Analytical Design of Radial Turbine Design-Nasa

11111111111 1111 II Ilil 111111 1111 1111 II 111111 II 1111111111 I 11111111 ,I
3 1176 00161 6656

,
------'
DOE/NASA/Ol06-1
NASA CR- Ih$/-'-.--:7._l?_
GARRETT 31-3653
NASA-CR-165170
19810012429
ANALYTICAL DESIGN OF AN
ADVANCED RADIAL TURBINE
Gerold D. Large, David G. Finger, and Charles G. Linder
Garrett Turbine Engine Company
A Division of the Garrett Corporation
! --;,.. --\..., '7 ....... -- ~ ....... i
.- -
MAY 5 1 ~ 0 1
February 1981
!. I_!
~ 'f I ~ _I ~ ! f : / )..A .. '2_ J
Prepared for
NATIONAL AERONAUTICS AND SPACE ADMINISTRATION
Lewis Research Center
Under Contract DEN3-l06
fo r
U.S. DEPARTMENT OF ENERGY,
CONSERVATION AND SOLAR ENERGY
Office of Transportation Programs
111111/11111111111111111111111111111111111111
NF02436
,
/
, ,
,
./
NOTICE
This report was prepared to document work
sponsored by the united States Government.
Neither the united states nor its agent, the
united states Department of Energy, nor any
federal employees, nor any of their con-
tractors, subcontractors or their employ-
ees, makes any warranty, express or im-
plied, or assumes any legal liability or
responsibility for the accuracy, complete-
ness, or usefulness of any information,
apparatus, product or process disclosed, or
represents that its use would not infringe
privately owned rights.
ERRATA
NASA Contractor Report 165170
DOE/NASA/0106-1
GARRETT 31-3653
ANAL YTICAL DESIGN OF AN ADVANCED RADIAL TURBINE
Gerold D. Large, David G. Finger, and Charles G. Linder
February 1981
The Contractor Report number on the cover should be 165170 instead of 16570.
ANALYTICAL DESIGN OF AN
ADVANCED RADIAL TURBINE
G.D. Large,D.G. Finger and C.G. Linder
A Division of the Garrett Corporation
February 1981
Prepared for
National Aeronautics and Space Administration
Lewis Research Center
Cleveland Ohio 44135
Under Contract DEN3106
for
DOE/NASA/Ol06-1
NASA CR165170
Garrett 313653
U.S. DEPARTMENT OF ENERGY, CONSERVATION AND SOLAR ENERGY
Office of Transportatron Programs
Division of Automotive Technology Development
Washington, D.C. 20545
Under Interagency Agreement DEAl 0177CS51040
i
This Page Intentionally Left Blank
TABLE OF CONTENTS
1.0 SUMMARY
1.1 Identification of Program
1.2 Objective
1.3 Summary of Major Results
2.0 INTRODUCTION
2.1 Background
2.2 Related Work
2.3 Scope/Purpose
3.0 DISCUSSION OF RESULTS
1
1
1
1
3
3
3
10
11
3.1 Single-Shaft Engine Cycle Selection 11
3.2 Configuration Definition 11
3.3 Optimization Study Parameter Range 11
3.4 Aerodynamic Performance Correlations for
Parametric Study 15
3.5 Projected 1983-1985 Performance Improvements 31
3.6 Preliminary Stator and Rotor Geometry 32
3.7 Parametric Study Results 33
3.8 Stator and Rotor Solidity Studies 55
3.9 Parametric Study for Two-Dimensional Mechanical
Analysis 60
3.10 Parametric Study of Detail Aerodynamic Analysis 77
4.0 DETAIL ROTOR AERODYNAMIC AND MECHANICAL DESIGN
4.1 Aerodynamic Detail Design
4.2 Detailed Mechanical Design
5.0 OFF-DESIGN PERFORMANCE EVALUATION
6.0 CONCLUSIONS
ATTACHMENTS:
Appendix A.
Appendix B.
Appendix C.
REFERENCES
Abbreviations and Symbols
Deswir1 Vane Loss Model
Rotor Flow Path and Blade Geometry
for ~ B = 0, 10, and 20
86
86
91
110
114
117
123
131
145
iii
TABLES:
1.
II.
III.
IV.
V.
VI.
TABLE OF CONTENTS (Contd)
Ceramic Radial Inflow Turbine Reference Cycle
Requirements
Parameter Range for Optimization Study
Performance and Nondimensional Geometry for
Selected Diffuser Configuration
Selected Configurations for Detail
Aerodynamic and Mechanical Design
Optimum Configuration Derived from Parametric
Study and 3-D Mechanical Analysis
Stress and Cumulative Probability of Success
Summary at 95,000 rpm
LIST OF ILLUSTRATIONS:
1.
2.
3.
4.
5.
6.
7.
8.
9.
