute El and E, at
each section fol' f'>E = 4 ft.)
.
15-13. Water flowing under a sluicc ga.te dischal'ges into a rectangular plain stilling
basin h.oNing the same wideh ~3 the gat.a. After t.he contra~tion of the jet the flow
has an average velocity of 8.0 ips and !1. depth of 6 ft. Determine (a) the s~quellt tailwater d'lpth, (b) the length of the b~"sin required to confine the jump, (e) the effectiveness of .the basin to dissipate energy (i.e., efficiency.of th'e jump), and (d) the type of
iump to be expectetl.
15-14. Design the stilling basin for EX'ample 15-2, using
jump control by (n)
sharp-crested weir, (b) broa.d-crested weir, and (e) abrupt l'ise.
15-15. Design the st.illing basin for Exu.mple 15-2, using a jump control by abrupt
drop.
15-16. Proportion an SAF stilling basin for Example 15-2.
15-17. Construot the jump rating curve at the 'toe of the spillwo.y designed in
Example 14-1.
.
. Hi-i8. Discuss the development of an energy-dissipatiDn structure at the toe of the
IOpillway under consider'ltwn in the preceding problem, providecl tile tll.illl'ater rating
curve can be r~presented' by (al Q = 40y', (b) Q = 100y', (c) Q = 2y', Dad (d)
Q = 9tJOy"',
15-19.. Design the energy-dissipation structure, if needed, for the different tailwater
conditions givcn in the preceding problem.
434
15-26. In a flow 23'08 fps in velocity and 0.70 ft. deep, an oblique hydraulic jump is
produc.ed by [l. deflecting vertical wall making a 15 angle with the direct.ion of the
flow. Detet'mine (u) tha wave angle, (b) the sequent depth, and (0) the energy loss .
REFERENCES
2.
3.
4.
5.
13.
7.
8.
9.
10.
11.
12.
13.
II
1
I
:r
'I
I
I
!I
435
-I,
-\
-j
',',,'
436
v,
;f
\ .
I
I
437
sua
r
\
438
II
'I
,.j
CHAPTER
16
>1
16-1, Nature of the Flow. The presence of curveI'; or bends in aligi1Hlent is unavoidable in the design of open channels. Difficulties in design
often ariee because of the complexit,y of the flow around a curved path.
The streamlines of the flow are not only curvilinear but also interwoven,
resulting in spiral cmrellts and cross waves. Furthermore, the cent::-ifugal force acting on the flow around a bend produces, a unique feature
known as su.perelevation, that is,a rise ill the water surface at the outer
bank with an accompanying lowering at the inner bank. Also, the
velocity distribution in the channel sec~ions in the bend is very irregular
and the coefficients a and fJ are usuu.lly far greater than unity.
In ch!1nnels of nonlinear alignment" flows behave differently according
to the state .of the flow. Generally spe:l.king, subcritical flow shows
smooth water surface and slight superelevation, ",haren,s supercriticr.l fio,w
exhibits characteristic cross-wave-disturbance patterns on the surface
and thus exaggerates the superelevation.
In the study of subcritica1 flow, the spiral currents are of primary inter- .
est. These currents are essentiaUy a friction phenomenon; hence their
analysis requires use of the Reynolds number as a pars,meter.
,
In the study of sl1percriticD.i flow, the formation of cross waves is of
major COMel'll. These waves represent the gravity effect of the free
surface upon the flow; hence their analysis willi'ely on the.use of Fronde
number as a basic parameter.
16-2. Spiral Flow, Spiral flo'w l'efersto movement of water particles
along a helical path in the general direction of the flow. Thus, in addit.ion to the major velocity component normal to the channel cross section,
. there are transverBe velocity components on the cross section. These
trarisverse components will create so-called secondary flow in the plane of
the croas section.
Spiral fiow in curved channels was first observed by Thomson [IJ in
1876. Since then many studies have been .made [2-7J. It is believed
that. this phenomenon is due mainly to (1) friction on the channel walls,
which causes higher filarnental velocities near the center of 'the channel
th.3.!l near the walls; (2) centrifugal force, which deflects the particles of
water from straight-line motion; and (3) a vertical velocity distribution
439
- \
(;
~. \
\
~
440
which
in the 'appl'or,ch chan~lel and thus initiates spiral motion in
the flow, The centrifugal force is responsible also for tl~e superelevl1tiOl1 in the flow surface,
.
It is generally knowr, that, looking downstream, a chM1Uel curve to
the right causes a counterclockwise spiral, \Yherei),~.a cht:l,llnel curve to
the left causes. a clockwise spir~1L. If the curve is followed by a long
tnngent, the spiral flow developed 1\1 the curve will
for somE: distance dowllstream.
The actual pa.ttern of a spiral flow is cornplicatedand three-dimensionaL In order to record th:o actual flow pattel'll, it is desirable to use a
specially designed instrument that can measure dil'cctlv the velocity
:lUch [is the pit~t.
used
components in different coordinate
by Shulery [2].
Spiral flow exists in straight channels (Art. 2-4) .3 :welt IlS in curved
channels. In a curved channel, ho\vever, the spi!'f:.l flow lnduced by the
centrifugal force is "Vary pronounced L1.nd irregulaF
the bend. The
co::nplicate~ pattern of flow is caused by thB interference of the spiral flow
ongmated m the strR.ightapproach channel with that generated in the
curve. The strongest lateral currents usually o.ppC\UI' close to the outer
1'[,\11 f1.t the mid-section of the cm"'le, where they h,w'3 n. general direction
toward the outside of the curve. Their direction rend position cha.nge
gradually through the second half of the curve until the exi.t is reached,
.
where they approach the inside wall with upward inclination.
In ordpl' to delilleate the magnitude and effect of the spilil\l flow in different curves under v(1.rying conditions of flow,Shukry [2] hn~ used a term
known as the slre:nglh
a sp'iral flow. This term is defined as th.~ pe\'~ent!l.ge ratio of the mean kinetic energy of theJateral motion to the total
ldn8tic onergy of flow at a given cross section. It is to be noted that, the
kinetic energy of flow depends Oll the sqtmre of the velocity. Re.fel'rin o
to the channel cross section represented
the xy plane in Fig. 2-6, th~
strength'of the spiral flow at this section is
(H;-I)
]~!l.eh ~ub~
j.
!
I .
441
442
443
'"
O.5~
Vi
.~ 0'10
c: 0.30
'"
.3 0.1 0
I---+,,-t~
0.2.0 \--_..L__"--_
\---''---'~-,--.!f----r-..,..--''-r--1
toO
2.
:3 4
5
6
7
B 0.50
ReyllOlds' number R (multipliec by 10')
.
200
2.50
,!
-"
., 050 r---1-''''''-1'''''"-,-'''-- ~-~-r--'->
':;;
'iii
1.50 .
0,40
0,30 !-I-I--~""
~ 0.20 f---+",..,f--
J.
15
a----'-J...----I
(b) PalOmal.' R; !=1.00 Clnd j~:0.50
L-~,__~~~b~,_~.__~o~~J'__~
023456780
4
Reynclds' number R (mult:plied by 10 )
0,25
0.50
0.75
Rotio of ongle of curval"r.
1.00
1.20
'\
1,10
1---
1.00
..:! 0.90
.,
!
1
!,
~l~"~
I,
I
'1/
l-J
Case :0
II
water slll'f3.ce l'is~s above the critical-depth line, a hydraulic jump will be
~~~.
g 0.80
~
~~
0.70
'E 0.60
to
~;:;
a
0.50
0,40
0.30
0,20
0,10
3
4
5
6
7
8 0.50
ReynOlds' numbe, R (multiplied by 104 )
2.
3.00
it
0.70
0.90
110
Depth-brealh rolio
1.30
of curve resist.a.nce.
445
444
V2
(16-2)
fc 2g
where l' is the mean velocity in the section and Ie is the coeffiCient of
Clll've resistance. The coefficient fe varies cOl1sider(tbly with each of the
parameters R of the approach flow, ?'e/b, y/b, and 8/180. Fumilies of
curves for these parameters based all the experimellt,~ made by Shuki'Y
[21 (see [9] (tnd [lOJ for other studies) are shown in Fig. 16-2. The curves
CM be used to determine approxim[,tely the value of to in sm.ooth curved
channels. For allY given c(tse, fe is obtnined by first fixing its v(tlue with
respect to two variables and then adju~tillg with respect to the third and'
fourth vmj(tbles.
:)
:I
:.'!
Example 16-1. Given R = 55,500, "c/b = 1.30, y/b = 0.8, and 0/180 = 0.556.
Determine j"
Sohltion. First, keeping y/b = 1.00 and 0/180 = 0.50, it is found that, for
R = 55,500 and ",/b = 1.30,f,'~0.200 (Fig, 16-2c). Then, keeping l'clb = 1.00 and
8/180 0 = 0.50, it is found that, for R = 55,500 and y/b "" 1.00, j, = 0.230 and ~hnt,
for R = 55,500 and y/b
0.80, j, = 0,275 (Fig. 16, 2b), Adjusting the condit,ion of
y/b = l.00 to that of u/b = Q,80, the conected coefficient = 0.200 X 0,275/0.230 =
0.239.' Now, keeping y/b
LOO and relb = 1.00, it is found that, for R = 55,500 and
8/180 = 0.556, j, = 0.245 (Fig, 16-2a), Similariy, adjus~ing t,lle condition of
8/180
0.50 to that of 8/180 = 0.556, the finally' corrected coefficient = 0.239 X
0.2'15/0.230
0.::\55, say, 0,26,
The above procedure of interpolation is shown in Table 16-1.
TABLE
16-1.
Step
y/b
II/180
r,/b
(1)
(2)
(3)
(4)
1. 00
0.50
. 0.50
0.50
0.50
0.556
0,556
55,500
55,500
55,500
55,500
1.30
1.00
1.00'
1,20
1,00
1. 30
(5)
1.00
0.80
0.80
1.00
(6)
0.80
55,500:
55,500
i
i
j,
~n
0/
~~-1Z----z=::J=
if.!I R~o,,~,;
I
Ii ("'s$'O{) \ I
1
:~
I\
,/~/
~\'\,
~
9 10
'
60--1-
,1
_~ _ ':I{\j-',!<:!
a 7~
(b)
FORWARD VELOCITIES
_~_Cer,terhr<eoffl\Jme
I
s
---PQ/:I':~~~~~~l~
,~
20 ,07'0==?O-!~--___t:\oO~
-----7'J~
.
-_ ______ :-__ _
,f-~:"L-1_- -~-~ep
1_ -==--_-~~
'"
"
701
ij<~/
.. ~~~-=-:-~-- ---=-~=:q. ~'iJO-f-...
L ~--
~I
Remarks
at,
0".: , --;;::'::::"_-:'::-,_
..... .
.
I' J
d forward velo0ities in flow around
Contour lin~~ ~~, e'lual [~Y1aC;tU~:~: ~~els measured in cm (= 0,3937 ill.)
a 180 be:l~. . (Ajt6'l' .... ::1L " y ~',
,: /b = 1.00; YA/b =; }.OO; and, with TT A =
and velOCItIes III em/sec (-R 0.~3;8315~S\"I'e~.e subscript ..1 refers to conditions at the
77.8 em/sec (= 3,55 fps), A - I ,
",
inlet-approach section of the bend,
FIG. 16-3.
0
throngh the separation poiut 3, For different bends ,under vm'~~us C~ll"
d't" .. of flow the- point of maximum surface depresslOn cl and th~ pcmt
I lOn::;
.'
1 t d (s given in To,ble 16-2, The positlOllS of
of separatlOll s ,v ere oc(t e , \ .
.
1 9
1 IJ
the two points n,l'C given in tenils of the ratios of theIr ang es , d (tl1C d"
. d f
th inlet of the bend to the total angle fJ of the ben .
measure rom e
' 1 1 ff t 1 b a i n g the
These positiolls were found to be only slig It y a ec ec y v r.Y h . ,
t 'U/ b Hence fol' pmctieal pmposes, Table 16-2, ",hlO gIves
par~~e e l l ' f . /1. ~ 1 00 can also be applied to other values of y/b.
pOSItIOn va uea or y v - . ,
.
446
In using the table for interpolation, the procedure i.'> similar to that for
f. values (ExlUllple 16-1).
16-2. LOCATCONS OF PorN~'S OF MAXUlUiI<i SURl"ACE DEPRE8SCON
d IN FIG, i6-3) "-ND ,POINTS OF SEP .... RA1'ION (POIN'l' 8 IN
}'IG; 16-3) FOR VARYING PARAMETERS, EXCEPT THAT y/b = 1.00*
V.
TADLE
(POINT
Parameter
9d/9
,i~ =
lD,500
I
(J,/6
.i 8d/ O [
t~
= 31,500
I ed/B I 0,/8
e,/e
(A) 0/180
r./b
0.500
1.00
2.00
3.00
21,000
so/a !8,/e
ILOOO I
I
[I. i
I
18
/8
I 0;/&
0.111
0.111
0.1'18
0.051l
0,222
0,250
0.333
0.778
= 1.00,
cons~ant
0.333
D.IS\)
0.185
0.111
I
Ti
--1
T,
~---ln-
To(
ri
~o -
(16-5)
. {" V dT
" (E -
CZ/2g1 2 ) dl'
=~-~- = :C.'.'~__~~_ _ __
'l'o -
ri
T-:;J -
-)
1';.
C2
(16-6)
In the above equations 7'. and J'i are, ret;pectively, the outer and inner radii
of the curve. NO'N, the discharge i5
-I
(16-7)
If Q, 7'." !'" and E are given, the r::onstant C can be determined from Eq.
(16-7). The velocity pond depth at any radius l' are then obtained by
Eqs. (16-3) and (16-4). Thus, the superelevation t.h of the water surface
can be shown to be
-1'
I'
(16-8)
where CIA is the energy coefnciept and II A. is the mean forward velocity.
The specific energy at, the radial section passing through point d is
(16-3)
l'
Ym
2(JT ,1';
.. ..
(B) I'c/b
~~---
_1~o -
= E - ---
f' (C/r) dT =
= 73,500
6/180
0.50
0.75
1.00
IR
0.50, a constant
_____
1 _ _ ....
0.Z5
I~ = ~2,500
447
!J 2
_..2-
2g
(16-4)
For an elaborate ma.thematica.l ,analysis by the l.aW' of free-vortex motion see 1121.
where L is the channel length between sectiCYl1 A and the section COll~aining point. d; Sf is the friction slope, which may be either determined
experimentally or computed by the M~nni:ng formula; and h f is the. energy
loss due to curve resistance, which may be determined a.'> showni in Example .16-1. The coefficient 0.4 was found to be practically constant
for any curve.
.11le above m~thod was found to be. reasonably accurate as long as
the angle of the c1.irve was greater than 90 .. For smaller angles, Sl~ukry
assumed that C varies lineatly 'with e from rV.. at e = 0 to its full
v::dl1e at. 0 = 90. Therefore, for an.y angle e less than 90, the Clrcu-
-\
,I
I
448
(
o+
(1
(16-11 )
TT 2
';'
1';
(16 .. 12)
Woodward )4,151 l.l.ssumed th,tt the velocity is zero at the banks and
has a 11l[Lximmu value
at the center, varying in between according to
. a parabolic curve .. Using Newton's seconcpaw, he obtained the fol1o\l'.ing formula for sllperelevation:
I
I
I Cross waves ill supercritical flow iJJ.wa~er are' aualogolls to the shock waves
In supersonic t!01V of gase.~. This analogy was first noticed by Prandtl [16J, Hiab~ucbinsky [17], and von Karman 1181 and then investigated experimeatally by
PreisIVerlc [19J and others. These scientists, however, were iIiterestecl ill'imrtrHy in
applications of this ~tinciple to sttpersonic flow of gases. La,ter, comprehensive
$tudies of .the subject! in hydraulic applications ware made by Knapp, Ippen, ilnd
othel's [20-26). To facilitate the analysis of shocld wave in gases, B\lSemann [27J
developed !l. gra.phical method known !\s the melhod of charactedstics, which was lall"r
applied to hydraulic ;problems by Preiswerk, Ippen, and Knapp. For detfliled
de$criptioll5 of this method, the reader should refer to [23J, [211, and [27].
449
I The disturbance line OJ' wave angle created by an oblique bYi<iralJlic itlrnp is considered positive in order to distingilish it from the negative distlirbance line c>f W!lvefront d\IC to nn oblique expansion ,i-ave. The negative disturbJl,nce Iille Ilctjlally only
marks the beginning of a. distuxbance region; it is not a. distinctly, defined Iiuc like the
positive one.
i
. -- 1
,
45,0
451
. ary line AB AI, the< flow is unaffected by.the curve and thus continues to
move in its original direction; that is, parallel to the tipstre!un tangent.
Beyond point B, the two wA.vefronts AB and AlB affect each othel' and
are 110
propagated in straight lines but in the curved paths ED
and BC,
TheOllt,cr COll.cave wall A C tends to deflect
the flow, .yhich would otherwise follow the tangential direction. Consequently, the water surface is raised higher and higher around the outer
wall up to a maximum at C. After G the effect of the inner wall, which is
to lower the water surface, begins to operate., Thus, the water surface
along the outer wall starts to drop. On Uie imler convex side, the flow
tends t~ depart from the wall, and the water surfa.ce is depressed lower
rmd lower around AID until the point D is reached. After D the effect
of the outer wall comes into pl!),y and the surface' begins to rise again.
The reflection oC disturbance waves from the outer B.nd inner walls will
not come j,o equilibrium 01' .stop when they meet near the center of the
channel. They will continlle to be reflected back and forth' across' the
Chal111el, crmsing lhe surface profiles
the walls t,o h1 ve a Beries of
maxima, and minima of surface elevation, approximately at f~ngles 0,
30, 58, . , . , from the beginning of the curve. The
e marks
half the wavelength of the disturbance pattern. For practical purposes,
. it may be assumed that the points of maximum and minimum for each
(J occur on the same radial line; such as DC for the first maxiphase
~um 'and minim um. 1 By geometry (Fig. 16-5), the cen tnd angle to the
first maximum can be shown to be
8
..
tal1-' (21",
2b
+ 51 tan f3
-1
I
I
(16-14)
curvature of the outer wa.ll starts at the beginning point. A and is propagated along the line AB, which
an initial wave angle /3 with the
tangent extended beyond point A; In the meantime, the initial distu~bance produced by the i~'ll1er wall is propagated along the line A'B.
The
. two p'ropagation fronts meet at point B. . Upstream
. from the bOWld-
1 Actuo.lly th~ locations of the first maximum and minimljm do not oceUl' exactly
on the ra,diaf line OC but slightly on the left side nnd right side of DC, respectively
(Fig, 16-5). 'However, the error involved in the l15surnption is negligible. This
assumption applies also to the locations of the subsequent maxima. lLnd minimn..
~ A det&iled study of the complete surface contours is possible by means of the
method of ch!l.t!l,cteristios (23].
.
.- ~I
-\
.. 1
.\
.~
. 452
r'
(.
~.
[
I
i
1
V
Fl = F = -.-- = - . sm ,6 . vfiiY
Substituting y for YI Rnd y
I
ay
(16-1.5)
. Ii sec~ {f tan e
=
. tan f3 .:- tan e
~--~----
For a small angle, tr,ri e may be replaeed by de, and ,tilll (; in the denominator is negligible compar'ed wit.h tan (f. Thus, Eq. (l6-16) becomes
11
(16-17)
V~
---
tan (3 de
(16-18)
de =
'2(E - y)
_
,/<.,'
-I
8 -.
-v u tan
Sillce 2E'
'
.,.
':<
..
. (16-19)
.,
vY
v2E - 3y
y(2
8 = '\/
;,
Q
(16-20)*
(16-21)
cons
tan- I
',/3
--==
.y F~ - 1
453
y= -
,6
8'
+-)
2
(16-22)
(16-16)
ely = -.-~,-- de
.
8m f3 eos fj
!r
I
t
1 On the basis of !l.Ssumption 4 and, hence, of the conservation o{ energy, the velocity
in any stl'eamline must chani!;e as the depth changes, since E='y + V'/2g must be
con,~tant.
The flow E1,l'Ound the outer wall, boing the deepest, should be the slowest.
Actual measurements, however, indicate on the contrary tl\r.t the velocity around,the
autc,. w(Lllli~lllains constant or even increases slightly, whereas f\long the inner W:L1l
the velocitj"'decreases. It is believed th:.t this dispD.rity is due to varyhig effects of
the channel irictian, which actua.lly render ass~ll1ption 4 invalid. Consequently, it
is rell.soned that a constant velocity may be assumed. Equ:.tion (16-22) was, therefore, derived from the assumption of constant velocity, which rep!E1,ces assumption 4.
Since the author cannot discern here the rigor of the origirull mr.thematical derivatioll,
he would rather consider Eq. (16-22) to be empirical.
.
454
455
or b/tan f3 by
that the angle AG' A' = (3. Thus, the wavelength is 2b/tull fl.
Frain the above discussion, it may be concluded that the disturbance
wave p.l1ttel'll, which osoillntes about the plane represented by
has a
wl:welength of2b/tl1n.6 and an amplitude of 112b/2TcU. This
can
be used to cstimn.te roughly the s~lpercritica[-flow profile ill simple curved
where FI a.nd F. are, respectively, the initi",ll1.nd fina.l Froude numbers of the flow.
The rehtion among F t , F,; and th.e tot!1-l deflectioll angle ~, can be represented by Eq.
(16-21), In that equation, e '" 0 when F = Fl and 0 = 8, when F.= F.. Thus, if F,
aud 8, are give!l, F, can be determined. The angular mnge in which the depth change
takes place may be determined by
(11)-23). Since the depth cha,nge does not
involVe 'appreciable energy
depths before and after the change ca.n be
rela.ted by 'assuming a constant
er,ergy, 'Thus, the mtio of the final to ,the
tnitia.! u(~pth can be shown to be
chn.nlle[s.
For disturbance continuing into thedowl1stream
channel, the
primal':\, wavelength is sti!! 2b/tan p. AI>, the curvature suddenly changes
to straight alignment, a r,ew disturbl1l1ce pattern, starting with a m<J.ximum height on the onter wall at the point of tangency, has a wo,velength
and magnitude equal to that of the original disturbance pa.ttern developed
in the curve. The l:esul'.;ing disturbance pattern in the tangent is the
sum of the original and new
Th~ new disturbanGe pattern
may, therefore, be elimilmtecl by adopting curve lellgth~ of 28, 40, . . . ,
which will have a minimum wava height at the outer wall at the efld of the
curve, just
to cllncel out the newly created maximum wave heif!;ht.
Example 18-2., Des()tibe ~he ehal'l1.ctBl'istics of all oblique expansion
channels of nonlinear A.llgnme::Jt.
W2N6
,1
(
in
In general, curves
Mm?q#:;:J;;77/7:/7//?//t::::Lm7///.7//~7/A; ,
Seclloo A'-A
expa.n~ion
wave,
i!
Sol-uliQn. The oblique expa.nsion wave occurs when the channel wllll is turned oute
ward from the flow at all o.ngle {Jc (Fig. 166). The depth of flow decreases in the fanshaped region of the wave disturbo.nce delineated by wave angles (1, and {:J" measured
with'l'eference to tbe initial and final flow dirac.tians, respectively. By Eq. (16-15),
and
(15-23)
I
t
I.,
loss and letta to dllnger of serious locn.l erosion due to spiml flow. Reduction of spira; flow is the mlljor conce.rn in the design of curved ch:J.n~lels for
subcritical flow, For a proper proportioning of the curve, a ratio of
To/V = 3 is recolnmended, because it will give the least, radius at whichthe .
effect clUB to spired flow is minimi.zed (Art. 16-2).
In erodible ch!l.l1!lsls, the action of spiral flow will develop a
tion in the bed. The size of the cOllfiguration is minimal if 'l'r/b is 3.0 OF
larger. When bl111k protection is necess::I.ry, it is needed most on the
outer bank at the do\vnstream end of t.he curve and to a lesser degree
on the inner bank at the beginning of the CUl've. This is suggested in
accordance with the behavior of spiral flow.
In an alluvial bend it seems that the configuration of the channel cross
sectioll is defined more or less in accordance with certain l1a,turallaws.
This subject has been
by many scientists and engiheers of
river hydraulics since
[29]. According to the studies by
Ripley [30], the configuration at an. alluvial river bend (Fig. 16-7) may
be represented
the following empirical equation:
I
x~ - 0.433) ( 1
Y = 6.35D ( ,\/0.437
- T2
Xl()
+ r;
(16-25)
where y is the ordiril1teot depth in ft, :1: is the abscissa in ft, Dis the
hydraulic dept,h in ft, T is the top width in ft, r" is the mdius of curvature
on the concave side of't.he channel in ft, and J( is a coefficient equal to
456
I
I
.17.52. The crigin of the cool'dillates for this equation is on the water
l>'Urface' at n pOlut equidistant. from the banks~ ~l1using :this equation
the foil.oviring I'emarks shoulcl' be noted:
.
1. For],. less than 40 tiInes the square root of t,he water area, no further
deepening of the channel seems ttl result from the increased curvature;
hence, in such cases, the ivalue of riJ used in t.he equation should be
40 VA. Consequently, 'bends are const,!'llctive and stahle when r. is
. greater them 40
wher~!l.S sharper bends are destructive, te,nding to
shift the channel.
2. For r. gren.ter thn.ll about 110 VA, the equation bacome.s il~ valid.
457
. VZ
gr
(16-26)
I
.r-
cro~s
o
FIG. 16-8. Ph\ll of sill insta:llativn in a curved ch3.nnel.
,vill
i
I.
'>
458
O!
is 30.
,I
(16-27)
I
I
0.30L'
!
i
,,I
(16-28)
where L' = b/tan p, that "is the half wftvelength of a channel disturb:?nce.
This method can be used as a remedial measure in
channels
which have been designed as simple circular curves or other un:=mtisfactory fOl'ms. The major disadvantages of this method are
mail1tenal1.;e cost, pronounced distUl'bitnCe at low flows, and possible
cavitation at extreme high-velocity flows. The pronounced disturbll.nce
l1t low flows may cause uneasiness and distrust in an unexperienced
observer liS he may not be ea.'5ily convinced that the distul'bance will be
greatly reduced at high d~sigl1 flows. .
PROBLEMS
1~-1. A 120 bend with Ii
10 In. and 1', = 15 in. is designed to c.onnect two
etra.lght channele of the same width, ca.rrying 1.3 cis at a normal depth of i in. Determine the bend loss.
16-2. Det~rm~ne the specific energy at the section containing, maximum surrace depreSSIOn In the bend given in the preceding problem. Assume 0: = 1.0 alld
n
0.01.
.
, 16-S. Compute the WH.IA'.['-sr:r
profile in the cross section containing maximum
surface depression, as
in the
problem, using the law of free YortE'x.'
What is the snpel'elevation?
16-4. Det.ermine the superelevation in the preceding problem by (a) Eq. (16-11},
(b) ;Eq. (1.5-12), and (e) Eg. (16-13).
.
16-5: Verify Eq .. (16-14).
16~6. DeLermine the approximate flow profile ill the curved channel given in Prob.
16-1 If the approach channel carries a supercriticl11 fio\I' at a depth of 0,5 in.
.16-7. Design the curvad channel for. a rectangular flume to turn,lj.n angle of 50
wIth t1 radms of 250.it.. The flume is'12 ft wide and built of smooth concrete. The
de.sign discharge is 350 cfs at a slope of 1 %.
'
. 1~-8 .. Using Eq. (16-25), compute the channel cross 8ec~ion at a bend in the MissisSIPPI Rl;erj where A. = 148,010 ft 2, T '"' 2,340 it, and 1.
18/300 ft.
result WIth the actuat cross section, which is as follows:
0
1,170
1,000
800
600
400
0.00
36.00
88.00'
108.00
111.00
ZOO
-200
-400
97.00
74,00
65.00
55.00
1I
y
'-660
-800
-1,000
-1,170
44.00
38.00
24.00
0.00
.{
!
\
l
459
REFERENCES
l Jallles Tholnson: Oil the origin a.nd winding of rivers in allu,rial I,laius, with
rema.rks on the flow around bends in pipes, Proceedings, Royal Society of L01!rf,on,
vol. 25, pp. 5-8, Mar, 4, 1876.
2. Ahmed Shukl'Y: Flow arour:d bends in an open flume, T1ansa.dionli, Ame1'ican
Society oj Civil Engineer.s, voL 115, pp. 751-779, 1950.
3. A. Hinderks:' NebenstrBmungen in gekrilmmten Kaniilen (Secondary flow in
Gun'eel can!l.ls), Zeitschdfl. des rre1~in.~ deldscher I ngenieuTe, Berlin, vol. 71, no. 5J,
pp. iTi9-1783, Dec. 17, 1927.
4. Bmllo Poggi: Col'l'enti veloci nei (lana-Ii in cmva.(8wift flow in curved channels),
L'Ell,ergia eleUrit;a, Mila/Ito, vol. 33, no. 5, pp. 465-480, May, !f)5G.
5. F. L. Blue, Jr., J. K. Herbert, a.nd R. I,. Lancefield: FibII' around a river bend
inveRtigated, Cl:"n En!j1:neenng, vol. 4, no. 5, pp. 258-260, May, 1934.
G. Herbert D. Vogd Ilnd Paul W. Thompson: Flow in ril'erbends, Civil Engineerinll,
vol. 3, 110. 5, pp. 2613-263, May, 1933. .
7. Herbert D. Vogel and Paul W, Thompson: Existence of helicoidal flow, Ci1iil
EnfJ1.needll(J, vol. 4, no. 7, pp. 370-371, July, 1\J34.
8. Robert Muller: Theoretische Grundl:J.gen del' Fluss- und Wildbachvcrbauungun
(Theoretical principles for mgulatioll of rivers and t{)rrent.s), Eixlgeniissi,che
Teclmi~che HClchsclntle; Z1ldch, jJ[itieii1,t~{le" de,. Ye.rsuchsanslalt fiLl' Was.,erbau
1.nd El'dbau, No.4, 1943.
9. SaHjiva. Puttu llaj\l.: Versuche i.iber den SCl'Omungswicierstll.nd gekrummter
offencl' Kaniile (St'.lcly on the flow resistance ill CllT\'cd open channels), lI{Uleibngen des Hyd1'f.mi::schlm [nstitu.ts del' Tec/',nischen H ochscknle Miinchin, no. 6, pp.
45-60, Munich, 1933. English tTll.nslaticn by Cla.rence E. Bal'cMey, R.",is~allcc to
fiow in curved open dmnnels,
A1I!,~,':c(m ,society of Civil Engineers,
vol. 63, pt. 2, p. 49 after p. 1834, November, 1937.
10. C. H. Yen and J. W. Howe: Efferti'; of channel shape on.losse5 in a canal bend,
Civ';: EnrJineering, vol. 12, no. 1/ pp. 2.8-20/
19':12,
11. H. Wittma,!l and P. B6ss: "Wasser und
in gckrihnmten
FIuss-strecken" (" Vi' a.ter and Eed-load Movement, in Curved River Reaches "),
Springer'''erlr.g, Bfll'lin, 1938.
12. C. E. Mockmol'e: Flo\l' around hends ill stable channels, TnlllSact':oTl.!, Amel'ican
Society of Civil Engineers, vol. 109, Tip. 513-618, 1944.
13. Armin' Schoklitsch: ".:Hydraulic Stmctul'es," trallslated from the German by
Samuel Shntitz, American SGciety of M'lchftnical Engine~rs, NEllV' York, 1937,
.
vol. I, p. 151.
. 14; Shenna.n ))..1. Woodward: Hydraulics of the Mi!l.rni Rood control project, ilIia.1ll,i
Cons6"unncy Distr':ct, Tel'hnica.l Repi,,t, Pt. V Il, Dayton, Ohio, 1920.
15. Sherman M. Woodwa.rd and Charles J. PORCY, "Hydraulics of Steady Flow in
Open Channels," Jobn Wiley & Sons, Inc., Nt'w York, 1941, p. H2.
16. L. Pral!::hl: Abriss de,. StrQmungslehre ("Outline of the Theory of Flow"),
VieIVeg-Verlag, Brunswkk, Germany, 1031.
17. D. Riabouchinsky: Sur l'analogie hydraulique des mouvements d'un ffuidp. compressible (On hydralllic anl.Llogy of flaws of' a compressible Jiuid), Com.p/.es renaw,
de I' Acad.!?nie des Scie1!ce~, vol. 195, pp. 998-999, 1932; vol. H)9, pp. 632-634, 1934.
18. Theodor yon Karman: Eill;; i)rai<tische Anwendung del' Ann.logie zwischen tJbernaILlSt:l'o:mumll.: in Gasen und ilberkritischer Striimuug in offenen Gerinl1en (A
ap;pUI)MIOn of nna.logy between supersonic flow in ga,seil and supercritical
flow in open channels), Zei/,1lchlift flZ' (l'lI.gelUandle M(lth~matik un.d Mec/tnnik,
Berlin, voL 18, pp. :1.9-56, February, 1938.
-1
460
ar.
0
20. Arthur T. Ippen and Robert T. Knapp; A study of high velocity flow in curved
channels of rectangular cross section, 1"ransaclions, American Geophl sical U '.
vol. 17, pp. 516-521, 1936.
Y
mon,
21. Robert T. Kna~~alld Arthur T. Ipj:>en: Ol:rvilineur fiow of liquids with free
sur~.\l.ces at veloCitIes abov: that of wave propagation, Proceedings of the 6th Interna/tonal (')on9 res s of Apphed Mecillznics, Cambridge, Mass., John Wile' & Sons
Inc., New York, pp. 531-536, H)38. .
Y
,
22. Arthur T. ~ppen: GM-wave Rnruogies in ope:l~challnel flow, Proceedings of the
2d Hydrm'/tcs
Conference, June 1-1, 1 19./,2
St(tie [/niU'Ts't'
of I~w'a J S'Lolluozes
,. '!"L
.
'"
-r:
~:y
E'ngtnee''',g, B1tl/etm 21, no, 400, pp. 248-265, 1943.
23. :~r~hur T. 1ppen: Mechanic.s of sllp~reritioal flow, 1st paper in High-Yelo~ity flow
HI open channels: A symposIUm, T"unsar;iions, A.1nerican Society of Civil E'lUI~'nee-rs
vol. 1] 6, pp. 268-295, 1951.
,
24. R~bert T .. KnapJ): I?esign of channel Cu\,ves for slipercril-iC!l.1 flow, 2d paper in
oS
....
.J
'-'
J:
sur
CHAPTER
17
PI - P 2
Pi - F/
Vsb,Ya.
Fol' convenience
1 Gradu'ally varied flow through nonprismatic channel sections has been covered in
Arts. 3-5, 3-8, and 11-,1 to 11-7.
.
The following discussion is based 011 a.tre!Ltmeut given by Jaeger [J-3].
461
J
FLOW THROUGH NONPRISMATIC CHANNEL SE<"'TIONS
462
Region 1.
Region 2.
su bcritical.
Region 3.
Region 4.
supercrit.icaL
(yJ/Vl)[ (VJ/Yl)2 - 1J
2[(Ya/Yl) - 1/(b a/b 1
2 _
)J
(17 -2)
. where F 12
V 12 /gYl' This equation call be plotted as. shown in Fig.
17-3, using bl/b l as a parameter. The family of CUl'yes thus plotted are
Horizontal
contrQction
463
Horizontal
expor.sion
Vertical
expansion
I
\
0.5 (~)
i(Ya) + 111.1
Yl L Yt
(Why?).
.
4. The upstream flow is supercdtical ill the regionaboye the horizontal
line F 12 = 1 or F I = 1 and sub critical below t.his line... The downstream
2
flow is subcritical in the region below the curve Fl~ = (!Ja!Yl)3(ba/ b1) and
supercritical above it. (Why?) Consequently, four regions in the
figure, representing four regimes, are created:
FIG. 17-3. Plot of Fli aga.inst 1)./1/1, using D.lol as a parameter fQ; the ana.!ysi;; of
sudden horizontal contractions (shown by the full curves) and expa.nsions (shown by
the dashed curves).
IlE
ar
IlE
Yl
+ -V2(1 - y. -V3
2g
1 + F12 - r~ +
2
1..l!1
2(Y3hh)2(b /b
Yl
-_.!L_-J
a
1)2
(17-3)
(17-4)
464
By applying this eq uatio!l to a certain part of the curves, the flow can
be found to be impossible if the computed value of /::,.E is negative.. It
should be noted, however, that this discussion is intended to present a
theoretical t'malysisof the phen6menon and to develop a classification of
the flow through sudden transitions. In real problems, the theoretically
impossible flow may become actdally possible becanse tbe iJ.ssumptions
made in the ab~ve derivation may not be true under actual circumstances.!
,
I.
I
405
. by Formica
' . [4],
. !LS shOwn in Fig, 17-4. .Tl1e t YdP11cn(
. I measured'
were made
_ '._ )
I (col1~rn.ctlon)
expansIOn
flow profiles and energy 1'Illes f
or
. hall
'fi encrn-y
1"'(-5. "
Th" enel'''Y
speCI ill
c e the
b
.
h
are sown
m
'" lines represent
. . t e11k
~\'hel'e the tl'(1nsltlOll t, 'es pc,
Y TI CI Vtj'/.
""V' Near the section
.
35
30
f-'-~
,
25
20
em
15
+._._---~t '
'I...-., ~-. .
10
5
-0
E. ~=-4
35
~
4
30
11
25
em
hori-
(17-5)
-1
o
~udderl tr!)nsition~.
1 Is-1
\7-6
co~B?-cient
lng a ==
shows the computed values of the energy
different section."3 of the ~hannel for various designs of transItIOns.
at
The
0:
rI
I
467
vhlues are apparently very close to unity immediately after the sudden
contractions hut are generally higher than unity after the .sudden expansions. This indicates that the flow in a sudden expansion is irregularly
diffused.
466
!
1.1 0
r---r--r----'-~
-+-~
1.08
1.l0
.-
__
.,..r--
!.08
--o--~
1.06
~~
'1.04
l'>.'
1.00
!.to
--
1.08
1.04
~,
1.00
1.10
,,
1.06
1.04
.......... ---~
1.08
Il
C'" 41
1.04
1.02 ~
1.00
1.10
....... --:---=-
F
t--
I.oe
!.os
n,
.0'" 49
1.04
1.02
1.00 1
~,~
""'-----
0, A
91- i---,
...
/'
E
-~
~r ,
6'
'"
0-25.2
\\
f"
~ '-~,
.--
I.
;-
taf~, ~
;':1~"
,-.=,~
---<
I ,,/ {g;/
<
-~
f--
'--
---r
:-..
tOO
,'I
-".
"'}-- - --..:....- V
.,..-:"'
.1.02
....
"lV'
o:~jt1irttJt~1
'::l~*lll-ttB
B 1.003
Sec1iqn
>-~-
<J 5
1.08
0" 25
(06
1.02
'y-
'"
.'
I-Q.J--'~
ts- I
Q = 15.6
---iJ---
"1',
,J:''';F--
1.02
LtO
--
.:j
5 0
bed
'~~-~
~
71c
---
+~ ....c::.
(After G.
10
-',
I~
L----
,......
20
~-
2S
....
r r"
W-~ ~d
~: ~~ V-w-
30 35
q [1s"1
f,t
/vp,
.IJf
[-
J--
.~:======
FIG. 17-6. Variation of the energy coefficient near sudden transitions.
Fo,mica [4J.) 113- 1 = 0.03532 cIs.
1:= ---"1-f-
Sf---
iI
40
45
50
55
113-;1
== 0,03532
I
I
.'\
-"l
f.
468
,",r'
expressed by
.
V3~'
E=K-'
29
and in a sudden expansion by
(17-G)
\
t
4G9
(17-7)
E=
wherel( and are coefficients, and VI and V3 are the upstream and downstren.m velocities, respectively. By applying these equations to open
'VI
c'
AJ
tOl
it
curved tranaitil;m.
(Cou?'iesy of A. T. Jppen.)
for sudden
~~~~Of,~e~igrL':"", .10,821 :,871 ~.681 :',~I :'27l~.291 :A51 :.4~Accorcling to the experimental data obtained by Formica, the' values of
K for sudden contractions seem to vary in a wide range, generally increasing with the discharge. The approximate median value of J( for design I
is 0,10 and, for designs II to IV, 0.06.
17 -3. Contractions in Supercritical Flow. When supercritical flow is
introduced thro\lgh a contraction with symmetrical converging wa.lls,
cross waves similar to those developed in a curved channel (Art. 16-5) will
appear. The cross waves in a contraction, however, are symmetrical
with respect to the centerline of the' channel (Fig. 17-8). Theanalytical
study of the problem can be performed by the principles of the mechanics
of supercritical flow described in the preceding chapter.
From an experimental and analytical investigation, Ippeu and Dawson
Along wall
AI on 9 ce n Ie,.'.;,;Ii;,;."e=--_r-_ _ __
h
hz
r-~"----l
F------------,;
I"
tel
FlO. 17-9. Designs of skajght-waU contractions. (After Ip~cn and Da,"'son ..[6].) . ~~)
Generai dis~ul'bance patterns; (b) minimum downstream distUl'bance, (e) .,chema Ie
profile.
experi'.
.
of curved contractions, the cross-wave patt ern ID!'.}'b'.e determil-ed
mentally by model test or !'.na.1yt.ically by the method of charactenstlcs
(see first footnote to Art. 1 6 - 5 ) . '
.'
In supel'critical flow through a straight contraction (Flg, 17-ga) ,symmetrical shock wl1ves are developed at points A and At at .the entrance.
These waves extend across the channel at. wave angles {h, mter~ect ~t B
on the center li~le of t.he channel, and, finally, af~er some modlfic~tlO~,
reach the opposite walls at C and Ci. In the regIOns A!'W and A BG ,
t,he flow proceeds through a new field characterized by the Froude number
F~. Note that, ot the end of the contraction, negative disturbances are
";
'1
\
470
created at points D and D" These disturbances will result in more complicated disturbances dOIVnstream. In a good contraction design, it is
possible to minimize the downstream distu:rballces. This can be done by
directing the shock waves to the opposite walls at D and D', thus car.celhig off theoretically the newly created negative disturbances. As 11
result, the flow will be calm in the downstream channeL This is the situation shown in Fig. 17-9b. From the geomeky of this situation, the letlgth
of the contraction can be shown to be
L
471
1
\
I
I
,,,,ill
(17-8)
(~)%(Fa)
Yl,
FI ,
Q,or
(17-9)
The above equations and either Fig. 15-23 or Eq. (15-24) can be use.d to
design a strn.ight contraction with minimum downstream disturbances.
Generally speaking, high values of FI and low values of Ya!Yl will give a
long contra.ction. To reduce the length of the contraction, the advisable
va.lue of Y3/YL seems to be bGtween 2 and 3, provided that F J stays well
above the critical value.
.
Example 1'1-1. Design a straight contraction con!1ecting two recta.ngular channels
12 ft and 6 ft wide. The discharge through the contra.ction is 200 cI3. The depth of
the ll,pproach Bow is 0.70 ft.
.
Sa/uliano Since Q
200 crs and A = 12 X 0.70 - 8,4 fP, the velocity of the
approach flow jg VI
200/8.4 = Z3.8 ips. The Froude number Pi = 5.0l.
Assume y,/y,
2: and take b, = 12, b, = 6, and FI
5.01j Eq. (17-9)
F3 = 3.54. This vI111.1.e should not be less than 1.0 or even too close to 1.
Now, assume a volue of 9, say, 15', and take Fl = 5.01; the diagram in Fig. 15-23
gives y';V' = 2.60 and F.
2.S.
.
A second determination using the same e = 15 and repla.cing F, by Fi
2.8 pro
duces values of y,/y" ... 1.80 and F;
1.77 from Fig. 15-23 (corresponding to vliv,
and F" respectively, in the dia.gram). Howe.rer, these values do not neces3arily
represent the actual flow condition in the required design, since,the flow condition
downstream ffia,y be complica.ted by the negative disturbances origina.ting from the
points D and D'.
.
Multiply 11./1/1 by 1/./y%; the first trial va,lu(l of y';y, is equal to 2.50 X 1.80
4.68.
Since this does not a.gree with the Il.Ssumed value (that is, 2.00), the procedure should
be repea.ted with lJ, new value of " until agreement is reached between the assm;ned
y,/y, a.nd the value obtained by the trial.
