Documenti di Didattica
Documenti di Professioni
Documenti di Cultura
PVP2013
July 14-18, 2013, Paris, France
PVP2013-97622
STRESS ANALYSIS OF PIPE SUPPORT ATTACHMENTS: A COMPARISON OF
ANALYTICAL METHODS AND FINITE ELEMENT ANALYSIS FOR CIRCULAR AND
NON-CIRCULAR ATTACHMENTS
Anindya Bhattacharya
Technical Head, Stress Analysis
CB&I, 40 East Bourne Terrace, London, W2 6LG,
United Kingdom.
Phone: +442070535668
ABSTRACT
4. Comparison of FEA, WRC 107, WRC297 and Kellogg
methods w.r.t the following parameters:
Type of loading (Radial, Longitudinal, Circumferential)
applied in a stand-alone manner is absence of pressure
D , t and d ratios
T
T
D
Combined loading including pressure
Different element types
NOMENCLATURE
, - components in the ( , ) coordinate of the main
shell
r,R - mid surface radii of the branch pipe and main shell
E, - Youngs modulus and Poisson ratio respectively
un - radial displacement
0 - diameter ratio = d
T
p
, ,z
D
thickness of main shell
Airy stress function
internal pressure
global cylindrical coordinates in 3D space
directions
w - vertical displacement
ET 3
12( 1 2 )
loading in vertical direction
thickness of attached shell
ET
foundations stiffness = 2
R
direction of longitudinal axis of cylinder
section modulus of the trunnion pipe
area of cross section of trunnion pipe
INTRODUCTION
Z t -
K x S A -
2.1 Approach 1:
This approach is popularly known as Kellogg method in
the piping industry. This approach has been so named as it
appeared for the first time in [4] and is based on ring loading
around a circular cylinder.
Governing differential equation [1]:
For an axi-symmetric loading on a circular cylinder, the
governing differential equation is the well known beam on
elastic foundation equation:
d 4w
(1)
H 4 + Kw = Z
dx
d 4 w ETw Z
(2)
+
=
dx 4 HR 2 H
1.28
3( 1 2 )
Introducing 4 =
, i.e. =
considering
2 2
RT
RT
= 0.3 .
1. SHELL THEORIES:
There are various shell theories and each one has its own
protagonist. Any shell theory has to be evaluated within the
postulates of Sanders-Koiters approach [12, 21] which can be
summarised as follows:
1. The equations can be written in general tensor form.
2. The deformations are described by six strain measures,
three of which are components of the usual membrane strain
tensor and the other three deviate from the components of the
geometrical curvature change tensor only by terms that are
bilinear in the components of the curvature and membrane
strain tensor.
3. The stresses are described by six stress measures that satisfy
the equations of equilibrium without approximation.
4. The theory has a principle of virtual work that is exact for
displacements obeying the Kirchoff hypothesis; in conjunction
with approximate constitutive relations between the stress and
strain measures. Well-set boundary value problems can be
formulated, and the usual minimum and reciprocal relations of
structural mechanics hold good.
5. The theory contains an exact static-geometric analogy. This
analogy can be formulated by replacing the static quantities by
corresponding geometrical quantities in homogeneous
equations of equilibrium and the resulting equations become
identical with the compatibility conditions.
6. When applied to the symmetrical bending of shells of
revolution, the stress and strain measures agree with those
generally used. They are consistent with those of the most
simple curved beam theory.
For the present purpose, we will discuss the issue of
cylindrical pipes with circular (referred to as trunnion) as well
as non-circular (referred to as pipe shoes) attachments. Hence
there is no puncture in the header pipe. The mathematical
problem of the main shell with cut-out is a boundary value
problem of partial differential equation. It means that the
cylindrical shell equation, whose general solutions have many
unknown constants, is suitable on the shell surface with or
without cut-out. In order to determine the unknown constants
the boundary conditions have to be used.
we therefore get
d 4w
Z
+ 4 4 w =
4
H
dx
(3)
1.17 P R
1.5
(4)
12
distance to centroid, longitudinally = A
moment of inertia, circumferential = A
distance to centroid = B
12
distance to centroid = A
12
+ 5 BL2
9
3
moment of inertia, circumferential = B
distance to centroid = B
distance to centroid = A
+ BL
12
3
moment of inertia, circumferential = B
distance to centroid = B
+ AB
in
2
the developed surface. In deriving the equations it has been
assumed that 0 = 0 (circumferential strain).