10.
iv
Optimum Tip Speed as a Function of Stage Work
and Blade Number
Effect of Rotor Tip Speed on Attainable
Turbine Efficiency
Illustration of Swirl Effect on Rotor
Downstream Ducting (ref. 2)
Turbine System Optimization Technique
Definition of Radial Turbine System for
Parametric Study
Specific Speed Correlation for Radial
Turbines
Reynolds Number Correlation for Radial
Turbines
Summary of Radial Turbine Rotor Clearance
Effects on Total Stage Efficiency
Exhaust Diffuser Configuration Used in
Parametric Study
Diffuser Duct Loss Data
12
14
23
87
95
108
4
6
8
9
13
,
16
18
20
22
24

LIST OF ILLUSTRATIONS (Contd):
11.
12.
13.
14.
15.
16.
17.
18.
19.
20.
21.
22.
23.
24.
25.
Deswirl Vane Loss Characteristics (ref. 3)
Using Arbitrary Air-Foil Shapes
Representative Rotor Exit Efficiency
Characteristics for Radial Turbines
Preliminary Correlation for Radial Turbine
Reaction
Turbine System Parametric Study,
N = 80,000 rpm, ~ B = 0
N = 80,000 rpm, ~ B = 10
N = 80,000 rpm ~ B = 20
N = 95,000 rpm, ~ B = 0
N = 95,000 rpm, ~ B = 10
N = 95,000 rpm, ~ B = 20
N = 110,000 rpm, ~ B = 0
N = 110,000 rpm, ~ B = 10
N = 110,000 rpm, ~ B = 20
Effects of Rotor Exit Area and Radius Ratio
on Turbine System Efficiency, N = 95,000 rpm,
~ B = 0
Turbine Vector Diagram and System Parameters
Along Individual Tip Speed Envelopes,
N = 80,000 rpm, ~ B = 0
Turbine Vector Diagram and System Parameters
N = 80,000 rpm, ~ B = 0 (Contd)
25
28
30
35
36
37
38
39
40
41
42
43
45
46
47
v
LIST OF ILLUSTRATIONS (Contd):
26.
27.
2S.
29.
30.
31.
32.
33.
34.
35.
36.
37.
3S.
39.
40.
41.
42.
43.
44.
vi
Peak System Efficiency for N = SO,OOO rpm
Peak System Efficiency for N = 95,000 rpm
with Deswirl Vanes
Tip Speed Envelope with Reaction and Deswirl
Vane Effects
Characteristic Rotor Flow Path Configurations
for Selected Parametric Study Rotational Speeds
Stator Solidity Study, N = 95,000 rpm,
fiB = 0
Rotor Solidity Study, Axial Length = 4.S cm
(1.9 in.)
Rotor Solidity Study, Axial Length = 4.3 cm
(1.7 in.)
Rotor Solidity Study, Axial Length = 3.S cm
(1.5 in.)
Typical Two-Dimensional Finite Element Model
Tangential Thickness and Maximum Principal
Stresses
Campbell Diagram for SO,OOO-rpm Design
Normalized Deflections (Modes 1 and 2)
Tangential Thicknesses and Maximum Principal
Stresses
Campbell Diagram for 95,000-rpm Case With
5S.1 cm
2
(9.0 in.2) Exit Area
Stresses
Campbell Diagram for 95,000-rpm Case with
45.2 cm
2
(7 in.
2
) Exit Area
49
50
51
52
53
54
56
57
5S
59
62
63
64
65
66
67
68
69
70
LIST OF ILLUSTRATIONS (Contd)
45.
46.
47.
48.
49.
50.
51.
52.
53.
Stresses
Campbell Diagram for 110,000-rpm Case
Effect of Speed on Maximum Principal Disk
Stress
Maximum Disk Stress Versus Number of Blades
Stresses
Stresses
Maximum Disk Stress Versus Tip Speed,
AE = 45.2 cm
2
(7 in.
2
), N = 95,000 rpm,
NB = 12
54. Effect of Rotor Exit Hub Blockage on Rotor
Loss
55.
56.
57.
58.
59.
60. -
61.
62.
63.
Effect of Rotor Exit Hub Blockage and Blade
Number on System Efficiency
Effects of Reaction, Rotor Exit Area, and
Blockage on System Efficiency
Effect of Reaction, Blockage, and Exit
Annular Area on Tip Speed Characteristics
Final Rotor Flow Path Configurations
Rotor Blade Angle Distribution Versus Percent
Meridional Distance
Velocity Distributions
Typical Three-Dimensional Finite Element
Model
Maximum Principal Stresses
Maximum Principal Stresses
71
72
73
74
75
76
78
79
80
81
82
83
85
89
90
92
94
96
98
vii
LIST OF ILLUSTRATIONS: (Contd)
64.
65.
66.
67.
68.
69.
70.
71.
72.
73.
74.
75.