. '
,
Alter several trials, the correct angle 9:is found to be 5. With 9 = 5 and'F,
5.01, the diagram in Fig. 15-23 gives II1./Yl= 1.50 and y,/y, = 1.35. The 'value
of 1/3/1/1 = 1.50 X L35 = 2.03, which is close to the assumed value,
By Eq. (17-8), the length of the contraction is found to be 34.3 ft.
17-4. Expansions in Supercritical Flow.. Channel expansion in supercritical flow occurs frequently at places where flow, emerges at high
Eddies and
circ.ulo1ion
C
--- --,
'
_ ....,) LSePOrOliOn
--.,
surfoce
---
,I
.J
FIG. 17-10. Separation of How in an expa.nsion of rapid, divergence.
I
1
J
I
h
472
value of b1iYl and for various valuesofF l l1re shown in Fig. 17-12. This
form of expansion boundary corresponds approximately to the shape of
the streamlines that" corifine about 90 % of the flow. N cite that the
begi;ming of this expansion is sufficiently gl'O.dual to reduce the effect of
nonhydl'Ostatic pressure distribution to a minirriun1, so that the factor
bdYI is no longer an essential vl1riri.ble. In thisiol'm, the general iliCrel1Se
. 1h
z)
(b1F.' b;,
(17-10)
1iIf.~0.90
00.801>0.40
"0.7000.30
eO.60 1iI0.20
"'0.50 ()O. 1 0
i'
I
473
----
1 "
-----
\1
\,
2.0
)
"
"~
r
I
")
I,
I:
VOiues of D,F,"
2.0
fOI"
. b;.
I( x
~ \b 1F 1
)~'
+ 72
(17-11)
contou~Q
for a mean
!I
I
"I
()
sllTfac;~
BI~oola,
474
4. The disturbance in the downstream chaimel may also be eliminated if the transition mentioned r,bove is designed with a well-pl:oportioned reversal of the wall curvature. In snch a design, the positive and
negative disturbances developed. by the reversid of the curvature may
offset each other so that the flow is restored to complete uniformity at
. the end of the transition. Figure 17-13 shows the boundary curves of
such a desigrl, derived by the method of characteristics. Although these
~--"--~----'---~~~~~~--~~-4---=3t'__~_~~5
I
.
~:
= 25
b,
..
1.1
---I' I
I----+-----+-a~b~+--+-~ = 2--+---/--b,
-1...---1---.-1
I
_I
2
:3
-=-_--L-_J
9
10
11
12
Values of _x_
b, F,
de~ermined by the
b, = dowl!stream
.;
!
,
, 'I..
475
-I
I
--I!
I
-I
.I
.~
. -.{
476
j.
by the tailwater. A critical section exists at or near the neck of the flume
only when the flow is free.
Illvestigations on the subject of flow through constrictions are many,
but most of them have dealt with subcritical flow. The contracted-opening
method, I v, method fOT determining flood discharges which has been very
popular with Americn,n engineers since 1918 or earlier, is based on the
application of the energy and continuity equations to the fiow through a
constriction in the waterw[1.Y. The first laboratory investigation in the
United States is believed to have been m[1.de by Lane [ll}. This investigation dealt with simple constrictions of flows having Froude numbers
slightly higher than those uSlmlly fOllnd in lll'.1;ural channels. In recent
years, :.L ne.w development in the study of constrictions in subcritic<Li flow
has been achieved by comprehensive fundamental research carried out at
the U.S. Geological Survey by IGndsvater, Carter, and Tracy [12-151.
The method they developed pl"(:duces more exact results for yariolls gi ven
conditions and so will undoubtedly repbcB the conventional contl'actedopening method in .e~lgineering practice. The following two articles will
cover the essentials of this new development,
17-6. Snbcritical Flow through Constrictions. 1Vhen an [,rea constriction is introduced to nn otherwise uniform, frictioll-controlled flow in 1).
prismatic channel of mild slope (Fig, 17-15),2 a backlvater of llfl-type
profile is first developed upstream from the constriction. The upstream
end point of the bnckwater curve is assumed to be at section 0. Near
the constriction at section 1 the central body of watel' begins to accelerate,
de\~elel'ation occurs along the outer boundaries, and separation zones nre
created in the corners adjacent to the constriction. An adequate approxims,tioll for the location of section 1 may be troken at a point one opening
width b from the center of the opening. Between sections 0 and 1, the
flow.is gradually va.ried.
At the constriction, the flow is rapidly varied, characterized by marked
acceleration in directions both norma] arid parallel to the stl-eamlines.
The 101<gitudin::t.l water surface drops mpicUy in this region. Within the
coilstriction, the live stl'eam contracts to <\ width somewhat)essthan the
nominal width of the opening, and the spnc8s bt:tween the live stream and
the constriction boundaries are separation zones occupied by eddying
I
Section
(OJ _ _ _
T -~
________________
'Ln"-- -~~
(4)
It
Live slreom
bo~ndory
Eddy zone
~__~ ___________-L~__~--I--------------~
(0)
Constriction
Oalum
(bl
. Backwater profile
Normal profile
ho'hj
hOn'hln
j
/.
Datum
HOlik [lOJ credited S. M: Woodward with the ~evelopnlent of this method for use
by the Mi"mi Consen'ancy District, The othel equn.lly important method reported
by Houk is tile slope-a.rea. method (Art. 6-9).
, In Fig. i7-15, it is'shown that the flow through the constriction is subcritical, e.s
in Fig. 17-14a.. This is t.ile case usually encountered in practical problems, If th~
water surface drops below the critical-depth line M the constriction, supercritical flow
will occtu, as shown ill Fig. 17-14b. To reestablish subcritical flow downstream,
a hydraulic' jum{will be developed near the downstream side of the constriction.
The method described here, however, is not applicable to this case.
.UC~'VW"
(e)
FIG,
[14],)
loss.
17-15. Definitioll sketch of flow through CDnatliction. (After l"}'(fey cmd Carter
((~) PIau; (b) elevation; (c) elevlltioIl, ad'lpted to ltsstunptioll of zero friction
478
"'[tter.
the water passes through the contraction, the contracted
f>tremn reMhes a minimum width at section 2, which corresponds to the
vena contracta in an orifice flow. After the vena contracta, the live
stre&111 begins to expand until it reaches downstrea.m section 4, where the
uniform-flow
is reestablished in the Jull-width channel. Between
sections 3. and 4, the flow is gntdually varied, Over the whole reach
from sections 0 to 4: encomp:1ssed by the backwater "ffect of the C012stricti on, the total energy loss .is the same as that for uniform flow.
FOl' a simple analysis of the flow characteristics, the following conditions m!1-y be assumed:
L The channel .ftOOl' is horizontal or nearly so.
.
2. Laboratory observatiol1s have shown that the level of the water at
the vena contracta can be closely approximated by the leve! at sect.ion 3,
which is at the downstream face of the constrictioll. H611Ce, the depth
Y2 can be taken as Ya.
.
. . .
3., A coefficient of dischr.rge is intl'ciduced in the a.naIJ~sis to include
the over-all effect of contraction, eddy loss, nonuniform distribution of
velocity, and nonh)rdrostatic distribution of pressure. The related notatiOll to be used in the analysis is defined as follo,,'s:
hd = eddy loss due to turbulence el1gend,~red in the upstream
separation tones. It
assumed that this loss can be
exuri8SE;ea in terms of the velocity head at section 3,01'
is
h. = k.(Y:NZg)
0<1
aI'
and 0'3
alld a..'
kl'(y~2/2g)
+ Yl
3:
1112
0'1 -
2g
,
al
VI
11 3 2
Oil
-2
g
;,
0'3 Yl
h,
+ hi
1, aaty,
(17-13)1
479
Q = Co;4JTT,
where ila is the water
for the disclu,rge,
~l.l~ea
at section 3.
I-------------~~;"
CA. ,\/2f)
.
where t.h
equn.l to
= Til
'I
(6.1& \
hi
.
(17-15)
al
.j
+ Yb)/2b, x/b,
i':,
t/(y'J
+ 6.h),j, Ljb]
(17-11)
i
r
-I
'.
Il
r,
.1
480
Jt
I
J
.,'"'"
.J::.
",-
.~ ~
"'0
-~
o
-'"
c~
.~.g
uc
=-10
~-
.,~
"
"
'u
Ca)
:3
-:
;0
:.::
(b)
1.2
tt
b
io
(0 -~
tD
tl)
~'---"--r-, --'"'.~~
:T
~!' "0,
f' I
""
c::
"
"0o.
Li
II
~~~
11 , )
I,
I
J
"II
I-
,',
N
d
f---'+-'\--"l-
'0
l-----+--+-!f-.--,
~
."
<D
e:0
';;
5!
v,
3!
<t
e:0
~~j'il
1 '"
'0
[ -~ 3:1
(e)
FIG. 17-16. Disclul.rge coefficient for constriction of type I opening, vertical embankments (E = 0) and vertical abutments. (U.S. Geol(J!lfe~1 Survey [12J.) (a) Base .
' curve for Coefficient of discharge; (b) vl1l'ia.tion of discharge coefficient with Fl'oude
nll.mbruj (cJ "Il;liation of discharge coefficient with entrance rounding.
481
482
483
"
D
~ o.ao j-----'-----'----+---+---=--..;...:...,=c-=?---=+-==l---l,...,i
Stone-erd conditions
:6
X=-+Yb
~'O.~O
'0 0.70
.!!
_t_,o
u
0.60
u
0
j ;0
~~ +LI'
IT
~-.
<? '0'
10
3D
20
qo
50
60
70
90
100
(a)
1
.1
tOO
<1>'0'
10'
(17-18)
I
I
-25'-
--_._----
0.90
_2Q._
Ko/>
\i~
0.80
--,-
(e)
.i
i
{
-I
41"-~---~i
d.70
10
20
40
30
m: Per cel'lt
of
50
so
c.honnel controclion
484
485
1
I
/TI =
(a)
m=Per cen1 of chonl'\el contraction
(((.)
0.05
K,~0.90
0.800:--"---'10--.--'20
30
50
60
70
90
100
(b)
'T~~~~ ---+-'--+----+-=.4
----1----1----,
0.90
K
(b)
0.80
40
80
(e)
,--."..,,_ _---'--c-_-'--_
0.06
0.10
0.12
0.14
{=Rotio
or
x tcwidth of opening
(e)
FIG. 17-19. Discha.rge coefficient for constriction of type III opening, embankmerrts
and obutments with 1: 1 slope (E = 1). (U.S. Geolorrica1 Survey [12J.) (a) Base .
curve for coefficient of discharge; (b) variation of discharge coefficient with angula.rity;
(I) variation of discharge coefficient with x/b ratio.
)-
486
nAP~DLY
VARIED FLOW
487
'0
>
"0
l;
~
~ 0.90 I----'~"
.
~
0.80
I--_~_-I-.
___-'-_~"",,-!..::::,.,,,,,,_
'il
lO.70
~
"
..'"~
50
rnP~r
cf)'nt
~l
60
70
60
90
ioe
channel tontrct:tion
at chon,,1
con1r~ction
(a)
(a)
11'\'
(bJ
(b)
1.15
1.10
Kw
1.05
I
;. ,r
\.00
0
~:r Rafio of x 10 wi_oth 01 opening
(c)
FIG. 17-20. Discharge coefficient for constriction of type III opening, embankments
and abutments with 2:1 slope (Iff = 2). (U.S. Geological Surull11 [12J.) (a) Base
curve for coefficient of discharge; (b) variation of dIscharge coefficient with angularity;
(el variation of discharge coefficient with x/b raj,io.
I
I
!
!
(c)
.!
489
1.00_=~=c
f
f
!
t
---r[---11
,.00\-----:;0",--..
osoo
0:,
0.2
0.;
0.4 .
(a)
(b)
(a)
t.OO'
I PI~E:S
0/>- O'
--'10'0-----
t:";WdZl
0.90
___
0.70
~rr
4"i---~-
. ~l
10
j!.~
A3
OI'OO~~~~'
g~
~<-'--l .i!
0,90
',DO
0.90
",-
~---
K~
O.BO
C
0
--20'------
~O
60
30
70 .
60
(bj
(c)
P'ERS
=
=
jo
1.00
l-----,~-.,..-'--'-..,---r===+===F=33;:::;1
I
I 0.05 I
'1
I
'
~lo.H5:::::i::::I:o:.lo:+I-::-:-:-::::::I===:J~=-_.-L-r
i
so
20
30
4Q
50
m~PI!"
60
70
.80
100
(d)
KO 1. 10
FIG.
I.003).,O~.-::::;J35'"'.,..---,4-:f;O":---.-:4f;5;;;---;;5';;;O'--,;'5"5.--7
60 ,
eAo~l.
01 wing ... 11
(d)
K; from the diagraro(Fig. l7-23c), enter the hOl'izontuj scale at the propel'
value of m, then move verticidly to the vaiue of Lib, horizontully'to the
line marked j = 0.10, vertically to the vulue of j, and horizontally to the
value of f{j. If j happens to be greater than O.lCl, an approximate computation may be made ~y taking C = C' for the given m and omitting Kj
490
It is
that ~rtaiIi combinations of the empirical coefficients
applied to C' may appear to yield a value of C greater than 1.0. In sllch
cases, howBvei, a value of C
1.0 should be used.
.
Tile friction-loss term h, in the dif;lcharge equation, Eq. (17-15), represent.s the total loss of head due to .friction between sections 1 and 3.
This loss is equal to the sum of the loss in the approach reach of lellgth
La from section 1 to the upstream side of the contracted opening and the
lo~s ill the contracted reach of length L. The total head-friction' loss
.may be cpmptited by.
L"(VI~d(3Y +L(I~3Y
stage .to the backwater stage at section I iskrlOwn as the backwater of lhe
c(Jnslriction. The'distance Lilt is the difference in water-surface elevation
between secti.ons 1 and 3. From Eq. (17-15) the value of tJ.h can be.
shown to be
V
~l 2~ + hI
(17-21)
The rat.io h*/ tAh is called the backwater l'atio, which is known to be a
function of the cluH1l1el roughness, percentage of chamiel contraction,
(17-19)
where [(land J(. are the total conveyances of sections 1 and 3, respec. (17-15) the above expression for hI, and
tively. Substitut!'l
Q/ A! for l'l; the
may be expressed as
Q
= 8.02CA a
(17-20)
0.040
Manning's un"
0.050
0060
0.070
FIG. 17-24. The effect Qf cha.nue! roughness on the backwat,er ratio for basic-type
constrictiolls. (;iji8" Tracy and CarieI' [141.)
of
.J
(
492
The backw[\,t(-~r I':1tio in Fig. 17-24 is for constriction of basic type., that
is fol' a vertica.l-raced cor,striction ,v-ith square abutments. The backw~lter mtio for other types of constriction may be obtained by multiplying the backwater ratio by an adjustment factor lea. This factor has
been found to be a function of the contraction ratio m and the ratio
CIC"",;" Cb "" ~lncl C are, respectively, the discharge coefficients for the
Taking an approach sect-ion 350 ft ~lpstream from the constriction, the friction slope
section and the contracted section ma.y be estimated by the Manning
equation S, =Q~n2j2.22rl'Rj,~. Now Q = 50,000 cf~, n = 0.035, A = 400 X 20 =
8;000 ft', and-R = 18.2 ft,; hence, Sf = 0.000-!48. The frictional loss h! = O.000-!48 X
350 = 0.16.
AsBuming h~ = 0.20, e:.h = 0.20/0.26 = 0.77. \Vith As = 350 X 19.'13 = 6,800
ft' and C = 0.87, V a'j2gC' = 1.1l. By Eq. (17-21), uh = 1.11 - 0.59 + 0,16 =
0.68. This vaiue is less than the t.ri:J.i value 0.77. The computation is to be repeated
until agreemeat is reached. Before this' is done, however, the value of C should be
adjust~d lor F,. Now Fa = 50,000/(6,800 X --v'32.2 X 19.'13) = 0.29. Using
Fig. 17-16b, C = 0.87 X 0_955 = 0_83, Thus, Va2/2gC' = 1.22 and e:./i = 0.79.
This is close enough to 0.77; hence,a r6peat.ed cumputation for successive approximation is not required.
between~his
;.
..I 0
\~~_I
\ "I . . . . . ~
V73--~--'--'---~I---'
'il73
1--,.... V73
0.9
o.s
0.6
(After
493.
I
l
J:
~.
~.
17 -8. Flow through Culverts. A culvert is a unique type of constriction, and its entrance is a speciallcirid of contraction. The culven a'cts ",<os
an open channel as long as the flow is partly full. The characteristics of
the flow are very complicated because the flow is controlled by many v8.riables, including the inlet geometry, slope, size, roughness, r,pproach and
tailwater- conditions, etc. Hence, an adequate determination of the flow
through a culvert should be made by laboratory or field investigations.
Yarnell, Nagler, and vVood ward [16] were notable pioneers who, made
more than 3,000 tests on flow througli different pipe and box culverts.
Later on, round smooth pipe culverts were tested by Mavis [17], corrugated and concrete pipe culverts by Straub and Morris [18-20] and by
Straub, Anderson, and 130\\i el'S [21,22], and standard box culverts by
Shoemaker and CbytGn [23]. In addition, a comprehensive experimental inve'stigation of the hydraulic behavior of commonly used pipe
culverts has been conducted at the U.S. Bureau of St::l,lldards, as reportr::d
by French [24-26].
.
A culvert will flow full when the outlet is submerged or when the outlet
is not submerged buttb.e headw[,ter is high and the barrel is long. According to laboratory investigations, the eJ1trance of an Ol'ciinary culvert willnot be submerged if the headwater is less than a certain critical value,
designated by H*, while the outlet is not submerged. The value of H*
varies fr0111. 1.2 to 1.5 times the height of the culvert, depending on the
entrance geometry, barrel charn.cteristics, and approach condition.
For a. preliminary analysis, the upper limit H* = L5d maybe used, where
d is the culved height, because computations have shown that, where submergence ,vas uncertain, greater accuracy could be obtained by assuming
that the entrance was not submerged;
Laboratory investigations also indicate that a culvert, usually vhth a
square edge at the top of the entrance, will not fiow full even if the
entr:1nce is below headwater level when the outlet is not submerged.
Under these conditions, the flow entering the culvert will contract to a
depth less than the height or-the culvel't barrel- in a manner very simila.r
494
"''"
r,-rr-rr-'---"'--~---'
to the contrac.tion of Row in the form of a jet under a sluice gate. Thii':
high-veloc.il,y let will continue through the barrel length, but its velocil;Y
will be reduced slowly as head is gradually lost by friction. If the culvert.
is not sufficiently long to allow the expanding depth of flow below the
contraction to rise and fill the barrel, the culvelt will never flow fulL
+-11
II t: !
.,
0
.; 0.05
a.
0
.,
".,
+-t--+-j-L
I
0.04
/' v
11
"
I
0.01
:II/,.V
0
0
1\'\ \
\
"'- I',",~)J)\
'"~~ l\\~
~~~'"
""r::: t::;::
t!
{'
~(j'c'
\
I
'"o
'"od
00
if>
o
c5
-t+-
.I
2 _,
~t- 1--1
(lon9_)
I ~Type
Vi~
L...-\--i
10
20
0.01
f--
"
f-.--
....!o
f---
;::J
0
oj
:::.:::....l--Ul
'"
I
0
.'
...Jjo
30
I\:
~ \
J\
...
..c:
.~
!,
>
r-<
;::J
<J
'"
bJ)
0;
..s
"d
Q
.~b
oj
\
+-1 1- -.-
pt~o~\O
Jg~\
r
"'0
\
I
f--+--
FIG. 17-26. Criteria for hydraulically short and long box ane! pipe culv~rts with concrete barrels :J.nd squ&re, rounded, or beveled flush entrances from a vertical headwall,
either with or without wing walls. (U.S. Geological 81J.rvey 1271)
'"
...,
.~~
..c:cD
:;::
-3
L'J
,.....'"
.0
0.02
F
r-=: ~
'"0-
.,00
~i(b~~"~ \.\
,
",-" ~~
~"
hydrau~ican!l
f::: t::=
S!
r--:::
I--
od
'C/)
'I
OJ
<t
0.
lD
..'"
C,J
I
Ul
0
,s...
:~
'"
00
'"
g
if>
I
~
495
. '!
<t
<D
Type). V
I
,/ )
I%i
::::;
~Io
\
I
'.(Shorl)
0.02 I--
I
'17
o.o~_ f--
II
I -'
O)
I----
IiI--
'"
0
I----
""
~
<if
..JIO
0.06
.::l
f--
<l
.~
'"
-,:-." -.
~.~
..
',.
_. -',_.
-- - - -".
->.:':."
lically short and 11. hydraulically long culvert;. Under suitable conditions,
n hydrttulica!ly short culvert with submerged entran~e may prime itself
automatically ,md thus flow full. AGcordi:1g to the laboratory
tiotls by Li and Patterson [28], this self-priming ,~ctiol1 is due to a rise of
the water up to t.he t.op of the culyert caused in most c\\ses by a
Jump, the backwater eff0ct of the outlet, 01' a st,lndilig surface Wiwe
developed i.nside the barreL
.
For Pl'!lctiooal plll'po:>es, cuivert flow 111\)' he clf\,ssified irlto six
as
shown in Fig. 17-28. The ic1enti5cfltion of en~h type may be explained
accQrding to the followi'ng outline.
A. Ontlct submerged ....
Type
(1) Outlet submerged
H >d
1/,
>d
Full
ftOlV
H>H"
y; < d
l?ull now
B. Outlet' unsubmcl'ged
1. Headwater greater than the critical value
4
5
6
If the outlot. is
the cllivert wi1lflow full iike 8. pipe, and the
flow will be of type 1. . If the outlet is not submerged, the headwrtter
may be either
01' less tlmn the cl'itic:11 vn,!ue': When the headwater
ii; greater than the critical
the cuh<ert lllay ,be either hydraulically
short. 01' long; these call be differentiated by l)1e&,I1S of the charts in Figs'.
17-2ij and 17-27. The fiow is of type 2 if the culvert is hydraulically
long and of type 3 if it, is hydraulically short. When the headwater is
less thrHl tQe cl'iticul value, the t.nil tI,i.ter may be eitherhigher or lower
th::m the critical deptb of the flow at the cuI vert outlet. For highel' tuilwatei' , the flow is of type 4. 11'01' lower tuilwater thefiow is of type 5 if
the culvert slupe is subcriticnl and of type () if the slope is super(,l'iticaL
III the above cln.38ification, there is an exception in that type 1 f1O\v
can OCClll' with tl1ilwnter slightly highel' than t,he cl'iticOJ depth or with
tail water .higher than the top of the outlet if the .bed slope is very steep.
The first, two ty pes of flow are pipe fiow, and the other types r..re openchannel flow, For type 3 How, the culvert act:! llke an orifice. The
coefficient of discharge .varies approximately from 0,45 to 0.75. }rOI'
type 4, 5, and 6 flows, the entmnce is not sealed by WiLter and it acts like
a weir. The discha~g:e coefficient v[l)'ies r..pproximn.tely from 0.75 to
0.95, depending on the entrance geometry and headwr..t~l'
As shown in Fig. 17-2$, type 4 fiow is subcriticu.l throughout
length. Type.5 flow is subcritical and, hence, the contl;ol section is at
H <H*
lit
< y,
SuiJcl'itical flo,,'
Control at outlet
H
'Yr
< U'
< Ye.
Supercritica,[ flow
Control at entra.nce
i
t
I
"
497
j
498
II
..
may be obtained by means of the charts in Figs. 17-29 and 17-30 for box l
culverts and circular culverts, respectively. Both charts supply information only fOl' (l1llvert.s having square-edged entrance.s. For culverts
having rOllnded entrances under average conditions, the value of H/d
may be roughly estimated
the following expressions, in whIch Hid
refers to the l'!1,tio of headwater to barrel height for culvert with squareedged entra.nce:
'"
'u
'0
'"
I HId,
Type
'+ 5
:3
10
20
30 4050
Circutar, ..
Box, .. " ..
'
.. ,
.5
..;
.
h d t.e on 00" 'ulver~~ with square-edgGd enefstlll'rna~lBng
de:.,.. \d~I,a.'of U.S.AB:reat,
~f Public Roads [29].)
U.
I. aJe
,
100 r--t-'-Jj,..-
50 f--+--lrl-t-+-t-;--y:
J.
J 0 I----t---'H...,..
N<
00
t")
':t
\i')
Discna rg e io cis, 0
the outlet.
~ntrfanche. flows
na YSlS 0 t ese
is aAtth1e
,~."
..
10
1.5
0.87 H/d
0;36 + 0.64l!ld
Hid> 1.5
1.09
0.02
+ O.WHlll
+ OA6H fa
of
g 40f--t--lrl-t-+-t-r-iTT
'i;
< 1.0
O.87Hld
1.00Illd
....
,
FlO. 17-29,
~harG
folr
tl'llnces,
flowmg
p!l.r~ y
499
~Y2(2
2)
+ F21)
= YJ(2
+ F32)
(17-23)
data for box culverts were prepared by analogy from experience with circUlar
pipes and are beli~ved to be conservative; tht\t is, they will give h!l!\.d as high as is
likely to occur uncier adverse conditions.
1 The
For important'referen<;es on the subject or' flo\v through bridge pi<lTS and pile
see '[301 t<i [52]. On trash racks, see [531 to [611. On piers andltbutments
on top of SPiUW3.yS, see Art. 14-7, [43]. and [62] to [65).
. 2
~restjes.
i00
.
V2b~Y2
.F22U2Y2~
or
when')
(f
bdB3.
= Y1Bay.
Fa 2 Y3 J:.
e1Fa2(2 --1
FzZ(2
F~2)3
(17-26)
+ Fa2)l
(17-25) .
1.
[
I
I
\
501
ingly, the flo'll is subcriticnl iUhe value of Fa faUs on the left side of the
plotted curve 9 an itSsumed energy loss andslipercritical if F J f<ill" on
the tight side of the curve.
. .
17 -10. Flow between Bridge
For flow between bridge piers,
Yarnell [42] has IDJ.de an extensive stl:dy of the rclaterllit;ero.ture and
conducted a large number oi tests on different kinds of pier commonly
used in American practice: He found that the bridge-pIer formulas
most COli-un only used in the United States were d' AubuissOll'S [30],
Naglel;'s [33], Weisbach's [31l, and Rehbock's [35-37]. The Weisbach
furmull1 was considered theoretica.lly unsound l because of the discordant
results obtained b:;r YarnelL In connection with the use of the Rehbock
formula, the Bow passing through the obstruction is classified a8 least,
moderate, llnd complete turbulelit. Yarnell indicated that this clo.ssificil.tion might be u:;eful but that classification us subcl'itical and supercritical is more logicaL Least and complete tUl'blllent flow were found
to conespond to subcritical and supercl'itical flow, respectively, and
moderate turbulent flow was found to be in the neighborhood of or equal
to critical flow. }"or least and moderate tui'bulent flow,. the Nagler
formula seemed to fit Yamell's experimental dD-ta better than the
d' Aubuisson formula. For complete turbulent flOiV, the d' Aubuissol1
formula seemed to fit the data better. In general, however, neither
formula applies too well at high velocities.
The ~Naglel' ,formula is
(17-28)
FIG. 17...aZ. Criterion for subcritical and mpercrit'cai flows through an iJbstructiQll,
2
27E 1F,L
(2 + Fn 2)3
(17-27)
For: a given IT, therefol'e, the flow through/ the obstru~tiol1 is critical if
Fa T F aL . Examination of Eq. (17-26) indicates thatithe flow through
thei obstruction is 8ubcritical if
< FaL and supel'critical if Fa > F aL
Eqt\ation (17-27) is plotted ~s shown ill
17-32 fori E
1 (no energy
losS), = 0:95 (5 % energy lo~s), and E
OJ) (10% energy loss). Accord-
mll.)' b.e ca.lculated as the Burn o( two portions, the lower p\,-rt, flow
orifice a.nd Lite upper part flowin~ over a weir . Ap~arelltly,
and has no theoretical basis. Houk [10] and Bhbeucley
it unWll.rril.L1ted and cOlltrary toestnblished hydra.ulk t!teory.
II
I
I
502
V 22
+ YI = CJ!z 2i + Y2 +
h
I
(17-29)
503
Conveyance
where Yl = Y2
10 2, l,he velocity-distribution coefficients ell and a~ are
assumed 'equal to 1.0, and the friction loss hI is Msrimed zero. Taking
Type of pier
0.9
0.8
.0.7
ra~io u
0.6
0,5
I---;--I::---l-,-~,--l---:--'-
"1-
..
f......-.. -I------l~"--.+-.-I-
::>
~ 1.4 ___...!i--........:.-_+_-+
1.21----1----+
00
02
0.3
0.4
0.5
0.6
Caoveyonce rotio
el' Aubuisson made no distinction between the depths Yz and Yz, but
assumed Yz
= Y3.
-1
1.8
"' 1.6
"'-
+ VIi.
(17-31)
where K.{ depends mainly on the degree of channel cOl'ltraction and OIl.
the shape and orientation of the obstl'l1c&ion. The assumption Y2
Ya
implies that the backwater h* = ha = hz in Fig. 17-31. Accordingly,
the backwater ratio has a va:ueof 1.0. This),,; of course, not always
true, although in many cases there is very little difference between Y2
and Yl' For this reason, the d'Aubuisson formula is genel'ally recognized
to be empirical and approximate.
The following' are extracted from the results of Yarnell's investigation
on bridge. piers of variolls shapes and sizes:
L 'The height of the back~vl),ter due to bridge piers varies directly with
the depth of the unobstructed channeL
. 2. The Nagler, d' Aubuisson, and Rehbock formulas give' approximately correct results for ordinary velocities when the proper coefficients
al'e used, but they do not. hold for extremely high velocities. From his
3, For flow of low yeladty and least turbulence, the more efficient
shapes arfJ len.'S-shaped nose and '"ail, lens-sh~"ped nose and sell1icircular
tail, semicirClilar nose. and lens-shaped tail, co'llyex no'se and tail, and
semicircuhll" nose and, taiL 1 However, the data are 110t sufficient to
differentiate RlnOllg these shapes for high
of contraction.
4. Twin-cylinder piers either with or without connecting diaphragms,
piers with 90" triangular noses and tails, a.nd piers with rec;::ssed webs
are less efficient hydraulically them those just mentioned, and piers with
sqnare noses and t?ils are least efficient,
5. Application of batter to the ends of
slightly increases their
hydraulic efficiency, that is, raises the KN and
values.
6.
the length of a pier from 4 times the widt.h to 13 time;::
. the width has compal'atively little effect on its 'hydraulic efficiency, II!
some cases the efficiency .is t.hus increased anti in other case.~ decreased.
The optimunl length-width ratio probably varies with velocity a.ndis
generaUybetwe:en 4 and 7. On the average, the values of KN and ](A
will increase about 3 to 5 % for the increase of the ratio from 4 to 13.
7. Placing the piers at an angle with the current has an insignificant
effect on the alUount of backwater if~the angle is
than 10. Placing
0
the piers at 20 or more m<t.terially increases the n.mo\lnt. of backwater;
the increaee depends upon the quantity of flow, the dept.h, and the degree
and ](A willdecrfmse
of channel contraction. In general, the values of
about 7
at 20.
and to fa~iIitate .
lens-shaped nDse or tail is fDl'lned by two convex C~ll'ves tangent to the si;:lcs
of pier l),nd of radills twice the pier width. Th", convex nose 01' I,ail is ionned:by two
curves ('illlgent t-o the sides of pier find circum$criberl on i'\n cquila.tenl triangle.
504
505
,2.4
1'-
2.2
-'Vi
II
/ vi r- -
Yo
I
lYi[ll
I ',ti't1 '-0==0 I
I. ~
J4f",
L
-~. ~ f11J
o k. ':'=
e==)-
--_. --+----
0.6
0.4
C=>"-l
0.8
1-I'<H-lr'r--j 1. 5 h"-ft
l!
, .... - f-
1.0
Vf;-
c:::::J~
h" 1.4
1.2 -
._-
0,2
0.6
r-
---
FIG. 17-35. Chart for determining baclC\\,ater of supercritical fiow through bridge
piers. (After D. L. Yarnell 142].)
.
F:IG. 17-34. Chart for determining backwater of subcritical flow through bridge pier~.~
are square noses and tails, 90" triangular noses and tails, twin-cylinder
piers with and without connecting diaphragm, and semicircular noses and
tails. If the backwater is given am). it is required to determine the discharge, it will be necessary to as:mme a value of V 3 and to carry out the
computation by trial and error until a aolution ia obta.ined.
The chart in Fig. 17-35 gives a graphical solution for supercritlcal
flow through bridge piers. With ,the propel' va.lue of (1', determine the
limiting value F3L from Eq. (17-27) or Fig. 17-32. Then enter Fa/FoL .
. on the chart and read h*/y~ for the given pier shape. Tliis value multiplied by Ya gives the backwater h *. If Yl, v!, bz, B 3 , and the pier shape
are known and it is required to determine ;the dischar:ge, h* /Ya may be
computed from h*/V! = (YI ~ Va)/Y3; Fa/F3L may betaken from the
chart; u = bdB3;'and F3L may be obtained from Fig. 17-2701' Fig. 17~32.
The vaiue of F3 can then be calculated, and from it V3 and then Q,
Yarnell's graphical solutions apply to piers with length-width ratio
1.2
I
.'
l/f<==)
Length -width
rolio =13
- - c..._
---
. ----
,- /
y-
/-'
'\7"-
-'
;,. :7'
/"
--- ---e:"L.
-.... - -- -- -f/
'"l-
-.-
--'
--
c::::::J
Length-wIdth
F./F 3L
[421.)
-'--
---
----
506
'
507
equal to 4. For ratios of, 7 and 13, the effect of increasing length on
backwater is I:Ihown in Fig; 17-36, which is plotted \vith Yarnell's data.
It seems that tM bZ,ckwater caused by the long piers is greater when
the pier ends are semicircular than when they are square. It is pl'obahle
that an, abrupt entrance in the case of square pier ends tends to'decrease
friction losses for (t short distance downstream because of its effect on the
velocity distribution, since the velocity along the walls is reduced. The
effect of bridge piers present ill a constriction has been considel;'ed in
Alt. 17-6 and in Fig. 17-23d.
17-11. Flow through Pile Trestles. Yarnell's investigation [41] indica.tes that the N{tgler formula may be suitably applied to subcrltical
flo w passing a pile trestle a,net the d' Au buissoll formula to superc.ritical
flow passing a pile trestle. The following coefficients are recommen.ded
for use in these formulas:
Value of (3
Square nose and tail, Lis = 5. _., ....... , ' ..... '"
Squ!1rtl nose and semicircul<U' t,ail, Lis =' 5..... ..,.
Semiciroular nose and tail, Lis = 5. ..............
RO'.Jod ... , . .. .: ........ ,.".... ,., ...... '....
Airfoil ... _. . . . . . . . .. . ..... "...................
2.42
1.83
1.67
1,79
-1
0,76
Spangler [55] has extended t.he experiment and determined the value of f3
for IX = 30 0 , ,45 0 , and 60 0 ,
Ac.cording to F'ellenius [54], an empirical formula for c can be given
as follows:
C =
. {j
+ Ii )2 sm=
'
S
Ji. ( S
(17 -3L~)
.j
Type of t"e811e
Single-track 5-pile trestle ben t
Parallel to curren t. . . .. . .. . ............. ,
At to" angle with current ......... '.' ..... .
At 20 angle with current .............. , ..
At 30 angle with current ................ .
Double-track JO-pile trestle bent ............ .
Two sillgle-track 5-pile trestle b'~nts: ..... , . , .
0.96
V2
= c2g-
T'alue of /J.
Square nose and tail
With sharp corners, Lis ~10 ...... , .................... 7.1
With sharp corners, Lis = 12 .......... , ...... _....... 6.2
,\Vith slightly rounded corners, Lis = 8 to 11. .......... 6.1
Semioircular nose and tail, Lis = 7 ................ .... 5. I)
0.88
0.86
(17-33)
I
I.
I
-- .-:<
V(tlue "fx
1.0
1.0
LO
i.5
In general, ~c = 1.0 for bars having sharp 01' slightly rounded corners and
x = 1.5 for ba!'s having rounded corners. For cross-connected and
clipped rack bars, the value of Ji. should be increa.'led by about 22.5 %.
Scimemi [56] and Kozeny [50] have provided values of c, /3, and Ji., and
other data for racks iustalled in several hydropower plants .
17 -13. Unqerfiow Gates. Certain control gates ir, cana.ls may be called
unde7'ftow gates 1 frcim the f act that. water passes underneath the struct ure.
Common examples are the sluice gate, Tain~er gate (or radial gate) I
and rolling gate (Fig. 17-37). In d.esigning such gates the hydraulic
engineer is most interested in two major .features: the 'head-discharge
relationship and the preSSUl'e distribution over the gate surfaces for
various positions of the gate ~nd forms of the gate lip. The form of the
lip will .not only affect the velol:ity and pressure distributions and the
energy loss in flow through the gate opening, but may also develop very
disturbing vibrations that should be avoided during gate operation. As
(17-32)
,,c = {3 (S)H
Jj sin 0
Ji.
.1
1 In contr~t ~o the underfk..r gate is the ,ovet'flow gate through which the water flows
over the stl1fctur~. The drum gate (Art., 14-9) is an exa'mple of au overflow gate_
, Hydraulically speaking, the overflow gate ,acts like! a weir as much as the underflow
gate act~ like an orifice. There are also designs for which the water flows above ,and
below the structure at the same time (Fig. 17-37).
508
Q=
where
aLh~-IJ-:(-Y1-+-a-';~'7. ~-;:::-)
(17-35)
For the vertical sluice gate, l experimentally determined curves representing the value of a have been prepared by Hflnry [68], aFl Shown in
. Fig. 17-38; The dashed line. A represents the result obtained by Eq.
(17.:.35) on the basis of t11e energy principle; the dashed line B is obtained
through the use of the. momentum principle. The value of :IfQ is the
Froude number of the flow through the gate opening.
.
Similat curves fo.r the Tainter gate (Fig. 17-39) were prepared by Toch
[69] on the basis of an experimental study" induding the pivot height as a
+Qt.
O.4l--I-+'<:""'1I---.J,c...,.61~';""'F-~F-;~I=--l
c
Sluice gate with combined
oyer- and underflow
Sluice gate
0.1
.'1,
~
Z /
Tointer gote
the height of the gate opening, Yl is tbe upstream depth of flow, and
ex. V 12/ 2g is the velocity head of the approach flow. The outflow of the
gate may be either free or submerged, depending on the tailwater depth.
For submerged flow, Yl in the above equation should be replaced by the
effective head, or the difference between the upstream and downstream
depths.
.
.
.
For the .purpose of experimental.studies, the ~elocity-h~ad term m
Eq. (17-35) may be omitted, and its effect may be mcluded m ~hecoefficient a. Thus,
(17-36)
Q = aLh, V'2i1h
04
..
.'
For
!--f----jJ-l--ll----jF:=.:I-/!='"'
For other studies, see [39J and [71]. For a theoretical:analysis of the flowthrough
a sharp-edged sluice gate, see [72] and [73].
~.
.
1
510
511
(a)
"
I
I
(hl
-----'--~--------4~
.I
. .FlO. 17-39. DischArge coefficient Ior Tainter gate.
'
(el
512
513
I
J
surface, the res1lltant of the 'pressures must pass through the trunnion'
axis O.
The outflow from the gate uswiJly constitutes 11 high-velocity jet,
capable of eroding the downstream channel bed. l The erosion thus
developed may be avoided by means of a stilling basin (Art. HHl). If
no preventive measures are provided, the erosion D:1ay result in a deep
seoUl' hole, the formation of which
depend' on the type of the jet.
According to the studies of Escande [71] on fiow through a. channel drop
downstream from [l, sluioe gate/ four types' of jet are possible: (1) submerged How with rising j Elt, ,(2) submerged flow with diving jet, (3) free
flow with rising jet, and (4) free flow with diving jet. Generally speaking, the diving jet will induce less pressure on the vertical surface of the
drop than the rising i et and will usually develop a scour hole with its
deepest point closer to the drop than the rising jet. In submerged Bow,
the depth of scour ,hole depends on the tall water depth, whereas in free
flow the downstream conditions have little or no effeot upon the depth of
scouring.
17 -14. Channel Junctions. Flow ,through a channel junction is a
phenomenon that involves numerous variables, such as the number of
the adjoining channels, the angles of intersection, the shape and, slope
of the channels, the'directions and discharges of flow, the rounding of the
corners at the junction, etc. The problem IS so complicated that only
a. few simple and specific cases have been studied. The conclusions
of such studies indicatE;. that gene1'8.l1zation of the problem is not possible
or even desirable. When the application of hydraulic theory t.o the
problem encounters limitations, !L model study will give the best solution
fOl' the How characteristics involved.
For stcbc1"itical flow passing through the channels at a junction, Ta.ylor
[80] has investigated the specifiC cases sholvn in Fig. 17-41. The channels
. are horizontal and of equal width.' In the case 9f combining flow, the
following assumptions were made: (1) the flow is from channels 1 and 2
into channel 3; (2) channels 1 and 3 lie in a straight line; (3) the flow is
parallel to the chn.nnel w<lJls, and the velocity is uniformly distributed
immediately above and below t.he junction; (4) ordinary wall friction is
. negligible in compn.rison,with other forces involved; and (5) the depths
in channels 1 and 2 ate equal immediately above the junction. With
will
1 The erosion of channel bed due to jet is a subject beyond the scope of this book.
For detailed informatiDn see [1] to [3] and [75] to [77].
2 Other examples of jets below hydraulic stnictures in channels have been found and
studied by Camichel [78]. The erosion due to iet through a bridge opening may cause
a deep scour hole generruly known as the bl1le hole. An experimental study of such a
problem has been reported by Hickenlooper, Guillou, and Chow [79J.
.k2
n.~(ni
- 1)
+ cos 0) + ny -
1]
(17-37)
=:
where k2 = ~22/2gY2' nq
Q2/Q3, ny = Ya/Yb, V 2 is the velocit,y in
~han~el 2, Y: IS the depth 0, flow in channel 2, Ya is the depth above the,
JunctIOn, Yb IS the depth below the junction, and 8 is the angle between
the. merghig chann~ls. Taking n. as. a parameter, k2 may be plotted
agamst nv for any gIven 8.
(oj
(b]
FIG. 17-41. Simple channel junctions (a) combining flow; (b) dividing flow.
exer.~ed by the flow in chann.el 2 is ~ol1nt.erac.ted by the pr~ssure on the opposite wall,
prov lded the water surface m the JunctIOn IS esser,tially flat. Thus, the only force
the flow from channel 2 can contribute to the flow in the straight channel 1-3 is its
mom~ntum component in the direction of 1 to 3.
514
I
/
1.30
1.25
~ 1.20
E
.<::
Q. US
'"
;7
Cl
1.10
1.05
2/
1.00
!Lo
t10
,--
Vi
rn
AiT
~W~cP
1.00
the division of flow of a given discharge QJ. First assume QI; then
Q2 = Qa - QI. The corresponding depths YI and Y2 can be obtained
from the rating curves of channels 1 and 2. For Yl/Y2, the ratio Ya/Y2
can be determined from Fig. 17-42.. By assuming other values of QII
the correspoildirlg ratios YZ/Y2 c<,n be obtained. Thus, Qz/Qa is plotted
against Ya/Y2, as shown by the curve. A in Fig. 17-43. The intersection
C)f this curve with the ka curves now
gives all possible combinatior.s of the
0.20
B
variables, among which one value of
.
k3 would correspond to the flow of,
I
say, 500 cfs entering the junction::
!
0.16
1----. Tbe llext step is to plot the ill terSected
k3 values against the corresponding
Ya. For the example shown by the
.., 0.12
line A in Fig. 17-43, ka = 0.10,
o
~
.
Ya/Y2 ~ 1.067! and Q2/Qa = 0.677 or
~
"
!I
Q2 = 0.677 QJ. From the ratiag curve
~ 0.08
of channel 2, the V2 corresponding 'to
,
V3z~
:-the Q2 just computed can be found
~kl=2Y
I
9 l
and, hence,. Ya= l.067Y2. The plot
0.04
II
of k, ~,gainst lis can be constructed,
as shmvl1 by the curve B in Fig.
o
17-44. The proper value of k. must
4
6
8
10
satisfy not only curve B but also the
V?IUes of YJ
solid. curve showing the relationship
FIG. 17-44. Factor k, D.S a funccion of
ka = V 3~ /2gY3~ The ill tersection of Ya. (Afler E. If. Taylm' [80).)
the two curves gives the required
values of lea and Y3. With this le" the corresponding value of Q2/ Q3 may
be determined from Fig. 17-43, and the divided discharges Ql and Q2 can
.
be determined accordihgly.