The force/moment system is shown in fig-(2) below
Fig-(3) [25]
qz = qn cos = qn 0
M x = 2c
cos , q y = qn sin = qn 0
sin
0
0
q z Rc sin d = 2c 2 R
qn 0
(5)
cos sin d
0
(6)
Fy = 2c
q y c d = 2c 2
qn 0
sin d
0
(7)
dm
Dm
Dm
2
T
( )
2.3 Approach 3: Post WRC-107 approaches WRC297 and works of Morley, Simmonds and Hwang et
al. [5, 33, 11, 16, 25]
Theory of thin elastic shells, in which T/R<<1 is
insignificant in magnitude is derived on the basis of LoveKirchhoff assumptions. A generally accepted fact is this
approach has an error of order of magnitude O T .
R
When a solution is derived by omitting some terms, which
has order of magnitude larger than O T
(such as
R
2
4
2
2
+ 4 i 2
( )
( )
( )
= 0
R
2
4 = 12 (1 2 )
T
4 2
= un + i
ETR
( )
( )
) from
R
Flgges equation [13]. This equation is quite simple and can
be expressed in complex-valued displacement-stress function
form (Lekkerkerker [15] and Steele [3]) as follows:
2
4
2
(11)
where, and are the same as in eq-(10) and (9). The right
hand side of eq-(11) is a load function dependent on the
surface force components acting on the shell.
The cylindrical thin shell equations derived by
Goldenveizer, Morley, Simmonds and Timoshenko (which
was used by Bijlaard) have the same inherent error in order of
magnitude O T . The solution has the order of accuracy
R
O T . WRC-297, which is based on Steeles work on
R
shallow shell equations covers a range of only
r sin < 0.5 .
R
6
For detailed analysis of the approach taken by Xue et al
refer [17, 31].
In essence, the approach taken is to use compatibility
conditions enforced on the geometrically correct curve of
intersection as opposed to an assumed curve of intersection
and using theories which are of order O T
which may or
R
may not involve using different shell theories for intersecting
cylinders.
To summarize, different cylindrical shell equations are
suitable to different ranges of the developed surface. Fig-(4)
below shows the different ranges of developed surface [25].
= P ( p , p , p )
( )
(8)
(9)
(10)
( )
Fig-(4) [25]
Donnel [10] showed that his shallow shell equations could
be suitable to the range of < < i.e. 0 < 0.5 . In
6
6
[15] Lekkerkerker showed that the shallow shell equations
could be applied to the range of 0 0.25 . The different
applicable ranges adopted by different authors are dependant
on different allowable intrinsic errors.
1.
2.
3.
Fundamental assumption of the shell theory based on LoveKirchoff hypothesis and zero strain in the through thickness
direction.
Expressing the base vectors of a surface located off middle
surface i.e. a general surface in terms of the base vectors of
the middle surface (both covariant and contra-variant
versions).
Expressing the metric tensor of a surface located off
middle surface in terms of the metric tensor (both
covariant and contra-variant versions) of the middle
surface.
Expressing the rotation vector.
Expressing the Cristoffel symbols and permutation
tensors (Levi-Cevita tensors) of the surface located off
the middle surface in terms of the corresponding tensors
of the middle surface.
Expressing the strain tensors of a surface located off
middle surface in terms of the strain tensor of the middle
surface (both covariant and contra-variant versions).