For 8upercritical flow passing through the channels at a jUllcLiQll, the
flow condition is more complicated .. Studies of a few cases of combining
flow of the general arrangement :~hown in Fig. I7- n have been made by
thel;.T.S. Soil Conservation Service [81]. In such cases, a hydraulic
jump m~.y forin in one or both of the inlet channels, depending lIpon the
junction design, the discharges, velocities, and related flow conditions
intheadjoinillg channels. Thus, a large increase in the height of the
side walls in the vicinity of the j unction may be necessary. When the
prihclple of momentum is applied to sllch a problem, it is necessary to
know the position of the jump or the depth of ~ow at the uj:oEtl'eam edge
of the junction in order to compute the momentum contributed by the.
inlet flows. If the flow passing through the junction is supercriticaI.
without the formation of any hydraulic jump upstream, the flow condi-
.
'I
~r~
1.30
1.20
Depth rolio, ~
.,
(Afler E. H. TaylO1'
"i~ 1.20!1--~-l---t-~
1.1 51----l------I---j-----
.t:
.Ii
.,
o
ti Oll---'--l----+---t
1.05
il
1
I
1
.1
lh
'\
Yz
. Ii'!
&;15
---
Quantity rolio, ~ 2
3
For the specific junction showll in Fig. 17-41b, Taylor has made an
experimental approach to the problem as follows:
. _.
For any given value of 8, it is possible to correlate the (hmenSlOnle~s
ratios Q2/Qa, Y3/Y~, yt/Y2, and ka = V.;2/2gYa, all derived from the expel'lmentally determined data. For g:= 900, such correlation curves ~re
shown in Figs. 17-42 and 17-43. These curves can be used to determme
,J
.,
(
. 1516
PROBLEMS
1'1-1. Verify Eq. (17-5), and interpl'et the dashed curves in Fig. 17-3 that represent
this equation.
'
. .
17-Z..
Formica's data. for sudden expansions, plot a Cllrve showing the relac
tiollship betwe(ln .aucl the angle of divergenc-e fI, and then aiscuss this rela.tionship.
17-3. COnllmte. the wave angles fJ, anG fl. of th? contruction designe.t in Example
17-1, and s\;:e~(h the position of the shock wf,vefronts.
17 -4.
a straight 'contmction to redu~e the width of \I. rectanguiar channel
fl'Olll Q. to 3
The channel carrips a discharge of 4;33 ds with an approach velocity
of 7.2 fl)S. ASSUl1l11 11;/11i
3.00.
17-5. Compute: the wiivenng!es of the design inProb. 17-4, and sketch the posi'tlQIl Q! Ill\) sll(ld: wiwefront..'.
17-8. ))c>lign the c-ontracl.ion in EXIIlJ1ple 17-1, !l.Ssllmbg 'Ya/Y, = 3.00.
17-7. A c.hnnllcl 6 ftwide (;afries a discharge of 100 cfs at a velocity of 20 fpS.
COIl~tl'Uct th.l surface contour of the !low through a sq<lare-f'dged abl'Upt expansion
of the !'hr.rllloL
17":8. If tho dmnnel described in the preceding problem has a gradual expansion
wil h its bOllTHlnril>3 ropresented by Eq. (l7-ll), constl'uct the slIrface contour of the
,l'
(
17 -9. !II 11 ke n preliminary design 'of a curved e.xpansjO!l for the channel in Prob. 17-7
with Iho !lid of the ,Ihnellsion\ess boundary curves detetmiHed by the method of
!,hnral'tcri~ti('l! (Fij!;. 17-13). AsslIm() hi/II = 3.
'
.
17-10..Solve EXam1)le 17-.2 fol' tlw following addition"t'conditions: (a) the center
lille. Of-thfJ bridge and highwn.y emballkmellt makes an angle of 70" with the directioll
of flow, (I,) the ~i1t.m!lCe is rounded fol' rib = 0.10, [lnd (e) the bridge has a pier 30 ft
loug nnd 10 rt wide at I,hl) middle of the waterway opening.
17-11. Dctermine the flood disdlllrge through 11 bridge conBtrintion having a type
III opel1illg with (;mb~lIlm1ont Ilnd abutment slope 2:1. The COl1strietion 30 rt long
r(!.dtl(t~ t'le nonnal wid th of w!lterWt1)' from' <JuG it to 200 ft... The brid~e has thre~
picr~, ctwh 30 it long ami !l ft wide. The average de.pths of water at the :Lpprofl.cb
scctiQl~ 200 ft upstream from the brid;;e and at. the contracted section are 20.5 and
H)'3 ft, rn~lllwtivalJ'. The upstreilln t.op edge of the embankmen~ is 1 ft above the
w!lf(:l"St,l'fl1ce elevation lit the approndl section. The value of n = 0.035.
17 -12: Solve the j)l'cl'cdillg problem if the bottom oi the bridge is 3 ft below the
top of 1.lie t.'!lll\mnkment. The bridge deck; however, is not overflowed.
17-18. A box ctllvert 8 X 8 X 60 ft is Inid Qn Il. slope of Ho. It has square--edged'
ent.lfllwe Ilncl Yl!rt.icnl he..'\dwlIlIs. The headwater and tail\\'ater eleva.tions are S.OZ ft
above the inv(:rt of the entrq.nce and 13 ft above the invert of the outlet, respectively.
. the culvert, (b) the type o~ flow through the culvert,
Dd.orn1ine' (a) the'
Itnd (c) the normnl depth .flow in the ba.rreL The vruu~ of n = 0.015.
17-14. Solve the precedin~ problem if the headwater i~ 13 It above the invert of
the en trance.
1'1-15. 8'01ve Prob. 1713 if the culvert is circular, having a diameter of 8 ft.
517
17~le.
the preceding problem if the headwater is 13 ft above the inv'ert of
the entrance jl.nd the barml is: rough, having n = 0.025.
'.
.
17-17. A llighway bridge 350 ftloIlg h!iS six piers with semicircular nos~s and tails,
ea.ch 8 It wide and equl'.lly spaced. At the peak of a flood, the backwat~r was found
to l:!e 0.30 it for the seven openings. The average depth of flow immediately down:stream from the opening was 8 ft, Compute the flood discharge, (a) by the Na.gler
formula, (b) by the d'Aubuiss.oIl formula, and (Il) by the Yarnell charts.
17-1.8. A. bridge has four piers wit4 semicircular noses and ta.ils, each 40 ft long by
10 ft
During a flood peak of 45,000 cfs, the total width
the strealIl was
390 It, and the
depth at a. downstream section was 19.4 ft. Estimate the
backwater.
' .
.
17-1.9. Soive the preceding probleIll by the USGS method d~8Cribed in Art. 17-7.
The chaonel roughness coefficient is ,It
0.03. Unavailable .data.are to be reasoIl.
abl)' a.s5umed.
17-20. If the bridge in Prob. 17-17 ill exposed to a flood discharge of 35,000 C[6,
<:5timate the backvrater. The rating cur"" of the chanIlel gives a normal depth of
. 8.5 it for this UllScmLrg'e.
17-21. A
pile-trestle I',dlroad bridge 840 ft long is supported by bents,
each 2 ft wide and spaced 11 n center to center. At the peak of a flood, the ayerage
ha.ckwat,er V{as found to be 0.15 ft. The aver1:1ge depth of flow immediately downstream from t.he trestle W3S 6 ft. Estimate the bnc.kwater.
1'l~22. Verify Eq. (17-37).
of
REFERENCES
1. Charles Jaeger: Der Mischungsvorgang bei plt'tzlichem QuerschnittsUbergang
(The mixing process in sudden transitions), lVasserkTaft 'l.!nd Was'lenoirtsckafl,
Mllnich,voi. 31, no: 24, pp. gOo-SOD, 193B.
Z. Charles J a.eger: "Er,gineering Fluid Mchanics," translated from the German. by
P. O. Wolf, Blackie & Son, Ltd" London and Glasgow, HI5S, l>p. 157-169.
3. Charles Jaeger: De I'impulsion totnJe et de ses rapports avec l'enel'gie tot.ale
d'un courant liqui<:le A. sUr:lace libre (The total :impulse and it..~ relations with the
total energy of a. iree-surface liquid flow), Relme ytnJral.e de ~;hydratlh'q!le, Paris,
vol. 13, no. 39, pp. 143-151, 1947.
4. Gianni Formica: Esperienze preliminari sulle perdite di carico Dei can ali, dovute a
cambiamenti di sezione (Preliminary test on head losses in channels due to C!OSSsectional changes), L'Energia eleltricll,iIfila.no, vol. 32, no. 7, pp. 554-568, July,
1955; reprinted as lstitu.tQ di I dr{Lulica e Coslruzioni I dmulichc, Milano, :nie'lr!crie e
~t'!.ldi No. 124, 19'55.
5. R. L. Daugherty and A. C. Ingeraoll: "Fluid Mechanics," McGraw-Hill Book
Company, Inc., New York, 1954, pp.19O-HI3.
.
6. Arthtlr T. Ippen and John H. Dawson:: Design of channel contractions, 3d paper
in High-velocity flow in open channels: A symposium, Transactions, American
Sociel.yof Ciflil Engineers, vol. 116, pp. 326-346, 1951.
.
7. Mtl.S.a!lhl;Hom-ma and Sukeyuki Shima: On the flolV in a gradu!illy dIverged open
channel,: The Japa.n Scienc~
Series yoL 2, no. 3, pp. 253-260, 1952.
8. Hlmter ;Rouse, B. V. Bhoota.,
Design of channel' expansions,
4th pap~r in High-velocity flow in op:en channels: A symposillm, Transactions,
Amt<riea~ Society of Civil Engi';'ws, vdL 116, pp. 326-346, 1951.;
9. Fred W; Blaisdell: Flo\v through diverging open channel transitions at super-
518
10.
11.
12.
13.
14.
15.
16.
. 17.
'18.
19.
20. Lorenz G. Stra.ub and Henry Morris: Hydra1.llic tests on corrugated metal culvert
pipes, Univer~ily of }/in'Lesota, St. Anthony Falls Hydra'.J.lics Laboratory, Technical
'
.
Pa.per No.5, SeriBs B, July, 1950.
21. L0renz G. Straub, Al...-in G. AndBrson, :md Chades E. Bowers: Effect of
Inl~t
a.
519
.oj
I
J
520
t,
piers of bridges), Com.ptell rCIu:i.ItS d.e l' Acadpmie des Sciences, vol. 108, pp. 1970197:/;, JIlne 19, l ! l 3 ! J . .
.
47. Leopold Esco.nde: Recherches sur l'ecoulement de l'eau entre .pll~. de ponts
(Resellrcl.es on the How of w:J.ter through piers of bridges), I;e G~nte C'tmt, voL 115,
'.
no. 5, pp. 113-117, Aug. 5; and no,. 7, p~. 1~8-1.40, Aug. 12. 1;39.
48. Etienne Crausse: Sur un phenomene d oscl\latlOn du plarr d eau provo que pa.
I\~coulement autour d'obstllcles en form de piies de ponts (On the phenomenon
of oscillation of watel' surface caused by flow a.round obstacles in the form ot
bridge piers), CDmp!es rendus I;: l' Academie des.8ciences, voL 209, pp. 197-199,
63.
64.
65.
66.
Pll. 546-549.
51, l)hili-pp Forchheimer: "Hydrt'!.'ulik" ("Hydraulics"), 3d ed., Teubner yerlagsg;~sellsch!lft, Leipzig and Berlin, Hl30, pp. 519-522 on backwater a~ bru:!g-e
piers and pp. 522-:524 on ioss of head at trash. ri1cks.
.
52. Armin Schoklitsch: "Handbuch des Wasaerbaues" ("Handbook of Hydrauhc
Engineering"), Springer-Verlag, Vienna, 1950, vol. 1, pp. 122-124 on backwater
oJ bridge piers and pp. 125-127 on loss of head :J.t trash racks.
.
53. Otto Kirschmer: Unters\\chungen liber den Gefallsverlust an Rechen (Studles on
the head loss through a rack), Mitteilungen des hydraulischen I nst-Uuts der technisc/llm HocMclmle JltIilnchen, no. 1, pp. 21--41, Mllliic~l, 1926..
.
54. W. ~'ellellius: Undel'sokinglll' betroffande faUf5riuster 1 skyddsgrmder vld vattenkraftanlOggninp;er (Studies of heu.dloss through racks in power pla.nts), Tran$aclirms, Hyr/"alliic Institute of the Royal Technical University of Stockholm, no. 5,
1927.
.
.
55 .. J. 'Spangler: Untersuchungen tiber den Vel'lust all Rechen beischr!i.g~ Z~stromullg
(Studies on the head tollS through a ra.l!k inclined to stream flow). r.bttel{un~en d~s
hydr~lllichen I nstituts 4er technischen H ochschttle M tincher., no. 2, pp. 46-60,
, I
Munich, 1928.
.
.. .
56. E. Scimemi: Jlilievi sperimentali sui funxionan:.ento idraulico del grandl. Impla~tl
industriali ('Experimental studies on the hydmulic function of large mdustrHl.l
plants, L' Energia ele/lrica, MilanI), vel. 10, no. 9, pp.7Q5-723, Sep~ernber; no. 11,
pp. B97.-924, November, 1933.
57. Armin Schoklit.sch: "Hydraulic Structures," translated from the German by
Samuel Shulits, American Society of Mechanical Engineers, New York, 1937,
.
.
. .
. vol. II, pp. 891-892. .
58. Leopold Escande: Expression de Ie. perte de charge a. la traversee des gnlles
(E:<:pressionof the head loss in water through grills), Com:ptu rendlls de I' AcCld~mie des Sciences va!. 218, pp. 179-181, Jan. 31,1944.
59. Leopold Escande;' 1!:turle experirnentalii de 'Ia perte de charge Ji. In. .tllra)verCsee des
grilles (Experiment!!.1 studies of head lo~s m water through gn s, ,ompws
rer,dlls de l'Acad~",ie des Sciences, vol. 218"pp. 266-268, Feb. 14, 1944.
60. Leopold EsCl1nde: Etude theoriq,ue, et experimentl1le
la perte de charge de
l'eal.l i Ie. traversee d'une grille (Theoretical and expemnentl1l study of tho head
loss in waLer through 0.. grill), Le G~rlie civil, vol. 122, no.. Z3, pp. 188-190,
Dec. 1; and flO. 24, pp. 197-HJ8, Dec. 15, 1945.
61. David H. Kent: l\lodt"ls of hydra.'alic stnlctures, pt. I, Water Power, London,
vol. 7, no. S, pp. 301-307, August, 1955.. Gives head losses through tailrace fish
?e.
'11
d
screens.
62. FI~yd A. Nagl~r and Albion Davis:.Experiments on discharge over Spl ways an
6;.
69.
70.
71.
models, Keokuk Da.m, Transactions, American' Society of Civil Engtneers, vol. 94,
pp.77F820, 1930.
Louis G. Puis: Spillway discharge capacity of Wilson Dam, Transactions, American Society of Civil Engineers, vol. 95, pp. 316-329, 1931.
E. Camichel, L. Escande, and G. SalJl;.th~~ Sur la similitude des barrages a COD.tractions latel'~les (On the similitude of dams with lateral contractions), Compte.
rendus de l' AcaMrnie deJl Sciences, voL 194, pp. 807:-809, Mar. 7, 1932.
E. Camichel, L. Escande, and P. Dupin: RemarqUB!l sur certains phenom~nes de
contractions laterales dans Ire barrages (Not~son cert.ain phenomena of late.ral
contractions ill 'dams), 90mples rendus de l' Acad6mie de.t Sciences, vol. 197,
pp. 722-725, Oct. 9, 1933.
O. Mueller: Schwingungsuntersuchungen an unher:;trornten Weh,en (Vihration
studies on underflow weirs), Mitteilungen aer preussischen Vel'suchsanstall fur
WM3erbau und Schiffba.u, no. 13, Berlin, 1933.'
J. B. Tiffany: Laboratory research applied to the hydraulic design of large d&.ms,
U.S.A.rmy Enflin~r8 Waterways Ezperi-m;enl Station, 'Bulletin 32, 194!f
Harold R. Henry: Discussion of Diffusion of submerged jets, by M. L. Albeltson,
Y. B. Dai , R. A. Jensen, and Hunter Rouse, Tra1l.8(JctiQ:/I.I!, American Society of .
Civil Enaineers, vol. 115, pp. 63i-69-i. 1950.
Arthur Tach; Discharge chara.cteristics of Tainter gates, Transactions, A'lL61'ican
Society of Civil Engineers. yoL 120, pp. 290-300, 1955.
.
Hunter Rouse (editor): "Engineering Hydraulics," John Wiley &; Sons, Inc.,
New York, 1950, pp. 540-541-.
Leopold Escande:lttude theoriq,ue et experimentale de l'e~o\llement pal' vanne de
fond (Theoretical and experimental study of fiow through sluice gates), Revue
g~n~rale de l'hydraulique. Paris, vol. 4, No. 19, I)P. 2.5-29, no. 20, pp. 72-79, and
00.21, pp. 120-128, 1938; vol. .5, no. 25, pp. 21-34, no, 26, pp. 65-77 and 131-139,
1~~
I
!
521
..
I
I
l
I.
522
to fI~vi!l.1 hydraulics); Revue generllle de l'hydral!lique, Pll.ris, vol. 1, !lO. 5, 'pp. 235242, and no. fI, pp. 293-2g9, 1935.
.
.
79. Irby J. Hiclccnlooper, John C. Guillou, and Ven Te Chow: HydraulIc stu~les of ~
highway bridge, University of lilinoi.s, Civil Engineerinfl Stl.die~, H ydraHhc Engtneering Beties No.4, June, 1957.
. '
.
80. Edward H. Taylor: Flow characteristics at rectangular open-channel Junctions,
Transactions American 8od~!y of Civil ,Engineers, voL 109, pp. 893-003, 1014.
81. Charles E. Bowers: Studies of open-channel junctions, pt. V of Hydraulic mode!
studies for Whiting Field Nl1vll.l Air Station, lI>lilton, Florida, Univ6rlli!y OJ
Minnesota, Sl. Anth~my Falls Hydraulic Laboratory, Technical Paper No. 6j Serie3
B, Januaty, 1950.
PART V
UNSTEADY FLOW
I
J
523
. ...
CHAPTER
18
1
18-1. Continu;ity of Unsteady Flew. The law of continuity, for
unstea,.dy flow may be established by considering the conservation of
m~ss in an infinitesimal space betw~en tWQohannel sections (Fig. 18-1).
In unsteady flow, the discharge changes with dista.nce at a rate fJQ/ax,
loilio1
wafer
surface
and the depth changes with time at a rate ay/at. r~e change in discharge through space in the time dt is (aQ/ax) d$ dt. The corresponding
change in ch['.nnel storage in space is T dx (ay/iJt) dt = dx, (aA/at) dt.
Since water is incompressiblel-~~!?Il!::!~lus thech@J~'l
in stomge should be zero; that is,
= 0
SImplifying,
+ 7' ay =
at
aQ + aA =
fJQ
ax
or
At a given section, Q
ax
at
(I8-1)
(18-2) , ../
fJ{VA)
ax,
525
+ T ay = 0
at
(18-3) ."",.-
j inJ
~6P ~c
'c/'d'v:mt-rt9
526
UNS'rEADY FLOW
527
or
+ V ~~ + T ay= a
A aV
ax
ax
at
(1&-'/
small, that ~he ac~e].eration is in the direction of x,' and that its vertical
c?mponel1~ IS neghglble, Thus, the work done by this force through a
distance
between .the two channei seotions shown in the figure is
(w/g)(a V/at) dx. ThIS amount of work is equal to the energy los,
due t? acceleration. Dividing by W,' the loss iil head is expressed b;
II
d:
I
/'
(1/ghav/3t) dx.
The above equations are all forms of the cont~'nuity equation j07' unsteady
flow in open channels. For. f1 rectl1ugubr channel of infinite width,
Eq. (18-I) may be written
iJq ay _ 0
(18-6)
ax T at -
~~d' _ _-->I
,Acc!!!lerot1on Irne.s.Jope :$0
where q is the- discharge per unit width. This expression Wfl,S first,
introduced by Saint-Venant. [21).
When the channel is to feed laterally with ::t. supplementary di.,chnrg.;)
of q' pel' unit length, for instance, into an area that is being flooded over a
dike, Eq. (18-2) may be written
'
aQ
ao;
+ aA
+ q'
CJt
= 0
ax
+ aA + aA'
at.
at
29
.:l-
-::"....
[n~y~:::::::::-==-~
e
J1 oPe
,--':-z-I
.r
)'
(18-8)
where A' is the water area of the side' section. This equation also
applies to the case of a ohannel containing groins; the water oircllIates
between the groins to guide the fiow in the main channel but not to
/L,M"'Rrr...~SH1tribute discharge.
~ 18-2. Dynamic Equation for Unsteady Flow. For simplicity,!
:(f;<() .+l1tJ/y' unsteady flow will be treated like two~dill1ensional steady flow except
- - - - - that an additional variable for the time element will be used. This time
variable takes into account the variation in velocity of flow and accordingly brings to the fore the aoceleration, ,'thich produces force and
causes additional enel~gy losses in the flow. 2
With reference to Fig. 18-2, the force due to acceleration J V! at acting
on a unit weight w of water is equal to (wig) avjt1t; that is, force =
roass X acceleration. , It is assumed that the slope of the channel is i
For rigorous analyses~of unsteady flow, see [5J, [6], [9J, [111, !l5] and [20J.
The acceleration is positive for an increasing velocity of flow and negative for a '
'
decreasing velocity of flo'Y.
1
.. -.-,":
"
I <IV
h,'S,dx
'
i . ,:+d(af-qZ)
ed. Slope,s
""
1
1
~.
Channel b
....L.,_ - - '_'"
1
L L' __ y~u~~o':.:. __ ~"--.l
z
y+dy
-i
---l
hiha,qaT dX
Sf/'---'-
WOle,
~ISodl
(18-'1)
'If the channel consists of a deep main section and an extensive, shallow
side section, it l1HW be assumed that., whereas the discharge in the main
section is relatively very high, the side section cOli tributes only to storage
but not to discharge. Thus, Eq. (18-2) may be written
aQ
Q::!..~
z+dz
D'V2
+ 2i
=z+
Simplifying,
d
dz
.
+y+
dy
(+ +. 2{j-
V
+~
+
2
2g
D'V2)
.2
-...'I,
d D'V2
2g
+!:. a v
+S
d.
gat x ~
----.
(18-9)
1 aV'" .
d:I; -
gat dx
d.
.J
(18-10)
-I
":vo terms o~ the l'lght. are the head losses due to friction and acceleration, respectively.
I
.:.,
'
,I
528
I-
529
. UNSTJ':ADT FLOW
in gra.dually varied unsteady flow depellds Oil the effects of friction and
acceleration. In gradually varied sten.dy flow, the chn.nge in head
depends entirely on friction, provided eddy loss is negligible.
Dividing through by dx in Eq. (18-10) and utilizing parti<'ll diffel'entials,
Q2. The depth of the wavefront is gradually varied from the upstreani
section to the downstream section. During a time interval t, the wavefront moves forward a di~tallce equal to V",t. The wave velo'city is
greater than either of th~ mean velocities V 1 and Vi' in the uniform-flow
regions. When the wavefront passell ~ver the flow in the channel, it
takes in a steady discharge
(Y", - V 1) A i at the front. Since the
"rave configuration is of stable form and coilstant volume, the front
leaves behind an equal st.eady discharge Qo = (V,. - V 2)A 2 to the
+ Sf a
+.i.
(0:2g.
V2) + .!g aatv = 0
x
av + --1 &Y
Jz
+ o:TT -::+ -~
+ Sf. = (I
-'--:-=
or
a.1:
IL8L-- 0;1-
(18-11.)
(18-12)
ThL'S is the geneml dynamic (J(pwtivn 101' gr'adnally tlaried unsteady flow.
The friction slope in the equation can be evaluated by the Manning
formula, the Chezy formula, or any other suitable uniform-flow furmula.
For prisnmtic channels, i.e., for -dZ/CJX = So, Eq. (18-12) may be
written
(18-13)
I,
upstream side. The steady dischlLrge Qo which flows through the wavefront in' the upstream clirE.'ction is called the OV/3T7"un. Equating the
values of the steady discharge,
Q.
= (Vw
Solving for V w,
( 18-H:)
(18-15)
Since in uniform-flow
may be written
(18-16)
The above equations can be used for the computation of the velocity
of the monoclinal
wave. Equation (18-15) shows that, if there is
no initial flow, that
if Vl
0 and
= 0, then V",= V z If there is
initial flow, then V w is always
than either TTl or V 2. Thiil is
true because the wave must move more rapidly than the water particles
; ------?>-if it is to make up the volume of the wave configuration at any time.
~
In natural channels, the velocity of an assumed mOlloclinal rising
flood wa.ve may be d~termined by the so~called Kleitz-Seddan principle.
.
Equation (18-16) indicates that the velocity of a monoclinalhsing wave
\
is a function of the water .area' and the discharge relationship for the
1
channel. Figure 18-4 shows a curve representing such a relationship.
For ordinary channel sections in which. the velocity increases as the water
530
UNSTEADY FLOW
0,
--------
PI
I,
,./ I
/17 I
/1 'j I
./ /1 L-t-Um,ting position of ~ Pz
// I
!. I
/
I
!
I
/
I
!
1
/
I
I
I
_L_
!
,
...,. I Pz
tori (max VJ
"
')A
FIG. 18-4.. Graphical interpretation of water area, discharge, allti velocity relations in
IDonodinal rising wave.
. .
the slope of the secant P 1P 2 approaches the slope of the tangent at point
PI as a limit. Therefore,
dQ
(V"')m~~ = dA
(18-17)
ay =
dX
where dQ/dy is the slope of the rating curve of the channel. If the top
width of the channel section is known, the maximum velocity can be
computed by this equation. Likewise, if the maximum velocity is
known, the average width of the channel can be determined. The term
d!)Jdy at a given stage rna;: be computed from the mean slope of the
..!-ating_curves at gaging stations in the reach under col1sicieratiojl. The
maximum velocity 17,. can. be determined from the distance and time of
the wave travel between th!:.l stations.
- .
Equation(18-18) may also be used to determine the ratio bet'iveen the
.531
1 -
8 0 - Sf
Q.Z/gA2D
(18-19)
1I
i
This is the dynamic eqttation for uniformly progn;ssive ftowin a p.1'ilimalic channel. Since the flow is unsteadY;-Uiepositfon of the wave
prOftlischanging constantly, but the shape of the profile at ariy instant
remains the same. Equation (18-20), therefore, describes a "snapshot ':'
view of the profile at any instant. For wide rectangularchannels J
the equation can he further simplified and becomes mathematically
integrable. It should be noted that ay/ax represents the insta.ntaneous
surface slope of the wave profile. At a. given instant of time, this is
-I
I
j
532
UNSTEADY FLOW
In the equation, the normal discharge Q" can be determined from the
rating cui've at the station, and other quantities can be measured in the
field. A similar. equation has been developed empirically by Jones [30J
and is used widely for correcting the normal discharge to obtain the true
discharge in a river when the stage is changing. In the Jones formula,
however, the wave velocity .in Eq. (18-24) is l;eplaced by the surface
velocity, .which is much easier to determine but lower than the hue wave
velocity. A more exact equation than Eq, (18-24) may be obtained from
Eq, (18-21) by letting dy/dx
-j/V" and solving for Q. However, the
derivation and the resulting form of the equation are too complicated for
practi.!l1l purposes.
18-5. Wave Profile of Uniformly Progressive Flow. The wave profile
of uniformly
flow can be easily der~ved for wide rect~1ng1.llar
channels by using the Chezy formula. For the unit width of a wide
rectangular channel, A
R
D = y. By the Chezy formula, ](2 =
02y3. Substituting these expressions in Eq. (18-20), the slope at any
point on the wave profile at a given instant of time may be expressed as
~f the velocity i~ very low, so that the second te~ of the denominator
v:hlCh covers the velocity-head. effect, can be neglected, the above eaua~
.
tlOn may be written
dy
. (18-22)
Q = Q"
(18-23)
, L~t j represen~ the ratf'l of rise in the water surfJ).ce at a given channel
,sectIOn,: 1?en, J =
V,,:dy/dx, or dy/dx
-JJT1... Substitutin this
expressIon 111 Eq. (18-23),
.
g
:Q = Q",
,
'"
(18-24)
1
(V",y - Q.)2/0as o
= So 1"
y
(18-21)
dx
533
the same as the permanent slope dy / dx and hence: may also be expressed
by the latter.
.
There
is
an
lLpparent
similarity
betwee~
(18'
-'10)
.
.
.'
., and th e 'd'
yna.mIC
.
e~ll~tr~n for steLl,dy varied flow [say,
(9-16) with a = I}. This
slllilanty can, be exp:ained as follows. Imagine that an observer on the
~ha.unel ban1l: ,IS l'~nnm.glljter the uniformly progressive flow of Fig. 18-3
In the same dn'ectwn and at the same velocity, V., as the "mvefront.
H~
will see 11 picture of a stead v flow
like that ill Fig. 18-5; the "water
surfa~e will appeilr statioD.ary .to
him,and the flow will. show a
discharge Qc and a velocity ,Vw
Y nt every point along
the channel. Therefore, if Qo is
FIG. 18-5. View appearil\g to the observar .regarded as a steady dischaq,~e,
who follows the wavefront of a uniformly the steady-v!l.ried-flow equation
progreJ;slve flow,
can be applied directly 'to this
.
. uniformly progressive flow except
t.~at the dIscharge used for computmg the frictional slope should be set
equal to .VwA ~ Qo. This discharge is responsible for the frictional loss.
The dlscha~ge [~t any section of a uniformly progressive flow is Q =
VlOA - Q.. ~ hl1s, Q.
Vwi! - Q, ll.nd Eq. (18~20) beco:Q:les
(18-25)
y3 - QNg
X '7
1
So
,;3 - y 3
":
<
dy
(y - Yt)(Y - yz)(y - Ya) .
11 = 18c'f'c.
4Cl"C,
+ C)IC~I
+ 0 1 !:
(18-26)
'l:'hen,
534
UNSTEADY FLOW
voir into a dry 'wide rectangular channel. Use I,he Manning formula for expressing
the velocity of uniform :t!o\v.
,
Solution. In this case the wavefront moves in a steeply inclined wall of water whose
profile is apparently uncilanging as long 'as 'the channel conditions remain fixed and
the source .of supply is constant. Therefore, this can be considered a special case of
uniformly progressive flow, known specifics-lly as t.he roll wave in which A!
0
Vz = 0, Q" = 0, and V .. "" v',
c -,/ii~s;. Subs~itutin:g the ~:alues of Qo alld of
V wand K = CA "/11 ill Eq. (18-20),
dy
Ja: == S.
535
(18-21)
LetYl ;, Y" which is the norm!!.l depth at the crest of the W/l,ve, where the flow is
pr:l.ctically uniform. Solve Eq. (18-27) for dx; then
1
1
)
V/Y.
(18-28)
, . Choose the tip of the wavefront, whe.e x = 0 and y = 0, as the origin; and integrate
. the above equation; the length of the wave profile from the wavefront to the s~cti()n of
de;>th '!/ is
x =
FIG. 18-6. Profiles of uniformly progressive Row (11:) when y,
,
1(b) Ivhen Y. > Y2 or V", > Va + V OYz.
.!.
r (1 _ =---,1'---,-~)'"
dy
1
V/'.}.
S. }O
1/
(18-29) .
Integrating,
x =
the wave attain,s equilibrium. The first traces of the wave are transmitted downstream by dynamic propaga.tion at a velocity V 2 + v'YiJ'lJ
which is much greater than the fin1ll wave velocity V". By this actiol1
the downstream end of an actual wave will approach the theoretic:),}
asymptotic form vel'y rapidly. If the velocity. V,. is greater than
V 2 Vfiiiz, it is evidently impossible for the wavefront to extend an
infinite distance downstream. This is the case 'when Yc > Y2. The
downstream end of the wave is terminated with an abrupt front which in
theory should cross the overrun-critical-depth line vertically .. The
front is essentially a rapidly varied ullstea,dy flo,;', known specifically as
the hydraulic bOr'e (Art. 19-2).
y.,
S;!z + In
(l
z)]
(IS-30)
Rhine :River at the German-Swiss frontier. By a finite-difference graphical calculation of the char>'Lcturislic equations (Art. 20-2) developed by Craya (33], Re was able
to compute the fiov.' t.o be expeded from the destruction of 11 do.m for one specific
~ and of Chezy's resistance fa,rtQr. -rul~lud;:r the p;es;;nce
of some water initially below the dam, cau~ing the formation of a bore in his problem.
In 1952, Levin [341 described I!. solu~ion of the problem by a graphical method~ In
the S!l.rtle year, Dressler [35] presented a solution in which the resistance effect was
evaluatd by the Chezy formula.. This solution gives good agreement with observed
data exceptin the region near the tip of the wavefront. Accordingly, it is believed
that the Chezy formula may not be adequate for application to bighly unsteady flow
or to t,he condition of turblJlence in the tip region. In 1953, Dressler and Whitham
[36J improved the solution by a more rigorous analysis of the, flQW at the tip region.
Dressler J36) &lso developed a method of analyzing the flow during the very small
time interval after motion begins.
, .
Experimental data of the darn-breaking problem are few, notably the early Ilxpcriment,s made by SSholditsh [37] and Egui:3.zaroff [SS] and the later coues by the U.S.
National Bureau of Standa.rds [36).
. ,
A problem similar ~o that of dam breaking deals with the flow due to sudden lifting
of Ii'. sluice gate. A theoretical analysis of the wave profiles t)lUS developed was made
by Pohle [39) in 1950, assuming a. velocity potential flow in Lagrangian coordinates.
Experiments on this problem aiso were made at the National Bureau of Standards.
A compB,risOD. of the theories and experiments on both problems WllS preSented by
I?ressler [36J ill 1954. A comprehensive treatment of the dam-breaking problem along
hnes worked out by Pohle was giver. by Stoker [5, pp. 513-522j in 1951.
Example 18-1. Determine the wave profile of the nood generated by the failure of
a dam. 1 The failUre causes a sudden release of the. impOunding water from i!.s reserI This is the so-called dam-break problem, which is well-known in hydra,ulics.
In
1892, Ritter [31) offered the first solu~loll of the problem, using thea.pproximate
Saint-Venimt equation but ignoring the effects of the frictional and turbulence resist
ance to the unsteady flow. Consequently, this unrealistic classical solution was not
8upportedby the experimental findings and more practics.l analyses had to be sought.
In 19B, Forchheimel" [5, pp. 267-271\ presented a ~ummary of most of the previous
work on this problem, including attem pts to include the'resistance effects.
Toward the end of World War
[32) was instructed by the Allied Command
to study the problem in anticipation of possible destructiQn of the large da.ms on the
n, Re
I.
-I
.1r
536
UNSTEADY FLOW
where z = V/V".
10 failure of the
ron
Example 18~3. Determine the wa.ve velocity and prome of a uniformly progressive
flow in a. wide open channel if YI = 25 ft, Y. = 10 ft, Qhezy's C = 100, and So =
0.0004.
Solu.tion. By the Chazy formula a.nd Eq. (18-15), V., = 12.45 fps. By Eq.
(18-14), Q, = 61.25 cis per unit width. Since the discriminant e. is positive, the
numerator of Eq. (18-25) has three rea! roots, which are Vi = 25, y. = 10, and
y, = 61.25 i /(100X 0.0004 X 25 X 10) = 3.76. Tht; overrun critical depth y" =
..y61.25~/32.2 = 4.89 ft. Since y. < Y" no hydraulic bore will be formed. Substituting all known values in Eq. (18-26),
.
Vw = )T,G
G = 1 - (Yl/y.)~2
1 - yJ/y.
where
(18-31)
:1)=
(18-32)
x =
A,V,--AV
Ax - A
. i
I
(18-34)
dy
dx = So - Sf
Since V
= C vYS; and V.
= C
S,
Vy,S"
~
C'y
(18-35)
S,z (G _ g -z !)'
(18-36)
or
G - 1)'
-z
So ( .
So - ~- G -
dx=Y2{I+
(GZ,-G+l)"
So
(% ..,.. 1)[.' - \G' - l)z
(G _ 1)'J
08-37)
}d.
(18-38)
By integration,
_ v, { I ln
z+3_za
x-S;
I'
_ (G- 1)'[(2G'
+ G,
(l-z)
+ I)(G + 1) -4J In 2z
+ I)' - 4
2.
+ 1)
(G -- 1)'(2G
2(3-2G)
.
-In{.'-(G'-I)z+(G--,I)j
- (G' - 1) - (G - 1) V(G
- (G2 - 1) + (G - i) v(G
+ 1)"
+ 1)'
- 4}
- ;I
(18-39)"
where z = vlv" 11> is the depth at high stage, V is the depth at any section along the
wave profile, G i~ defined .by Eq. (1832), Yl is the depth at low stage, and C1 is a constant tha.t call be det.ernuned from the condition that x = 0 when Y = (YI
y.)/2,.
y' - 116.7
d
(y _ 25)(y _ lO)(y _ 3.76) . Y
+ C1
(18-40)
f (1 .
dy
+ C1
(18-41)
= 0.0004
[
Y
+ In (y
(25 - y) .
]
_ 10)9.431(y _ 3.76)',<19'
'
T
CI
(18-42)
".
where Z = y/y" The wave profile under consideration is long and flat; hence, the
mean velocity V will change very slowly with respect to both time t and distance x,
Thus, aT';ax = 0 and DV IB/ = O. From Eq. (18-13), using ciifferentials,
By integration,
Eliminating V w from Eqs. (18-31) and (18 e33) and solving for V,
1)
_1_
0.0001
(18-33)
Gz
- - V.
V= ( G - -
1
0.0004
By Eq, (18-15), replacing Al and V l by A and V, respectively, for any given section
V" =
537
c'
* Derivation
--I
538
UNSTEADY FLOW
pictul'e of steady flow (Fig. 18-7b) in which the wave appears to stand
still while the flow moves at a velocity equal to c in magnitude. Neglecting friction and assuming a small slope and a = 1 at all sections, the
energy equation between the tlormal section of the flow and the section
539
=Vg(y
+ h)
(18-48)
small heights, is
c
~
~.
foX t
\j21r an
h 21ry
(18-49)
j
..l
>;7////9///)#/9/////)///
(b)
FIG. 18-7. Generation of asotitary WaNe. (a) Unsteadj- flow; (b) the flow that
appears steady to un observer following the wave crest.
c2
Y+2g=y+h+
CZ (
Y )2
y+h
(18-43)
Solving for r,
c
where h is the wltve height above the normal water surface.
of moderate height, Eq. (18-44) may be approximfl,ted by
c=
(18-44)
For waves
(18-45)
FlG. 188. Wave patterns createq. by disturbances, (a) Still water, V = 0; (b) subc.ritical flow, V < c; (t;) critical flow; V = c; (d) Bupercritical flow, V > c.
deep water, where y is large comp!u'ed with A,the above equation becomes
c. -./g;.,/'Er. For small-wave heights, Ais large compared with hand
tanh (21rY/A) may be replaced by 21rY/)... Thus, Eq. (18-49) becomes
Eq. (lS-46).
The equation for celerity, either Eq. (18-46) or (18-47), may be used.
for studying the propagation of gravity waves. If a pebble is dropped
into still
the wave pattern can be represented by the concentric
circles shown in Fig. 18~Sa. The waves travel away from the source
of the disturbance in all directions with a velocity or celerity equal to c.
If the water is fio.w-ing, the wave pattern produced by a disturbance will
be displn.cedin the direction of flow. When the velocity V of the water
is less than the celerity, the wave pattern is as shown in Fig. 18-8b.
Since the velocity of flow is less than the celerity, it is possible for the
wave to t.ravel upstream at a velocity equal to
V tD
"'"
C -.
(18-50)
Its velocity
(IS-51)
",
540
Note that thb celerity represented byEq. (18-47) is identical with the
critical velocity of the flo;~. (1-10)].; The fiow under consider-\Ltion is,
therefore, subcl'iticaL ')
.
.
When the velocity of the water is equal to the celerity, the wavefronts .
in the l+pstream direction ILre stationary, or V", = 0, and' those in the
downstream direction have a velocity equal to V w' = 2c. This wa.ve
pattern is showl{ in Fig. lS~8c. The flow is, app~relltlYl criticaL.
When the velocity of the water'is greater than the cel<;1rity, the waves
will travel downstream only. This wave pattern is show11111 Fig. 18-8d.
The flo"r is, apparently, supercriticaL The lines tangent to the wa.vefronts lie at an angle to the original direction of flow. The angle fJ is the
(16-15), 01' .
wave angle, and its magnitude is given by
Sill
fJ
~_
v~ ~ ~
These Yectol'S are, as 8. general rule, parallel to the channel axis; so Eq.
18~53 mny be reduced to a simple algebraic sum
=
yc
of
(18-52)
v...
541
UNSTEA.DY FLOW
FIG. 18-9. Definition sketch for the method of finite increments, The subscript 1 is
for upstree,m section at initial state; 2 il> fOI' upstream I!e()ti~o oiter At; 3 is for downstream section at initW state; and 4 is for downstream section !lfter tJ.t,
is L:.t.
A =
A4 = -;r
2:A
'~~--=-~_,::" ..C--_:
T = Tl
+ Tz +4 Ts +. T4
R = Rl
+ Rz + R. + R4
V = VI
+ V + VI + V\
=2:V
(1S-54) ,
I
18-7. Solutii:ln of the Unsteady-flow Ejquations. Owing to complexity
of flow conditi~nsJ the general unstes.dy-flow equations defy exact mathe. matical solution. For demol1stI'atiYe p1u'poses, two approximate meth- .
= ~T
4
2:R
ay
ax
.!. (us 2
L:.x
VI
+ Yt -
L:.x
Y2) =
/
i!
fJQ = aCAl') =
(AlVa - A 1 Y 1
fJx
ax
2
L:.x
A1V I
+A
2V 2
YI
+ Y2
VI
+ Vz -
Va -~
2~
+ A4V4 -. AZV2)
, L:.x
AiVs - A,V, .
2L:.x
- Ya - V4 .
-~&E~-
542
UNSTEADY FLOW
uy = !2 (Y2 iU
aV
Yl -I-
- VI
6.t
t:d.
f:,.t
=! (172
y~)
Y4 -
+V
4 -
173)
fit
Y4
llxT.T(YI- Y2
=,
Substituting Sj
and simplifying,
So
;:=:
+ Y3 -
Y4)/46.t
A,
+ A1V1 +
A 2'V z - AaVJ
(18-55)
. .
(1:17)2
1h + y~ - y~ - Y4 _
4C2 T.R 2llx
(18-13)
a:::Z~~'-..'..~~__,,-__.:':
(18-56)
In a given proble~, tIle quantities Yl, A 1,Vl , Yl, A. 2 , 172 , Y3, AJ,.and Vz
are known from the initial conditions and from the compuk'~lOns on
previous reaches. The unknown quantities Y4 and 17<\ can be o.btained
by solving Eqs. (18-55) and. (18-56) simultaneously. Repeatmg :he
computation for subsequent reaches, the complete prohle and veloc1ty
variation of the flow can be determined. It is clear, however, that the
solution by the method of finite increments is extraordinarily ~aborious.
In the trial-aud-error method, the wave profile through the given reach
is first sketched arbitrarily and then revised if the sketched profile f~ils
to satisfy all necessary conditions. The procedure for the computatlOn
of a positively progressive flow is as follows:
....
.
. L For a given reach, draw the wave profile at the 11lltialmstant and
sketeh the wave profile at the final instant by extending the known wave
proS.l,,,: fromtlle upstl'eam reach. The extended profile is shown by the
dashed line in Fig. 18-9.
2. Measure the depths Yl, Y2, YJ, and Y4 and the surfa,ce siopes Sl, S2,
S anc1 S at the sections and instants under consideration.
3,
3. Compute
the water areas A l , A z, A~, and A4! the hy dra~ l'lC ra d"
11
R 1, R 2, R.! and R" the velocities V l ! 17 2, 17" and 17 4 , fl,nd the discharges
Ql, Q2> Q" al'ld Q4.
. .
+' .
4. Substitute the above quantities il?-to the contInUIty equaulOn.
Using finite increments, Eq. (18-2) gives
0
CQI
Q2 -
Qa - Q4) III
= (-.4.1 +
A2 - A.
A4) llx
(18-57)
543
It should be noted that the above procedure" does not include the
velocity-head and acceleration-head terms. If these terms were to be
included, the procedure woald become so tedious as to be almost prohibitive for practical purposes.
One of the most important problems in gradually varied unsteady
flow is the routing of a flood wave through a channel. Many practical
solutions of this problem have been developed, Because of its unique
nature, the problem with its pr<J,otical solutions will be dis.oussed separately in Chap. 20.
18-8. Spatially Varied Unsteady Surface Flow. The general differential equation for spatially va.ried unsteady flow can be obtained by introducing a term for acceleration effect into the equation for spatially varied
steady flow. According to Art. 18-2, such a term is (l/(l) (aY/at) dx.
Including this term in Eq. (12-2) and using partial differentials,
ay
or
= -
So - SI = ay
ax
+.!
g
Sf) ox
(v iW + Av aQ + a~)
ox
at
all. ilx
(18-58)
(18-59)
A hydrograph may also be a durve of the stage or velocity of the flow plotted
544
1.0
0.9
~~+
/'"
0.8
0.6
L-
,-
'-
0.5
/ J
0.4 I---
0.3
---
,/
0.2
0.1
..,/
0.1.
0.2
I
0.4
0.5
v.6
01
0.8
fJ
0.9
to
-.
r
sionless curve, as shown in Fig. 18-10.
D.
tn.
(I8-61 )
60q.
'---+~
_.
0.3
J( ='
O.OOO7i+ c
. S~'i
(18-63)
This equation was developed for slope not steeper than about
0.04. The roughness factor c was evaluated as follows:
545
2D.
t.
'---
.~
CT
0.7
0'
.....
iL
q. = 43,200
(18-60)
Value of c
0.0070
0.0075
0.0082
0.0120
0.0170
0.0460
0.0600
. When rainfall C8ases, the runoff decreases. The time t, from the
beginning of the recession hydrograph to the point where qlq. = r is
=
T
DoF'(r)
50q.
(18-64)
(18-65)
Equation (18-64) is derived mathematically from the finding that detention on th'e recession curve is proportional to the one-third power of the
discharge, Le.,
!!...
= (~)~
Do
q.
= r14
545
UNSTEADY FLOW
Using the dimensionless hydrograph and the above equations, it is possible to construct II hydrogmph for surface runoff due to a rainfall of
given intensity and duration: It is understood that the experiments
conducted for deriving the dimensionless hydrograph and the above
empirical equations were made under the condition that the flow was
lamiilar at all times. Thel'efol'e,the method is most suitable for laminar
flow and, according to Izzard, should be limited to cases where the product of the rainfall intensity in inches per hour and the length of surface
flow in feet. is Jess than 500 ..
. B. Tt~rbu.lent Flow. With reference to the type of flow ranging from
fully turbulent to laminar, Horton [53J has developed' the following
equation for the discharge of a spatioJly varied tlDsteady surface flow
due to a uniform rate of rainfall excess or supply rate: .
q
utanh"'
lmm+1
t ]
(uK) lim 60
are recommended:
Type of
i
I!'
u tanh 2
Ij
./
t
I
t
~
I.
'IF'<:ro,,
,/>:in/hr
(18-67)
overland flow due to 11 uniform rate of rainfall lasting indefinitely. However, the hydrograph for rainfall of finite 9,uration t can be easily derived .
In Fig. 18-11, the top figure shows t,he hydrograph for rainfall of infinite
'. duration, and the middle figure shows an offset hydrograph for which the
same rainfall starts at a. time of t later. The difference between these
two hydrographs will, evidently, produce the bottom figure, which is the
hydrograph for rainfall of 'duration t.
.
(18~68)
;/
. J
0.02
0.10
0.20
0.40
0.80
/ConJinuous
rainfall excess or i in/hr
tr
(18-66)
.q
Value of G
The Horton equatioll is most suitable for turbulent flow with high
discharge, The use of this equo'tion will produce a hydrograph for the
- 1,020 VS
K .IcL
~1!rfaGe
where q
.
547
E:umple 18-4. A smooth aspbalt highway pavement is 72 ft long and has a slope
of 0.005. Determine the hydrograph at the downstream end of the pavement when a
minfall of 1.8!l in./hI' is applied for 10 miD..
UNSTEADY FLOW
'
increru;e in discharge, aiter which the curve drops rapidly. This inorease in discharge
is due to the fact that during rainfall the amount of'detention on the; payement is
grea ter than the amount required to cause }he snme discharge after rainfall has ceased.
For i = OJ Eq. (18-6.3) gives K == 0.041 and Eq, (18-62) gives D~ = 0.434 ft. The
excess detention is, therefore, 0.515
0.434 = 0.081 ft, discharging at'a rate equal
to or greater than q, ~ 0.00315 da. The length of time ~ required to discharge the
;,xcess is, therefore, 0.081/0:00315
26 aee or 0.43 min. Accordingly, the beginning point of the actual
ourve is 10 + 0.43 = 10.43 min; at this time /]
0.00316 cfs and the detention
0,434, ft'. The recession time Of the hydrograph
(Fig. 18-12) caJl eMily bl) computed by Eq. (18-64).
1
2'.00 '
2.50
3.00
4.00
5,00
5.45
6.00
7.00
8.00
9.00
10.00
10.43
I >,
0,367
0,458
0.550
0.733
0.917
,1.000
0,0002
0,0010
0,0015
0.0021
0.0027
0.0030
0.0031
0.0032
0.0032
0.0032
0.0032
0.0032
0.0032
0.0019
0,0013
0.0009
0.0006
0,0003
0,30
0,48
0.65
CQ5
0.97
,[)4
1.00
1.00
1.00
1.00
LOO
1.00
0.60
0.40
0.30
020
0.10
1O.9l
11 40
11 ,86
12,64
14.57
0.21
i 0.48
0.42
0.97
1.43
0.02
0.'0
1.80
2,21
4.14
PROBLEMS
18-1. Assuming that the flow condition does not change with time, derive from
Eqs. (I8-1) and (18-13) the continuity and dynamic equations for gradually yaried
steady flow in prismatic channels.
,
18-2. Show that the overrun of a mon:lclinal rising wave may be e)(pres~ed 118
Q.
lOmm.
and
'
0.004
....
..
.,
Q,E
., ,.
..
-u
0.003
:v
0"-
.E'O
'II
0.002
II
~~
0'0
.c
,u,]t
~~ .~
0.001
Dr:;
'"
0
'L
0
II I
VI
\.
I I "-
.10
ii
'I
V: i
2:
12
Nt
16
(18-7Q)
When the minbll ee:J.ses at 10 min, it will be noted that there is a. momenta.ry ,
'- r:;
(18-69)
18-3. Show that the overrun Md velocity of a Inonocliual rising Waye iu a wide
rectangular challnel of uDit width may be expressed, respectively, 8.8
549
I
!
I
(18-71)
55,0
UNSTEADY FLOW
IS-14. The initinl and final depths of a. monoclinal flood wave are .10 ft and 25 ft,
respectively. Given C = 100 and S. ='0.0004, determine the percentage inel'eases
in the true'dischHrgc over the normal discharge at depths of 15 and 20 ft.
18-15. Compute the Reyn<:>lds number of the
discharge from the pavement
described in EXllmple 18-4, aud show that the 1I.0\V laminar. The discharge immediately a.fter the rainfnll ceasE'S is assumed to increase momentarfly about 10 %.
18-1S. Solve Example 18-4 if the rainfall is 3.78 in./hr.
1B-17. Solve E:<ample 18-4 if the rainfall is 1.89 in./br for the first 2 min and then
inl:'/"ea.ses to 3,78 in.jhr for the remaining 8 min. In solVIng this problem, it is nellessary 'to convert the first 2 min to il. time., say to, such that rainfs!! at 3.78 in./hr lasting
/0 min would have built up the same absolute volume of <letention !J.S was built up in
Z min at 1.89 in./ill'. The computation of the hydrograpb alter 2 min deals simply
wittl a const.ant rainfall int,ensity
3.78 in./hr lasting I, + 8 min.
18-1B. In an airfield, .0. strip of turfed area is 510 ft long and has a slope of O.g %.
Construl:'t the hydrograph of the overland flow on the are" due to ll. rainfall excess of
4.35 in,jhr lasting 10 min. For turfed area, assume c = 0.32 in the Hort<J/l equation.
of
REFERENCES
1. J. J. Stoker: "Water Waves," vol. IV of "Pure and Applied' M:athematics,"
Interscience Publishers, Inc., New York, 1957.
2. G, H, Keulegan; Wave motioil, chap. XI of "Engineering Hydraulics/' edited by
Hunter Rouse, John Wiley & Sons, Inc., New York, 1850, pp. 711-768.
3. Sir Horace Lamb: "Hydrodynamics," 6th ed., Dever Publications, New York,
1932. For tidal waves, see chap. VIII, pp, 250-3(}Z; for surface waves, see chap.
IX, pp. 3(}i3-475,
,
4. U.S. Army Corps of Engineers: Bibliography on tidul hydraulics, CommiJ,lee on.
Tic/.{).! 'IIlIdmulics, RepQrCNo. 2, Februa.ry, 19M; Su.pplement No.1, June, 1955;
Supple.men/. No.2, May, 1957j and Supplement No.3, May, 1959.
5. Philipp ForchheiUler: "Hydraulil, " (" Hydr!l.ulics "), 3d cd., Teubner Verlagsp;esell.
Bchaft; Leipzig; and Berlin, 1930,
o. Charles Jaeger: "Er.gineering Fluid Mechanics," translated fr?1l1 the Germ&n by
P. O. Wolf, Blackie & Son, Ltd., London and Glasgow, 1956,
7. Takeo Kinosita: Hydrodynamical study on the ftood flow, in Floods, vol. III of
Symposia Darcy, Intemal{onal Association of Scientific Hydrology, Publica~ion No,
I
I
I
551
13. D. N. Dietz: A new method for calculating the conduct of translation w>tves in
prismaticcannls, PhI/sica, vol. 8, no. 2, pp. '177--1\)5, February, 1941.
14. Giulio Supino: Sur la propagation des ondes dans les Cll.naUl( (Or,. the propaga.tion
of waves in channels), Revua !len~"'ale de l'hydra:uliquB, Paris, vol. 5, no. 29, pp.
260-262, 1939.
'
,
.
15. Ha.rold A. Thomas: The hydraulics of flood movements in rivers, Carnegie Inslitute
of Tecknology, Engineering B1Lllelin, 1934; reprinted in 1936 and 1937.
16, L. Cagniard: Hydrodynamiqt1e fiuviale: Regimes variables (Fluvial hydrQdynaIllic~: Variable regimes), ReVILe gtnerale de I 'hlldrG'U.iiql1e, Paris, vol. 3, no. 15,
May-June, 1937,
17,
M:;.sse: D~ {ntumescenl~es dans lea torrents (Tra.nslatory W!l.ves in torrents),
Revue iJdnerale de l'hydrGu1iqlle, Puris, vol. 3, no. 18, pp. 305-305, 1937.
18. Ph. Deymi{j: Propa.gaLion d'une intumescence allongee (Propagation of a grn.duaUy varied translatory wave), RevM genb:ale iie l'hydrauliq!Le, Pal:.is, vol. 1-2, no. 3,
pp. 138-142,1935-1936,
19. Pierre Masse: L'amortisMment de." intumescences (The damping of tmnsla.torr
waves), Revue gen~T!ll.e de l'hydrauliq!L6, Paris, voL 1-2, no. 6, pp: 300-308, 19351936.
20. Josef Frank: "Nichtstationare Vo;ga.itge in den Zuleitungs- und Ableitungska.n1i.len von WIl.Bserkraftwerken" (".unst.eady Flow in Heo.draces ",od Tailral;es
of Hydropower Plants"), Springer-Verlag, Berlin, 1957.
21. A. J. C, Barre de Sa.int-Venant; "E:tudes th.eoriques et
roent'des eaux courantes" ("Theoretical a.nd Practical
Paris, 1848.
22. A. J, C. Barre de Saint-Venont: Theorie du mouvlmlent non perma.nent des eaux,
avec application aUx cruea des rivieres et A I'introduction des marees dans leur !its
. (Theory of the nonpermanent movement of waters with o.pplication to the fl~ods
of rivers and to the introduction of the tides within their beds), Comple.i rend1LS
des seanc(lS de I' Acad6mie des Sciences, vol. 73, pp. 147-154 and 237-240, 1871.
23. H; Reineke: Die Berechnung der Tidewelle im Tideflusse (Compmation of tida.i
wave in a tidal river), Besondcrc MifJ.e~7:ungen, J ah.,.buch Jur die GelOtisserklmde
Narddeutschlands, Berlin, vol. 3, no. 4, 1921.
24. J. Frank and J. SchUller: "Scliwingungen in den Zuleitungl'- und Abldtung.i- '
kaniilen von Wa.sserkm.fta.nlagen" (" Oscilla.tions in Her.draces and Tailra.ces of
Hydropower Pla.nts"), Sprir-ger-Verlag, Berlin, 1 9 3 8 . '
'
25. H. Favre: "lttude theorique et experimentale des ondes de translation' dans.les
canauJ( deco\lverts" ("Theoretical and Experimental S!.udy of Translatorv Wa.ves
in Open Cbannels"), Dunod, Paris, 1\J35.
.
.
26, Carlo Drioli: Esperienoe sui mota perturbato nei canali industriali (Experiment
on
in industrial canals), U ElI erg ia. eleitrica, }I;[ilano, voL 14,110.4, pp. 285305,
appeng,ice, pp. 306-311,Apdl, 1937.
27. M. IGeitz: Note sur la tluiorie du rnOUvement non permanent des liquides et sur
applicati?u a la propagation des crues des rivieres (Note on the theory of unste~dy
How of liquids and on application tC' flood propaga.tion in rivers), Anna.les des
'fonts at chauss.es, ser. 5, vol. 16, 2e semestre, pp. 133-196, 1877.
28. James A., Seddon: River hydraulica, Transactions, A meriron Sociely oj Civil
E111linecrs, vol. 43, pp. 179-229, 1900,
American.
29. J. H. Wilkinspn: Translatory waves in natural channels,
Socidy oj Civil EnginCer8, voL 110, pp. 1203-1225, 1945.
'
30. Benjamin E. Jones: A method for COITBctio.g river discharge for !J. changing stage
U.S. Geological Survey, Wat;:r Supply Paper 375 (e), 1916, pp. 117-130.
"
"1
'f
.1
,I
I
-I
1
~'II
./
11
552
jJ
l. ")
I
UNSTEADY FLOW
!!
(,
553
555
+ V"u:.t12 -
V",A 1
(19-4)
A2
19
vr
( v .. -
or
.
Flow. If th
19-1. Uniformly ProgressIve
. e front of a :rnonocliniil
h O~ ~n abrupt change iil curvature or a sudden change. 111
wave ''''' '"'
Th' ff t
be p"O
. depth the flow in the front is rapidly varied, . . IS ~ ec m~y h ~
duced' for example by a sudden increase in gate opelllng at tee anne
,
' 111
"F"19. 19-1, The
velocity of the mass of W<J.ter
entrance,
as sh
OVYll
<
(19-5)
V",
(19-6)
G . ,
~GO"
II
U
or
'-2_-+1T"
....
--WAi'h
W/////m)W#407;;;:
11}
between the gate and the ...vavefl'Ont is increased from 1'"1 to V z, an~,the
.
. accord'lllgIy. By J.Newton's second law of"'t
motion.
mOPlentum is increased
t' .
the 'unbalanced force required to change the momentum p~ ~m Hl1e
is the product of the mass and the
ill velocity per um une, or
F
(V",
V 2 )A 21V(V 2
(J
V 1)
c=
wA1Yl
where fj: and iiI are the centroida.l depths of the areas.
above values of F and simplifying,
'
(V.. - V 2 )(V 2
. Al _ )
V 1)= ( ih - A2 Yl g
55~
(19-8)
(19-1)
= wA~y~
(19-7)
(199)
(19-2)
Equating the
. i
J
(19-10)
(19-3)
~,l
556
UNSTEADY FLOW
,
v'" =
I"I
(19-11)
c - Y1
which is identical with Jiiq. (18-50). It can also be shown that Eq..
(19-10) applies to type C and Eq. (19-11) to type D. It is theoretically
~
I
'I,_Y,
Y, --Vt
r
;;"?'~ml////;;/I';"-//h-:--"';:r-;'-/,/,-
557
r-=F:::[
Y, -V,
I
J/;;7///74//~r/h'l/}"
top waves will retreat faster than the bottom ones. The result is that
, the wavefront becomes sloping and eventul111y flattens out.
Owing to the presence of channel friction, however, the wave profile
of art advancing front n.ctmilly changes. This. effect may not be noticeable in relatively short artificial chQ.unels of small friction effect, but
it is significant ill long reaches of natural channels, where ~riction is
considerable.
.
19-2. The Moving Hydraulic Jump. When a rapidly vfl.l'ied unsteady
How involves a change in stage from subcl'itical to supercritical, a so-called
moving hydl'(J,1tlic jump will result. An observer on the bank. running
n.long with the wave crest. at the same velocity as the. wave will see a
stationary hydraulic jump. This is illustrated by the figures at t.he
bottom in Fig. 19-2 for four types of flow. All equations derived in the
preceding article apply also to these flows. However, another method of
.
deriving the equation may be described below_.
For a stationary hydraulic jump,
(3-21) can be reduced to
(19-12)
,
Tipe B
7,!pe C
Typ~
The right side of the equation is equr,l to the celerity expressed by Eq.
(19-9). Applying this equation tothe steady hydraulic jump of type A
in Fig. 19-2, as seen by the observer, V 1 should be replaced by V.. - 'f\;
that
FIG. 19-2. li'ol11' tYlJes of rapidly varied uiliformly progressive flow. (Top) Unsteady
flows; (bottom)' the corresponding flows that' appear stea.dy to an observer following
the wavefront~
(19-13)
or
(19-14)
V .. =
2Yl
+ 1;2)
- VI
= C- V1
(19-15)
558
UNSTEADY .FLOW
559
:Type B usually occurs in tidal rivers. Typical examples' are the Hangchow bore at Raining on the Chien Tang River, China (Fig. 19-3); the
bore on the Seyern River near Gloucester,' England; anel the 'bores 011
rivers connecting the Bay of Fundy, Nova Scotia. In' power caI'mls
serving hydraulic turbines, type B flow, known as rejection surge, will .
oecm' as a result of suddell dzcrease in power output.
.
- r
I'.
..,
.,
.~
~
~ \\;:-"r";:-----....--...
---l .
---l.
. --.
~~
Bup~
-=
~
\~
, /// ,.//~V1 --..
1
j
~-
'1
I
!
I
..
.//~~ Vz~
adVancing
--..... V ...... -
-r~-==
",
Type
FIG. 193. Hangchow bore at Haining em the Chie?- Tang Rinr;. China.. The wave
tront was about 16 it h,igh, traveling at high velocIty, Seven mm after It could first
be distinguished on the horizon the wave had passed. The water rellched a final
height. of about 28 ft within 30 min. The width of the river at the place of obser'mtion
was about 1 mile. (Caurle.~y of Mead and Hunt, I n c . ) .
.
~
'///-7.:;:
\
1
/~
1
;j
1
!
V-
"h~-07-~
'
i~
'
v2 ---....
--
Vw ___~ __ ~_
~up'stre~
.
;;,-..
'/
~~
.~
Type D-retrealing
inclined channels.
Increased.
In Fig: 19-2, tl;Ie initial. and final stages of a po.ssing surge ~tre assumed
to ?~ ~lllform .. Such surges OCcur frequently io channels of sma.ll slope.
. In lIl,chned channels, the four types of sur"e will Occur as shown in Fig
19-4 [111.
'"
.
560
Consider the positive surge of type A (Fig, 19-2); the absolute wave
velocity can be expressed byEq. (18-15). For a rectangular channel, the
velocity is
YI
Y2
+ Y2)g
2YIV;--
(Yi
(19-17)
This equation represents the relationship among the initial and final
velqcities and depths of the surge .. Similarly; it can be shown that'
Eq. (19-17) applies also to surge of type B (lfig. 19-2). Note that, if
Eq. (H)~17) is used to determine Yi or Y2, it must be solved by trial ::tnd
error.
MUltiplying Eq. (19-17) by the square of Eq. (19-9) lma simplifying,
it CMl be shown that
(Hl-IS)
where h =Y2 - Yl) or the height of surge, and where c is the celerity.
On the right side of the equa.tioll, the positive
applies to type B
surge a.nd the negative sign to type A surge.
When the height of surge is smn,ll compared with the depth of flow,
111 "'" Y2. Thus, Eq.(19-18) may be written
I
i
I
I
(19-20)
21/1
+ Y2
~g
(19-23)
h=~V
(19-24)
l'
1
and
c(
{/ ,Yi
) V
(19-22)
If the height of the surge is small relative to the depth of flow, Eqs.
(19-21) and (19-22) may be approximated,
.
respectlvely, by
=
(YI - yP (VI
y)g (19-25)
2Y1Y
(19-1\))
(19-21)
(19-15)
Y1
561 .
UNSTEADY FLOW
(01
(P)
( ll 1'+ V)2
(y' 1
y)~ (y{
= Y/,
y)g
2y{v
(19-2G)
. Solving the above two equations simultaneously, the unknowns l' and y .
can be determined. The absolute wave velocities of the new surges
can be determined from Eq. (19-16). Thus,"
V,. =
and
V,.'
11
Vl~Yl/
Y-
VI
+ Vy
yr'
(19-27)
(19-28)
When the height of the surges is small relative to the. depth of flow,
Thus,
(19-29)
,)//7/7777//7/tl
(19-33)
Applying
.
. ) (Ut + Y2)g
(Yt- Y2 2"':'=--::::2-'-=-"':"'::"
,
YtY2
-[
(19-35)
and
where c and c' are the celerities of the right-side and len-side suI'ges,
respectl vely.
_
(Vi -'- V 2 P
By geometric continuity,
By hydraulic continuity,
The positive sign applies to the right-side surge and the negative sign
to the left-side surge:
C. Surge Cro:>:>ing a Step. In surge analysis [. step on the channel'
floor may be real or fictitious. The fictiFricUorl;
tious step is, usually assumed to simulate
borrier
the effect of channel slope.. The lliethod
I:
Pro ..'d ed for
of analysis, however, is the same in both
{rlc'ion effect !
cases.
.......--_......:\l.,I-,t~
In previous discussions, the slQpe of
T!
cha.nnels has been assumed small. For
Y, ~VI
I
V;_ y;
channels of large slope, the ana.lytlis of
surge moyement will be illustrated by
l:-3'//?/)'T.r7I'/7/
Example 19-2. In simplified calculations,
Ne9ligible .
{oj
difference in
the effect of slope ma.y be treated by conveloeily head j.J
verting the slope into a series of steps.
Vw~ In this way the c,lmnnel length is divided
t-;~ J il. ,,'t
~.
into a number of reache.3. In ea.ch reach,
I
VZ_Y.ZI'y'-.+-V
Y'
the
fioor is considered horiwntal, and the
~' - VI
! .(
v; __ (
actual drop in floor elevation due to slope
.:h7?777d771d i
I
effect. is represented by a step (Fig. 19-6)
f!);'7//7T/?/hzhw:
placed either in the middle or at one end
IO!
of the reach. The height F of'the step is
FIG. 10-6. Arrival of surge at a
equal to the slope of the' channel multistep rmd friction barrier. (a)
Before the arriva.l; (b 1 after the plied by the length of the reach. This
simplified calculation is usually used in
arrivaL.
'
analyses in which the effect of channel
friction is also considered (Art. 19-5).
When a surge arrives :J.t a step, the result is to produce two new surges,
one traveling upstream and the at-her downstream. Figure 19-6 illustrates the an'ivaI of a surge at the step and the condition immediately
after arrival. Applying Eq. (19-17) to the left-side Ilew surge,
563
UNSTEADY FLOW
562
Example HI-1. A steady flow s.t 6.80 fps and 42.90 It deep is suddenly stopped at
the downstream end of the 9,700 ft reach oJ a. frictionless rectangular channel. Determine the surge thus produced.
Solulion. A reflected suq;e advancing upstrea.m is produced. The velocity at the
downstream end evidently cha.nges from V I = 6.80 fps to V. = O. By. Eq. (19-17) or
(19-20), with Yl = 42.90 ft, the depth or the oovallcing surge y" CllJl be computed as
51.09 f t . '
,
Note that Eq. (W-I7) or (19-20) must be solved by trialllnd error. A simplified
procedure is to use Eq. (18-46) together with Eq. (19-24). Thus, Eq. (I8-4cG) g1\'e;;
c = V32.2 X 42.g
37.2 ips, and then Eq. (19-24) gives h = :17.2 X G.80/32.2 =
7.85 it, which makes 1b = 42.90 + 7.85 = 50.75 it. This value may be used as the
best estimate /Dr the trial-a.nd-error solution oJ Eq. (l9-17) or (19-20).
By Eq. (19-10), the wave velocity V..;
35.6 fps. The time requir~d for tb.e wavefront to arrive at the upstrea.m end of the reach is 9,700/35.6
272 sep_ The celerity
is 35.6 + 6.8 ... 42.4 Ips.
Example 19-2. A gradually va.ried steady flow is decreased a.t the lower end of an
inclined rectangular channel by the sudden parti'a.l closing of a gate. Describe!!.
method for determining the. surge thus produced (also. see Art. 19-5)_
"
Solulian. Referring to Fig. 19-7, the channel has a slope S. and carries a steady discharge Q. The depth y and velocity V of the steady flow at any. section, say nn, can
be computed.
.
For the determination of the surge due to sudden decrease 01 flow at section '00, a
step computation may be eI)1ployed.' The channel is divided in~ 9.- number. of
reaches, each ha.ving a length equal to t.:r.. The computation may start at section
00 and. then proceed by steps from reach to rea.ch. Consider the step computation
from section n'n' to section nn. The flow condition at section 11,'.,.' is assumed to
ha ve been determined, and it is now requifed to determine the l3urge height h. and
velocity
at section nn
'
(19-30)
.v...
I
i
! I;
'~
1-
554
I
(1~37j
where 11 is the depth iUld l' is the velocity of flow at section nn. If the surge height
is re\ll.Uvely small, say h.;fi, a.I!I, the above eqtnltion may be replaced by
, :
(19-38)
The time required for the smgc to move from section n'1l."to nn. mILY be <:lxpressed as
(I(l-39)
where
(19~40)
)
\
~~~~ZXZ'<~~lola'
I
I
,I
Thus,
, W '" t;Q AI
(19-4.1)
where t.Q is th~ reduction i'[j discha.rge clue to partial closing of the grlte.
El i~linating 1'.. and' At
the abo"e thre~ equations, the velocity of surge may
be written
'
!I.
A~ Ax
_ l',/
(19-42)
he
I,
II
i
of
(7.00
= y,
V:)2
+ 2.3&.
(41.65
(48 71
)2
(53.47 + Yl)g
(0 - Y , ')'- y", 102.l8(YJ + 2.38)
Ys!l~ =
Then, compute the values of Yv by Eqs. (19-31) and (19-42), respectjve\y. If the
two values of V .. so computed do not agree,assume another 11. and repeat the computation, until the two values of V", agree ~nd hence the correct values
Ii. and V'" Me
obtained. To start Lhe step computation at section 00, Eq,. (19-36) may be used to
replace Eq. (19-42), since in Eq.(19-36} the va.lue of AQn at section 00 is equal to IlQ.
Example 19-3. A step with F= 2.38 It is located at the upstrea.m end of the reach
considered in Example 19-1. Determine the flow condition immediately after the:
sru.ge arrives at the barrier. The flow upstream from the reach llUder consideration
has fl, sk;ady velocity of 7.00 fps'in the downstream direction shd a depth of 41.65 ft.
Solt,tUm. At the step the sur'ge produced in Example 19-1 is spli't into two com: ponent surges, one tra.veling fa,rther upstream and. the other, II- fe-fleeted surge, travel,ing dowllstream. For the surge tra.veling upstream, Eq. (19-30) gi\'es
(19-44)
V,,_
565
-y
. i
"
UNSTllJADy FLOW
'l,
Th~s, W may
determined by geometry,
horu;ontal water surface in the channel after the passing' of the surge.
V,'(Yt
+ 2.38)
(19-45)
(19-46)
666
KE.
wV 'YT'
2 g
height~
(19-47)
where y is the depth of water snd V is the velodty of flow. By Eq. (1,9-23) or Eq.
(HI-19), !liS the case may be, the shove equation, may be reduced to
'
l{.E. = 'Ulh'gi(
2c'
Assuming
(19-48)
of y. By the momentum principle, the corresponding change in hydrostatic pressure should be equal to the force required to change the
momentum of the vertical element between y and y + dy. Considering
a unit width of the channel and assuming PI = /32 = 1,
W
Z y2
2'
(y
dy
. V
=
P.E,
+ ILE.
wh'T
(19-50)
r
Y2
Assumed
profile
ofter 2.1 sec
-
-:,,---
Assumed
profile
one: t sec
~,-
Assump.d
surge profile
immed,otely
dy 1
V2
~
[
' ;
l:!
VW1
-<--'
I I...
,
FIG.
-,
VI-
YI
)//.1,/'//'yhhw/7/,//?//mJ/?,.;;:v/.?7
one on top of the other. The velocity of the small wave a.t tho point
under consideration may be expressed ar'.()ording to Eq. (19-11) as
Vw
FIG,
19-8. Propagation of
II.
vgy -
(19-53)
steep front, it will soon flatten out as the surge moves through the
channel (Fig. 19-8). If the height of the surge is modemte or small
compared with the depth of flow, the equations derived for a positive
surge can be applied to determine approximatdy the propagation of the
negative surge. If the height of the surge is relatively large, a more
elaborate analysis is necessary as follows:
Figure'19-9 shows a type D surge (Fig. 19-2) of relatively large height,
retreating in an upstl'eam direction. The surge is caused by the sudden
lifting of a sluice gate. The wave velocity of the surge actually varies
from point to point. For example, V .. is the wave velocity at a point
on the surface of the wave where the depth is y and the velocity of Jlow
through the section is V. During a time interval dt, the change in y is dy.
The valu~ of dy is positive for an in,crease of y and negative for a decrease
...... !'--: .. ; .. -.
. I
.1
w
. ,?X?i/,-//?7-0?V/.?7/,T/.//.//?
ofter creation
-- -=- - - --
",
, f
___ ~w~_
---
- r
(19-52)
. g
~~--=-.dV
(19-51)
c=-../iY
567
UNSTEADY FLOW
where w is the UlliL weight of water, T is the top width, and h is the surge
The kinetic energy of the element is evidently equal to
'
V .. 2
...}gY2 - Vz
(19-54)
When the surge is not ,too high, a straight-line relation between V"'l
and V",2 may be assumed. Thus, the mean velocity of the wave may be
considered to be
.
(19-56)
Now, eliminating V", between Eqs. (19-52) and (19-53),
!!JL
--/y
dV,
- ...}Y
(19-57)
569
RAPIDLY VAlUED UNSTEADY FLOW
568 .
UNSTEADY FLOW
Velocity
----2.r .heod
",
V",t
i
I
= :;
.ygy-; -
2 'VgY2 - V 2
I.
(Vw2 - V~'I)t
Profile of
the origiMI
steod Y flow
. -
--...
I
;
Iplont
- .../~
tal
---~-~--h-l-e-
v.+
-------jd
)(
E:tample 19-5. Show that the equation of the wa.ve profile, reSUlting from the
failure of. a drun is in the form of
x = 2t
v'iiii. -
31
[bj
~cenler line
I,
)
I
I
1
I
/.
tel
of
'I ::.:-
,Rolf wove
...,. V.... i =2.,..I9Y2,
'19-5. Surge in Power Canals. Engineers are interested in the determination of the maximum stage of water that could be developed. 8J.I a.
_--.=-....I
l'
Maximum .tage
01 th'l rejection
\1U,Q9
~~
~
(;;""
Equation (19(i2) represents a. par~bQla with:vertical axis and vertex on the' cha.nnel
bottom, as shown in Fig. 19-10. At the site of the dam, x = 0 and the depthy.
4y./9, Qwing to the channel friction, the actual profile takes the form iudiCl!.ted
by the dashed line. . This profile has a rounded ,front at the downstream end, forming a
bore (see Example 18-1). At the upstream end, the theoretical profile thus developed
has been checked satisfMtorlly with experiments by Schoklitsch[121.
.
r--------~
--
l"XG.
V=0-+1
/
~-l.v
'it:,
Ihe dam
L.-!I
":..
Jm
Vmax--""'"
.....
Yz
-vgy
v.=o
-vIiii .
where::t: is the distance measured from the da.m site, y is the depth of the wa.ve profile,
lit is the depth of the impounding wa.ter, and t is the time a.fter the dam broke.
Solu/ian. SiuGe the impoundiug wa.ter
velocity, or V. = 0,. the. wave
velocity by .Eq. (19-59) is Y", = 3
2
Since V 18 is in the negative direction of z, x = - V"I, which gives Eq. (19-62).
;C~:.=lW/&~~
---..
of
To power
.......Jr,..
~;
(19-61) ,
The
. type C.
Reservoir.
(19-60)
Let l be the time elapsed since the surge was created or, in thif! easel
since the sluice gat.e 'was opened. At t = 0, the wavelength A O.
After t sec, the wavelength is equal to
, \
\
!
FIG.
.
.
. ' on sur' e advancing upstream is produced, :
is suddenly thrown off, a reJectl d' g t
sual observations, the water
as shown in Fig. 19~1lb. Aceol'
0 +~s approximately level. Thus,
surface ab downstream from the wave ro~" the water surface throughout
when the wavefront reaches the re(s;:volf
). However, a steadily
11C
the entire canal becomes level
Ig.
m1.
i9_
570
UNSTEADY JrLOW
571
increasing volume of water must pass through the wavefront to fill the
space corresponding to abc. Because of this action and the effect of
friction, the wavefront is being greatly reduced in height and velocity
on its way to the reservoir. Furthermore, as the surge reaches the
reservoir, the velocity of ft.ow in the canal must show a progressive
increase from a maximum at the reservoir to zero at the downstream end
where there is no outlet for relief. As a result, water will be built up
to a maximum stage at the downstream end, as shown in Fig. 19-11d,
sometime soon after the arrival of the surge front at the resenroir.
For the determination of the maximum stage produced by the rejection surge,l Johnson [14,15J has applied an analytical method 2 to the
design of the supply canal of the Queenston-Chippewa plant in the
vicinity of Niagara Falls . . In this example (Fig, 19-12a), the entire
length of the canal is divided into four rea.ches each 9,700 ft long.
At the very beginning, t = 0, the flow is steady, and the flow proftle
can be computed. The drop in water-surface elevation in each reach is
indicated at the division section between reaches, for exampte, 1.13 ft
for reach 4in }I'ig. 19-12b (see Example 19-3). The water surface in each
reach is assumed leveL Note that the friction in each reach at the steadyflow condition is considered in the computation of the flow profile. The
slope effect of the channel is represented by a drop on the channel floor
at the division sec Lion between the reaches, for example, 2.38 ft for reach 4
(Art. 19-30). The water depths ill feet are represented by the figures
wriHen vertically. The velocities in feet per second are r.epresented by
the figures written horizontally. In reach 4, for example,. the depth is
42.90 ft and the velocity is 6.80 fps.
When the load is suddenly thrown off, a surge is developed at the
dowllstream end Qf the canal (Fig. 19-12c). The computation of this
initial surge has been shown in Example 19-1.
After 272 sec have passed since the surge was produced, the surge
arrives at the barrier at the upstream end of reach 4 (}I'ig. 19-12d). The
water depth throughout reach 4 is 51.09 ft; and the velocity of flow is
zero. At the barrier, the surge is soon split int.o two component surges,
one traveling upstream and the. other downstream. The computation
for these surges has been given in Example 19-3.
The component surge traveling downstream will soon reach the downstream end of the canaL After it reaches the end, it will be reflected.
The reflected surge will travel upstream again.
The component surge traveling upstream will soon re:ach another
friction barrier farther upstream. At the barrier it will split into two
components, one traveling farther upstream and one traveling down-
"'"
......
-1'>1I
~.
.1.
ci'"
1'1
I""l
;-,--fr~-~-'--IF----~- --ir=-----~-~r
. ~I
1
I,
(b J
-.
I"'l
~I
~I
'
' ';':.;;''
'"
::.11
';1
'"
oon
"'I
"!,
...,
;;;
<:$
l-
;
,..,
'"
ao
'"ci
<D
oJ
".
~!,
ci
~I
N
7.00
'!l
0
1
--~--'''''~I~.---~~-~--~,--~~--~
7.20
"'I.
.n
27<' sec
.,.
.oj
.....
"'
ci
ci
.-
.,
'",..;
"!
,..,
.,.
0
N
0
"'
ci
Velocify
10
oJ
_....,
.D
_.
recoIJe ry
/1~ .1.
-1
oJ
oJ
on
S~
7. 65.mox,. rrH
I,B51
sec
III
will {ollow' accordingly until th8 maximum stage is obtained l1.t the downstream end of the canaL In the example given by Johnson, the maximum stage is found to be 7.65 ft above the reservoir pool level (Fig ..
19-12e), This will be reached 1,851 sec after the load is rejected.
It is apparent that the Johnson method is very laborious. The proc.edute becomes very involved as the computation conHnues, for numerOWl surges will be produced an.d propagated.
The surge in a canal can
1 \
572
l
I
U~STEApy
FLOW
l For !tn analysi:; of surges deve]opedby partially or fully opening a !:niter gate,
see [27].
.
573
! '
1 Simllar methods with further approximations are used by tIle U.S. Army Corps 'fJf
Engineem [28].
:
.
574
Col. 3.
Initial velocity of flow in fps before the incremen.tal surge arrives, equal to
maximum height of the surge is found to be about 2 ft, which is'generally required in
<design for the determination of the freeboard and tho water pressure against the channet walls. The surge height at the en.d of the equalization. time should be zero_ The
value -0.01 as complJted is dlls to the approximation of< the slide-rule computation..
<
TABLE
19-1.
COMPUTATION OF SURGE
PROFILE
o 0
a
15.00
12.5 0
0.83 15.00
25.0 0.83 1.(10 15.57
37.5 1.50 2.33 16.11
50.0 2<33 3.01 16.63
46.9 3.01 2.74 17.12
43.8 2.74 2.5916.92
<40.6 1 2.592.4216.81
37.52.422.2416.69
34.4 2.24 2.08 16.56
31.3 2.08 1.90 16.45
28.1 1.90 1.72 16.32
25.0 1.72 J .55 16_19
18.8 1.55 1.17 16.07
12<5 L17 0.79 15.80
6.3079 0.41 15.53
0.411015<17
0
____~
1 __~__~.....
Col. 5.
Col. 6.
_~
o
600
0
15.00 22.0 22.0
0_57 I 570
0.57 1I:L57 22.4 22.4
0.54 15.11 22<8 23.6
1.11 540
0 52 16<63 23.2 ~4.8
1.63 510
2.12 480
0.4917.12 23<5 25.8
-0.20 16.92 23_3 26.3
1. 92 450
--0.11 16.8123.326,0
1. 81 420
-0.12 Hi,fi9 23_225.8
1.69 390
-0.1315.5623.1 1 25.5
1. 56 360
-0.11 16.45 23.0,'25.2
1.45 330
-0.13 16.02 2:L'9 25.0
1. 32 300
-0.13 16<Hj 2~L8 24.7
L 19 270
1.07 240
-0.12 16.07 22.8 24.5
-0.27 15<80 22.6 24.2
0<80 180
-0.27 15.53.22.4 23<6
0.53 120
0.27
60
-0.26 15<27122<0< 22.8
-0.01
a
-0.28 14.99 O.
0
__~_ _ _~
113,200
12,800
12,700
12,500
12,400
11 ,800
10,900
10,100
9,200
8,300
7,500
6,700
5;900
4,400
.L-__~__
'"'
.,
8.000
!r
6.000
~~
(12)
(11)
12.000
II
10,000
IN A NAVIGA'l'ION CANAL
.l:'!
30
60
90
120
150
180
210
240
270
300
330
360
420
480
540
600
575
UNSTEADY FLOW
'5U>
I . ,
4POO
i
4
6
5
Time i., min
{a)
.!: 2
I' - -
-- -.J...--
V
o ~~
o
2,800
1,400
Initial depth in ft, equal to the final depth y, of the previous step
Height of the incrementa.! surge in ft, computed by Eq. (19-19) or
-r-~
<-<
5,000
......
;---
------
-.:;;..;;-
-T
Ii
}.
.la, 00 0
D1stom:e in ft
j
<'
lb)
(19-63)
As an approxima.tion, tlie value c (col. 8} of the previous step is used in this equation.
Aiter l = 150 sec, V, becomes less tha.n V 2 ; tha.t 15, the nega.tive incremen.tal surge
occurs. Thus, the computed his TI".''''.I''''_
Col. 7. Final depth of flow in ft, equal to the iniLi"l depth in col. 5 plus the height
of the incrementa! surge in col. 6
Co\. 8. Celerity in Ips, co}nputed by Eq. (19-9)
Col. 9. Absolute velocity of surge in fps, computed by Eq. (19-16)
Col. 10. Cumulative. va.lue of the height of in.cremimtal5urge (coL 5)
Col. 11. Travel time in sec until the discharge stops, equal to the equa.li~ation time
(600 sec) minus the time in col. 1 since the discharge started
Col. 12. Trovel
in ft for ea.ch increment.aLsurge until the discharge stops,
equal to the product of the time in col. 11 and the absolute velocity in col. 9
Plotting the travel distance against the cumulative height of the incremental
the whole configuration of the surge is shown in Fig. 19-13b. This is
in the downstrea.m caDal after a period equal to the equalizatiou time is
.(
j
I
i!
576
UNSTEA.DY FLOW
T Ih1Cl = T 2h 2 c2
+ T 3haca
(19-64)
h",
(19-69)
(19-70)
and
1I
I
I
?WT/T/ffffO?7il
,
( oj
7-1"77. .7;;'-:7,///;;.??/~
Ii>l
fa?WJ;//.i7//h//Y"';;'l
~Q...:''''a'7?/rr:/.''""z'''''/;;'''''/$,
'
n
n
n
<1
11.
>1
19-14a),
hS)Ci = T 2h2c2
h3 2)Cl = T2hz 2C2
(19-66)
(19-67)
11
Reflection
coefficient
Reflection
Cr 1
Positive
Partial positive
Or > 0
None
Or - 0
Partial negative
Or < 0
-1 Negative
Or
n= >
T1(h i
TI(h 12
c.; so
(19-72)
(el
""
"
Width rfl.tio
577
Chaullel condition
A dead-end barrier
Contraction
No tmnsition
Expansion
Entrance to a lake or sea
....;gg-
h2
'""
L,
""~~'---"
ThuB, h.
h. h. - O.11B ft.
By Eq. (19-71), Or 0.333.
.2
X 300 X
?OO + 150
0.353
"
= 0.471 it
578
UNST'EADY FLOW
Forebay
579
The disc}targe entering the canal is I:1Q, =c,T.h. = 17.0 X 150 X 0.471 = 1;200
ds . . The reflected discharg;: is D.Q, = 1,800 - 1,200 = 600 ds.
.--_F_o_re_b_o-'-y_ _ _ _ _ _ __
-.-~--~---+I----~~--..:~S-j~:-n~-c~
FIG. 19-15. Analysis of surge through the trunsitio!l between forebay and canal.
J. Kozeny [31J.)
the two adjoining channels 'with celerities (:2 and /;3, respectively. Assuming that surge heights are small compared with the depth of water ill
the chaunels, thB celerities can be computed by Eq. (18-'16). In the
meantime, a reflected surge will travel along channel I at a celerity
hpproximately equal to Cl. Figure 19-16b shows the condition at the
junction immediately after the arrival of the incoming surge.
(Afler
_I
II
v31r-c,
Now, if the tmnsition between the callul and the lock is assumed gndual, as shown
in Fig. 19-15, the computation can be mude by letting h. = 11.,
dh, and c.T, =
cIT,
d(c,T,). Accordingly, Eq. (19-69) may be written
Ch~nnel
II
V4
-f=
Woter orea=A2
V3=O
h3=O
Channel
Weier arHo=A,
CI
(10-73)
hi
= const,
Vc';;r.)