Strain tensors are expressed as the difference between
metric tensors and curvature tensors in the deformed and
un-deformed states
Writing expression for stress and moment resultants.
Using appropriate constitutive relations.
Table-1 (Contd)
5.0 RESULTS
Radial
Force
Longitudinal
force
Circumferential
force
13
13
FEA continuum
element Cylinder
FEA continuum
element Trunnion
12
Loading Type
Table-2
Radial
Force
Longitudinal
force
Circumferential
force
45
12
NA
NA
NA
51
16
56
16
Kellogg Cylinder
Kellogg Trunnion
0.5
0.2
0.2
Loading Type
Table-1
30 inch header, 24 inch trunnion, wall thickness = 9.52 mm for
both. Magnitude of Force = 10KN, length of trunnion = 100
mm, d = 0.8 , t = 1 :
D
T
Radial
Force
Longitudinal
force
Circumferential
force
21
45
16
15
NA
NA
NA
17
50
22
11
54
20
20
Kellogg Cylinder
14
Kellogg Trunnion
0.6
0.4
0.4
FEA continuum
element Cylinder
19
10
FEA continuum
element Trunnion
15
15
Loading Type
Table 3
Table 4 (Contd.)
Radial
Force
Longitudinal
force
Circumferential
force
WRC 297
Trunnion
90
13
44
Kellogg Cylinder
Loading Type
Radial
Force
Longitudinal
force
Circumferential
force
Kellogg Trunnion
48
10
31
19
10
NA
NA
NA
20
54
30
41
17
103
30
75
19
Kellogg Cylinder
15
11
22
19
10
Kellogg Trunnion
20
46
16
29
FEA continuum
element Cylinder
17
11
48
16
31
FEA continuum
element Trunnion
19
42
14
26
43
15
27
45
16
28
47
15
30
FEA continuum
element Cylinder
44
13
27
Loading Type
FEA continuum
element Trunnion
46
14
29
Loading Type
Table 5
24 inch header, 8 inch trunnion, and wall thickness = 9.52 mm
for header and 8.18 mm for trunnion. Magnitude of Force =
10KN, length of trunnion = 100 mm, d = 0.36 , t = 0.86 :
D
T
Table 4
24 inch header, 20 inch trunnion, and wall thickness = 9.52
mm for header and 6.35 mm for trunnion. Magnitude of Force
= 10KN, length of trunnion = 100 mm, d = 0.84 ,
D
t = 0.67 :
T
Radial
Force
Longitudinal
force
Circumferential
force
47
21
53
NA
NA
NA
69
31
77
74
34
78
Kellogg Cylinder
16
20
40
Kellogg Trunnion
Radial
Force
Longitudinal
force
Circumferential
force
48
26
46
44
20
43
21
43
WRC 107
Trunnion
NA
NA
NA
44
22
40
44
23
38
19
36
Loading Type
10
44
22
40
39
19
37
FEA continuum
element Cylinder
46
24
44
FEA continuum
element Trunnion
41
19
43
Circumferential
moment
Longitudinal
moment
Circumferential
force
Shear Force
(Circumferential)
Longitudinal
force
Shear Force
(Longitudinal)
Radial
Force
Loading Type
Radial Force
Torsional Moment
Table 6 (Contd)
Loading Type
Table 5 (Contd)
FEA shell
element
(STRI65)
Cylinder
47
10
20
106
361
FEA shell
element
(STRI65)
Trunnion
45
23
104
402
FEA
continuum
element
Cylinder
44
18
106
360
FEA
continuum
element
Trunnion
47
20
104
398
WRC 107
Cylinder
48
13
99
310
WRC 107
Trunnion
NA
NA
NA
NA
NA
NA
WRC 297
Cylinder
54
13
153
413
WRC 297
Trunnion
FEA shell
element (S8R)
Cylinder
Circumferential
moment
Longitudinal
moment
Torsional Moment
Shear Force
(Circumferential)
Shear Force
(Longitudinal)
Radial Force
Loading Type
Table 7
258
NA
Kellogg Cylinder
87
Kellogg Trunnion
103
46
10
19
21
295
108
752
363
FEA shell
element (S8R)
Trunnion
48
FEA shell
element
(STRI3)
Cylinder
46
21
106
359
FEA shell
element
(STRI3)
Trunnion
46
24
103
403
23
105
401
11
121
63
120
59
125
66
126
70
Table 8
Table 10
307
NA
Kellogg Cylinder
100
Kellogg Trunnion
0.8
146
75
143
Longitudinal
Force
Circumferential
Force
181
18
61
Kellogg Cylinder
30
18
40
Loading Type
Kellogg Shoe
75
20
36
76
145
77
30
32
76
82
18
35
148
74
82
18
35
Table 9
75
22
37
80
30
33
81
24
33
78
35
32
321
NA
Kellogg Cylinder