= canst
(19-75)
(b)
FIG. 19-16. SUrge at channel junction. (a) Before arrival of surge; (b) ufter B.i:rival.
of surge.
Since the surge heights are assumed small compared with the depths of
flow in the channelS, Eqs.(19-1 9) and (19-2'1) can be applied. Thus, the
following equations ca.n be written for channels I, II, and III, respectively:
With T,lT. = 2, the above equation gives hol"l = 1.414_ . Thus. h, = 1.41411., =
0.499 It and h, = 0.146 ft. The corresponding C, = 0.413, I:1Q, = 1,~70 cis, nnd
t.Q, = 530 cis.
This p.xample illustrates t.he application of the theory of surges to the hydraulic
st.udy of ship lock and forebay. Such applications have been described by many
hydrnulicians, among t,hem Dantscher [29,30] and !{o~eny [31J.
Chonnel I
Woler area = Al
(a)
(19-74)
Channel [
Woler orec=A,
-1
For channel I
hi - h = ~ (17 4 - V 1)
(19-77)
.~
V1
(19-78)
h = ~ V2
(19-79)
h ..:..
-
(19-80)
. For channel I
hi
For channel II
-C3
V3
A.V. + A3V3
(19-81J*
where
~!lAi
let.
I = A1
C1
(19-82)
+ A2 + A3
C2
(:3
(19-83)
-.J
580
UNSTEADY ,FLOW
,
(19-85)
581
582
velocity, that
or
UNSTEADY FLOW
when
L5V ~ V
Vic ~ 2
+c
(19-86)
n9-87)
Since Vic is equal to the Froude number and since roll waves can be
formed only on supercritical slopes) the range of Froude numbei' for the
formation of roll waves is 1 & F & 2. Solving
(6-43) for S, it can be
8how'n that 8 = 3F2/R. Thus, the range of slope for the formation of
roll waves is 3/R :::2 S ~ 12/R.
Slug flow can beformed when the surface velocity is great.er than the
wave velocity. Theoretically,
the formation of slug flow will
occur when F > 2 and S > 12/R.
PROBLEMS
19-1. Prove
(19-11).
19-2. Prove Eg. (19-15).
.
19-3. Assuming the energy loss in thp. moving hydraulic jump t.o be negligible,
Koch an'd
l27J have derived the following equation ior the wave velooity
or.a. surge;
(19-88)
where the positive sign is for type.A and C surges and the negative sign is for type B
and D SUrges (Fig. 19-2). Verify this equation.
.
19-!!. Assuming atl initial depth of 20 ft, an initia.l velocity of 15 fps, a.nd a rectangttlar channel cross section, compute the velocity and overrun of thE'. bore shown in
. Fig: 193 and estimate the distance trs.vel>ad by the bore in 7 min.
19-5. Compute the height", depths, and wave velocities of the two component
surges produced in Example 19-3.
.
19-6. A positive surge 0.63 fthigh and 51.72 ft deep is moving ill still water 5l.0a.
ft deep with a wave velocity of 41.70 fps toward the daRd end of a channeL Deeermine the height and wave velocity of the reflected S'Irge after the original surge hits
the dead end.
19-'1'. Solve Example 19-1 if the original steady flow has a velocity of 10 fps and a
depth of SO ft .
. 19-8. Solve Example 19-3, if the step has a height of F = 3 ft.
19-9. Two positive surges 3 and 2 ft in height, rc~pe~tively, move in opposite
directions toward each other in a frictionless borizontal channel where the water is
initially stationary at a depth of 30 ft .. The high surge has a wave velocity of 30 fps
and the low surge a wave velocity of 20 fps. Determine the flow condition after the
surges meet.
. 19-10. Solve Example 19-1 if the chanuel has a slope of 0.01. The initial steady
flow is assumed uniform.
19-11. Show that the celerity Iorrn4la in the form of Eq. (18-45) applies only to
positive surges and that the corresponding celerity formula. for negative surges is
(19-89)
583
.,;gy;,
lS-12. Referring to
shotv that the crest velocity V ... =
that
the trough velocity Vw!
thl3.t the velocity of flow through the !lam site
is %
and that the
through the dam site is %7 yg !l,li.
19-13. Contmuet.he
oi the example illustrated in Fig. 19-12, Bnd
determine the flow conditions a.t 420, 504,552, 743, and 767 sec, \spectively, after the
load was suddenly thrown off.
19-14. Soive Example 19-6 if the equllJi2'ation time is (a) S min and (b) 12 min ..
The dil!charge from the lock increase. linearl)"from 0 to 10,000 cfs for the first 20% of
the time and then decrease$ linearly to 0 for the remaining part of the time. Study
the effect due to cbange in eqllalizati'm time.
19-16. Solve EX!l.J11ple 19-7 if the discharge is 1,000 cis.
. 19-16. Three horizontal frictionlE'lis channels '}, 6, and S ft wide meet 'at a jnnction.
The W!l.ter in the channels has an initicl statiana.ry depth of 10 ft. When an incoming surge 0.4 ft bigh approaches the junction fro~ the narroWI..'St ci]!l.nnel ddermine
the flovi condition after the surge enters thl:! junction. Compute the relle~ticn lloeffi.
cient of the junction.
vYY.,
REFERENCES
l
l'
584
15.
16.
17.
18.
1
\
19.
ZOo
Zl.
22.
'23,
24..
25.
UNSTEADY FLOW
!:
t.
I
iI,
II
I
1
585.
26. Josef Frank: "Nichtstationiire Vorgange in den Zuleitungs- und Ableitungskaniiien 'von Wasserkraftwerken" ("Unsteady Flow in Headraces and Taiimces
of Hydropower Plants"), Springer-Verlag, Bcrlin, June, 1957.
27. Alexander Koch and Max CllrstanjeLl: "Von der Bewegung des Wassers und dabei
auftretenden I{riiften" (" Movement of Water and Associated Forces"), SpringerVerlag, Berlin, 1926, pp. 132-150.
28. U.S. Army Corps of Engineers: Hydraulic design: Surges in canals, Civil Works
Construction, Engineering j{amtal, March, 1949, pt. CXVI, chap. 6, 13 pp.
29. Kll.Sper Dantscher: Wfillderwellen in Schiffahrtskan1ilen (Tra,eling waves in
navigation cp_naIR), Wasserkraft 1md Wasserwirtschafl, Munich, voL 35, no. 7,
pp. 145-147, July 15, 1940.
.
3Q. Kasper Dantscher: Die Wanderwelle in Schiffahrtskanal (The traveling wave in a
navigation canal), Wasserkraft und lVasserwi?tschaft, Munich, vol. 35, no. IO,
pp. 226-229, Oct. 15, 19'10. .
.,
31, Josef Koteny, "Hydraulik" ("Hydraulics"), Springer-Verlag, Vienna, 1953,
pp. 263-265.
.
32. Francis E, Swain: Determinat.ion of flolYs in interconnected estullrine (:hannels,
U.S. Bureau of Reclamation, 2'echnical lYIemora71dt"'~ 640, February, 1951. .
33. Francis E. 'Swain: DctcrmintLtion of flows in intercoIlnected estuarine channels
produc~d by the wmbined effects of tidal fluctuatioM and gravity flows, Transactions, Anw'ican Ge1tphysi~al Union, voL 32, no.' 5, pp. G53-672, Oct.ober, 1951.
34. D. J. Hebert and F. C. Lowe:. Progress report on model studies of the Sr.cramentoSan Joaquin Ddta, Central Valley Project, California, U.S. B"reau of Reclamation,
HvdralLlic Laboratory Report No. Hyd-142, Apr. 10, H)44.
35. D. J. Hebert and J. E. "Warnock: Skeleton outline of a plan for developing t.he
Delta-Mendota Irrigation Water Supply and Salt-Water Repulsion In the
Sacmmento-San Joaquin Delta Region, 'Central VaHey Project, U.S. B1LrealL of
Reclamation, Hyd-rUliliee Laboratory Report No. Hyd-145, July 10, 1944.
36. Cherry Creek Dam arid Reservoir: Report or'model Gtudies, spillway and stillin::o;
bn.sin, Hydraulic LaboratDry lteport No. Hyd-146, prepared by U.S. Bureau bf
Reclamation for U.S. Army Cofps of Engineers, July, 1944.
37. R. E. Glover, D. J. Hebert, and C. R. D9.um: Application of an hydraulic problem,
in Electrical am.logics D.nd electronic computers: A symposium, T'ran8act'ions,
American,Society oj Civil Engineers, 'vbL, UB, pp. JOIO-101G, 1953.
38. Harold A. - Thomas: The propagation of waves iIi steep prismatic conduits,
Proceedings of Hydraulic Conference, State University of IOlVa; St1tdies in Engineering, Blllletin 20, March, 11'10, pp. 214-229.
39. Paui G. H. Mayer: A study of roll waves and sIng flows in inclined open channels,
doctor"l thesis, Cornell University, Ithaca, N.Y.,' S~ptember, 11)57.
."
FLOOD ROUTING
CHAPTER
20
FLOOD ROUTING
586
f
I
ill
'
/
587
J)fh,J~~
4~J? ------7'
IF
ay
a;r;
C j.;}1:-~ D aTf
O'n iC-r)I.UJ
a.1;
F alT
+ ~ af!
+ Tf
a.);
!I
ay d.-r;
ike
at
+. i!..!L
d- t
d
at!J
aT'd'"- ...L
aT! dt
iI!
So
dY
S,
(20-1)
(20-2)
(20-3)
(20-4)
chan:~e of .depth of flo,v with respect to time, aT' lax is the cho,nge of
velOCity wl~h respe?t to distance, avfat is the
of velocity with
respect to tllne, SO IS the cfll:l.l1l1el slope, .91 is the friction slope, dy is the
totlll change of depth, and dV is the total change of velocity. Note
that Eq. (20-1) is a dynamic 'equation repl'e'3entiug
assuming
a: = 1 and that Eq. (20-2) is the continuity equation identical with
~q. (18-5). lfsyation (20-3) indicates thllt the totnl change ill depth
equal to the sum, of the partial changes in depth due to distance a.tid
~hne) re~pe~tively. Similarly) Eq. (20-4) indicates that the total change
10 velOCIty IS equal to the sum of the partial changes in velocity due to
distance and time, respectively.
IS
..
(
I
588
UNSTEAD-r FLOW
-D(So - S)
+ Dd~ _
J
g dt
------~~~--~
g dt
I'
"dy +..!:dydx
gdt
.~ (dX)2 _ 2 V dx + V
g ,dt
(J
dt dt
(20-5)
d:t
eV
c) dl
2c)
deY
+ 2c)
(20-7)
dx
dt
(20-8)
g(So - Sf) dt
V - c
(20-9)
(20-10)
(20-11)
It CD.n be seen that Eqs. (20-8) and. (20-10) D.l'e identical with Eq. (18-54),
expressiI1g the velocity of wave prcipagation. These equations can be
represellted graphically on an ;rt plane, as sh01vn iIi Fig. 20-1. . For a finite
time increment ilt, the point p represents the position of the channel
section; under: consideration at timet
ilt, and the pOInts u and d
represeJilt, respectively, the posi~ions of c~rtain upstrean\. and downstream ;sections at time t. The. velocity of wave propagatio.n can be
=g(So - Sf) dt
represented by the slopes of lines constructed on. the xt plane. When the'
flow is subcriticaJ,1 as it is in most streams, that is, when V < c, the
slope of the line up, a positive quantity, represe~lts V
cof Eq. (20-8),
8.nd the slope of the line dp, a negative quantIty, represents V - c of
Eq. (20-10). These lines up anddp al'e hereafter re:e.rred to as r.har~cter
..1.S<ZC
" s It is evident that point tt represents
the pOSItIOn of the
.
'II upstl
. eamt
section from \vhich' an infinitesimal sUI~ge, once developed, WI arnve a
(20-6)
589
FLOOD ROUTING
(aJ
[b]
FIG. 20l. Graphical: representation of, :vav~ prop~gation for sub critical flow.
Flow condition at inflow; (b) Row condItIOn In channel.
.
(a)
section p after the time interval 6t. Similarly, point d represents the
position of the downstrea~ section fwm which a surge once developed
. '
will arrive at section p after At.
The flow conditions at the upstream and downstream sectIOns are to be
indicated by the subscripts 'U and d, l'es~ectively. In~egrating Eq. (20-9)
from time t to t
ilt, which interval is the time reqUired f~r the :'[l.ve to
travel from the upstre9iID section to the section ull;del' conSIderatIOn, .
+ 2c =
V"
+ 2c + gSa M
u
(1+-'>.1
- g it
S, dt
(20,12
I+A1
th~
S, dt
(S,,,,
1 Theoretica.lly,
method of cha.ract.eristics is similarly
to
flow .. However, the possibility of disturba.nce due to formatIOn of standmg waves
would have to be taken into acrioun~ [23].
+ 2c = G + K
(J", = V + 2c" + K"
Ie {J(So-2Sf") Ilt-
(20-13)
(20-14)
(20-15)
= . /39,5008 0 Ilt -
.\1
(20-16)
(20-25)
x in ini
I '.
(20-17)
where
(20~18)
and
(20-10)
J(
]( cH
1,"t1O,000n 2 tlt
and
Eliminating
591
. F.LOOD ROUTING
UN,sl'lllilY FLOW
590
'200
A
I.
c'P
'/
I .I
I
\
i
~O.D
I
~~5'
jhrL~q
(20-20)
1'<::--.,J:-'-7p",-+..:.,w.'-b-.~l--+.---'j'
-i
16.6 mi .......
This equation can be used for computing c at the time t + Ilt if the
upstream !l.ud downstream flow cOllditions at the time t are given.
Eliminating c from Eqs, (20-13) and (20-17) and solving for l' V _ J(
Y-K=
Gu
+ G~
(20-21)
+ 2c
(20-22)
\j
1
.I
Substituting this
(20-23)
From Eqs. (20-22) and (20-23J, the values of Y and IC can be determined.
,
The solution for c and V, however, can be simplified by a graphical
procedure. Taking the wide channel as an example, R = y. c~!g.
Thus, Eq. (20-23) becomes
(20-24)
The units in t-his equation are expressed in terms of feet ane. seconds.
Fw
2(~:la..
chamctcristic.~:
plot of c vs. V for K" (Fig. 20-2e), and (2) ~L plot ofT! - J{ against V
using c as a parameter, which is called "the plot of T1
]( vs. V for.
c" (Fig. 20-2j) , The use of these two plots
simplify the problem, .
which is to solve for c and Yat time t + Ilt if the flow conditions c and V
at the upstream and downstream sections at time t are given.
In illustrating the application of the gmphical procedure, it is assunred
(1) that the channel
infinitely long and very wide, (2) that t.he initial
flow is uniform and steady, IJ,.nd (3) that the time interval Ilt is COlllltant.
The data required for the solution include the initial flow condition, the
I.
"vill
is
-1
592
FLOOD ROUTING
UNS'l>EADY FLOW
593
as
x in rnl
:::
.5
.
~
t- +3200
I
x in rni
intervals.
A. Preliminary Com1mtation
1. Compute the initial velocity V" and depth y" by the Mannillg
formula, and the initial celerity by c,.
2. Using Eq. (20~24) or (20-25) construct
plot of C VB. V for J(
(Fig. 20-2(3) and the plot of V - K VB. V for c
30-2f).
3. Construct the referen.ce lines OB and AB
20-1 and Fig. 20-2a).
, Fo~ oth~ types of hydrogra~h, for example, in the form of the discharge q ~ f(t),
a modification of the procedure IS neCeS5ll.ry. For a detcdled disC\I~~iQnl :;;ee [23].
J(1n mi
FIG.
594
595
UNSTEADY FLOW
FLOOD ROU'l'lNG
ill the preceding step, and plot them 011 the ex plane at x = 0 (Fig. 20-2b).
ALso, plot the vahle of V,. on the Vx plane at x = 0 (Fig. 20-2c).
through the time I (Fig. 20-'10.), the values of Cd and Vd are equal to the
values of c and V 2,t time t.
3. From theploi.J of C VB. V for J( (Fig .. 20-2e), find the rca corresponding
to Va and Cd, al1d then compute Gd by Eq. (20-18) .
. 4. Compute Y
I{ by Eq. (20-22) with the value of c obtained froin
the inflow hydrograph (Fig. 20-2(1).
i I
I I
-1
j'
,I
I I
~
2.5
v in
FIG.
.. .
--'4
3.5
4.5
mph
20-2f. Routing by the method of characteristic.s: the 'V - J( vs. l' fore plot.
3.5
Vir mph
C VB.
V for K plot.
-"~
..,', -: "- -
---,--~'"-,~
..
"-,"
.",,;.'
59B
UNSTEADY FLOW
22 -
!
'--
/""
2I
Givert
inflow,,-
'i
2 01----
8----,- l -
.!;
...
16
A. Preliminarycomputalion
/
I
f.
I-
\
1
.J:
0.
-- I---~--
c,
---
: I
/.
14
1V
10
\ -L
\
\ -~,
30
40
50 60
in hr
70
80
90
100
Vgy
= -V0!21.5
=
---,
l"IG.
By this equal.ion, the auxiliftry plots (Fig. 20-2e and fJ are computed ,as shown in
Table 20-1, p. 600.
3. On the xl plane (Fig. 20-2a) construct the reference lines OB and AB, the slopes
of which are 2.60 + 14.0 = 16.6 mph nnd 2.60 - 14.0 = -H.4 mph, respectively.
4. ,construct the inflow hydrogr'1ph (Fig. 20-2g) by the foJlowi~lg equation:
I-----
'
(20-20)
- --
1\
/
20
(20-25) gives
,f--1:~ompvted
outflow
1
15
Example 20-1. An infinitely long and wide channel carri~s an initial uniform flow
of 50 cfs per foot of width.' If a hydrogrnph l'epresented by y = 21.5 - 8.4 cos
"(1l't/48) ft is imposed on the upstream end of the channel, route the flood through the
first 50 miles for a pp.riod of 24 hr. Given 11 = 0.03 and So = 1/5,280.
Sohdion. The computation follows the procedure described above.
597
FLOOD ROUTING
I
I
(20-27)
5. By Eg. (20.27), compute the values oi C Il.t the upstream curl of the channel (at
x = 0) for ,t = 0,3, . '. _ ,24 hr, and plot them in Fig, 202b. Also, plot the value of
V = 2.60 mph for x = 0 and t = 0 in Fig. 20-2c. .
.
R. ComputatioT! Jar flow condiliQn at inflow. The step of computation from t = 12
to t = 18 hr ~vill be t.n.ken for illustration, assuming that the values of c and V along
the channel at t = 12 hr have been determined in a previous step. The following'
procedure rlescribes the determination of c and V at the inflow at t = 18 hr.
1. Draw a trb.l characteristic at point 100 in thp. generlLl dirp.ction of AB (Fig.
20-Za). The slope of this characteristin should be somewhat flatter than 'that of the
previou." step, that is, flatter th~n the slope of the chllructeristic!at 12a, because the
inflo\~ hydrograph CFig. 20-2g) indicntes that c is increasing during the interval from
t = 12 to 15 hr. The downstream end of this 6ho.racteristic is at t: = 12 hr and Xd =
36 miles.
.
2. With Xd '"' 36 miles and t = 12 hr, find Cd = 14.38 and V~ = 2.72 from Figs.
.20-2Q and c. Theae value" are ent~red in eols. 9 !lnd 8, respectively, of Table 20-2.
3. IWEth Cd = 14.38 and V. = 2:72, ,find Kd = -0.04 from Fig. 20-2e, and enter
the value in col. 10 of Table 20-2. By Eq. (2018), Gd = 2.72 - 2 X 14.38 - 0.64, =
-26:68, as shown in col. 12 of Table 20-2.
'
1
598
4. From tbe inflow bydrograph (Fig. 2()..2g), c= 15.84 for t = 15. By'Eq. (20-22),
V - K, = -26.68 + :.! X 15.84= 5.00, a.~ shown in col. 13 of Table 20-2.
5. With V - K = 5.00 ande = 15.84, find V = 3.19 from Fig. 20-2f and enter
tbe value in col. 2 of Tobie 20-2.
6.Tbe slope of the charoct,eristic should be (2.72 - 14.38 + 3.19 ~ 15.84)/2 =
-12.16, which gives Xd = 12.16 X 3 = 3.6.48 miles. This value is dose to x~ = 36
miles for the trial characteristic. If f,he agreement is not dose, a new trial characteristic must be drown and the computation repeated.
file do not change. In other words, the flow mfl.y be considered \)niformly progressive
aiterthe flow conditions at points a and.b are determined .. This may be explained
in Fig. 20-3 as below.
Oil thE; ex plane, draw a hor'izolltalline passing through the / = 9 hr curve at x = O.
This line intersects the t = 12 hr curve at x = 13 miles, Similarly, a horizontal Jine
I
"j
1. Draw two trial cha.racteristics at 15& in the general directions of OB and AB.
Note that tile characteristic in the direction of OB starts from the point 12a. Thus,
x. = 0, x = 52 miles, and x~ = 86 miles.
2. From Figs. 20-2& and 0, find c. = 15.22 and V. = 2.96 nt x" = 0, and Cd = 14.02
and V ~ = 2.59 at x" = 86 miles. Enter thes"l values in cols. 5 to .3 of Table 20-2.
3. From Fig. 20-2e, find Ku = -1.08, as shown in col. 7 of Table 20-2, for t:u =
15.22 and Vu = 2.96; and also [{~ = 0, as shown in col. 10, for Cd = 14.02 and
1'd = 2.59.
4. By Eq. (20-14), G. = 2.95
2 X 15.22 - 1.08 = 32.32, as shown in col. 11
of 'Table 20-2. By Eq. (20-18), Gd = 2.59 - 2 X 14.02 = -25.45, as shown in
col. 12 of Table 20-2.
5. By Eq. (20-20), c = (32.32 + 25.45)/4 = 14.44, as 'shown in col. 3 01 Table 20-2.
6. By E<\.. (20-21), l' - K = (32.32 - 25.45)/2 = 3.44, as shown in col. i3 of
Table 20-2.
7. From Fig. 20-21, find V = 2.75, as shown in col. 2 of Table 20-2, for c = 14A4
and V - K = 3.44.
8. Use the values' of c and V to check the slopes of the trial characteristi~s. Thus,
x = (2.75 + 14.44) X 3 = 51.57 miles and Xd = -(2.75 - 14.44) X 3 + 51.5'7 =
86.64 miles. The trial slopes are satisfactory. Otherwise, new charactc:ristics should
be drawn and the computation repeated.
9. Proceed with the computation step by step, as shown in Table '20-2.
10. Draw a vertical line
x = 50 miles on the ex plane (Fig. 20-2b). The linll
intersects the curves, giving yalues of c at various times. Plotting these values of c
against time, II. hYdrograph at x = 50 miles for a period of 24 hr is obtained, as shown
in Fig. 20-2tp. This hydro graph call be converted to y = J(t) by means of the relation
599
FLOOD ROUT1NG
UNSTEADY FLOW
!
. \
at
y = c'/g.
It may be noted that, after the Bow conditions at points (l and b for each time.interval are' determined, part of the curves on the ex plane and Vx plane for ~he first 50
miles or so can be tmced in approximately and then revised or refined by further
determination of more points, such as c and d, in the channel. It can be seen that the
graphical procedure offers a practical solution of Massa.u's equations; Nevertheless,
the computation involved is stilt complicated and requires practice for routine applications. For accura.te solutions, the time interval .c.l should be reduced, all graphs
should be prepared on large scales, and the computation should be carried out to
sufficient significant figures. For Table 20-2, the computation was done on a slide
rille, since the example was given only for illustrative purposes. The values in cols.
7 and 10 were roughly estimated from the curves; they are not accurate even tD the
decimal places showD. For accurate computation of K, Eqs. (20-15) and (20-19)
should h~ve been used.
;
In this particular example of a long uniform channel with init.ially uniform flow,
it may be assumed that the values of c ~nd V ata particuli1r section of the wave pro-
-':".:
ee'
1
l.
)
.600 .
UNSTEADY FLOW
1J.~1:l ~ is need~ only at point (J, at the inflow a.nd at point 11 in the channel for each
tIme mt~rvaL The va.lues of C IUld.V a.t ot~er points in the channel can be detcrmin.;d
by the SimplIfied procedure shown by the dashed linea in Fig. 20-2a.
20-1.
TABLE
I)
I
K
= . .,..7
'_~
____
1.0,1
0.01
I\
-0.5'1'
L' )
-1.0
-~.51'
I'
r)
-~.O
-2.5
I,
-3.01
)
1
0.0750
0.0166
0.0776
0.078,1
0.0192
0o.'OO~O~~g
u
00816
TABLE
20-2 ..
V [: -
= 47.2
Iv - K
V F -
c~;
c~~
54.3
61.6
;-;;-3j2.0;13.54 Z.54
3.093.09
3.13 3.63
3.16 4.16
3.20 4.70
\3.Z3 5.23
'3.26j 5.76
3 29 6.29
3.62!
3.66
3.70
3.74
3.78
3.82
13.851
1(,
---1----
3.62
4.16
4.70
5.83
&.38
5.24
5.80
5.78 4.34 6.34
6.32 14 .39 6.89
6.85\4.4317.43 5.0S
I
V
.
I
.'
C.
'K
~~I~!~
IT d
(8)
Cd
6.93
7 All
8.03
Kd
I
1
Gu
Gd
I'
30'64'1'--24'4?
I' .....
"I' ....
')
)
)
)
)
.~1
~~Ii!.!l_~, __--,--
~b ~.611406-0.1026\)[14.31-0.48~.5814.00
cH
xl
= 40,3
--
3tl
6a
Bb
\.
.2O-i
OF CHARACTERISTICS
)
\
di
2.53, 1.53
2.58 2.58
1
262,3.12
2.64,3.64
j2.61' 4.17
2.701'4.70
\2.73 5.23
2.75 5.75
1
j)
=~31~!-'-'[-C-=-1-6,-,-I-c-=-1-8,-,I-c----2-0-,--~""-2-2-,
3810
])
FOR
COMPUTATION
ai'59 "T?5':~
1::::~6'
-0:;433:0811
FLOOD ROU1!ING
{)01
Ohio; (2) movement of the 1947 flood through the confluence of the Ohio
and Mi;ssissippi rivers at Cairo, Ill.; and (3) movement of the 1948 and
1950 flood waves in the 184-mile long and na.rrow Kentucky Reservoir
on the Lower Teunessee River. The second problem is the same as that
described in Art. 11-10 and Fig. 11-] 5 except that the flow is treated as
unsteady in the present problem. .The computations for floW profiles
were thus made for times t = 0, 2.5,4, 10, and <:tJ hI', after the beginning
of the flood 50 miles up the Ohio River (see Fig. 11-15). The flow profiles
for t = 00 are identical with those computed for steady flow.
Solutioll of practical problems by the numerico.l method is generally
very tedious. In order to speed up the computation work, the Corps
has employed an electronic digito.l computer, such as the RemingtonRand UNIVAC I (Universal Automatic Computer I). For small quantities of work, however, the Corps has found tha.t a po,rt of the computn.tion can be performed less expensively and more efficiently by 3.
portable computer, such as the Burroughs EIOl desk-size computer;
the use of larger machines is justifiable only for very involved problems.
20-3. Method of Diffu.sion Analogy. An approximate hydraulic
approach to the problem of flood routing in natural channels has been
deyeloped us-illg the ciD.ssicn.l stc.tisticl).l theory of flow diffusion [35].
According to this theory, IL differential equation may be written for the
diffusion of an unsteady flow of particles as follows:
(20-28)
2.72
17
9.
2.61
;:;~
4.04
3.08
5,00
3.44
5.29
93
53.65
4 .'0 4
'~~l ::;~
where Nis the number of p:1l'ticles, t is the tim9, x is the dista.nce, and
K is a coefficient known l),S ditrw~b}ity. When the pl1rtides are flOWi11g
in a direction along the X axis, this equation gives the particle distributi()U in the dii'ection of flow as a function of time and position. This theory
is commonly applied to problems of heat tl'il.l1sier r3B]; there Eq.(20-28)
represents Fourier's ge-ne'rallGw of heat conduction, ih which N desigllates
temperature and K is known as thermal d~tJusiuity [37].
.
In natural streams, the disturbances of flow caused by local channel
irregularities have definite magnitUde at any time and position. They
are mixed, dissipated, and diffused as the flow moves along the channel.
In ll.pplying the theory of flow diffusion to the flow of water, it may be
assumed that the diffusion of the dist,urbanc~s is analogous to the diffusion
of the particles. If the over-all effect of the disturbances on flow is
represented by the variation in the flow depth y, Eq, (20-28) may be
written
(20-29)
"
602
J:.,.j-...:...........
603
UNSTEADY FLOW
FLOOD ROUTING
:..!
(20-30)
It is assumed further that the channel is relatively wide and that the
flow on the average is uniform and steady. Thus', the discharge per
unit width of channel may be represented by the Ch~zy iormula1
q = C . ./8 0 Vi,
Substi~uting
(20-31)
'0.\
a2y
ax':
(20-32)
This is the basic differential eqt;{ion for flood }1:ow in nat7);ral streams. It
can be seen that the coefficient of ayi~/ax in this equation depends on the
~ l'esistance and slopt and that the coeffIcient of a2 y/ax 2 depends
on the channel iW'Igularities .
.:.. __ This equation was orig'inally developed by Hayami [381 through an
r
'I
elaborate mathematical derivation.. A similar method of routing based
on diffusion analogy was also presented by Appleby [39]. The value of
diffusivity K in ordinary streams was estima.ted in the range from 10 6 to
10 7 cm 2 /sec. In large rivers, such as the Mississippi River: in the United
States and the Yangtse River in China., the value of K would be of the
order of lOS cm 2/sec.
. A solution of Eq. (20-32) by Hayami results in the following equation
for the propagation of a flood wave:
. ~y - Yn
Yo
,/
= 1 _~' r,./2VJ[i
rV",~; _ X2 _
-y'; Jo
exp l2K
(VwX)'2] d' K
4KX.i
"-j
" 1
(20-33)
where y is the depth at a point a distance x from the upstream end of the
reach under consideration, y~ is the normal depth of flow at the same
point before the flood arrives, yo is the depth at the upstream end, t is the
time, J( is the diffusiv:ity, V", = 1.5V, V is the mean velocity, and X is the
variable.
A fictitious wave having a constant depth and lasting a llilit time
interval tit may be assumed(Fig. 20-4). This wave is called a unit flood.
Elcpsed time I in hr
Elapsed lime I in hr
(a)
(b)
FIG. 20-4.. Propl>gl>tion of Cal a solitary unit lluud Md. Cb) two successive unit Boods.
(After S. Hal/ami [38].)
as the wave propagates dowllstream, its shape gri\.dual1y becomes asymmetrical and fi:J;ttens out.
In routing an actual fiood, the hydrograph of the fiood may be divided
into a number of unit flood hydrognt.phs (Fig. 20-5). All hydrographs
are plotted with the time against the sta.ge. The propagation of each
unit flood can be computed by Eq. (20-33). By the principle of super-
"'-~'
I,
..
'~'
..
.. '.
'.
",",'
..
)
604
I \
UN:;lTEADY FLOW
FLOOD
position, the propaga.ted height of the given flood:is equal to the sum
of the propa.gntod heights of all the unit :floods. This method has been
used to compute the propagation of a unit fioodartificiaUy produced
in the Yedo Rivet ill Japan; the theoretical computations were found ill
good agree,mellt with the observations. In order to simplify and expedite the I'outing pr9cedure by this
method, electronic' analog com!ai
puters have been developed on
the basis of
(20-32) and applied
to several rivers in Japan with
satisfactory results [40-44J.
ROUT~NG
605
where ilS/ilUs the change in storage during a period t::.t, I is the !lver~~ge
inflow during tit, and 0 is the lwerage o1itilow during t::.t. The value of
AS!t::.t is positive when the storage is increasing and nega.tive when the
storage is decreasing. , This equation constitutes the basis. for a ~ydro
logic procedure of routing in which ilt is known as the nmhng PBNQd.
The rate of storage can be plotted against time, as shown in
2O-6b,
from which it can be seen that the, storage iB increasing before the time
'(;l,t which the inflow equals theoutflovr and decreasing after that time.
The cumulative 'area below the
curve (shaded :l1rea in the
figure) represents the volume of storage at !l. time t after the beginning
of the flood. By plotting this volume
against the time, a storage-volume
/
curve can be obtained, as shown in
Fig. 20-6c. This curve hilS a pev.k
representing the maximum volume of
storage that occurs at the time when
the inflow equal:> the outflow.
If the storage is plotted against the
outflow discharge, the resulting curve
will generally take the form of a loop,
such as that illustrated in F'ig. 20-7.
Siorage
As shown ill this figure, the stornge for FIG. 20-7. The storage-outflow reb.a giv~n discharge on the rising (or tionship.
falling) part of the flood, wave will be ,
'
greater than (or less than) the sl;orage c'orrespollding to the condition of
steady fiow. The storage-outflow relationship for the condition of steady
flow is represented by the dashed curve, which is approxinlately !l.t the
average position of the two limbs ot: the loop.
'
The'storage in a channel reach for ullstendy flow depends prim!j.l'ily
on the inflow and outflow discharges and on th~ geometric and hydrauli.:;
, characteristics of the channel and its control features. It can be assumed
that the upstl'eam and downstream end sections of the reach have the
same mean discharge.and storage relationships with respect to the depth:
of flowy. Then the following equations may be written:
(G)
li.1'
lime
Time
I
I
I
I
"'!
[e,
, "
ll~
f
I'
,
ilS
ilt
:t
and
I =
fly'!!
(20-35)
ay"
(20-36)
(20-37)
Si d:. by,:,
bym
S.
(2038)
606
and
UNSTEADY FLOW
(20~37)
FLOOD ROGTING
a(i)""n
(20-39)
Eo b(~)""n
(20-40)
E,=
and
K[X I~
+ (1
- X)O-']
(20-42)
where K
bla"'''' alid x
min. In prismatic rectangular channels,
discharge varies with the five-thirds power of the dept.h on the basis
of the Manning formula, and storage varies with the first power. Since
n = % and m
I, the exponent x == 0.6. In natural channeJs, m may
be considerably greater than unity and hence :t; is larger than 0.6. Mady,
hydrologic routing procedures have been developed on the general basis
of Eq. (20-42). For simplicity and for practical purposes, x is commonly
assumed to be unity.
The hydrologic approach to the problem of flood routing is based 011 the
storage-discharge relationship described above. It is assumed that.
dynarniceffects of flow are negligible and that storage is a single-\'alued
function of discharge. This assumption implies that the flow is changing
slowly with time. Effects of 'abnormal surface slope in modifying discharge and in changing channel storage are, therefore, neglected.
This procedure is approximately correct for ordinary st.reams with
small slopes. .When storage is plotted against discharge, the resulting
loop is usually narro\v and an average curve may be fitted ill to represent
.the storage a.s a single-valued function of the discharge. If the loop is
wide, it can be reduced to a single-line relationship by an adjustIi:lent
such as tha.t emploY,cd in the Muskingum method [45]. In this method
the adjustment is made possible by varying the values of K and X in
Eq. (20-42).
.
607
. In streams having steep slopes, the . dynamic effects of flow are pro~
nounced and call1lOt be neglected. Consequently, for such streams the
hydrologic method of flood routing maybe found unsatisfactory.
20-5.' Methods of Hydrologic Routing. Numerous hydrologic methods
of flood routing h.we been developed. Descriptions of these method4J
can be found in the literature on engineering hydrology [45-50]. In this
article, therefore, the important methods will be mentioned only by
name and references will he provided for furt.her study. A simple
hydrologic method, however, "rill be described in the next article.
In general, hydrologic methods of, flood routing may be classified
into h'TO groups: t.he analytical audthe instrumental.
There are many analytical methods of routing. For routing through
reservoirs, the well:.known Rippl ma,ss curve [51] is widely llsed [52-54].
Other graphical m~thods of routing have also been developed, sllch as the
method of Sorensen (55 i 561. For routing through rivers, the storagedischarge relationship is simplified in the methods developed by Meyer
[57j, PuIs [47,58,59J, Wilson [60], Oheng [61], Johnstone and Cross [54),
Knappell, Stratton, and Davis [56J, and Chow {52]. Semigraphical procedures include the methods of Goodrich fo3J, Rutter, Graves, and Snyder
[64J, Wisler and Brater [65J, and Steinberg [66J. Other simplified meth~
ods use a nomograph, as suggested by Linsley [67], a str~ght. slide rule
as suggested by Posey [68,69], and circular computers as suggested by
Shepley and Walton [70J. A simplified method of successive average was
developed by Tatum [48,71]. One rather popular and satisfactory
method, known as the Mt~I;!";in{}um method was developed by :!I.1:cCarthy
[46,72]. In this method, Eq. (20-42) is used as the working equatioll,
assuming x
L; that is,
S = [([Xl
(1 _. X)OI
(20-43)
608
1
UNSTEADY FLOW
techniqu~
ROUTING
609
i
FIG. 208. Eleatronic f1oodronting a.nalog.
Bureau.)
r
\
610
UNSTEADY FLOW
I1 and 0 1
Then
o=
~ + [2) =
2
(0 1
~S = 8 2 -
+
O = average outflow during l:!.t
2.
2
)
81
611
FLOOD ROUTING
At
. 8t
2 tJ.t - 0 1
+I +I
J
S2
= 2 l:!.t
+a
.1
(20-44)
J
-,1
tt
UlUl
+ +
-0
-\
-s
S-<>-
(20-45)
When tJ.t
. This equation constitutes the basis of the present method. If tit is not
equal to 1 day, then the values of Sri tJ.t and 82/l:!.t should be used in place
of Sl and S2, respectively, for the computation and construction of the
curves in the procedure to be described a..') follows:
A. Construction of Characteristic Curves. These curves are constructed
on a plane of S vs. 0 + S (Fig. 20-10). The abscissa represents the
storage in acre-ft, and t.he ordinate represents the sum of the outflow
in cfs and the storage in acre-ft. The curves can be constructed given
any two of thesetbree items: informat.ion on inflow, information on,out~
flow, and storage data of a historical flood in the reach under consideration. The third item can be calculated from the two given items by
Eq. (20-34).
).
612
UNSTEADY FLOW!
\
I
flow will form ., loop, and so will the plot of S against 0 + S. If the
loop of the S curve is not wide, an eye-fitted aVemge curve may be drawn
to represcnt [l. line::ll; relationship between S o.nd 0 + S. If the width of
the loop is fairly wide, the average S curVe ml'.y be constructed as:shown
in Fig. 20-11. In this figure, a curve ofS against 1 + S also in t,he fonn
of a loop is plotted on the left side of the ordinate axis and the 8 curve in
loop fo!'m i~ plotted on the right side of the axis. On both sides' of the
ordinate axis, vertical lines are constructed at thc equ.al ~bsci8sas' corresponding to. a certain storage 8'. The lines intersect the loops at
poiuts A, B, A', and B'. The points Aand A' are on the limbs of ~he
rising stage of the loops, and B rmd B' are on the limbs of falling stage.
Draw stl'o.ightlines AA' and BE', which intersect at C. Draw a horizonto.llille irOlil C to the right to lTIf::et t.he verticul interc(~pt AB at point
C', C'is the r.equil'cd point on the nyerage S curve corresponding to the
given stomge 3'. It can be shown that the ordinate of C; repl'esen~s
the sum of S'and a certain discha.rge Q whose value is the average for
the rising and fulling of the flood stage.
3. Construct the image curve (1 curve in Fig. 20-10) whose abscissa
is equal to 8 t
0 1 Tlus is an image of the 45 line reflected horizontully on the left side of the curve. The curve may be constructed
easily by making the hOl'izontal intercept between the curve and the's
curve equal to the horizonto.lintercept be~\v~en the 8 cur~e and the
45 line.
.
B. Determination of the Outflow. This is based on the char:l.cteristic
curvtjs (Fig. 20-10) and the inflow hydrogl'nph of the flood to be routed.
The routing period is taken as 1 day.
1. The initial outflow at the beginning of the first routing period must
be known or assumed. Not.e that, if the value is assumed, the error
involved in assuming th~ vruue will not be magnified enough to produce
appreciable effect on the result.
2. Locate [l. horizonto.lintercept AB equal to the initial outflow
.between the 45 0 line and the S curve.
3. Extend AB towo.I'd the left to meet the 1 curve at C.
4. Extend AB to,vard the" right to point D, making CD = 1 1 + Iz.
5. Draw the vertioalline DE from D upward, to meet the 45 line at E.
6. Draw the horizontnlline EF from E to the left to meet the S curve
~~
. .
.7. Measure the length of EF, which is eqU9l1 to the outflow at the end
of the routIng period, or Oz.
. '
;
8. Continue the cycle of the above steps :by sta.rting from point F,
613
FIIOOD ROUTING
I
i
HD
and
Therefore,
. HA -- CA + CD"", 8 1 - 0 1 + II
GE = GF + FE ,= 8 2 + FE
HD =GE
S~ - 0 1 + 11 + I .. = 8 2 + FE
12
(20-47) .
20-3.
(4)
(2)
(1)
I,
I,
!I.-larch 23
24
25
25
27
28
29
30
31
i
April
22,8
23.6
57.4
59.5
161.1. 158.5
279.2 269.6
277.7 .268.2
195.9 189,2
133.9 129.4
92.8
96.1
71.9
i'L5
58.9
56.s1,
45.1
-37.3
32.8
29.6
26.6
3
4
24.1
l
I1.
(6)
(}utfiow
Inflow
Adj,
ds
(5)
aid
Given,
cis
I, +;,
o.
Ay,
80.2
215.9
428.1
53i.:S
457.4
318,6
222.2
164.1
128.7
101.9
82.4
70.1
62.'4
56.:2
50.~
(8)
(7)
Storage
Av,
Ch!mge>:
Accum,
aid
!lfd
!I.e-it
8,
S,
Ot
18.5
29.5
94.5
211.2
270.8i
229.3
168.5!
119.7
9E;'.8
64.5
50.2
40.0
34.0
30.4
28.0
26.2
,512.1
0 1 +0, S. - S,
48.0
124,0
305.7
482.0
500.1
397.8
288.2
216.5
16l.3
114.7
90.2
74.0
{i4.4
58.4
54.2
(9)
27.5
59.7
151.6
274.0
329.8
287.1
207.9
141.9
90.1
57.5
44.7
32,2
9l.9
122.4
55.8
-42.7
-79.2'
-66.0
-51.8
-32.6,
-12.8
-7.8
-3.9
-2.0
-2.2
-3.5
36.9
i
33.0
31.0
46.0
89.2
246.1
485:2
600.6
516.4
376A
261.6
1813.9
122.0
94.9
76.9
67.0
28.8
61.4
56.8
25.3
51.5
614
-"I
U-\'lSTElADY FLOW
FLOOD ROUTING
20-4.
600
(2)
(3)
Inflow
Date
(day)
A'll,
dd
I,
+ I,
.. ..
<fI
,I
(5)
Given,
ds
Adj,
0,
0,
01
01
2S1.0
111.S
20.3
65.0
129.0
126.0
70.0
72.5
40.0
1st
2d
20.0
113.2
3(1
4th
180.0
71.0
Sth
6th
7th
40.S
32.(}
2S.0
8th
22,5
20.0
47.S
19.0
17.5
39.0
9th
10th
lith
Total ...
.1
Outflow
Given,
cfs
II
I,
~;
400
.t 300
(4)
~,
500
(1)
615
133.2
293.2
.
07.0
42.5
36.5
29.0
24.0
21.0
19.5
18.9
562.7
cfs
20.3
134.8
128.8
125.8
59.8
i
39.8
~
~8.8
23.8
2{;L8
19.3
18.7
560.7
each other. As is usually the case, however, the sums here are not equal to each other.
Assuming that the outRoy, values are closer to the tr)le values tha,n the inflow values,
the inflow is adjusted by multiplying the values in col. Ii by the ratio 1,512.1/1,566.3 0.9654. The adiusted inflow values are listed in col. 3. The a.verage inflow and outflow in afd a.re equal to I j + I z a.nd 0, + 6" respectively, where I" I" 01, and O.
are in efs. Value~ of I I + I, and 0, + 02 are given in cola. 4 and 6, respectively.
The' difference between t,he values in eols. 4 and 6 is equal to the change in storage
listed in ~ol. 7, liinee (II + I.J
(0, + 0,) = S, - S, = AS. A positive value in col.