100
Kellogg Trunnion
0.8
157
Table 11
30 Pipe, wall thickness 9.52 mm, Shoe design corresponds to
3 Gusset, A=450, B=500, Shoe plate thickness=10 mm, L=350
mm (refer fig-1), Magnitude of load=40KN. Pressure is not
applied. 1 = 0.67 , 2 = 0.60
108
155
102
Loading Type
159
98
154
103
12
Radial
Force
Longitudinal
Force
Circumferential
Force
173
21
63
Kellogg Cylinder
27
17
38
Kellogg Shoe
60
15
18
80
22
12
62
14
18
75
22
13
60
14
20
82
22
14
63
18
25
82
20
16
Table 12
312
NA
Kellogg Cylinder
161
Radial
Force
Longitudinal
Force
Circumferential
Force
174
23
65
Kellogg Cylinder
24
15
34
115
Kellogg Shoe
131
35
12
22
Loading Type
73
22
10
34
12
22
Kellogg Shoe
12
126
118
128
113
132
119
50
22
10
35
13
22
53
17
10
298
39
15
24
NA
57
21
13
Kellogg Cylinder
136
Kellogg Shoe
12
80
85
84
89
82
83
330
88
NA
92
Kellogg Cylinder
186
Kellogg Shoe
12
180
155
184
156
182
153
188
159
13
Tables 1,2,4 show that WRC 107 and WRC 297 results
show significant differences with respect to FE results for the
radial load case. This is because of the high d ratio and
D
radial as opposed to vertical load representation of the same in
WRC 107 as explained in section 2.2 of this paper. Tables 3, 5
and 6 show that the results are comparable (even for the
Radial load case) indicating the criticality of the d factor in
D
WRC 107/297 approaches. For the Kellogg Method, the
significant difference is for the radial load case. This is
because of the basis of the method being axi-symmetrical ring
loading which significantly deviates from the actual
mathematical model in the radial load situation. The Kellogg
method also underestimates the stresses in the Trunnion. This
is due to the use of simple beam theory as opposed to shell
theory and the non-consideration of the compatibility
requirement between the header pipe and the Trunnion in this
method. Kellogg method also in most (but not all) cases
predicts lower magnitude of stresses in the Longitudinal and
Circumferential Force applications. However the allowable
stresses in the Kellogg method as long as they are specified as
the [28] allowable for local primary stress, the error will not in
general make the analysis non-conservative except for the
Radial Load scenario. For Tables 7, 8 and 9 which are for the
combined load scenario, WRC 107 results show significantly
higher magnitudes of Pb + Pl + Q with respect to FEA. Even
though the Pressure loading has not been modeled as a Radial
loading for these Tables, which would have resulted in even
higher magnitudes of Pb + Pl + Q if the direction of this load
would have been in the same direction as the additive radial
load, but the simplistic way of computing pressure stresses
also (as in Tables 7, 8 and 9) induces higher stresses in the
WRC 107 type of analysis. Pressure induced loading at a
cylinder to cylinder interface with or without other external
loadings is complicated and WRC 107 analysis which
considers the loading on the cylindrical surface as a
rectangular loading cannot predict the stresses correctly and
will err on the conservative side for most cases. WRC 107
/297 analysis has shown lower magnitudes of Stress for Shear
Forces and Torsion moments (Table 6 where the loadings have
been applied at the Shell-Nozzle Interface) with respect to
FEA. However, these loadings, in general are not the
governing loads in piping applications. When using WRC107/297 modules of a Pipe Stress Program, an analyst should
review the program document to see how pressure is modeled
in these modules.
14
( )
( T R)
12. Additional tests need to be done for Pipe Shoes for varying
effects of D and combined loadings. In authors opinion it
T
is futile to expect usability of WRC-107 for shoe attachments,
as based on typical dimensions of Pipe Shoes, these geometric
parameters will in most cases be not satisfied.
13. WRC 107 /297 analysis has shown lower magnitudes of
Stress for Shear Forces and Torsion moments (Table 6 where
the loadings have been applied at the Shell-Nozzle Interface)
with respect to FEA. However, these loadings, in general are
not the governing factors in piping applications.
15
ACKNOWLEDGMENTS
[17]
[18]
[19]
REFERENCES
[1]
[2]
[3]
[4]
[5]
[6]
[7]
[8]
[9]
[10]
[11]
[12]
[13]
[14]
[15]
[16]
[20]
Timoshenko, S: 1940, Theory of plates and shells, MCGraw Hill, New York
Bijlaard, P.P., 1955, Stresses from Radial loads and
External moments in Cylindrical pressure vessels,
Welding Journal, Miami, FL, US Vol. 34
Steele, C.R and Steele, M.L.M 1983, Stress analysis of
Nozzle in Cylindrical vessels with external Load,
ASME Journal of Pressure Vessel Technology, Vol. 105
M.W.Kellogg Company,2011,Design of Piping
Systems
Mehrson, J.L, Mokhtarian, K., Ranjan, G.V and
Rodabaugh, E.C., 1984, Local Stresses in Cylindrical
Shells due to External Loadings on Nozzle Supplement
to WRC bulletin 107WRC Bulletin No. 297
Goldenveizer A.L., 1961, Theory of Elastic Thin shells,
Pergamon, Oxford,
Koiter, W.T, 1959. A Consistent First Approximation
in the General Theory of Elastic Shells, Proceedings of
the Symposium on the Theory of Thin Elastic Shells,
Delft, the Netherlands, W.T.Koiter ed., North-Holland,
Amsterdam.
Donnell, L.H., 1933, Stability of Thin Walled Tubes
under Torsion, NACA Report No. 479
Morley, L.S.D., 1959, An Improvement on Donnells
Approximations for Thin walled Circular Cylinders,
Q.J.Mech., Appl. Math., Vol. 12
Donnell, L.H, 1976, Beams, Plates and Shells, McGraw -Hill, New York, Chapter 6
Simmonds , J.G., 1966, A Set of Accurate Equations for
Circular Cylindrical Elastic Shells, ,Int. J. Solids
Structure., Vol. 2,
Budiansky,B and Sanders , J.L, 1963, On the best first
First Order Linear Shell Theory, Progress in Applied
Mechanics (The Prager Anniversary Volume),
Macmillan, London.
Flugge, W.1967, Stresses in Shells, Springer, Berlin.
Hoff, N.J., 1955, The accuracy of Donnells equations
J.Appl. Mech Vol. 22
Lekkerkerker, J.G., 1972. The Determination of Elastic
Stresses near Cylinder-to-Cylinder Intersection. Nuclear
Eng. Des., Volume 20.
Xue, M.D., Du, Q.H., Li, D.F. and Hwang, K.C., 2006.
Theoretical Stress Analysis of Intersecting Cylindrical
[21]
[22]
[23]
[24]
[25]
[26]
[27]
[28]
[29]
[30]
[31]
[32]
[33]
[34]
16
[35]
17