7 indicates that the storage is increasing; a negative value indicates that the storage
is decreasing. Column 8 gives the cumula.tive stora.ge oomputed from the va.lues ill
col. 7.' The first entry in col. 8, i.e., 27.5, rep:resents the storege in the reach on
March 23 just before the flood began, which should be given or estimated. The sum
of thevlLlues in cols. 5 a.nd 8, or 0, + S" is ent.ered in col. 9. From the values in
ools. 8: and 9, the cha.racheristic curveS are constructed, a.a shown in Fig. 20-12.
Dole
The computat,ion of the outflow from the inflow of Il. given flood is shown in Table
20-4. In cols. 1 and 2 are inflow data. Column 3 gives the average inHow inafd, or
II + I 2 In coL 4, the initial outtlow is gi ven as 20.3. Other values of outflow are
obtained from the characteristic curves in accordance with the procedure described
previously. The outflow values, except for the given initial outflow, are ad;usted
and entered io col. 5 so that the total inflow agrees with the total outRow. The
"~
i
!
616
UNSTEADY FLOW
inflow .and out.f1ow hydrographs of the given flood are constructed as shown in Fig.
20-13 from the vtl.lues of cois. I, 2, and 3.
'FLOOD ROUTING
617
change in stage at Sta. B of the flood under investigation as a. result of the proposed
pla.u, Assuming that the discha.rge is a single-valued f unction of the ataga.
"REFERENCES
PROBLEMS
ZO~l. Extend the routing of the flood in Example 20-1 to a period of (a) 36 hr !l.nii
(b) 120 hr,
20-2. Route the following hypothetical solitary flood through the channel reach
described,in Example 20-1:
Celerity, cfs
14.0
15.0
15,,j
15.9
'14.0
3
6
9
12
The flow conditions before and after the flood wave Me uniform/having a celerity
.
'
of 14.0 cfs.
20-3. Verify the un'it-flood computation sho\vn in Fig, 20-4.
20-i. In a level-pool reservoir in which discharge is not controlled, the pl!ak outflow
must occur where the' outflo,w hydrograph intercepts the inflow hydrograph; that is,
P' in Fig. 28-6 must be at L Why?
'
20-li. In the development of a flood-control project for e. river bEtsin, the largest
flood on recol'd is investiga.ted. The discho.rges of this flood at an upstrea.m station
A llnd a downst.rc!t1U station B were observed, as shown in the accornpnnying table.
Main,
cfs
-------"
;[llly 29
13,600
30 20,100
31 106,000"
Aug. 1
:2
3
4
5
6
7
1,
(
92,800
49,600
22,700
11,000
8,050
13,800
19,000
14,500
I 'Iributo.ries,
Local,
cis
cia
1,900
46,300
3,000
63,800
72,lQO
21,100
50,100
I5,900
0,200
2,900
2,200
2,10.0
3,200
5,300
3;200
8,900
6,200
4,100
3,100
5,200
8,100
3,800
Outflow at Sta. B
Discha.rge,
ds
20,800
33,000
80,800
110,000
112,000
lIZ,OOO
'102,000
68,200
26,100
20,800
21,700
Discharges held
"out at Sta. A,
Gage ht, due to proposed
ft
reservoirs, cf.
14.4
19.6
35.1
43.0
43.5
43,4
41.1
31.5
16.8
14.4
14,8
0
60,000
149,000
96,400
46.100
17,900
2,000
2,100
3,200
4,800
2,200
The total inflow at Sta.. Ais equal to the sum of discha.rges from riIn.in stream, tributaries, and local areas a.djacent to the reach. The storage in the reach between StaG.
A and B" at to,; beginning of the flood )5 estimated at 50,000 acre-ft.,
For the purpose of flood control, a system of reservoirs is proposed on the tributaries above Sta. A. It is found that the effect of the reservoirs is to hold out discharges from the flow at 8ta. A, ~ shown in the 11l.5t column of the table. Predict the
1. N. J. Dahl; On non-permanent flow ill open cana.ls, Proceedings of the 6th Gene,-al
Meeling, International Asso,ia.lionJor Hydraulic Resea.rch, The Ha.gue 1956; vol. 4,
pp. D19-1 to D1916, 1955.
2. T8.keo IGnosita: Hydrodynamicn.i study on the flood flow, in Floods, voJ. III of
SYTnposia Darcy, International A~sQciaHon o.J Scientific Hyd"ology, PubliCa/ian
No. 42, 1956, pp. &5-63.
"
3. ,D. N. Dietz; A new method for calculating the conduct of translation waves in
prismatic canals, Physica, ~ol. 8, no. 2, pp. 177-195, Febru/l.Yy, 1941.
4. Junius Massau: Appendice au M:emoire sur l'integration graphique (Appendix: to
Memoir on graphical integration), Annales de I' Association des Ingd'lli~u1's sDrti~
des Eco.les Speciales de Ganc, voL 12, pp. 185-444, Ghent, BalgiuIll, 1889. '
5. Junius l'Ihssau: Memoire sur l'integration graph,ique des equations aux derivees
partie lies (Graphical integration of partial differential equations "lith special
applications to unstelLdy Bow in open channels), Annale8- de l' Association .des
[-ngJnieurs sortis des Ecoles Speciale;; de Grmd; vol. 23, pp. 95-214, Ghent, Belgium,
1900.
6. MlJ.rc Hemy: Propagation des intumescences dans un canal rectangulaire'-( Propagation of, translatory waves in a rectangular channei), RellUe g~n()I'ale de l'hydrallZi'lue, Paris, vol. 4, no. 19, pp. 17-24; no. 20, pp. 65-71, 1938.
7: Louis Bergeron: Methode graphique generaie de calcul des propagatiolls d'ondes
planes (General graphical method of computation of the propagations of plane
waves), Memc)1:res, Societ~ dell Ingenieurs Civiis de la Fra.nce, pp. 407-497, JulyAugust, 1937, ,
8. Louis Bergeron: Methode graphiqllIJ pour te calcul dell ondes de tro.ns~ation
(Graphical method for the complttation of translatory waves), Sw;~te Fran.~aise
des Mecaniciens, Bulletin No, 7, Paris, 1953.
9. S. A. Kh1'istianovich: Neustanovh'sheiesia dvizhenie v kanalakh i rekokh (Unsteady "motion in channels a.nd rivers), in "Nekotoryie Voprosy Mekhani\d
SploahnoI Srady" ("Several Questions on the Mecha.nics of ContinuouH' Media"),
Academy of Sciences, U.S.S.R., 1938, pp. 13-154.
10, J..eon Levin: Methode graphique de calcul du mouven1ent Iion permanent dans'
lescanailJ( a ecoulement libra (Grapbic method of computation of unsLeady Bow
in open channels), Le Gtnie civil, vol. 119, no. 11-12, pp. 109-113, Ma.rch 12-14,
1942.
11. A. Craya: Ca,lcul graphique des regimes variables dans les canaux (Graphical
computa.tion of varia.ble regimes of flow in channels), La Hou.ille blall,che, Grenoble,
new series, lilt yr., no. I, pp. 19-38, Novembel', 1945, and no, 2, pp. 111-130,
March, 1946. '
I
12. V. A. Arli:hangels!dl: "Raschety NcustaDovivshegosia Dvizheniia. y Otkrytykh
Vodotokakh" ("Cnlculation of Unsteady Flow in Open Channels"), Academy
of Sciences, U.S.S.R., 1947.
.
13, H. Holsters: Le Ca.lcul du Illouvement non perma.nent dans les rivieres par Is.
methode dite des "Iignes cl'influence" (The computation of unsLeady fiow in
rivers by the so-called "influence-lines" method), Revu.e generale de l'hydra!tl'i2tIB,
Paris, vol. 13, no. 37, pp. 36-39, no. 38, pp. 93-94, no. 39, pp. 121-130, no. 40, pp.
20Z-206, and [l0. 41, pp.,237-245, 1947.
.
618
UNSTEADY FLOW
14.. H. Holsters: Le Calcul du mouvement non permanent dans les rivie .."s par la
methode dite des "Iignes d'influence" (Calculation of nonpermanent flow in rivers
by t.he method kflown as "influence lines"), La Houill~ blanche, Grenoble, 8th yr.,
no. 4., pp. 49fi-5U9, August-September,1~.53.
.
15. J. J. StokE>r: The formation of breaker,,; and bores, New York Univer.lily, Comnmnications on Applied ltfathemalics, vo!. I, no. I, pp. 1-87, January, 1948.
16. J .r. Stoker: Nunierical solution of flood prediction and river regulation problems, I: Derivativ'n of hasic theory and formulation of numeriea! methods of
attack, New York University, Institute or Mathematical Sciences, Report No.
IMM-200,1953 . .
17. J. J. Stoker: II ';Vater waves," vol. IV of "Pure and Applied Mathematics," Interscience Publishers, Inc., New York,1957.
.
.
18. Henri J. Putman: Unst,ead:r flow ill open channels, Transactionll, American Geophysical Union, vol. 29, no. 2, pp. 227--232, April, 1948. Discussion by Pin-Nam
Lin, vol. 30, no. 2, pp. 302-306, April, 1949.
19. J. Lamoen: Tides and current velocities in.1 s%-lev~1 canal, E71!linceri:I(I, vol. 168,
no. 4.357, pp. 97-99, July 29, 1949.
20. G. J. Dmitriev: Vychislenie kharakt~ristik cst[l.noviTfshegosia pJavllo izmeniaiushchegosia dvizheniio. v prizmaticheskikh ruslnkh (Computat,i'Jil of characterist'cs
of a steady gradually varying movement in prismatic channels), Comptes rend".
(Dokludy) de l'AC(1d~mie de~ Scienceli de l'U.R.S.S., Akademiia NatLk S.S.S.R.,
Leningrad, vol. 68, no. 5, pp. 825-827, 194.9.
21. Francis F. Escoffier: A graphical method for investigating the stability of flow in
op~n cbannels or in closed conduits flowing part.ly full, Transactions, America11
Geophysical Union, vol. 31, no. 4, pp. 583-586, August, 1950.
22. G. D. Ransford: Contribution to first order theory of translation waves, La
Houille blanclw, Grenoble, 6th yr., no. 6, pp. 761-763, September-October,
1951.;
.
23. Pin-Nam Lin: Numerical analysis of continuousunsteady flow in open channels,
'l'ransCtctions, Americ.an Geophysical Union, vol. 33, no. 2, pp. 227-234., April,
1952. Discussions by J. C. Schonfeld and Fin.. Nam Lin, vol. 34, no. 5, pp. 7nnl5, October, 1953.
2'1. Shigeo Uchida: On the analysis of flood wave in a reservoir by the method of
c1111.racteristic"s, Proceedings oj the Ed Japan National Congress for Applied lIfechan~
ics, pp. 271-276, 1952.
25. J. N ougaro: Recherches experimenta.les sur les iritumescenoes dans les canaux
decouverts (Experimental researches on translatory wave" in open channels)
SociJt~ Fm1<~aise des M~caniCiens, BulldinNo. 9, Paris, pp. 23-35, 1953.
26. J. Nougaro: 'Theoretical and experimental studies o.f the propagation of the transI~"tion waves in open channels, Proceedings. oj.the Minnesota International Hydra11r
lics Convention, Joint 'Meeting of International Associationfor Hydraulic Reliearch
and Hydraulic,s Diuision oj American Society of Civil Engineers, pp. 555-559,
August, 1953.
27. Jean NotlgarD: Etude theorique et experimentale de la propagation des intumescences dans les canaux decouvert.<; (Theoretir.al and experimenta.l study of the
propaga.Uon of translation waves ill open chaunels), PtJblications scientifique.~ et
techniques d1J Ministere de PAir, France, No. 284, 1953.
28. J. Nougaro: Methode graphique pour Ie caJcul de 1iJ, propagation des intumescences
dans les canaux cle~()uverts (Graphical method for the computation of the propagation of translatory waves in open channels), Proceedings of the 6th General Meeting, International Association jor Hydraulic Research, The HaglLe 1965, vol. 4,
pp. D5-1 to D5-15, 1955.
FLOOD ROUTING
G19
29. Yu,ichi Iwagaki and Tomitaro Sueishi: On the unsteady.flow in open cha.nnels with
uniform lateral inflow (in' Japanese), P1'Occcdings, Japan Soci'-tll oj Civil Engineers"
vol. 29, no. H, Tokyo, November, 1954.
30. Yui,~hi Iwagaki: Fund~.mental studies on the runoff analysis by charaderistics,
Kyoto University; Disaster Prevention Research 1 nstilule, Bulletin No. 10, Kyoto,
Japan, December, 1955.
.
./ 4
31- Tomitaro Sueishi: On the run-orr analyds by the method of characterist.ics (in
Ja.panese), Transactions. Japan Society of C7:"il Engirwers, no. 29, pp. 74-87,
Tokyo, December, 1055.
32. E. J. Isaacson, J. J. Stoker, and B. A. 'Troesch: Numeric.al ~oIution of flood prediction and river regula Lion problems: Report 2, Numerical solution of flood
problems in simplified model" of the Ohio River and the junction of the Ohio and
Mississippi Rivers, New Y orlc Uni,'er.su1l, I nslitule of lVIathcIII.atica/ SC'iences,
Report No. I11JjI-I-205, 1954.
as. E. J. Isaa:cson, J. J. Stoker, acd B. A. Troesch: NUIIlerical solution .of flood prediction and river regulation problems: Report 3, Restllts of thr, numericll.J prediction of the 1945 and 194.8 floods in the Ohio River, of the 194.7 flood tin'ough tIle
junction of the OIJio 2"nd Mississippi Rivers, and of the floods of 1950 and 194.8
through l(eutuGky Reservoir, IV ew Y orlc University, Inst,:tute of Mathell1(!tica.L
Science~, Report No.IMM-NYU-235, 1956.
34. Edward A. Lawler and Fra.lLlr V. Druml: Hydraulic 'problem solution on electronic
computers, paper 1515, Proceeding", American Society of Civil Engineers, J o!!rnal,
l-Faterways and Harbors Division, vol. 8~, no. \VW1, pp. 1-38, January, 1958.
35. Georg Joos: "Theoreticai Physics," 2d ed., Hafner Publishing Company, N~w
York, 1950, pp. 590-594.
36. Shib-I Pai: "Viscous Flow Theorl',"yol. II, "Turbulent Flow," D. Van Nostrand
Company, Iuc., P,inGetcll, N.J., 1%7, pp. 1/9'::183 and 186-187.
37. Alfred Schack: "Industrial Heat TransJer," translaLed from the 'German by
Hans Goldscilmidt and Everett P. Partridge, John Wiley & Sons, Inc., New York,
1933, p. 29.
38. Shoitiro Hayami: On the propagation of flood WD.ves, Kyoio Unive.,sity, Disa.~ter
P)'evention Research lnsl/}ute, Bulletin 1, Kyoto, Japan, Deeember, 1951.
39. F. V. Appleby: Runoff dyno"mics: A hee,t conduction 2"na!ogue of stora.ge flow in
channel networks, A.s.scmbUe G~n~rale de Rom~, 1954, InternationaL Association of
8cientific Hyd1'oloflY, Publication No. 38, yol. 3, pp.338-3<18, 1954.
40. 'fojiro Ishihara, Shoitiro Hayami, and Shigenori Hayarni: On the electronic
analog computer [or flood routing, Proceedings of the Japan A.cadem!l, /O!. 30,
no. 9, pp. 891-895, Tokyo, 1954.
41. Tojiro Ishihara and Yasuo Ishihar!i: On the elect,ronic analog computer for !lood
routing (in Japanese), Transactions, Japan Society of Civil Engineers, no. 24,
pp. <a-57, Tokyo, April, 1955.
.
42. Tojiro 'Ishiharo; and Yasuo Ishihara: Electronic analog computer for !lood flows
in the Yodo River (in JapaneBe), Proceedings, Japa.n Society of Civil En!lineers,
vol. 4.1, no. 8, pp. 21-24, Tokyo, August, 1956 .
43. 'rojiro Ishihara: Applicaticin of electronic analog computer for flood routing to
actual rivers (in Japanese), Transactions, Japan Society of Civil Engineers, no .. 4:),
pp. 4.3-47, Tokyo, ]Tebrua,ry, 1957.
44. Tojiro Ishihara, Shoiti!'o Hayami, and Shigenori Hayami: Electl'onic analog
comput,er for flood flows, Proceedin!ls of the Regional Technical Conference on Wa~er
Resources Development in Asl:a. and the Far East, United Nation.~ Economic Comrnwsion jar Asia and the Far East, Flood Control Series No.9, Bangkok, 1956,
pp. 170-174.
J
"1
'1
620
!
I,
I
I
UNSTlj:ADY FLOW
FLOOD ROUTIN,G
6(1. L B. Steinl;>erg! A method of flood routillg, Civil Enyineering, vol. 8; no. 7! pp.
476-477, July, T938.
...
.
67. R. K. Linsley: Use of nomogl'l!.r5i1s in solving strea.mflow routing problems, Oiilil
" Engilteerin{f, vol. 14, no. 5, pp. 201}-210, l\1aYi 1944.
68; C. J. PQsey: SIicle rule forl'olltillg floods throug!l stomge reservoirs or In.kes,
Engineering News-Rec(wd, vol. Hi, pp. 580-581, Apr, 25, 1935.
69. Sherlll11.n .11. Woodwa.rd ami Chesley J. Posey: "Hydraulics .of Steady Flow in
Open Cha.n:neis" John 'Wiley & Sons, Inc" New York, 1941, pp. 133.,.145.
70. J. M, Shepley ~nd C. B. "Val ton: Solving reservoir problems with cirel,l!!.r pointby-point COlllput~r, Civil E'I!lineel'ing, vol. lZ, no. 3, Pl'. 154-155, IVlarch, 1942:
71. F. E.Tatum: A SiUlplificd method of !Outing flood flows through natum! va.lley
stora.ge, unpublished m~mo"mdum, U.S. Engineer's .office, Rock lsland, IlL,
1IIa.y 29, 1940,
.
72. O. T .. MoCr,rthy: The unit hydrogrrtph and flood routing, unpublished manllscript, presented at a conferCHlce of the North AtlAntic Division, U.s. Army,
Corps of Enginee!'s, June 24, 1938.
.
73. E. W. Lane: Predicting stages for the Lower Mis.sissippi, Civil Engineering, vol. 7;
.no. 2, pp. 122-125, February, 1937.
'
'.
:'.
H. I'vb;..: A. Kohler: A forecu.:;ting technique for routing m\d cOll\hlllll\g flow'ln t.eflHS
of stage, Tra.nsactions, American Geophysical Union, voL 25, pt, VI, pp. 1030-1o:J5,
, 64.
65. C,
0;
Transactions,
H!44.
621
75. Ralph E. King: Stage predictions for flood contra! opero.tions, T,'Msactiolts,
American Society a/Civil Engineers, \'01. 117, pp. 600-GIl8, H)52.
76. William E. Ray and Herman F. Mondllchein: A method of fnrecn.stillg stage.~ on
Jl~t rivers, Tran..sa.cli01l.s, Ame1'ica.1! GeophyS'ical Union, vol. 38, no. 5, pp. 698-707,
.
.
October, 1957.
77. J. F. TllXpleY, Jr.: A new integrilting machine, MiWary E71ginee,, voL 32, no. 181,
pp. 39-43,1 \)39.
..
:
.
.
78. Frank B. Harkness: HarkMss flood rout~l':, SpecificatIOns of construction and
operat.ion, Pa.t~nt File No, 2,550,692, U.S: Patent Ot1i;:e, Washington, D.C.,
May 1, i951.
. .
.
7\l. R. ic Linsley, L, W. Foskett, and M. fl.. Kohler: Electronic device speeds flood
forecasting, Ellyintering News-Record, vol; 141, no. 2?, pp. 64-66.< Dec. 23, 194~,
80. R K Linsley, L. W. Foskett, and 1--,1. A. Kohler: Use "i electronical ant!.loe;:r lfl
flood wave analysis, Comptes rerUius et rapports, de. l'AssembMe G~ne'rale d'Oslo,
19-28 aiJid 194.8, International Association of Scientific Hydro/agy, Publication No
29, 11)48, vol. I, pp. 221-227.
81. M. A. I{o,hler,: Application of electronic flow routing 1l1l;l.log, Transactions,
American Society 01 C~'vil Engineers, vol. 118, pp. 1028-/045, 1953.
I
l
APPENDIXES
'; j
I
I
,I
1I
:j
,I
.1
;1
;1.
r,
',"j .
ApPElND~X:
A
d o1
.Y..
d.
\
I
I
d,'"
0.0066
0.0132
0.01!)7
O.OO6~
0.0134
0.0202
o 0268
0.0;;36
0.0001
0.000-1
0.0326
0.1990
0.2800
0.3412
0.::'!)1\)
0.4359
0.05
0.0105
o 0147
0.06
0.07
0.08
0.09
0.10
0.0192
0.024:3
0.0294
0.0350
0.0409
O.494::J
0.5355
0.5735
0.6094
0.0435
0.0389
0.0451
0.0513
0.OS74
0.0635
0.4.750
o 5103
0.542(;
0.5724
0.6000
0.0106
O.(H74
O.01i42
0.0612
0.0682
0.0039
0.0053
0.0069
0.0087
0.0107
0.0470
0.0534
0.0<300
0.OGG8
0.0739
0.6761
0.7075
0.7377
0.7670
0.79M
0.0695
0.075-1
0.081a
0.0871
0.0929
0.6258
0.6499.
0.6n6
0.61;)40
0.7141
0.0752
0.0822
0.0892
0.1034
0.0129
0.0153
O.Oli\)
0,0217
0.0238
0.0811
0.0961
0.1039
0.1118
0.8230
0.8500
0.8763
0.9020
0.9273
0.0985
0.1042
0.1097
0.1152
0.1206
0.7332
0.7513
0.7GS-t
0.7846
0.8000
0.1106
0.1178
0.1252
0.1324
0.1398
o 0270
0.030*
0.0;339
0.0378
0.0418
0.0220
0.0247
0.0273
0.9521
0.976 ..
1.0003
1.0239
.1.0472
0.1259
0.1:312
0.1364
o~ 1416
0.14136
0.8146
0.8285
o 8-H7
0.25
0.Wl9
0.1281
0.1365
0.1449
0, HiSS
0.8542
0.86(10
0.1472
0.154[\
0.1622
0.1096
0.1774
0.0460
. 0 0503
0.0549
0.0597
0.0646
0.0301
0.0333
0.0.359
o 0394
0.0427
0.2()
0.27
0.28
O.2D
0.30
0.1623
0.1711
0.1800
0.1890
0.1982
1. 0701
1.0928
1.1152
1.1373
L 1593
0.15\6
0.1566
0.1614
Q.1662
0.1709
0.8773
0.8879
0.8980
0.9075
0.9165
0.1850
0.1925
0.2004
0.2084
0.2162
0.0697
0.0751
0.0805
0.0862
0.09Z1
0.0464
0.0"97
0.0536
o 0571
00610
0.21
0.22
0.23
d,
O.O'!;
o 24
\
d.
02838
0.3<!S2
0.4027
O. '15lO
0.15
0.17
0.18
0.10
0.20
cIo
00013
0.0037
o.n
I-
001
0.02
0.03
0.12
0.13
0.14
0.15
P
d.
O.OOLIO
O.OSB5
O.
0;OZ5Z
625
O.O<J64
O.OUIO
0.0017
0.0027
0.0002
0.0005
0.OU09
0.0015
0.0022
O.OO;H
0.0010
0.1]052
0.00()5
0.0079
0,001)5
o.oua
0.0131
0.0152
0.0173
O.OHJ()
~I
626
APPENDIXES
627
APPENDLX A
ApPENDIX A. GEOMETllIC ELEMENTS FOR CIRCULA.R
A.
AI'I.'EN.DrX
CHA.NNEL SECTIONS
A
do'
.'!!..
d,
R
da
.do
D
d.
T
d.
--0:31
0.32
0.33
0.34
0.35
0.2074
0.2167
0.2260
0.2355
0.2450
0.36
0.37
0.38
0.39
0.40
0.41
0.42
0.43
0.44
. 0.45
0.46
0.47
0..48
0.49
0.50
1.2025
1.2239
1.2451
1.2661
0.1755
0.1801
0.1848
O. ]891
0.1935
0.9250
0.9330
0.9404
0.9474
0.9539
0.2546
0.2642
0.2739
0,2836
0.2934
1.2870
1.3078
1.3284
1.3490
1.3694
0.1978
0.2020
0.2061
0.2102
Q.2142
0.9600
0.9656
0.9708
0.9755
0.9798
0.3032
0.3132
0.3229
0.3328
0.3428
1.3898
1.4303
1.4505
1.4706
0.2181
0.2220
0.2257
0.22(H
0.2331
0.3527
0.3627
0.3727
0.3827
0.3927
1. t907
1.5108
1.5308
1.5508
1.5708
O. '1027
0.51
0.52
0..53
0.54
0.55
0.4127
0.4227
0.4327
0.4426
0.56
0.57
0.58
0.59
0.60
0.4526
0.4625
0.-1723
0.4822
0.4920
0.61
0.62
0.63
0.64
0.65
0.5018
0.5115
0.52t2
0.5308
0.5404
:i
d.
0.2242
I) :232:l
O.240.f
0.2486
0.2568
0.0981
0.1044
0.1107
0.1112
0.1241
0.2652
0.273(3
0.2822
0.2908
0.2994
0.1310
(),1381
0,1>153
0.1528
0.11303
0,1050
0.9837
0.9871
0.9902
0.9928
0.9950
0.3082
0.3172
93262
0.3352
0.3446
0.1682
0.1761
0.1844
0.1927
0.2011
0.1100
0: 1147
.0.1196
0.1245
0.1298
0.2366
0.2400
0.243'1
0.2467
0.2500
0.9968
0.9982
0.9992
0.9998
1.0000
0.3538
0.3634
0.3730
0.3828
0.3928
0.2098
0.1348
0.2186
0.1401
0.1452
0.2275
0.2366 . 0.1505
0.2459
0.1558
1. 5908
1.6108
1.6308
1.6509
1. 6710
0.2531
0.2561
0.2591
0.2620
0,2649
0.9998
0.9992
0:9982
0.9968
0.9950
I 0.4028
0.4130
1.6911
1.7113
0.2676
0.2703
0.2728
0.2753
0.2776
1.7926
0.2797
1.8132
0.2818
1.8338
a 2839
1.8546
0.2860
I
.1.8755
0.2881
1.1810
I
I
1. 4101
A.R~~
do" s
1.73~5
1. 7518
1.7722
0.0650
0.0864
0.0909
0.0955
0.71
0.4448
0.2553
0.2650
0.2748
0.2848
O.2!H9
0.1610
0.1664
0.1715
0,1772
0.1825
0.9928
0.9902
o 9871
0.9837
0.9798
0.4558
0.4670
0.4786
0.4902
0.5022
0.3051
0.3158
0.3263
0.3373
0.3484
0.1878
0.1933
0.1987
0.204'1
0.2092
0.9755
0.9708
0.95513
0.9600
0.9539
0.5144
0.5270
0.5398
0.5530
0.5666
0.3560
0.3710
0.3830
0.3945
. 0,4p66
0.2146
0.2199
0.2252
0.2302
0.2358
0.4234
0.4340
A
do'
0.0736
0.0776
0.0820
0.1020
0.66 !
0.67
J
0.68
0.69
0.70 !
o.OMn
I!
0.72
0.73
.0.74
0.75
0.76
0.77
0.78
0.79
0.80
I
\
(conlinued)
T
D
R
d~
d.
---'--I---___~I-~-~ - ; - - - - 1 - - - - - -
d.
0.4188
0.4309
0.4437
0.45136
o -lG94
0.5804
0.5948
0.54()9
0.5594
0.5687
0.5780
o 5872
1.8965
1.9177
1. 9391
1.9606
1.9823
0.2891
0.2917
0.2935
0.2950
0.2962
0.9174
0.9404
0.9330
o 9250
0:9165
.0.6096
0.5964
0.6054
0.6143
0.6231
0.6318
2.0042
2.0254
2.0438
2.07H
2.01)44
0.21)73
0,2184
0.2095
0.3006
0.3017
0.9075
0.8980
0.8870
'0..87i3
o 86GO
0.6572
0.6742
0.6918
0.7104
0.7296
o 531l:!.
0.6404
2.1176
2.1412
2 1652
2.1895
2.2143
I)
3025
0.3032
0.3037
0.3040
0..3042
0.8542
0.8417
0.8285
0.8146
08000
o 1498
o 7710
o..'5G!)5
0.7934
0.8170
08420
0.3044
0.3043
0.3041
03038
0.3033
0.7846
0,7684
0.7513
0.7332
0.7141
0.8686
0.8970
0.9276
0.9606
0.9964
0.6489
0.6573
0.6655
0.6736
.'
0.6250
0.6408
0.2407
0.2460
0,2.510
o 2560
o 2DU8
0.4831
0.jJJG4
0.5100
05248
O.2IJ.5:j
0.2702
0.2751
0.27H4
0.2840
0.5540
0,5850
0.6011
0.G177
0.2888
0.2930
o 2060
0.3008
0.3045
0.G347
0.6524
0.6707
0.6897
0.7098
0.3082
O.aJl8
0.3151
0.3182
0.3212
0.81
0.82
0.83
0.34
0.85
0.6893
0.6969
0.7043
0.7115
2.2395
2.2653
2.29l()
2.3186
2.3462
0.86
0.87
0.38
0.89
0.90
0.7186
0.7254
0.7320
0.7380
0.7445
2.3746
2 4038
2.4341
2.4655
2.4981
0.3026
0.3017
0.3008
0.2996
0.2980
0.6940
0.6726
0.ti49\l
0.15258
O.GOOO
1.0354
1.0784
1. 12fH
1.1800
1 2408
0.7:307
0.7528
0.7754.
0.8016
0.8285
0.3240
0.3204
0.3286
0.3307
0.3324
0.91
0.92
0.93
0.9:1
0.95
0.7504
0.7560
0.7612
0.7662
0.7707
2.5322
2.5681
2.6061
:2 .6467
26906
0.2963
0.21)44
0.2922
0.28913
O.28(l4
0.5724
0.5426
0.5103
0.4750
0.4359
L3UO
1.3932
1.4918
1.6130
1.7682
0.8586
0.8917
0.9202
0.9725
1,0242
0.3336
0,:334.5
0.3350
0.3353
0.3349
0.96
0.97
0.98
0.99
1.00
0.7749
0.7785
0.78t6
0.7841
0.7854
2.7389
2.7934
2.8578
2 9412
3,1416
0.2830
0.2787
0.2735
0.2665
0.2500
0.391\)
0.3412
0.2800
0.1990
0.0000
1.9770
2.2820
2.7916
3.9400
1.0888
1.1752
1.3050
1.5554
'"
""
0.3.'34:0
0.3322
0.3291
0.3248
0.3117
0.68~5
- 1
-I
I
i
II
!'
,i
"
Apl'ENDL"
II
B.
GEOMETRIC
I
I
I
1
II
!
I,
I
\
'I
1
,
I
r
y;
629
~/~
..
0:
::!
'-6
'"
'"""o
", ~
1; 1
side slopes.
:t:
0:
'"
"
'is
l?
'"
....
"
~
"0
>-
:L
S 7 8910
0.2
0.3
A,;!Q. A, liz
1 side slopes.
...
'.
,
~
...-L-..
....
0::
.2
."
.,.
to;)
'-'"
Areo, A,
1,2
FIG. B-3. Hydraulic elements oftrapozoidal channels with 2:1 side slopo!!.
0::
"".w
w
.;
"
'0
l2'
.2
"3
0
.f5
".,
J::
Areo,A, tlZ
FlO. B-4. Hydraulic element.s of t.rapezoidal channels wit.h 2~i: 1 side slopes.
t-
l!w:",--+--+.--J-j-+-i
.!!
O.')~~--!-,-+-1---1--1-4-
:c
0.7
g
-g,
0.8
r--r--+-+-++-b
I
---\.---i--.".j"~";::
0.6
0.5 ----i-'7"'g-,,4--+-7"i"-
10
. 15
Area, A 0 2
;:::
Ii.
.;
"
15
Ql
.;1
:;
<!
'"J:
>.
(-'~-
-'--
~(--
FIG. B-7. Hydraulic etemehts.of trapezoidal channels with 5:1 sid" siol?es.
,.
....
rl
.;
-0
c..
e.->
-1
(.3
30
t
:r:
..____
.---,~u.w
15
Afea,A, HZ
FIG. B-3, Hydranlic elements of trapezoidal cbannels with 6:1 side slopes.
'!
""'"''!''''!''''l'''
2.5i=---,...+-----,-;---~+_H_+---!-
-..i~1--r-::7F7""+__::.:r'9-::l_-
:1'
" 1--
a:
'"::>
'i5
0>
<;;:J
'"
.!'!
:;
0
.;;;>..
::c
:';
.~,
.,
'.'J
i,
I:
.........
Oai'l!'
..., 1.0
:;1(;>
m:::;,,",
.s::::
.......
, iTi
ApPENDIX
4)-
>=
'"
'"
.,.
b.J
-:
n:
'";..
<1)-.~
1
\
'E Ei'
\
I
"Z
(LI
00
\
liej ';\~/('Ot9'i\1
'"E-<
:r:
g; ~ ~ i3 88
0'
o' dd dd
'0
...
.~
(0. ~ ~ ~
1lJ
ro
i.'!~m_
;::l
'0
o.
>0
a:::.:
'"
~
:!
.5
;;;
p<.
'"
::a"
0
rrp rT""T"Tn-r'll;
<oW
cici
~tf'!
00
'"<:5
it l'i n I
_II> "'
.,. ....
jl
d~ ~ J'~ ~
'1"""'"'1
N
..
--:
'-
,,
,T
f:l
>-oJ
<1l
P '"
G;
0:
c:
Cf.l
.
So.
.. -
., ..0
CL.~ ~
. a
I
T
I
,,
P<
<=l
'"
.~~;
iii
o ....
.-
i:iA
~.2m
rri .~.,3
c: 0'0
'"
Q
::l 0._
"ro >';:: ~ ~
....A_
~8
u. t..O:';l'JIJOw
t:
..0>
.....-.
~-
~
..,
!\)oIE-<
:::
rl
~ t..?
.....
~~I
;:;
ro
-"00
E::
:s
'" 0
gad ~ ....
CJ
..:::0:21
i""I"'lN'T'T'TITm=J"'1'~'rT'l'l'l'ITI'"'1""r'"":rrITlTT'JT"""F!"'T'l
:>
en
0-::;:-
-<
:;E
11"'
~;;::
.., _
0~
, .a
:&sj
.-
.~.t:
"CO
It
-::l
g ~ f!
:)~"O
'J.
::;cilI:
a. ...
'" ~ 0
~
.. . .g..,
..
::I
::20
J.
I
\
640
UN
AND F(iJ.,N)_s.
IoU
~.,.
:t
2.2
2A
2..6
2.8
0.000
0.020
0.040
0.060
0.080
0,000
0.020
0.0,10
0.D60
0.080
3.0
3.2
3.4
3.6
4.0
r=.,
....
~ -""
..:I
b
IoU 1
;0..,>-:>
~ i:!
:::> .. "0
.;:
.0.-0
~"'V~
..,;t
0'-
D.
F(u,N) =
gOO::;
0.000 0.000
0.020 0.020
0.01.0 0.040
0.060 U.03O
0.080 0.080
0.100 0,100
0.120 0.120.
0.140 0.110
0.161 0.161
0.181 0.181
0.202 0.201
0.223 0.222
0.244 0.243
0.26.'5 0.263
0.286 0.284
O.SO 0.307 0.305
0.32 0.329 0.326
0.34 0.351 0.:148
0.3? 0.372 0.369
0.395 0.293
0.38
0.40 0.418 0.414
0 .. 42 0.442 0.431
0.44 0.465. 0.460
0
0.489 0.'183
0.514 0.507
0
0.539 0.531
0,565 0.557
0.592 0.582
0.619 0.608
0,648 0.635
0,676 0.663
10.691 0.678
0.7061 0.692
0.722. 0.707
0.738 0.722
0.754 0.737
0.65
0.66 0.771 0.753
0.67. 0.787 0.769
0.68
804 0.785
o 69 0.822
0.804[
0.00
0.02
0.04
0.06
0.08
0.10
0.12
0.14
0.16
0.18
u.20
,p.22
0.2-1
0.26
0.28
0.000 0.000
0.020 0.020
0.0,10 0.0;10
0.060 0.060
0.080 0.080
0.100 0.100
0.120 0.120
0.140 0.140
0.160 0.160
0.1800.180
0.20010.200
0.220' 0.220
0.240 0.240
0.260 0.260
0.280 0.280
0.301 0.300
0.321 0.321
0.341 0.341
0.362 0.361
0.382 0.382
0.403 00402
0.423 0.423
0.444 i 0.443
OA650H14
o .486 0 .485
0.508 0.506
0.529 0.528
0.551 0.550
0.574 0.572
0.596 0.594
0.620 0.617
0.631 0.628
0.643'0.640
0.655 0,652
0.667 0.664
0.680 0.676
0.692 0.688
0,70,,) 0.701
0.718 0:71.3
0.731 0.726
* The table of the varied-flow fUlletion fOI' i)ositive slopes F(II,N) is 1'\}pl'oduced
from Ven Te Cho\v, Inteljratin' g the equntion of gradually n.ded fiow, PToceedin(lo,
A JIIerican Society of Cil,ii I~!li"eej's, vol. 81, paper no. 838, pp. 1-32, Novembel', 1955.
Tpe ta.ble of the varied-flow [tinction for negative slOpes F{1L,N) -8, isrepl'oduced from
the author's closing discussion of this pU.per in roceedings, voL 83, JOli1'llut of
Hyd1'auUC8 Di.vision, no. HYl, paper no. 1177, pp. 9-22, February, 1957.
641
643
APPENDIX D
642
APPENDIXES
TABLE
TABLE
D-l. T.Hf:
2.4
2.6 1 2.8
0.819 0.802
0.836 0.819
o.85S 0.B3r}
0.874 0.854
Oy892 I 0.868
0.787
0.804
0.820
0.837
0.854
3.0
0.775
0.791
0.807
0.823
0.840
3.4
3.2
F(u,N) (continued)'
3.8 I
3.6
0.766 0.757
0.7810.772
0.796 0.786
0.811 0.802
0.827 0.817
0.750
0.764
0.779
0.793
0.808
0.744
0.758
0.772
0.786
o 800
0.890 0.872
0.909 0.890
0.930 0,909
(I.gSO,0.929
0.971 0.949
0.857
0.874
0.892
0.911
0.930
0.844
0.861
0.B78
0.896
0.914
0.833
0.849
O.SGG
0.883
0.901
0.823
0.839
0.855
0.872
0.889
0.815 J 0,808
0.830 0.823
0.846 0.838
0.863 0.854
0.879 0.870
0.994
1.017
1.041
1.067
1.094
1.121;
1.153
1.182
1.228
1.255
0.970
0.992
1.015
1.039
1.064
0.950
0.971
0.g93
1. OHi
1.040
0.934
0.954
0 .. 974
0.916
l.OW
0.919
0.938
0.958
0:979
1.001
0.907
0.925
0.945
0.965
0.985
0.895
0.914
0.932
0.952
0.9:12
0.887
0.904
0.922
0940
0.960
1.043 .1,024
1.048
1,095 1.074
1.124 1.10[
1 155 1.131
1.007
1.031
1.055
1.081
1.110
993
1.015
1.039
1.064
1.091
0.980
1.002
1.025
1.049
1.075
1.140
1.173
1.210
1.251
1.297
0.8 5
0.8 6
0.87
0.8 8
0.8 9
1,201
1.238
1.272
1.314
1.357
1.157
1.192
1. 223
1.262
1.302
1. 091
1.119
1.149
1.181
1. 216
1.065
1.092
1,120
1.151
1.183
0.9 0
0.9 1
0.9 2
0.9 3
0.9 4
1.401
1.452
1.505
1.564
1. 615
1.218
1.257
1.300
1.348
1.403
1.189
1 225
1.266
1.311
1.363
1.163
1.197
1.236
1.279
1.328
1.120
1.152
1.187
1.22G
1.270
1,103
1.133
1. 166
1.204
1.246
1.582
1.665 i
1. 780
1.853
1.946 i
2.056!
2.212
2:478
3.097
1.467
1.,515
1.644
1.707
1.783
1.423
1.497
1.590
1.6'19
1.720
1.296
1.355
1.431
1.47\1
1.536
11.737
1. 833
1.652
1. 741
0.9 70 1.969 1.866
1.945
75
2.055
09
0.9 80 2.164 2.045
0.9 85 2.294 2.165
0.9 90 2.477 2.333
0.9 95 2.792 2.621
0.9 99 3.523 3.292
LO 00
" I ~
LO 01 3.317 I 2.931
LO05 2.587 2.266
1.0 10 2.273 1.977
1.0 15 2.090 1.807
LO 20 1.961 1.711
1.779 1.531
1.0 3
1.651 1.410
1.0 4
1.552 1.334
1.0 5
LO 6 1:472 1 250
LO 7 1.404 1.195
0.9
O.g60
(](J
1.346 1.139
1.08
1.295 1.089
1.09
1.1 0 1.250 1.050
1.1 1 1.209 1.014
0.981
1.1 2
1.172
1.518
1.601
707
1.773
L855
1.959
2.106
2.355
2.931
11.
1.0G8
1.880 1.812
2.017 UNO
2.250 2.159
2.788 2.6133
0:>
'"
2.640
2.022
1.757
1.602
1:493
2.399
1.818
1.572
1.428
1.327
2.184
1.649
1.419
1.286
1.191
2.008
1.50[l
1.291
1.340
1.232
1.150
1.082
1.026
1.186
1;086
1.010
0.948,
0.896
1.060
3.967
0.896
0.838
0.790
0.978
0.935
0.897
0.864
0.833
0.851
0.812
0.777
0741;
0.718
0.749
0.713
0.681
0.652
0.626
0.565
0.631
0.601
0.575
0.551
1.16G
1.078
1.14
1.15
l.It'
1.17
1.13
1.19
1.20
1.22
1.24
1.26
1.23
1.30
1.32
1.34
0.485
0.457
0.433
0.411
0.392
2.4 ,
2.6
I 2.8
1!~3.2
'.3.4 ...
3.6
0.602 0.5290,468
11.10710.9210.7800 (l69 0.5810.50910.450
1.078 0.892 0.756/0.64710.561 1),490 0.432
1.052 0.870 0.734 0.627 0.54.2 0.473 i 0.417
1.027 0.850 0.713 '0.608 0.525 0.458 ' OA02
0.825
0.810
0.787
0.755
0725
0.694
0.676
0.659
0.628
0.600
0,591
0,574
0.559
0.531
0.505
OA17
OAOO
0.384
0.369
0.356
(3.8
4.0
0 .'374'
0.358
0.343
0.329
0.317
0.509
0,494
0.480
0.454
OA31
0.305
0
O.
O.
O.
.0219
0.23310.205
0.219\ O. HJ3
0.207 0.11l1
0.191i 0.171
0.185 0.162
1.003
0.981
0.950
0.922
0.1387
1.46
1:48
l.50
1.55
L60
1.65
1.70
1.75
1.80
L85
1.90
1.95
2.00
210
2.20
2.3
0.325 0.235 0.173 0.120 0.09810.075
2.4
0.308 0.220 0100 0.1:1\10.0890.058
2.5
0.2920.2070.1500.110 0.082 0.062
2.6
10.277 0.197 0 140 0.102 0.07G 0.0.57
2.70.264 0.188 01310.005 0.0700.05'3
11.
Q.535
0.506
0.480
0.457
0.436
F(11,N) (conlin'l!ed)
0.595
0.563
0.536
U.511
0.488
2.2
1.36
1.33
00
00
0.769
0.752
0.766
0.780
o 794
0.8 1
0.8 2
0.8 3
0.8 4
0.80
~",1
4.0
0.940 0.913
0,961,0933
0.98510.954
1.007 0.976
1.031 0.998
LOli6 1.022
1.083 1.046
l.UO 1.072
1.139 1.099
1.171 1.129
0.7 5
0.7 6
0.7 7
0.7 8
0.7 9
D-l.
0,441
0.415
0.392
0.372
0.354
2.8
2.9
3.0
3.5
4.0
0.2520.1760.124 0.08(10.01350.0-180.0360.0270.0200.015
[0.241 O.WG 0.11.7 0.083 (J.050 0.044.0.03:3 0.024 0.018 0.014
;,0.230 0.159 0.1l0,1 0.078 0.056 0.0.J.1' 0.0::\0 0.022 0,{)17 0.012
0.1900.1260.085'1.0590.0410.(21) 0.0210.015 0.01l 0.008
0.1610.104 0.0690.0460.0310.022 0.015 0.010 0.007 0.005
4.5
5.0
6.0
7.0
8.0
0.1390.0870.0571.037.0025
0.122' 0.076 0.048 0.03110.020
0.098 O. aGO 0.0313 0 022 0 . 014.
0.081'0,048 0.028,0.017 0.010
.0.OGg 0.040 0.022 0.013 0.008
9.0
10.0
20.0
.J.
00.58
0.052
0.0.,!7
0.043
0.030
O.Oli
0.013
0.009
0.006
0.005
0.01110.008
O.oun 0.006
O. naG IJ. 004
0.004 O. 002
0.003.0.002
0.005
O.OOt,
o. 002
0.002
0'001 '
0.004
0003
0.002
0.001
0.001
.~
),
,
'\
I
1
(J44
TAlILlll
D-1.
4.2
0.00
0.02
O. [11
0.00
0.08
0.10
0.12
O.H
0 III
0.18
\).2f)
0.2'2
0.2-l
o.'ili
0.28
0.30
0,32
0.34
0.31;
0.38
U,4ll
0.42
0.4-1
Vi
0.000 0.000
O.O:W 0.020
O.O.fU 0.040
O. OM O.OUO
0.080 0,080
0.100
0.120
0 1
. ..10
1 0.160
0.180
. 0.200
0.220
0.240
0.260
0.280
5.0
a 000
0.020
0.040
O.OGO
0.080
6.2
0.000
0.020
0.040
0.000
0,080
(I,son
0300
0321 0.320
0.341' 0.3-l0
O.3lil! 0.301
U.:j81 0.381
OA02.0AOI
o , 422 ' 0.421
o H31 0.442
0..163 I tL -!()2
o , ~ 84 . U. 483
D .300
0.320
0.340
0.360
i 0 .. 381
TADLE
F(lt,N) (contintled)
G.B
0.000
0.020
o 040
O.OGO
0.080
0.000
0.020
0.040
O.OGO
0.080
0.000
fJ.020
0.040
0.0.60
0.080
0.80
0.81
0.82
0.83
0.84
0.100
0.120
0.1'10
O. HlO
0.180
0.100
0.120
0.140
0.160
0.180
0.100
0.120
0.140
O,lliO
0,180
0.100
O. J20
0.140
O.WO
o 180
0.100
0.120
0.1'!0
O.WO
0.180
0.85
0.86
0.87
0.88
.0.89
0.950
0.970
0.990
1.012
1.035
0.935
0.95-1
0.973
0.99-4
1.015
0.200
0.220
0,240
0.360
0.280
0.200
0.220
0,2!O
0.260
0.280
0.200
0.220
0.2-10
0.260
0.230
0.200
0.220
0.2'10
0.260
0.280
0.200
0.220
0.240
0.90, 1. 08711.060
1.116 1.088
0.91
'0.92 .1.148 1.117
1.184 1.151
'0.93
1.225 1 188
0.9t
0.950 1.272 1.232
0.060 . 1.329 1.285
o .970 \1. 402\1. 351
0.975 1, 447 1 1 ,393
0.980 . 1.502 1.443
1.039
1.064
1.092
1.123
1.158'
0,2~0
0.380
0.985
0.990
0.995
0.999
1.000
I . -,
0.711
0.71
o.n
o.n
0.74
0.75
0.7il
U.77
0.78
0.71)
0.735
0.7'18
0.761
0.774
0.788
0.802
o 817
O.8:n
0.847
10.8\l2
0.7'21
O.HO
0.752
0.765
0.779
0.722
0.7'34
0.7-1u
0.759
0.711
10.798
0.71710.714
0.729 I o.nu
0.741 0.737
0.753 0.749
0.7GG 0.761
5.0
g:~~~1 g:~~~
D-l.
7.8
645
APPENDIX D
APPENDIXES
0.712
0.723
0.73-1
0.H6
0.757
0.759
0.771
0.782
O. 71B
0.805
,)
0.845 0.8.38
0.860 0.852
0.875 \ 0.866
0.890. 0.S81
0.906.0.897
1
0.923 0.912
0.940 \ 0.030
0.95!) (1.9-17
0.97810.966
0.999 0 986
G.G
6.2
5.8
0.893
0.913
0.929
0.1146
0.964
1.045
1.072 i
1.1?11
1. li;14
1.029
1.054
1. 081
l:1l3
1.016
1.039
1.065
1.095
",
1.033
0.737
0.610
0.537
0.'186
I
..,
""
0.951 0.870
0.669 0.612
0.551 0.502
0.483'0."1-4.0
0.436 0.394
L03
1.04
1.05
1.06
1.07
1.13.
1.14
1.15
1.16
1.17
0.305
0.291
0'.278
0,26D
0.255
1.18'
1.19.
1.20
1.22
1. 24
0,244
0.235
0.226
0.209
0.195
0.3-11
0.319
0.299
0.282
0.267
Q,';!g2
0.272
0.254
0.239
0.225
0.252'
0.234
0.218
0.204
0.192
0.253!
0.240
0.229
0.218
0.208
0.1!)\}
0,1111
0.183
0.168
0.155
0.212
0.201
0.191
0.181
0.173
0.18'1
0.170
0.161
0.153
0.145
0.165
0.157
0.150
0.138
0.127
0.887 0.88~
0.901 0.896
0.91() 0.911
0.932 0.927
0.949 0.943
0.891
0.90(l
0.922
0.938
0.956
o lUO
0.986 1 0.979
1.006 0.999
1.029 1.021
I 1.054 1.0H
1.417
1. 036
0,873
0.778
0.111
0.4.03
0 379
.
1 0.357
0.338
0.321
1.264 1.138
0.915 0.817
0.766.0.681
Q.68010,602
o li20 0.54.6
,
.7.8
0.905
0.921
0.937
0.955
0.974
7.4
1.199
1.248
1.310
1.34.8
1.395
7.0'
o . 832 .
0.84(l
0.8GO
0.874
0,889
1.001
1.00f!
1.010
L015
1.02
1.08
1.09
L 10
1.11
1.12
0.705
0.716
0.727
0.737
o 748
.5.4
F(It,N) (cohtimled)
o . 220 O. 194
0.204 0.179 i
0.18910.165.
O. 17G 1 0,] 5,1
0.IG50.143
1.003
1,027
1.050
1.080
0.905
1.016
1.040
1.066
l.Ul
1.149
1.197
1.227
1.262
1.09.7 1.084
1.133 1,119
1.179 1.162
1.207 1.190
1.241 1.221
1.309
1.313
1.479
1,725
1.2B4
1.3<1&
1.451
1.678
L073
1.106
1 148
1.173
".201
1.23S 1.243
1.319 1.297
1.416 l.3SS
1.635 1 596
eo
00
'"
"
0.803
0.553
0.459
0.S\19
0.358
0'.746
0.526
0,422
0.366
Q.327
0.697
0.481
0.389
0.33()
0.300
0.651
0.447
0.360
0.310
0.271)
0,228
0.105
0.1740.154.
0.136
0.172
0.158
0.146
0.135
0.125
0 153
0.140
0.129
0.119
0.1l0
0.137
0.125 '
0.114
0.105
0.097
0,123
0.112
0.102
0.094
0.086
0.155
0.H6
0.137
0.130
i 0,123
0.135
0.126
O.l1S
0.111
0.105
0.111'
0.109
0.102
0.096
0,090
0.102'
0.095'
0.089
0.084
0.078
0.090
0.084
0.078
0.072
0.068
0,080
0,074
0,068
0.064
0,060
0.1113
0.110
0.W5
0.095
0.08ti
0.099
0.094
0.OB8
0.080
0.072
0.085
0.080
0.076
0.M8
0.060
0.073
0.068
0.064
0.057
0
0,063
0.059
0.056
0.049
0.044
D.OSS
0.051
0.048
0.042
0.038
TABLE
TABLE
D-L
,~Z I
4.2
4.6
I 5.0 I'
5.45.~
6.2
0.1l71~.095
~O;
6.6
7.0'
7.4
7.8
1.36
1.38
1.40
] .42
1.44
0.134
0.127
0.120
0.114
O.lOS
0.103
0.097
0.092
0.OS7
0.082
0.081
0.076
0.071
0.067
0.063
0.064
0 060
0.056
0.052
0049
0.052
0.048
0.044
0.041
0.038
0.042
0.038
0.036
0.033
0.030
0.034.0.028
0.032 0.026
0.028 0.023
0.026 0:021
0.024 0.019
fl.023
0.021
0.0]9
0.017
0.016
1.46
1.48
1.50
1.55
1.60
0.103
0.098
0.093
0.083
0.074
0.077
0.073
0.069
0.061
0.054
0.059 0.046
0.056 0.043
0.05310.040
0.046 0.035
0.040 0.030
0.036
0.033
0.031
0.026
0.023
0.028
0.026
0.024
0.020
0.017
0.022
0 ..021
1).020
0.016
0.013
0.018
0.017
0.015
0.012
0.010
0.014 I 0.012
0.01310.010
0.012 0.009
O.OlD 0.008
0.008 0.006
1.65
1.70
1.75
1.80
1.85
0.067
0.060
0.054
0.049
O. 045
0.048
0.013
0.038
0.034
0.031
0.03,5
0.031
0.027
0.024
0.022
0.026
0.023
0.020
0.017
0.015
0.019
0.016
0.014
0.012
0.011
0.012
0.010
0.009
O. 008
0.011
0.009
0,008
1
0.007
a .00B
O.OOS
0.007
0.006
0.005
O. 004
0.006
0.00.5
0.004
0.004
0.003
1.90
1. 95
2.00
2.10
2.20
0 . 010
a .008
0.007
0.006
0.00.5
0.007
0.006
0.005
0.004
0.004
0.005
O. 004
0.004
0.003
O. 002
0.004
O. 003
0.003
0.002
a .001
0.003 0.002
0.002 0.002
0.002 0.001
0.0010.001
0.001 0.001
2.3
2.4
2.5
2.6
2.7
0.022
0.019
0.017
0.015
0.Ot3
0 . 003
0.002
0.002
0.001
0.001
0.002
0.001
0.001
0.001
0.001
0.001
0.001
0.001
0.001
o.oeo
0.001
0.001
0.000
0.000
0.000
0.014
0.012
0.010
0.009
0.008
I 0.014
0 009 0.0060.004'1
0.008 0.005 0.00:3
0.006 0.004 o.ooa
0.005 0.003 0.002
0.005.0.003 0.002
1
0.019
0.017
10.016
0.014
0.013
1
0.005
a 004
0.003
0.003
0.002
0.001
0.001
0.000
0.000
0.000
2.8
10.012'0.0070.00410.002 0.001 0.001 o.oor 0.0000.000 0.000
2.9
0.0100.0060.00410.0020.0010.0010.0000.000 0.0000.000
3.0
10.009, 0.005 0.003 0.002 0.001 0.001 0.000 0.000 0.000 0.000
3.50.006 0.003 0.002 0.001 0.001 0.000 0.000 0.000 0.000 0.000
4~0
,0.004 0.002 0.001 0.000 0.000 0.000 0.000 0.000 0.000 0.000
'il.0
10.0
20.0
0.000
0.000
0.000
0.000
0.000
0.000
0.000
0.000
0.000
0.000
0.000
0:000
0.000
0.000
0.000
D-L
F(u.,N) (continued)
4.5
5.0
6.0
7.0
8.0
647
APPENDIX: D
646
0.000
0:000
0.000
0.000
o. 000
0.000
0.000
0.000
0.000
0.000
0.000
0.000
0.000
0.000
0.000
a . 000 0.000 0.000 0.000 0 . 000 0 .000 o. 000 I 0.000 o. 000 0: 000
0.000 0.000 0.000 0.000 0.000 0.0000.000) 0.000 0.000 0.000
0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0 .. 000
1
8.2
0.00
0.02
0.04
0.06
0.08
0.000
0.)20
0.040
0.060
0.080
0.10
[l.12
o 14
0.16
O.lS
0.20
0.22
9.0
S.6
9.8
9.4
----
----1----
0.000
0.020
0.040
o 060
0.080
0.000
0.000
0.020
0.020
0.040
0.040
O.OGO .. 0.060
0.080
0.080
0.100
0.120
0.140
0.160
0.180
O.lDO
0.120
O.UO
0.160
0.180
0.100
0.120
0.140
O.HiO
0.180
0.200
0.220
0.240
0.260
0.280
0.200
0.220
0.240
0.260
0.280
0.300
0.320
0.340
0.360
0.380
0.000
o 020
0.040
0.060
0.080
0.100
0.120.
0.140
0.160
0.180
0.100
0.120
0.140
0.160
0.180
0.200
0.'220
0.240
0.260
0.280
0.200
0.220
0.2'10
0.260
0.280
0.200
0.220
0.240
0.260
0.280
0.300
0.320
0.340
0.360
0.380
0.300
0.320
0.340
0.360
0.380
0.300
0.320
0.340
0.360
0.380
II 0.300
0.320
OA8
0.400
0.420
0.440
0.460
0.480
0.400
O. '120
0.440
0.400
0.480
0.400
0.420
0.44.0
0.460
0,480
0.400
0.420
0.440
0.460
0.4CiO
1 DADO
0.420
0.440
0.460
0,480
0 . .50
0.52
0.54
0.56
0.58
'0.500
0.520
0.540
0.561
0.581
0.500
0.520
0.540
o ..'iB0
0.581
0.500
0.520
0.540
0.560
0.580
0.500
0.520
0 . .')40
O.5uO
0.580
0.60
0.61
0.62
0.63
0.64
0.601
0.611
0.621
0.632
0.6'12
0.1301
0.611
0.621
0.631
0.641
O.GOI
0.611
0.621
0.631
0.641
o 500
0.520
0.540
0.560
1 0.580
0.600
0.6J1 .
0.621
0.631
0.641
0.69
0.652
0.662
0.673
0.683
0.694
0.652
0.662
0.672
0.683
0.693
0.651
0.662
0.672
0.682
0.692
0.651
0.661
0.672
0.682
0.092
0.651
0.661
0.671
0.681
0.692.
0.70
0.71
0.72
0.73
0.74
0.704
0.715
0.726
0.736
0.747
0.704
o 714
0.725
0.735
0.746
0.703
0.713
0.724
0.734
0.745
0.702
0.il3
0.723
0.734
0.744.
0.702
0.712
0.723
0.733
0.744
0.75
0.76
0.77
0.78
0.79
0.758
0.769
0.780'
0.792
0.804
0.75i
0.768
0 .. 779
0.790
0.802
0.756
0.767
0.778
0.789
0.800
0.755
0.766
0.777
0.788
0.799
0.754
0.765
0.776
0.787
0.798
O.U
0.26
0.28
0.30
0.3~
0.34
0.36
0.38
0.40
0 .. 42
0.44
0.46
0.65
0.66
0.67
0.68
(continued)
---
t
0-
.l
l
0.340
0.360
0.380
.!
0.600
O.GlD
0.621
0.631
0.611
1
I
/ 1
.,I
-~-~
~.-~
-.I-~-.--~-
648
APPENDIXES
APPENDIX D
r
F(>II ,N) (COl1iin.l~ell)
TABLE
D-l.
0.8
0.80
0.81
0.S2
083
0.84
0.85
:'
II
0,80
0,37
0.88
o.S!)
o,no
a 815
0.827
0.8:l\)
0 852
O.SUS
0.878
0.SU2
rUJ07
a 921
0,937
0 (,)54
{UJI
0, !r72
O.!)2
() II:!
O.lllll
O.HI
I.U:Hi
,I
0.\)50
'(J,Hi5
(I. \l80
0.\i85
0.\)\11)
1.130
!J!i!)
0,\);0
n,D!).)
()
I'
t,
,
I/.
O,ntH!
1.O!H)
1.001
1.005
1. U10
:1.(jJ5
I.U:20
1.03
J 0-1
1,05
'j
I
J
I 2:2-1
1 275
1,3133
1.5UO
'"
0(\14
0 -1:.'0
0.::137
0.:'>8U
0,257
0,2U
O.ln
1. 13
n.07l
1.][
1. 15'
I 157
I 187
1.12
1.07
108
1.1J!1
1.10
1.11
,,'
1.002
1.0117
0 158
0.140
0.J:.l3
0.111
0, 1111
0. Uti:.'
0.084
li.077
LOU
1,012.
1.16
1.17
1.18
j,
1.19
"j
1 ,).)
1.24
1.20
O,OUS
0 001
Q,OS!i
0.052
0.048
O.!H5
O.OH
0,037
0.032
0:813
0.825
0 8;j7
0.S4\)
0.80:.!
0.875
a 389
0,\)U3
0,918
0,\)33
0.\)4\)
0 (.)()7
0 980
1.()OU
1,0211
105:;
1.085
1.1:.1
I 147
1.175
1.210
1 :WO
L:H2
1.530
""
0.&77
o .3ll1
0.313
O.:W!J
O. :.>:;17
0.1~5
0, !li5
0: l-l3
0.]'17
0.11:'>
0. J(j I
O.O!)1
0.08;3
0.U75
O.OUU
0.OU:3
0,058.
O.Oi>4
0.050
a.uw
0.042
0,0;~9
0.037
0.032
O.0::!8
0.811
0.823
0.8:35
O.S4/
0.8LiO
O.
O.
O,!)OO
0.0i4
O,('):l~)
0.914
0.801
.0.USO
G,I);)\)
1.02:2
I>OH
1.074
i 112
1 1:3-1
1 lUO
1.
1
1
0.810
0.8:,\2
0.8:l3
0.815
0.858
0.870
0,883
0, sn:'
0.011
0.925
0.010
0.057
0,1),15
0 Ufil
LOlli
1 . 010
1.0Li3
1.100
1. 122
1.150
1.
1
1.3U2
1.470
1.
co
o,SW
O.:lliS
0.2!14
0
O.
0.1
O.
U.
o.:ns
o 2:37
O::?UU
O. liD
0,14:3
0.12-1:
0, IlL!
(J.IOU
0.10:.'
0.0\11
O.
0
0 Oli7
0,0()3
0.08:2
O,UU
0,()()7
00li2
0.051)
0,05:.'
0.018
0.0.J5
0.0-11
0,037
o 0;;4
0,032
0.0:'>8
0.024
O.OliO
O,USS
0.050
O,OH)
0,043
0,040
0,1)30
0,0;33
O,O:jO
U,023
0,024
0.0:.n
I.ZG
0.820
1.28
0.8:31
0.8H
l.30
1.32
u.85iJ
0,S08
0,881
0.894
1.3-1
1.36
1.38
O.ll:.!:.!
1 42
lAO
O,DOS
1,4,4
0.93;"
1.'16
1.48
%3
0,070
1.50
1.55
O. ~'Sli
1 >UlO
LO:>::!
1. GO
1 65
1.70
1.053
1.087
1. 108
1.132
1.165
1,208
I 280
IA47
""
0.519
0,350
8.2
---
0.808
1.7.5
1.80
1.85
1. 90
1.95
2,00
2.10
2,;20
2,3
:2 4
2,5
'"
0.494
0,:,3]
0.262
(I
:223
2,1\
2.7
0 190
O.tSu
O,(:H
0.lI5
2.8
2.9
3.0
3.5
4.0
4.5
5,0
6.0
7.0
8.0
IJ.O
10.0
20.0
0.OD8
0.086
0,077
O.OLiv
0.OW2
0.055
0050
0,045
0.041
0.0:38
0,035
0.0;;2
0.029
0.027
0.025
0,021
0 018
f
r,
,..\
649
0.028
0.025
0.022
0.020.
0,018
0.016
0,014
0.013
0.011
0.\)10
O.OO!)
0.009
0.008
O.OOG
.0.005
0.004
0003
0.002
0.002
0.002
O,OOI
O.OOl
0.001
0.001
0.000
0,000
0.000
0.000
0.000
0.000
0.000
0.000
0.000
0.000
0.000
0.000
0.000
0.000
o.obO
. 0.000
U.OOO
0,000
O.OOU
8.6
-------~-
0.02-1
0.021
O.OH)
0 .. 017
0.015
0.013
0,012
0.011
0.009
0.008
0.008
0,007
O.OUO
0.005
0.004
0.003
0,002
0,002
0.001
0.001
0.001
0.001
O.OOl
0.000
0.000
0.000
0.000
0,000
0.000
0.000
0.000
0.000
0.000
0.000
0.000
0,000
0.000
0.000
0,000
0.000
0.000
0.000
0.000
I.
9.0
F{1!,N) (continued) .
9.4
9,8
0.018
0.016
0,014
0.012
0.010
0.009
0,008
0.007
0,006
0.00(3
0.005
0.004
0.00",
0.003
0.002
0.002
0.001
0,001
0,001
0,001
0,001
0.000
0.000
0,000
0,000
0,000
0.000
0:000
0.000
0.000
0.000
0.000
0,000
0,000
0.000
0.000
U.OOO
0.000
0.000
0,000
0.000
0,000
0.000
0.016
O.OH
0.012
0.010
0.009
0.008
0.007
O.OOG
0.005
0.005
O.OOt
0.004
0.003
0.003
0.002
0.001
0.001
0,001
0,001
0,001
0.000
0,000
0.000
0.000
0.000
0.000
0.000
0.000
0.000
0.000
0.000
0.000
0.000
0.000
0.000
0.000
0.000
0.000
0.000
0,000
0,000
0.000
0.000
---
0.021
0.018
0.016
0.014
0.012
0.011
0.010
D.OO?
0.008
0.007
0.00(3
0.005
0.005
0.004
0.003
0.002
0.002
0.002
0.00]
I 0.00]
! 0.001
0.00]
0.000
0.000
0.000
0.000
0.000
0,000
0.000
0.000
0,000
0.000
0.000
0.000
0.000
0.000
0.000
0.000
. 0.000
0:000
0.000
0,000
0.000
'.
II
650
'I
~N
''''1
APPENDIXES
D-2.
TAlJr.E
Tlll'l
V A.RIED~Ft,OW
.-~-
~~l~~--=- ~:~I~I~I~~I~I~!~~J~
:1
1:'J
i,
0.00
0,02
0.04
0.06
0.08
0:i20io:~2010:~~lO
!'
'1'
8:igg
0.30
0,32
0.34
0.36
0.38
0.29110.2930.29410.295 0.2[)5IO,297
0.308,0.311. 0.313 J 0.31410.31610.317
.0.326 0.329 10.331 0.333 0.33510.337
0.344 ,O.347! 0.350 0.36210.354 0.356
1
0.362,0.355
0.3GS 0.371 0.373 0.374
0,40
0.42
0,44
0.46
0.48
0.38010,384 0.387,0.3901.392
0.397 '.401.0.405,0.407 0.409
0.414 10,419 0.423; 0,426,0.429
0.431,0.437
0.440 0.44410.447
0.447,0.453 0.458 OAfH 0.464
0.50
0.52
0.54
0.56
0.58 .
0 479
0.494
0.509
0.524
0.485
0.501
.0.5H
0.533
0.60
0.81
0.62
0.63
0.64
0.65
0.66
0,67
0.68
0.69
0.540
0.547
0 .554
0.562
0.569
0.576
0.583
0.590
0,597
0.603
O.GW
0.617
0.624
0.630
0.637
0.548 1,0.555
0 556 10 55S
0.553 0.571
0.571 0.679
0.579 0.586
0.585 0:592
0.593 0.600
0.599 0.607
0.607 0.e15
0.613 0.621
0.620 0.629
0.627 0,636
0.634 0.1l43
0.641 0.650
0,648 0.657
0.70
0,71
0.72
0.73
0.74
0.75
0.76
0.77
0.78
0.79
0.491
0.507
0.523\
0,539
0.494,0.499
0.512\ 0,516
0.523 0.533
0.54.5' 0.550
0.298,0.299
0.318! 0,319
0.3381' 0,339
0.357 0.358
0.376,0.377
0.339
0.358
0.377
0,298
0,318
0.333
0.357
0.375
0.394 0.39510.3'J6
0.412'1'0.41310A14
0 432,0.433 \ 0.434
0,451' 0.452 0.453
00467 0.469 0.471 0.472
O.39G
0.4U
0.435
0.154
0,47:,
0.393
0./J,11
0.430
0.449
0.502
0.520
0.537
0.554
0.48910,491
0.5OS 0.507 0.509
0,522 0.525 0.527
0.54.0 10.543 0.545
0.558 0.561 0.563
0.5Gl
0 .'569
0,513
0.585
0.592
0.599,
0,607.
0,614
0.622
0.629
0.537
0.544
0.651
0.659
0,665
0.644
o 6.51
0.658
0.665
0.672
0.649
0.657
0.664
0.672
0.679
0.5M
0.661
0.669
0.677
0.684.
0,578,0.581
0 58710 58<:1
0:595 0.598
0.603 0.607
O.61I 0.615
0.519' 0.623
o 626 0.630
0.635 0.039
0.643 0.647
0.651 0.655
0,659 0.653
0.666 0.671
0,674 0,679
0.682 o 687
0.689 0.694
0.299
0,3Hl
0.402
0.511
0.529
D.5,n
0.567
0.583
- 00.6059g
0.60,
0.618
0,626
0.63
0.643
0.651
0.659
0.667
0.674
0.682
0.691
0.698
0.643 0.655 0.664 0.671 0.679 0.686 0.691 I O.G96 0.701 0,705
0.649
0.656
0.662
0.fi68
0.661
0.667
0,673
0.680
0.670
0.677
0.683
0.689
0.679
0.G85
0.G92
0.698
0,687
0.693
0.700
0.705
0.693
0.700
0.707
0.713
651
APPENDIX D
0.G09
0.705
0.713
0.719
0.704
0.711
0.718
0.724
0.709
0.715
0.723
0.729
0.713
0.719
0.727
0.733
2.2
2,4
2:6
3.8
0.685
0.H91
698
o 703
0.709
0.095
.701
0.707
0.713
o . 719
0.703
0.710
0.717
0.722
! 0.729
0.741
0.749
0.755
0.762
0.769
,.
0.80
0.8!
0.82
0.83
O.S-!
0,074
0.680
0.686
0.692
0698
0.704
0.710
0.715
0.721
0.727
0.732
0,738
0.743
0.7,19
0.n4
0.715
0.721
0.727
.0.733
0.B8
0.739
0.89
0.744
0.90
0.750
0.91
0.754
0.92
0.761
0.93
0.767
0.94
0.950 0.759 C),H2
O.gGO o. 7ii41 0,777
0,970 0.770 0.782
0,975 0,772,0.785
0.980 0,775 0 , i87
0,985 0.777! 0.790
0.990 0.780 0.793
0.995 0.782 0.7!l5
1.000 0.785 0.797
1.005 0.788 0.799
0.754j 0.i6,t
0.710 0.770
0.766 0.776
0.77210 782
o 777 0.787
0.783 0,793
0.788 0.798
0.793 0.803
0.796 0.805
0.79S 0.803
0.801 0,811
0.804 0.814
0.806 0.816
0.808 0.818
0.810 0.820
1.010
1.015
1.020
1.03
1.04
0.812
0.815
0.818
0.822
0.829
0.85
0.81\
0.87
LOS
1.00
1.07
1.08
1.09
1.10
1.11
1.12
1.13
1.14
1.15
1 16
1 17
1.18
1.19
1.20
1.22
1.24
1.26
1.28
0,790 O.80~
0.793,0.80",
0,795 0,807
0,800 0.811
O.80li 10,816
0.72510.735
0.731 0.741
0.7381'0:747
0.74310.753
0,749Io.i58
0.822
0.824
0.828
0.832
0.837
0.765
0,744 0.752
0.771
0.750 0.758
0.777
O.7oG 0.764
0.783
0.762 0.770
0.789
0.7671o.77G
0.781 0.789 0.795
0.787 0.795 0.801
0.7!)3 0,800 0.807
0.709 0 807 0.812
0.804 \ 0.813 0.818
0.801 0.80li 0,819 0.823
0.807 0.815 0.824 0.829
0.812 0.820! 0,82(j 0,834
0.814 0.822,0,828 0,836
0.81810.825 0,830 0.839
0.820 O.827i08~3IO,841
822 0 830 0 8.~7 0.844
0.82408:)210.84010.847
0.826 0.834 0.842\ 0.849
0.829 0.837 0.845 0.862
0,831 0,840 0.847 0,855
0,833 0.843 0,850,0.808
0.837 0.84.; 0.8~3IO.8~0
0.841 o . 850 0.8.)7 i O. 8/)4
0.845 0.855 0:8621,870
I) .864
0.871,0.879
0.869! 0.8761 0.88:)
0.873 0.880 i 0.887
0.877 0.885 io .892
0.895
0,899
0,902
0.906
0.910
0.913
0.917
0.925
0.9;:!')
0.945
0.921
0,929
0.935
0.942
0 .. 948
0.928
0,932
0.940
0.948
0.954
0.9,14
o.o:>s
0.945
0.954
0.%0
0.770 0.775
0.777 0.782
0,783 i 0.788
0.789! 0,7~J4
0, 7'..l51 O.BOO
0.8010.807
0.807 0.812
0.81310.818
0.818 0.823
824 0.829
0.829 0.835
0,835 0.841
0.840 o 84G
0.84:J (),848
0,845 0,851
0.847 0,853
0.850,0.856
0.75:) j 0,859
0.85(; 0.852
0.858 i 0.864
0.861 . 0.867
0.80 4 10.870
0,8GG i 0.872
0,871 0.877
0877 0.883
0,881 0.887
0.88.5 0.8!B
o .S89 0.896
0.893 0,900
0.898 0.904
0,903 0.908
0.907 a.BI2
0,910 0.916
0.n4. 0.919
0.918 0.923
0,
0.928'
0
0.931
b.
0,9:H
O.
0.939
0
0.942
0.940 0.945
0.944 0.949
0.950 0.. 955
0,900 0.9fl4
0.965 0.970
I
I
II
.... :1
652
TABLE
APPENDLXES
D-2.' .Till':
VARIED-FLOW FUNCTlOT,
. FOR
2.4
l.H
lAG
LIS
L75
:5.1)
;l.5
4.0
4.5
5.0
!:i.0
7.0
8,0
\J.O
10.0
5.0
5.5
0.000
0.020
0.0'10
0.060
0.080
0.000
0.020
0.040
0.060
0.080
1. 000
1.004
1. OO!)
1. 0)
1:1.0Ui
0.10
0,12
0:14
0,100
0.120
0.140
0.160
0.180
0.100
0.120
0.140
0.160
0.180
0.100
0.120
0.1;10
O.HIO
o 100
0,120
0.140
0.160
0.180
0.100
0.120
0.140
0.160
0.180 .
0,200
0.220
.0.240
0.260
0.280
0.200
0.220
0.2':10
0.260
0.280
0.200
0.220
'0.240
0.260
0.280
0.300
0.320
0.340
0.3130
0.380
0.300
. 0.320
0.340
0,360
0.380
0.300
0.320
0.340
0.360
0.380
' 0.073'
0.979
0,984
0989,
0.9\H
0.066
() UTi
0 :978
0.,983
O.US9,
., ( . ,
L05i
1.0U5
1.071
107'
1.085
1.015.1.017
1.0241' 1.020
1.0:H 1.035
1 " a41 /1 . (042
1.049 1.049
1.019
1.028
1.035
1. 042
1.0-18
1.0r.fj
1 A"
v
iJu 1 . 0-5
<:I
I. 053
L064 1 00') 11 ono ' 058
1.0li9
. .} .
1.077 Lg~! i i:2~~ i:g~~
1.081 1.0791 1.075 Lon
1.08211'.073111,075
1. Oil'; 1.0~0
1101 l.OO'!
1.121
4.5
I O.!l'i' j 0.950
O. 90ll
,0.957 0.OG5 0.072
G. Iln4 I 0.072 0, l!7!)
,0.tlm','O.!!77 0.983
0.07710.98310.989
0'\)0 90 0093
.!Hlt L 02 1.007
1.012 1.017 1.0'20
1.020 ,1.029 1.032
1.03011.042 1.0-14
:Ui
2.6
~.8
4.2
0.000
0,020
0.040
0.060
r 0.080
I. Hll
L17li
2 9
0.000
0.020
0.040
0.060
0.080
0.9Gl
0.\)(\7
U. \J7:~
U.iJ79
0.979 0.985
2.~
'l.7
4.0
0.000
0.0"20
0.040
0.060
0:080
0.055
0,\)61
0.907
Ilia
:L:1
~
0.00
0.02
0.04
0.06
0.08
U~~ ~: ~;~
L.W
3.8
0.1175
0
0: ~~6
0.\)91
0.996
1.056
1,.005
1.075 1.074 1.074 1.073 I'.072
L08G 1.085' 1.08-1' 1.082 LOllI.
1.0n7 1.0!)5 i 1.092 1.090 1.087
L85
LDO
l.U5
2,00
2.10
0.981
\J7G
0:
0.987
0.993
105211.05311054
1055
LilO
THE
\J8~
O98~
1.55
1.60
l.(;5
1. 70
D-2.
0.918
0.\).55
0.9(52
0.90B
O.D74
1.50
TABLE
0.941
0.948
O. \130 O. OJ!) i {j. 9JS O. U55
0.9:17 0.9461.954 09Ul
10.IN.! 0.952, O.ouo I. 0.967,
0.\)15
1.~0
NEG'\'TIVESLOI'E~
F(11 N)'-SCI (ci'
.1)
-.
;;1'",
'VtJ..ln'lleI.L,
2.6 :
o.\J:n
L~2
/353
APPENDIX D
1.184
1, 206
1. 22:)
L:2:l5
1,245
1 HiS L15-1
1 185 LllJ7
1 11>8 L i7U
I.?O~ 11.18;3
1.212 1, 188
1,.,/001.3-121202
~ . ~3~ 11. ~G~ I LlO::l
...1
.,,7:, 1.313
LiG! laS4 1.:119
1-171 1.3!),! 1.324
1,252
I. 200
1,261\
1269
1.27:2
i.221 L195
. 1. 225 1. 1gO
1 22[) 1.201
1 2:n 1.20;~
1.233 1.203
L 2!J:l
1.140
l.151
1.158
LHi2
I.HlG
U~g
Ugg
1.128, L117
l.138 I I 125
1. 142 1.129
l.HGl.131
L 149 1.1:14
1.107
1.113
1 117
1.119
1.121
1.136
L l:s0
1.137
1.137
1.137
1.1:12
1.122
1.122
IT'?
1:122
i'Ub
O,HI
0.18.
0.20
0.200
0.200
0.220
0.220
0.22
I
0.2-i I' 0.240
0.240
0.260
0.26 ! 0.260
0.28 , 0.280
0.280
0,30
0.300
0.300
0.32
0.320
0.320
0.339
0.340
0.34
0.36'
0.359
0.360
0,378
0.38
0.379
0.397
0.398
0.40
0,42
0.417
0.418
0.436
0.437
0.44
0,46
0.455
0.'156
0.48
0.474.
0 .. 475
0.50
OA03
0.494
0,512
0.52
0:513
0.54
0'.531 ' 0.532
0.549
0.550
0.56
0.58
0.567
0.569
0.60
0.585
0.587
0.61
0.594
0.596
0,605
0.62
0.603
,
0.IH2.
0.63
0.615
0.620
0.61
0.623
0.65
0.629
0.632
0.66
0.637
0.640 .
0.67
0.646
0.649 .
0.65.4
0.66
0.657
0.69.
0.662
0.665
0.70
0.670
0,673
0.71
0.678
0.681
0.72
0.686
0.689
0.73
0.694
0.698
0.74
0.702
0,705
0,75
0,.709
0.712
0.70
0.717
0.720
0,724
0.77
0.727
Q.78
0;.731
0.735
0.79
0,742
i,738
!
I
II
o.180~
0.398
0.418
0.437
0.457
0,476
0.400
0.420
0.440
0.459
o 479
0.400
0,419
0.439
0.458 .
0.478
0.495
0.515
0.S33
0.552
0.570
0.497
' 0.517
0536
0.555
0.574
0.498
0.518
0.537
0.558
0,576
0.589
0.598
0.607
0.61G
0.G25
0.593
0,602
0.611
0.020
0,629
0.595
0,604
0;013
0.622
0.631
0.634
0.643
0.652
0.660
0.668
0.638
0.647
0.656
0.665
0,674
0.640
0.650
0.659
0.668
0.677
0.677
0.685
0.694
0.702
0.710
0.682
- 0.690
0,G99
0.707
0,716
0.686
0.fi94
0.703
0.712
0.720
0.717
0.725
0.733
0.14,0
0.748
0.724
.0.731
0.739
0.747
0.754
0.728
0.7;36
0.744
0.752
0.760
,
.
.~
654
TABLE
APPENDIXES
~~I
0.80
0.81
0.82
0.83
0,84,
0,85
0,86
0.87
0.88
0,89
0,90
0.91
0.92
0,1l3
0.94
0,950,
0.960
0,970
0.975
0.980
0.985
0,990
0.995
1.000
1.005
1.010
1.015
1.020
1.03
1.04
1.05
1.0G
1.07
1.08
1.09
. 1.10
1.11
1.12
1.13
1.14
1.15
1.16
1.17
l.18
4.0
4.2
4.5
5.0
5.5
0.762
0.770
0.777
0,784
U.791
0.798
0.804
U.811
0.818
0.825
0.831
0.837
0.. 844
0,850
0.856
O.8Cl
0.8137
0,1)72
0.875
0,878
0.768
0.770
0,78:3
o 79Q
0.798
0.805
0.812
0.819
0,826
U.8;{2
0.839
0.845
0.851
. 0.857
0.8G4
0.8!}!)
0.875
0.881
0.883
0.886
0.8S9
0.891
0,894
0.897
0.899
0.90'2
0.904
0.907
0.911
0.916
o 920
0.924
0,928
0,932
TABLE
0.746
0.753
0.760
0,766
. 0,773
0.780
0,78G
0,793
0,799
0.805
0.811
0,817
0.823
0.829
0.835
0.840
0,846
0.851
0.854
[J,8S7
0.859
0,861
0.86,1
.0.SIl7
0.870.
0,873
0.875
0.877
0.882
0.888
a.892
0.896
0.901
0.905
0.909
0.913
0.917
0.921
0,925
0.928
0,932
0.936
0.~3g
.1. 1f)
0,943
0,947
- 1. 20
1. 22
1.24
i,2G
1. 28
0.950
0,95G
0,952
0,9!l8
0,974
0.750
0.757
0,764
0.771
0,778
0.784
0.791
0,797
0.803
0,810
0,816
0,821
0,828
0,833
0.8.,10
0.845
0.8G1
0.8G6
0.859
0.861
U.863
0.867
0.8G9
0.873
0,874
0.878
0.880
0.883
0.887
0.893
0.897
0.. 901
0.906
0,910
0.914
0.918
0,921
0.926
0.929
0'.933
o 936
0.041
0.944
0.947
0.950
0.953
0.057
0.91l2
0.971
0,9,77
0.755
0.702
0.709
O.i7G
0.783
0.790
0,797
0.803
0.810
0.816
0,822
0.828
0.834
0.8.J.0
0.84G
0.852 .
0.857
0.863
0.86G
0.8G8
0.870
0,873
0.876
0.879
0,881
o 884
0.886
0.8S9
0.893
0.898
0.903
0.907
0.911
0.91G
0.920
0.923
0,927
0,931
0.935
0.938
0.942
0.945
0,948
0.951
0.954
0,958
0,964
0,970
0.975
O,I)Sl
0.880
0.883
0.885
0.887
0,890
O.SfH
0.89G
0.898
0.902
0.907
0.911
0,915
0.919
0,923
0,927
0.931
0,935
0.939
0.943
0,947
0,950
0.953
C.957
0.%0
6 9G3
O.96G
0.972
0.077
U.ll82
0.987
0.936
0.94.0
0.941
0.948
0.951
0.954
0.957
0.9GO.
0.903
0.965
0.968
'0.970
0.1)76
0,981
0.986
O.Y\JO
~-I
.~
655
. APPENDIX D
,\
.1
I
i
I,
I
1.30
1.32
1.34
1.36
1.38
1.40
1.42
1.44
1.46
1.48
1.50
1.55
l. 60
1.65
1.70.
1. 75
1.80
1.85
1.90
1. 95
2.00
2.10
2,20
2,3
2.4
2.5
2.6
2,7
2.8
2.9
3.U
3.5
4.U
4.5
5,0
(} 0
7 0
8,0
9,0
10,0
4.0
4.2
.Uj
I- -5.0- - - -5.5- -
0.979 I 0.978
U.nS5 I U.991
U.985 I 0.98G
U,OllO i 0.995
0.9\12
.
0,\195
U.09U
U.999
0.9\)4
0.!1\)6
U ()!)9
1.002
./).998
1.00U ! 1 ,00:3 ,1,00C
1.001 I 1,004
1.000
1.006
1.005 I 1,008
1.010
I.U12
I,U09 1 J ,013
1,014
I.U16
1, U18
1.014
1.016
1.017
1,01\)
1,020
1.016
1.020
. J .022
1,020
1.U22
1. U21
1,02!)
J ,020
I,U28
1.029
[,0:34
1. U:35
1. U:35
J .032
1.041
J.04U
1.039 I I.U:)6
1,04<';
1, U43 I 1.0:30
1047
1.0S2
1. 051
1.047
1.042
1.057
1.05G
I.U51
1.045
1.061
1 U,59
1,054
i .1l17
1. UOS
1. 0(j0 '
1.U57 ! 1.0 ..llJ
1051 .
1.008
1 UG4
1.05\1
1,071
1.0(12 I 1.053
1,068
1.07G
1, U71
I.U5G
1.065
1,058
1. DSa
I . U7:3
1.0G8
1,084
1,079
LU71
I.U60
1.087
1.081
1,073
1,001
1.)00
1.0831.075
1.062
1.U92
1.085
1 OG:3
1.076
1. Ug-!
1,087
1,063
1.077
1. DOG
1.0SS
1.078
I.UG'l
1, U!)8
1 08f)
1.U65
1. Ui\)
I,OY9
1.U!)O
1.0GS.
1.080
1.10:3
1. UIl:3
1.082
1.06(1
1.10G
1 Ulli
1.067
1.U84
1.108
1.0DS
1,085
1,067
1,110
J ,oun
1.068
1.085
1,1U
1.100
1.085
1.068
1.111
1.100
1,086
1.068
1.086 .
1.111
1.100
1 OG8
1.111
1 100
1.086
I.OG8
I. III
1,100
1.068
1.U86
I
I
0.994
0.997
I,U01
1. 005
1.008
J .Ull
1.014
1,0IG
1.018
1,020
1.022 .
1.028
1, (J:3U
I,U:H
I.U:37
1.0:39
I.U41
1.04:3
.1.045
1.U46
1.(]47
i.O-!!l
1.050
1.051
1,0.";2
1.053
1.054
1. U5-!
I,U5!
1.U55
1.055
1,055
I.05!l
J, USG
1 05G
--)
J
i
I
.1
-t
-\
..l
_J)
)\
,(J5G
,()51i
05li
J 05(;
I.U5!i
._' !
,
\
I
ApPENDIX
E.
* Prepared and suppli<>d for publication through the courtcsy of Mr. C. J. Keif()l'
and Mr. H. H. Chu 'of the Department of Public Works, City of Chicago. The step
'lines in the' table show the location of n,ormal depth. They should not be crossed
by interpolation.
657
TABl.E
E-1.
V"aIED-FLOW FUNC'rlON
it.'
.~
0:59
0.58
O.!17
0.56
. 0.55
0.(\.1
0.53
O.5~
0.51
0.50'
OAD
O.~g
..
O.H
0.~6
0.~5
0.44
0.43
0.~2
OAI.
0.40
0,39
0.38
0.37
0.36
0.35
0.3.
0.33
0.32
0.31
<:n
<:rt
<0
~.30
0.20
0.28
0,27
O.2G
O.lUi
0.g4
0.23
0, 2~ .
0.21
0.20
0.19
O.IH
0.17
O.li
0.15
O.I~
0,13
0.12
Q,l1
'.
6~00,O 1809 Lom 3514'0 2110'0, \6931.125.1 {) 10220.0875;0 0139 0 06420.0593,0 0542,0,0529'0.0514'0 050-11,0450 0406'0 033. O.
1rn,IO 0510
0.5641 o.al54jo 0902 0 1>1961--;0 2781;,185110 13920.1113 0 0916 a am Q ,006S 0,058.lI0 0535,0 0492 .048010.0468,0.0459;0.0410,0, om'0,o308 O.
0223'0 0193
1
1
0.583010.0388 0.7246 n.31lo,ruo 4438 0.2260,D.l593 0 1225,0.009010,08t9 0.0711 l0 0804 05281,04811.0441'.04370,"" 0.OH710.0374,O 0339'0,0581
0,0204'0 0177
0.6023 o.f!I:lHIO 76401l.oof!l:l 0.31 H jlU8U5 0 1380 0.1080 {) 08820,01350,0640'0,0545 o.om 0 0,1420.010010.03911.03871' Ol80i,l.OlI0 0.0~09'0,0251
O.0l1l7 l\O 0163
0,02271069260.81-1 1.825 0,21,,7,0 Hl!l1 O.12113 0.09(,-1 1007860.0657 00512'004920.04330.0400 0.0369 0,0360 0.03510.0345 0309 O.
0.0171 0.0148
1
.01800 01500.0136
0,6446,0 7~U:O.S817'03ll(iO:O, ID71!O.131S 10.1051 0,0844 .0701 0.0589 0.051"10 04 H 0 03911.03611.0334 03270 0318\.031310.02811.
0.66S~.0 702:!,~ 0 2112 .1641,0. H8710 092310 01480 00340,05260 04580,04000,03530.0321 0 03030.02950,02880.0283 0025.0,
.01630,01421.0123
1
1
1
o 6947,0 80891
10 21060.1335 0 102010.081\ 0 06630.05570.0412
0 0.10 03590.03181.02960.0214 0 02670.0060 025110,02300.02
0149001290.0112
O. 12~~,O 8717 0 3J2u,O.1707!0, lIS 110 089~IO,0714 1),05871,04960 0421,0310,0 0:,23 0.0230 0.D2M {) 02470.0241.02350 023110,0203 0.01
.0131 O.Olli 0,0102
0.7591,0.97670 2380!0. HHIO.101410.07nllo.OS25 0 05210.04420.0377 0.03311 0200 025710 02390.02210.0217 0.021110 020&10 0\87 I) 01
,OI2~ (} OHIO 0 0002
o S0l51"i9S6-i;o IS3I,O.1136,O.OSHiO 06830.05540,04010 03910.0337002960 02590 02301 02130 0200 O.OI~2 0 01900 0I860.0I68Io.0IM,0,
.0100 0 00~5Io.ooa3
o 6118'1.2033 o. 1115 10060 01510.05970 013710 OlOB 0.0347 (l 03001.0205 (I 02320 02000
01920.0[78'0 0113 0 017Q 0.0
01510 01380
.00991 0085 (1,00'15
1
1
0,95150.2069;0 12150 085810 06.5;6.0522.0 04~910 03610 0308 00261 (1,0235 0 020710 OlaM 0171 Q 016010.015.<10,01521 0
.01351 012110
00S9 (I 007710.00~7
.01210,01!00
00800 00u9 o.0061
\3 l67110.159~IO.I014 0 o73510.(!5fii}io.045110 0317003190,0273002371,021000,84001511001520 DHH1.0139 0, 0136 (10
1
1
0 2650[V 1215;0 os,; 0.0332.0 04~lfO'O 100 ,03:120 0281 0.02400 0210 01860 01640 01460 01350 0127 0 OI~3 0.01211 0
,01OS 0
0071 0 0062 005..
1O.I~17 0 1042 O.U7~' 0 05UO .iH3G Q.0150 10.0201 0 02~7 0.0212 01S60 01640 01450.01300 0121 0 0
.00OB 0
.00050
0014 0 006:! 0 00550 0048
o. 13981 036~,O.
.o,68,a.0'l7'ljo.o30510 02550.02170 0lS1 0.011\.! Q,QIH a.olnla 01150 01070.0
.0095 o.
OOgSIO'
00650 0050 0 004 0 0043
o H07,O.Gna.o
04(l/'O 0324,0,02660,02230,01011 016510 OIH 0 012H 01131.01010.0004 O.
0 00840
00750
,OMS 0 C.Q50 000.3 0.0038
0.O&l5;O.06!}l;"
.0;,8:0.0281;0 02310.01950.01670,01,130 01261,01110.00990 00890.0082 0.007S O.
0.00740
0067 O.
0 00010.00" 0 003& a 0033
0.07330 Gro90
0299;00243;0.0"01,0 0170 O.OHSlI.OI26 0 01100.00970,0087 0.00780.0072 00680
0.00650
,00580.
000151 00390,003{1) D029
1 00580.00030.00000 0068000510.
O.0205:~ 04JO:0.
,02,To.nwo.O!iIIO,01i70,OI26IO,0IlOO 00%0,00850,00160
.00510.
0.OU400 00340 OQ2910 0026
10 0503 " Ol63IQ,027S,o.i)22110 0131!0 01510 01280 OIlO lJ il(l96 0 0084 0.007.1 0.00661.0059 0,0055 0
.00500 00500,
OOH O,OOH 0 0035000300.00260.0023
1
1
0.OH9.0 0307,O.023ZV.O!89jO 0155,0.01300.0110 9 00950,0032 0,0012 0.00640 00570.00520.00470,
0014 0,0013 O.
00380.003510 0020 0 00260 00220.0020
0.O~5!!,O.025O;Q.o:ml G.016! 0 OI3?,0.0I1! 0.0094 0 !l!!S21t1.0U1110 0062 a.oosso 00500.004500041 \I
.00360.00310
MS.J 000310 Uil26 il 00220,00190.0017
o.,L9" 0 Q:).'fi,O 0170,0 0131,o.omlo.00!l. O,OOSl 0 0010 fi. 006! 0 005.10.0043 0 oo.~ 0.0038 0 003610
0023 a 0032 O.
.0029 0.00'18 O.OO~210,OOl~ 0 0017 0.0015
,00:1;( 0 0021\0 00250.0022 ij,OOWIO 00\110 DOli \) 0013
:0.0214 0.0183 1.01<4;0 QI16,0,0096,O 0981 () ~'O69 0.0000 0 00520.0046 0.0041 0 0036jO,0033 0.003110
'O.020:l;1/ O[&!lo.om;O.0098 0.0082'10,006910. 00511 0 0051 0 00.(50,00390.003510,Q031 O.
0025 O.
.00231.00230.0021 0 0019 (),OOl6 0: OOIl 0 OiH~ o OOll
.0033 0.OU39 0 002Hl
.0020 0.00100.00180 00160 0014 0 0012'0 OOW 0 0009
0.Ol59,O Ol~3,O 0102;0.0083'1.0068,0.005810 00,9 Q 00'3,i!.
0.0139,0 OI071Q.OO,~5 0 00690 0051 10.0049'\0.00.!! 1fi.OOS6'O.OO:
C'Wl 0.00251 G!l220
.0017 0.00160,0015 0 00l~ O.OOI~'O.OOIOlo 0009 0:0008
o OilS Q.OOSOIO,OOll 0.005710 0048 IO.00n 0.0035 0.0000 0 0.1 00230
00310,001910.
.0014 0 00Hlo.0012 0 0011 0.001010 00080.00070.000(\
1 ODIM
\0.000400373 Q.0058,O,0048 a 004010 003-10.
00250.VO
00190
00!610
0012'0.00120.00120.001000010 0 IJ008 0.0007000050,0005
0.0077,0.0060 0 OO~910 0039\0.0032,0 0028 O.
.00210 00
OOloIO.OOHjO, 0013 O.
OOH)j;:;.OOIO 0,0010 0 00090.00080.00070 OU05'0 00050 0004
10.OOfi2,~ 00.8
00391.00320.00271.0023.0
001711.,,3
001300012 n oml 0.
.OOOE 0 @08\0 00081.00070000.0.0008 o.OOiJ510 000400004
Q.0050,0.00391O 00320 00261\).0022 o.OO!SI Q
.('214 0 OOlZIJ 0011 {I 0010 00009 O.
0(1)11 000610.00080 0006 O.
.00051.0005 (l 0004 0 00030.0003
o 0040,0,00310 00250,00210.0017 O.OO;S 0
.DUll 0 OOW 000090.0008:0 000'/0
.00060.00050.00050 0005 O.
,000.0.00040,00030 OOuS o.a002
o 00321 00250 002010 00181~.OOH 0 0012 v
0.:>19 000081) 0001 (I
00050
.0005,,.0004 000110 OOU1[o 0004 00030.000310 000> 000020 0002
0.00250 0019000160001300011 0 000!l 10 ifJilS 0 OOO~ 0 ilOOSjo 0005 D.
0904 0
OOO~ 000030.0003000030,00030.00030.00020.0002 U 00020,0001
10.6472 0.5935
0.0019!0,001510.0012jO 0010,0.
<) 000110 OOOB (} 0005
0005'1.00040
00030
00151\),0012,0 00091 00080
0005 G0050 000. 0 OOO~ O.
,00030
1 !TOOtlo 00030
o OOll,O ooo9jO 00010 OIloa!o
.00040
9003
,00020
1
1
0,0008 oooa 0 00050
00010 00010
1 OOO{'O 0004 j Q,tlOO3 0 0003,0.0002 10 {.Q02.0
0000 <> 000510 0001 0
0003.0.
00020 0002IQ.OOOllO 0001 0001000011.00011
1
o 0004 00030 0003 1O.
.0002 10.0002 .000110 0001 0001100001 O.DOC! 000010 0001
1 0001 0001\0 0001,0 0001\
0.00030.0002
(I 00020
.0001 1{L0001 00010 00010
1
1
0.00021 0002 0 000110 0001 .0001 .0001'0.000 1 0,0001'1 0001 1
1
!l 0001 .00010 0001000010.0001
I
1
I '
1
1
o
10
0
IQ
;0
10
10
10
~,000\g.oooirUOII0
,0
0002
0001
0001
0001
\0
'0
I'
10
10
'I
10
10
'I
0
00020000100001
00010 0001 0.0001
00010 0001 0 0001
00010 0001
10
10
."----'---'---
\_--',
]0
TAELE
~'_~I~i
l.OO
0.99
0.98
0.97
0.96
0,95
O,g,
0.93
0.92
0. ~l
. 0.90
0. 89
0.1l8
0.81
0.86
0.85
0.8l
'"
0.83
'" .
0.82'
o
O.BI
O. B!l
0.79
0. 78
0.77
0.76
0.75
0.74
0.73
C.72
0; n
O. 70
0.69
0.68
0.~7
0.68
0,65
0.64
0.63
0.62
0.61
0.60
0.59
0.58
0.S7
0.56
0.55
0.54
0.53
0.52
0;51
0.50
O.iS
0,48
047
04.
o.j,
0.41
0.43
0.42
0.41
0,40
0.39
0.38
0.37
0.36
0.35
0.31
0.33
0.32
0.31
0.30
0'.29
0.28
0.27
O.Z6
0.25
0.21
0.23
0.22
.0.21
O.W
0.19
0.19
0.17
~. 0 .16",.
0.15
0.1I
0.13
0.12
0.11
0.10
0.05
0,.00
50
0.
1 0.5.,
0;60
1.~I~7.I~I~'I~~I.~1
'100
0.0'
FUNC'1'ION'
;--.' O.ll2~50.46i!OO.30i!Ol(j.2IQ2;O.13530.11280.01l51:0.06i!OO.05520.0482
00
0
0
O.
0
0
jO.OOillO.Oool;O.OOOI 0.0001;0.0001 0.0001 0.0001 1.00010.0002,0.0003 0.000.'0.0005:0
lo----IO.OZ:H 10,4622!O.30a. 10. 009210.1380'0.1126'0.084910
1 0480
0.0003,0.0003,0 .0003:0.00030.0003;0.0003;0.0004 0.0005'0.0006,0.000.7;0 ..Q009'a.aa13:0.0010 Q0050:0.0180'0 4572,0.30080.0077'0.1374 O. 1120:0.0B46i':
:0419
O.
O.0005rOOOS;O.oo06,O.0006,O.000710 0008;0.0009 1.0010 O.OO]3.0.0016:0oo22Ioooa~,00004 ~O"508 0.2974'0.20560.1366 1l.1l13 O'()8~1 O.
.0418
I
,0.
O.OOOB 0.000810.0009/.0010,0.001110.001210.001410.001510.0019 O.00~4'O.0033 0.00;;3 0.01320
i004.28'0.289g 0,2025 0.13M 0.110510.0881.0671 '10'
.0~77
10.
.0012Io.00J2,0.0013 0.0014,0.0016 .001710.0019 O.OO!UiO.0025' 0.0033:0,0045:0.007510.01711'.0523:..429610.278510.1985 O. J339 0.10931' 08300.066S O.
.0474
0,00150.0015 0.0016,0.00170.0019,0.0021 10.0023,0.
.0028 0.0034 0.0013 0.0061 Il!lS8 o {l213 0.08241.3563:0.2501.1919 0'1321 0.1019 O.callS 0.0663 O.
.lJ.l70
0.001/}jll.ooI910.oo20I0.0022 0.0024,0.0026 0.0028 0.
.0036,0.00430.0054 0.00750.0122,0. U358 0.l!560.2932,O.2220,0.IS38 0.1300 a.lOGS 0 OB16 () 0659 O.
0168
0 0023,0.0024,0.00260.00270.0029 0.00320.0036:0.
.0045,0.00530.0067.0.0093;0.0150.0.0310,0. 16H :O.2465,O.20I4'O.1736!1);!278 0.1001 0:
:06&20,05400: 0466
O.0028t002910 0030,.0033 0036;0.00890.0042:0.
.0054
OJ
12
O.
0181,0.0368
~iO.
21641.18611.16561.1252.0.103410.
.
OM8
0.05;17
0.0462
.0064,0.0081,0.
j
1
1
1
.0641 0.
.0458
0.0034.0.00350.0036,0.00390.0042'000450 0000 0.0056,0.00630.00760.0096 0.0132;0.02J5I)043610 2500.0 19400.1729,0.1577 0.12240.1028 0.
2OO39iO.OOlO,Q.0~12IO.004510:001aJo.0053 O.005s!0.OOG5,o.007,!0.0088,o,0114,o.OJ51'lo.G25s1o.QS22:0. 2070,0. 1793!0.16380. 15120.11981.1000 0.07 .0634 O.
0456
U.lIDH !L0046 O. 00480.0052 0.0056iO.ooGI 0.006710.0076,0.00860. OJ03,O.Cl32,0.1UllJ O.O.03jU0848Io.1S52,O.13B5O.I5fiO;o. 11470.1168,0.
fl7
0627 ()
.0452
:06190:'
.OHa
0,00510.00530.0055
.0017,(1.00860.0099:0, 01l8.0.01S',U.Q21a,O.036Z,O.0S33 :0.1715;0.157:1:0.1478,0.13941.1139\0.
: liT
1
I
1.0059,0.0084 0.
0.00531.0061.00031.00671.00721
,ooS7,O.0097 IO.0112,0. 0]35,0. 011n,O. 02H 10.04il8:0.1193 :0, 1604,.1492 1413 .1344 O.1Il20.
.01 .0613 0.11515 O. 04H
1
1 J29710.1085 0.0931'0,01
0.0004 O.
.0071,0.O~76 0.0081;0.
.0098 O.OllO 0.0127,0.0153 0.02000.028:1.0.0534 10.2360 0.1514,0.1422 0. 1356.0.
.06010 0510'0 0439
0.0071:0,
.0079,0. oORlla.Gogo O.
.0110.0.,0123;0 m4~,O.017(0.02JO 0.0325:0.066,1.)919;0. H39,O. 1366;0. 130ij 0.1251 10 1058:0 (Y.)l3io.0729 OJ)598 0:05(1.1'0 0435
0.0079,0.
,00S8,0.. 009'1.0101\0.01
.0122,0.0138,O.0]61'0.or96 0.0262 0.03171 0849: 0.1135 .1376 O. 1313,O.125~;O .1213 0.1034'0.0898 O.
.0590 O.OliOO 0.0430
1
0.0087;O.1lOO2,O.1loo1;O.OI040.0112,1I.0I
.01360 01640.01800.0221'0.0299 0.0445
~0,]594 0.13211t1.1206O.1218~.1l78. 0.10081.
o583Io.0j94 O,0421i
0.00961.0101 iO.O:O'io.01l5 O.Ql24.0.0136O.01~OO 017Jio.00020.02W'00342 0053010.201510.1500;0.1272 0.1223'O.lJ?S'O. 1HIO,0985 O.
:0576'0.,0488!0,0.20
0.OI06,O.0l12;O,OI18,0.012U!O.0l36'O,0118!O.0IUG:O.OI900,WZS'0.02S~10. 03S9'0.06491:0.1806'0.],SM.1230!O. !185:0.1l-l3!O.II1C!0,OS63:0.
. 01ill5l0.0568lo.0182 0.041&
0.0111 iO.0123;.0. OI30.0.0139iO.OI49l0.01M;O.OIM,O,0202,0.03S2 0.0310'<1.0]500.0880 0.16SSI0.1363)0.1 191,0. J 150,0.1112 0.1080:0.0941 '0
. 11614 10. 0559 O.O.l?6'0.QHO
1
0.012alo.OI34!O.0I4210.0152iO.0l64,0.011010.000110,023~,0,0282.0.0358,0,0;242:,1003 0, ]S15,0. 1310,0. 1153,O.Ill5,O. JIk"lllo.1052
0.0019 0.
. 0663.0.11552,0. 0459;0.0~05
0.014fl,O,OH1,ILOI5.liO.OIM;il.01S!i,O 019~.O.022~ 0.02600.03140.01080.062510,2182,0. H61.0.1260.0.1lJ? 108;0.1052,0.1004 O.QS99 0.
.0652'0.0544 0.04!i2 C.Ul99
I
1
0.0l5G,O.OI09io.01l18{OlBlo.Olsalo.0216,O.0217,O.0211!!,0.0349,0.Ol031I.0768 0, 1871 1 139l,il.1l17 10.lOSS 0.I056 0,1021,.10000.08810.0184 0.0642;0.0537 0.0~55'0. 0396
l
O.0103,O.0I72jO.0114,O.ill97!O.0216,O,0238 ,0Z73 JO 03IS,0.0393,o,om o. 1l11!O. 171010. la3610.117710.]05710. 1026:0. 100fl O 00750.0$611.0168 0.0631 !0.03'18 0.044911.0389
O.0l17,O.Ol88,O.!l2OO.002!O,O.0236,O.0260 .0300 0 0352.0. OHO 0.063811.8457,0. 1590,0 12SS O.1I43.0.IU32 O. woa 0.O<J78 i O.
.0844-0.0755 0.0620'0.
.OH:J 0.0384
O.0192.0.02llJ;O.U2J7.0.1)2.33,O.0257,O.
.0330 0.0383,0.O~OO O. 07B8 0 ~192,O.I504 0.12H 0.11100 . 1004'0 , 097G'O.OYlij 0,
.0826/.07400.0,](110.
.043710.0S1Q
0.0200.0.0218 il.0'l84,O.02S2O.0278,O.
.035110.0429,0,0571'0, IOHlo.l05o,o.I437lo.120o O.1078!0.0980~O.09S;:0.003210.
.080S 0.0727 0.050010.
.0430 G. 0315
o,o:m,o.023T6.<lZM,O.0374!O.0306,O.
.0100 0.048.5 0676.!l~ 0.179,.13781.1161.10491.09561.09321.0009,0.
.0792,0,0713 0.0500 0.0198 .0424'Q,035P
1
0.0240'O.O~M .02150.0300 0.0333:0.
.OHS 0.0S4s;0.nS]710.2500.0, 1050,0.1325,0.113 .1022;0.0932;0.0909,0.088210.
.0776'0.0099 il.0580'O.04Sil 0.011S'0.0363
0.0250;0.0214 .11298,0 0326;0.0362'0.
.0494;n.OG3J o. J0301.o. 2105,0, 15.471,1280,0.]099 10.0095;0. O~1210 ,0890,0. 0870~0.
.016010,088110. (lti';OI(LO.j8! fo.om 10.0358
0.0280'10'
.03211,0852 O.
.05560.07.44 0.1741,0.19200.1465,0.1235 ,1066 n.0969 O.OBOI O,0867 0.0g,n,'0.
:074610.00730.05601,.0173 0,0404:0.0354
1
I
0.0301 0.0324,0.0350 O.
.05060.0633'0
OllSO'
0.3650,0.1706'0.
Hoa:O.1I97
O.
.094M.0869
0.0848
0.0831
O.
.07291'0,0660il
0&50 0.0454 0.0398'0.0318
1
0.0324 1.0349 1.0378'0.
0.0563 0.0724;.11000.2320 0.16000.13D2 0.1161 11
.092.1;0.0851!0.08300.0813 O.
.0714 0.0648 0:054010'0458IO.039ZI0.03~3
1
a.03OO,o.0319jO,Om;0.
.05241.0646 0.0849,0. 2250]O.2015;O.15U5Io.130410, !l210.
. .00020 ..0832,0.081210.0.910.
.06990.05350.0530/0.0449 0.0388 JO 033B
(l,0311.0.a-l03,o.OH4 0.0495,0,0576 0.0715;0 oo7B 0.3200 O.18EZ 0.1519,0.1253:0.]094 .09500.oB8110 {lSI3'O.0194 0.0118 0.076] 0.06841.0622 (I ()500 00142'0.0379'0.0332
0;0410.0. OHa 0.0488 O.05S~ 0.0850:0.0B24,0 12U '0. 2590'0.1763'0. 1400,0.1223,0.1
.093710.0851!.).0193,0.0775Io.015910 ON~lO.0671 0.0600 O:05lOIO:lJ.l32Io.0372'0. ons
O.O~38,0.OHB.O 0531,0.0600,0 07?] ,U 0ll3~O 23 Il,0.1680,0.lloo,O. 1189,0.
.0916 0.OR40,O.G777lo.0159Io.0H610.0729 O. 0658'0 ,0599 0.0500'0.0426 0.0367 .0322
1
0,0415,0. 01>19,0. OS7.B 0. 0Il66.0.081fi,O.1I091., 0:a4116,O.2120.0. ld06,0. 13l8,O .llf,2, 0.10091.08921.0820 0.0159,0.0142,0.0128.0.0713 0.064810058GIO.o<oolo.O.J18 0.0360,0,0317
O.OSII 0.OSS9 0.0627 O.07Jr, 0.0915 0. 1376'0.2700 O. 197Jj 0.1545,0 130B.O.ll20 n.0982,0.0870,O.0S!illil.0'T~3;O 0721,0:0712'0.0697 0.0629 O.057~IO'048!lo.04ll ,0.0353 0.1)$10
O.05M,0.OGOBO 0Ga/.0 08"..2,0.1080, 0 2"610 2,.10.0 1865..1490,0.1207 1 IOS5:0.0959,O.OR60,0 ..o182. 0.0725,O.0?!O,0.0597Io.0082.0.0618:0 ..o56Z:{t.O.17I,O.0402,O.0348.0.0304
'00
~I
.)
NAME INDEX
Abbot, H. I,., 94n., 126
Ackers, P. i 32.9, 352
Aridison,.H., 85
Agroskin, 1. 1., 401n., 436.
Ahm:lcl, N., 521
Airy, G. B., 553
Albertson, .M. L., 204, 213, 521
Allen,J, 17, 217n., 246
Allen, J. W., 329, 352
l\.mbTGSe, H. H., 212
Anderson, A. G., 493, 518
Appleby, F. V.,.602, 619
Atavin, V. I., 398n., 436
Arghyropoulos, P. A., lSi}
ArkhangelskiI, V. A., XlI., 587, 617
Auroy, F.,g91
Averianov, S. F., 36
..
'.
~.:','.
, -1
-. 'I
.(
<1l
(
'
'~
.. ', ':
\
I
664
I.
;-
\'
NAME INDEX
665
NAME INDEX
M. le, 391
Garot, r., 352
Gauckler, Ph., 99n., 126
Gentiiilii, B., 329,351, 572, 584
. Gh.:ll'!l.l'delli, L., 247
Ghotankll.f, S. T.,. 431
Gibb, H. M., 36
Gibson, A. H., 37, 393, 434
Gilcrest, B. R., 550, 620
Glover, R. E" 191, 585
Goldschmidt, H., 619
Goodrich, R. D., 607, 620
Gotaas, H. B., 85
Gl1lshof, F., 253, 294, 448
Graves, Q. B., 607, 620
Grimm, C. I., 296
Groyer, N, C., 31
'Qrzywienski, A., 390
GUil101.1, J. C., 51'211., 522
Gumensky, D .. B., 391
Gunder, D. F., 225, 247
G!1n~uli,
C., :xi, 5571.., 61-62, 83, 231, 246358,461, 517, 550, 584
A. H., 1 -!on., 84
Jansen,
B., 246, 391
Jeffrey", Ii., 150, 156, 214
Jegorow, S. A., 16, IS'
Jensen, R. A., 521
John$o!l, C. F" lOOn., 127
Johnson, J. 'N., 2H, 31
Johnson, RD., 510-571, 583
Johnstone, D., 607, 620
Jones, B. E., 533, 551
Jones, L. E., 296
G.,61i)
3. D., 361, 374, 389-390, 4.36
Kalinske, A. A., 435
Ko.ndaswa.rny, P. IL, :)HO
Klrr, M. H., 436
I{eifer, O. J.,
I(eIl.Iledy, R. G.,
Kennison, K. R, 393,
666
NA.ME.INDEX
NAME INDEX
Nebbia,
Neliclo v, I.
Nelson, H. C.,
NeWnl(lll,
21)5
Newmark,
M., ::i
Nikuradse, J., 9, 17, 201n., 202, 212, 214
Nimmo, W. H. R, :3:2\), 352
NDsedl.l., G.,
339,.., 352
No ugar'J , J.,
618
Obolensky, S., 391
O'llri.en, III. P., 2$), 37, 213
O'Connor,
583
elser" O ..T.,
101
Qrklley, J,. C.,
Orth, J., 35.3
Owen, W. M., 17
Pa,ger, G., 521
Pai, S. I., 619
667
,'l
I
J
\
668
NAME INDEX
Shchapov, N. M., 37
Shepley, J.
607, 621
Shima, 8.,
'171, 517
Jr., 493, 518
37, 431, 440-441,
45!J
Shulit.s, Samuel,
437, 459, 520, 620
Silber, R, "iin.,
Sil1:icrm:'..!l, E., 17
It., 295
Skdnde,
A. J 393, 409-412, 435-436 :
Smetana, J., 390,.393, 435
Smith, D., 328, 351
Snyder, F. F., 607, 620
Sorcrtllen, K E., 607, 620
Spangler, .1., 501, 520
Stanley, C. M., 437
Stall~on,
T. E., Hi
Stain, M. F.,35l
Steinberg, L E., 2SOn., '296, 607, 621
lVah!man, P., 570n.
Stevens, J. C., 13n., 18, 78, 82-Sa, 85,
'WaltOll, C. B., 607, 621
.436, 438
Warnock, J. K, 437, 585
Stokei, J. J., xn., 325, 53511.., 550,587,
Weaver, R. M., 436
'
600, 618-019 '
Wcisbach, J' I Sn., 16,40, 401n.,436, 501,
StmtWll, J. H., 607, 620
519
Stmub, L. G., 17, 4!}3, 518
, ,;ells, E. A" Jr., 85
Streck, 0., 520
,
Werner, P. W., 346, 353
St-reeter, V. L., 189, 212
White, M.
246, 429
Strickler, A., 99n., 126, 206, 213
,:llhitham, O.
535n.
Sueishi, T., 587, 619
Wien, \V'I 18
Supino, G., 550-551
Wilcox, E.
10671.., 127
Swain, F. E., 5aO, 585
Wilkinson, J.
531,551
Wilm, H. G.,
Wilson, W: T., 607, 620
,TD.k!lS~o, T., 553
Wisler, C. 0., 607, 620
Talwani, B. S., 437
Wittman, R., 459
T>\fpley, J. F., Jr., 608, 1)21
Wolf, P.O., 61, 83, 246,.358, 517, 550,
Tatum, F. E., 607, 621
, 584
,Taylor, E. H., 512-515, 522
Woodburn, J. G" 61
Terrell, P. \V., 190
Woodw9.rd, S. M., vii, Won., 121, 246,
Thomas, H. A., 214, 541, 551', 581, 585
2ge, 393, '134, 448, 459, 493, 518, 621
Thomas, R, 521
'
,
W6ycioki, K, 3gSn., 436
'l'holllp~on, P.W., 459
Thomsoll, J., 439, 459
1:iffany, J. B., 436, 521
Yarnell, D. L., 106n., 127, 425, 493, 501- '
Tison, L. J., xi, 61, 247
i 50.6, 518-519
,
Toch, A., 509-510, 521
Yassin, A. M" 155
Toebl!!l, C., 92, 125
Yen, C. H., 459
Tolkmitt, G., 253, 294
Yih, C. S., 214, 359
Tracy, H, J., 53n., 61, 476-477, 491-492,
51B
SUBJECT INDEX
A profiles, 230
Abrupt drop for jump control, 412-414
Abrilpt riEe for jump control, 411'-412
Accaleratioll line, .S27 ..
};.dverse slope, 223, 259-260
Aeration of nappe, 352-363
Air entrainment, :-13-36
Airy celerity equation, 539
}Jl-American Canal, 188-189
Alluvial riVer bends, 455-456
Alternate depth, 41, 551',
Angle of repose, 171-172
Artificial roughness; 202n.
Atterberg soil test, 173n.
Backwater, of bridge piers, 501-506
of cOllstriction, 490-493
of dam, 319.
effect of, 72
endpoint, 319
.
Backwater curve, 71, 222n.
Backwater envelope carve, 319
Backwater mtio, 491
Baffle piers, 415
Banking, 4513
Bazin da.ta, for broad-crestcd weir, 53n.
for channel roughness, 11, 95, 99n.
for overflow spillWays, 363-36'4
for rough channels, 204
for sharp-crested weirs, :;163, 382, 390
for wavy surfaces, 204
Bazin formula, 95
Bazin pro file, 363
, Ba~in'f! rr., 95
Bed load, 106
Bed-Iond fUnction, 166n.
Bends in alluvia.I rivers, 455-456
Bernoulli Emergy equation, ,40
Best hydraulic section, 160:"162
:
Blllaius equation, 8
Blasius-Prandtl-voll Kd.rman curve, 10
Blue hole, 5121l..
Bottom ro.ck, flow through,: 337-340
Bounda.ry layer, computation, 198-200
development, 192-193
C profiles, 230
Canal drop, 415
Canal faU, 415, 432
19
delivery of, 297-306
navig!l.tion, 572-S75
outlet. a.nd entrance, 30'1-306
Ca.pillary 'I'I'aVI!, 12
Cascaoos, ladder of, 433
Celerity, 13, 538-540, 582
Channel li:'lignment;lOii'(106-109
Channel contrachiO!\, 41-49, 258; 468410,479,
Channel expansion, 57-59
Channel geometry, 20-24
principle of, 16611..
Cha-anel irregUlarities, 103
Channel junctions, 321-323, 512-516
Cha.nnel sections, 20
best hydro.tllic, 160-162
circular (see Circular channels)
of constant: critical flow, 82
of eonstant'hydraulic udius, 152
geometric elementa, 22-24
669
uoo
NAME.INDEX
~,
436
NAME INDEX
667
RIl.IUBer,
,\
'';-.
670
SUBJECT INDEX
SUBJECT INDEX
Continuity equation, 5
of unsteady flow, 525-526
Continuous flow, Ci"
Contracted--opening method, V6, 476
Con traction, coeffll:iellt Of, 374-377
Contraction distance, 258
Contractiun ratio, 479
Contractions, -17-49, 468-470
Control, of !low; 70-74, 237
of hydnmlic jump, 408-414
Control scction, 70, 234-237
in laterru spillway chl>nnels, 342
Conversion lcss, 310
Convex flow, 30-31
Conveya;lCe, 128
Conyeyancfl ratio, 479
Co,iolis coefficient, 27
Cowo.n's method for Manning's n, 106109
Creager prums) 364n.
modified, 36411..
Critic!!.l depth,41
in circular sec tion, 82
curves for computation, 65
overrun, 533
in rectangular sections, 81-82
Critical discharge, 64
. through rectangular sections, 66
through sections of various shapes, 81
Criticru flow, ~pplicl1.tion of energy principle,4'2n.
.
appliaution of theorem of momentum,
54n,
computlltion (RM CriticpJflow computat.ion)
criterifL, 42-43, 55, 59-60, 63
definition, '13, 43, 55
Critical-flow oomputation, 63-70
hydraulic exponent. for, 66'--68
section fa"tor of, 64
Critical-flow flnme, 74
for closed conduits, 82-83
San Dim!l.S, 78
Critical-flow profiles, 23&-241
Critical roughnes;;, 195
Critical sectio n, 63
Critieal slope, 53
. at given normal depth, 142, 154
Critical-slope cur"e, 243
Critical state of flow, 13
Critical tractive force, 172
Critical velocity, 13
Cross waves, 448, 468
Cui verts, 20 .
flow through, 493-499
flow proJiles in, 265-266
flow types in, 496-499
671
"
, .'
672.
SUBJECT INDEX
;;hape,210
Fall increaser, 394n.
F&.w<)r tbeory, 358
Fish skimmer, 337
Flip b.ucket, overflow spillways, 85-36
Flood discharge, cllmputation, 146-148,
476-490
types, ix
Erosion below gates, 512
Escailde profile, 364n.
Expansion in chanllel, 57-59
Expansion wave, oblique, 449, <f54 ,
Expansions, 470-414
.
rapid, 13
rapidly varied, 6
rapidly. varied unsteady, 7
regimes of, 14:-16
rough-surfMe, 196-198
sec?ndary, 12, 439
shooting, 13
skimn:>lng, 196
slug, 581
spatially varied (see Spatio.lIy varied
flow)
.spiral, 25-26, 439-440
state of, 7-14
steady, 5
steady uniform, 5
streaming, 13
. subcritical, t3
supercritical, 13
673
SUBJECT INDEX
Flow, torrential, 13
tranquil, 13.
turbulent (see Turbulent flow)
uniform (see Uniform flow)
Ullstcac\y (see Unsteady Bow)
sp.ddle, 241
spiral, 241
vortex, 241
typical examples, 229
in wide horizontal channels, 292
Flow~resistll.Llce fa.ctor, 91
Flumed drop, 415
Flumes, 19
Parshall, 72--81
Force, drag, 168n.
plu:! momentum, 54/1..
shear, 168n..
specific, 53-56
of stream, 5<1n.
tracth-e (.ee Tractive force)
Force coefficient, 50
Fort Collins test,g, 364n.
Foriier-Scobey table of permissible
velocities, 165-166
Fourier's law, 601
Free overfall, 44
Free surface, 3
Free vortex, 446
Friction factor,
8~ 13
..
G. K. formula, 94
graphical solution, 95
Ganguillet and Kutter formula (see
G. K. formuia)
Gates, drum, 380-382, 507n.
overflow,S07n.
now, (}
!I'
,
6'74
SUBJEC'l' INDEX
260
[,
r
i
I
(
'
SUBJECT INDEx
weak,39S
Rydro.ulic ra.dius, 23
Hydra.ulic routing, 586
Hydrogra.ph, 543, .592
Hydrologic routing, 586, 804
Hydrostatic catena.q, 22, 35, 161
Ioe-covered channels, 137-138, 153
Impact-type energy dissipator, 422n.
Inflow-discharge-rating curve 305'
Initial depth, 46
'
relative, 396
Inlet loss, 311
Instability of flow, 204, 210
Inverted sipholl, 317-319
Irrigation canals. 175n.
Islands in rivers; 320-321
lsoclinal method, 346-347
Lamiu!Lr flow, 7
criterion, 7-8, 150
unst,eady, spatially vaded, 543
Lnminar suble.yer, 194
Laminar surf!!.Ce flow, spatially varied,
348:"349, 543
'
uniform, 149-150
Laue-Davis profile, 364n.
Later8.1 spillway channels, 342-346
dis,~harge in, 350
Lateral tra.ctive force, 176
Leach diagram, 271-272
Length, cn!>racteristic, 7-8, 13
mixing, 200
Limit, liquid, 173n.
{lla~tic, 17371.,
Limit slope, 142,243-244 ,
Lining of channels, 151-158, 160
Lintearia, 22
Liquid limit, 173n.
LocD-l phenomenon, 6, 43-'Hi
Losses, energy; in channels, 21871..,
due, to conversion, 310
to eddy, 267
at inlet, 311
at outlet, 311'
in pipes, 8
o.t transitions, 310-311
Low stage, 41
Lowel' normal depth, 231
Lucite, 10811.., 110
M profiles, 228
M value (see Hydraulic exponent At)
Maddock-Leopold's principle of nhannel
geometry, l66n.
Manning; formula, 93
exponent for, 99n.--l00n.
. for friction factor I, 124
international adoption, lOOn.,
nomog;ra'Pruc solution, 840
lVhnning's n, 99
in circular conduits, 106, 134-135
composite roughness, 135-140
determinatipn oC, 101-123, 206-210
Cowan'll method for, 106,.,109
dimensions, 9an.-!l9n.
of drain tiles, 135
equivalent value, 136
faciors afieqting, 101-1m~
on flood plains, 104;
for friction factor I. 124
for ice-covered channels, 137-138
for Panama Canal, 105-106
photographs of channels for, 114-123
relation to roughness height, 206
675
34G
of permissible velocity} 1[14, 167-168
of rela.Xll.tion, 358, 37411.., 384
of singula.r 'Point, 237-242
of tr!l.ctive force, 164, 175
of tria.l and error for unsteD-dy flow,
5i1
Mild channels, 223
Mild slope, 63
Minimum-energy th!:!ocem, 42
l\'linimum penniEsible velocity, 158
Mississippi River, Ba.~in'll1'1, 97
Chezy's C, 97
. diffusi vi ty, 602
gagings,94
junction problem, 322-323
Rutter's n, 97
Ma.nning',s n, 105
roughness datil., 206-207
Missouri River, ftQw~prQlile computa.tion,
27&:-292
junction problem, 321
Mixed progressive flow, 540
Mixing lengtll, 200
Momentum of o!len.channel no,,', 49-53
Momentum coefficient, 27
Momentum flux of stream, 54?t.
Momentum principle, 49; 56-59
Monaclin::.! rising. wave, 528
Moving hydraulic jump, 557
Muskingum method, 60(J-(\07
676
SUBJECT INDEX
Overrun, 529
Overrun critica.l depth, 533
Panama. Canal,
for, 105
Parabolic channels,
188
best hydraulic section, Hll
geometric elements, 631
of high order, ZOn.
hydraulic jump in, 432
Parall"l flow, 30
pressure distriou tio!! in, 33
Parsha.ll flume, 72-81 .
PavlC'vskii forlllula., 100,
Permissible maXimu.lTI veloc.itie:o, 1657166'
Permissible traetiv!,\ force, 165, 172~174
(SeE, also Tractiva force)
Permissible velpcities, ror cohesive ma- '
terial, 165--167
corrections for dElpth and sinuosity,
167
'
in grassed channels, 184-185
maximum, 157, 165-167
method of, 167-168
minimu..-u, 158
U.S.S.R. data, 166-167
Piers, baffle, 415
bridge, 501-1506
in gated spillways, 370-380
Pile trestles, 506 '
Pipe flow, 3
PitO& sphere, 440
Plastic limit, 173n.
Plasticity index, 173n.
Positive slope, 223
Posi',ive surge, 559--566
Positively progressive flo,;\', 540
Powell formula, 95
Powell's f,.98
Power canals, 568-572
Prandtl-von Karman curve, 12
PrandtI-von IC&rmAn equation, 9
'
Prandtl-von Ka.rman universal-velocity- ':'
distribution la.w, 201
~
Pressure on gates, 510-512
Pressure coefficient, 32
Pressure distribution, 30-34
coefficieuts, 32, 50
in curvilinear fiow, 34
effect of slope on, 32-34
hydrostatic 1e.w of, 30'-32
in parallel How, 33
Pressure-dlstribu tion ooefficien ts, 32, 50
Pressure-head correction, 31
Prismatic channel, 20
Progressive flow, 540, 554-557
Puisa.ting flow, 580
'
,
".
SUBJECT INDEX'
677
r
{
I
\
,I
.)
"
II
678
SUBJECT INDE.'4
204
'
SimJECT INDlllX
Sllpercritjcru slope, 63
,
Supercritieal state of flow, 13 :
Superelevlltion, 439
'
,
Surface curve, 222
Surface flow, .spatially varied, 347-349
spatially varied unsteady, 543
uniform, 148-150
,.
Surface irr.eg\ll!1rities, 106-109'
Surface roughness, 101-102
,~oncept, 104
explanatio1l; 194
hydraulically rongh, 1-95
nydraulic!l.lly smOGtll, 195
wa.vy, 195
Surface tensiou, 7,
Surge. 5S7
demand,559
energy in, 565-566
at junctions, 578-580'
in navigation caT.lnls, 572-575
negative, 566-568
positive, 559-566
in power ca.nals, 563-572
rejection, 559
t.hrough transitions, 575-578
Suspended load, 106
Sustaining slope,223
I
L
I
I
f
I,
To,:nintill.l flow, 13
Total force nf stream, 54n.
Tractive force, 168
, critical, 172
distribul;ion, 169
: lateral, 176
. method of design for, 175
p~rmissible, 165, 172-174
unit, 168, }(i9, 173
Tr~ctive-forc"l ratio, 170-171
Tra.nqui! flow, 13
.
'
Tra.nsition loss, 3LO-311
Tlranaitional depth, 2;!7n., 238, 2'!2-244'
Transitional discharge, 242
Trllnsil;ional pro.file, 238
Transitionailitate of flow, 7
679
'.~
.!
,
;~
680
SUllJECT INDEX
UNIVAC, 601
.
Univemal-velocity-distribution lq,w, 201
Unsteady How, 5-6, 523 .
continuity, 525-526
dynamic equation, 526-528
rll.pidly varied, 1
Unstendy uniform flow, 5
Unsteady varied flow, 6, 1, 523
Upper normal depth, 2:31
U8BR stilling basins, 415, 411-422,427429
Varjerl flow, 1)
gradually I 6'
gradua.lly unsteady, 7, 523
rapidly, 6
rapidly unstell.dy, 7, 523
spatially, 5, 327
Varied-flow fUllction, 254
for adverse slopes, 259
far circula.r conduits, 262
tables, 255, 259, 641-561
Varwick'a data for rough channe18, 1112
Vedemikov number, 210
Vegetll.l retardance, 171l, 184
Vegeta.tion roughness, 102-103, 106-109
Velocity, critical, 13
friction, 195, 201
measurement, 21
nonerodible, 165
nonsilting. 158
and non eroding, 166
shear, 201 . .
uLtmrnpid, 89
Velooity distribution, 24-26
t,urbulent.flow, 200-202
Velocity-distribution coefficients, 21':"30
for compound sections, 139-140
equations lor, 211
Velocity-distribution law, 201
Velocity gradient, 200
Velocity-head ratio, .l3n.
Vebcity-head recovery, 311
Vena contracta below sluiee, 401
Venturi flUme, 14
.
Vrilemonte weir sill, 80
Viscosity, dynamic, 8
effect on flow, 7
kinematic, 7-8
of water, 8
Vortex flow profiles, 241
Wake-interference flow, 12n., 196
Wash-water troughs, 34.9
Water area, 22 .
Wat.!r seal, 317
Wave angle, 429
Wave pro Olea, uniformly progressive
flml', 533-537
Wave suppressors, 422n.
Wa.ves, capillary, 12
.
graVity, 13
.
monoclinal rising, 528
. oblique expansion, 449, 454.
oscillatory, 523n.
propagation, 13, 537
roll, 210, 535, 581
. shock, 42\1n., 448n.
solitary, 531-538
translatory, 523
uniformly progressive, 528
Weak jump, 395
Weber nu.mber, Ix
Weirs, broad-crested (see Broad-<:rested
weirs)
sharp-<:rested, 360-362
Villemonte sill for, 80
Weisbach formula, 501
WEB Ge~al Spillway Tests-CW 801,
~64, 37011.. , 374
WES st~ndard spillway shapes, 364'365
iQietteo-perimeter, 22
Wide horizonta,l channels, 292
Wide open cbapnel, 2f1-27
.
critical a.nd normal depths in, 258
Width-depth ratio, 162-163
Ya.ngtze River, diffusivity, 602
,1