76.
viii
Blade Shape, Original Thickness (Rake = 0)
Blade Shape, Original Thickness (Rake = -30)
Maximum Principal Stresses, Original
Thickness
Blade Shape, Modified Thickness
Maximum Principal Stresses, Modified
Thickness (1)
Thickness (2)
Thickness (3)
Maximum Principal Stresses, Original
Thickness
Thickness (3)
Maximum Power Design-Point Selection for
Off-Design Performance Evaluation
Off-Design Performance Characteristics as a
Function of Design-Point Exit Swirl (from
NASA-DOE Advanced Gas Turbine Power-Train
System Development Program)
Equivalent Conical Angle Correlation of
Diffusing Cascade Data (ref. 3)
Deswir1 Equivalent Area Ratio for Maximum
Diffusion Efficiency (ref. 3)
99
100
101
102
103
104
105
106
107
111
112
126
128
l'
1.0 SUMMARY
1.1 Identification of Program
This is the final report for the Analytical Design of an
Advanced Radial Turbine study program under NASA Contract
DEN3-106, to define an optimum ceramic radial inflow, single-
stage turbine system, for an advanced automotive gas turbine
engine. The turbine study program is based on an advanced gas
turbine cycle, 100.67-kW (13S-shp) engine for use in a lS88-kg
(3S00-lb) automobile, with a specific fuel consumption of
0.227 kg/kW-hr (0.373 lb/hp-hr).
Utilization of high turbine cycle temperatures of 1370C
(2S000F) and ceramic materials are prime features of this study.
1.2 Objective
The program objective is to investigate the effects of tip
speed, non-radial rotor blading, inducer-to-exducer work split,
and deswirl vanes on the system efficiency and mechanical reli-
ability of a single-stage, ceramic, radial turbine designed to
meet the performance requirements of an advanced automotive gas
turbine cycle.
1.3 Summary of Major Results
Analysis results indicate that, based on projected 1983
aerodynamic performance and ceramic material properties, radial
blade rotor configurations with a tip speed of 701 m/sec
(2300 fps) are feasible and satisfy the goals of 87. O-percent
system efficiency and 0.9999 cumulative probability of success
mechanically.
Nonradial blade rotor configurations allow rotor inducer
tip speed to be reduced for equivalent performance. However,
the inherent higher blade stress associated with this type of
configuration requires a drastic increase in blade thick-
ness to restore reasonable cumulative probability of success
(CPS) values for the ceramic material blade. The required
increase in blade material thickness, in turn, results in higher
disk loadings and reduced CPS values for the wheel. Therefore,
from aerodynamic, mechanical, and off-design considerations,
the radial-blade rotor appears to offer the lowest overall risk
1
2
and is the optimum configuration considered in this study. This
design includes the following: ~
Shaft speed
Inducer tip speed
Rotor inlet blade angle
Rotor blade number
Rotor exit swirl
Rotor exit critical velocity ratio
Rotor exducer-to-inducer tip
radius ratio
95,000 rpm
701 m/sec (2300 fps)
o degrees
12 to 14
-10 degrees
0.449
0.632
2.0 INTRODUCTION
2.1 Background
The need for more efficient use of available fuel supplies
has led to increased interest in gas turbine engines for auto-
motive applications. Both government and industry have con-
ducted studies to determine optimum cycle parameters and mechan-
ical configurations. A high-temperature, regenerated, single-
shaft engine has created considerable interest for automotive
application due to economy, cost, packageability, and drive-
ability. In a current NASA-DOE contract design study effort for
an improved gas turbine engine, a single-stage radial turbine
has emerged as the leading candidate compared to a single- or
multi-stage axial turbine.
The single-stage radial inflow turbine has attributes par-
ticularly sui ted to this size engine because of ruggedness,
lower sensitivity to tip clearance, and higher stage work levels
when compared to axial turbines.
For a single-shaft automotive application, a single-stage
radial turbine must effectively cope with the basic requirement
of high-stage work levels at maximum power and still maintain
acceptable performance over the entire automotive duty cycle to
achieve significant reductions in fuel consumption. However, by
reducing fuel consumption on the order of 30 to 50 percent, the
gas turbine will offer a distinct advantage in fuel economy over
current spark-ignition engines and provide an automotive engine
that is adaptable to a wide range of alternate fuels.
2.2 Related Work
For a single-shaft configuration, studies conducted to date
show that work levels from 419 to 512 kJ/kg (180 to 220 Btu/lb)
are required at maximum power for the radial turbine. Numerous
experimental studies conducted at AiResearch have shown that
maximum radial turbine performance correlates well with the cen-
trifugal compressor slip factor developed by Stanitz (ref. 1).
That is, peak radial turbine performance occurs at a rotor inlet
work coefficient (rotor inlet tangential veloci ty /rotor inlet
tip speed) that is consistent with the slip factor cr iter ia
developed by Stani tz. On this basis, a relationship between
stage work and inducer tip speed can be derived for maximum
radial turbine efficiency.
This relationship is presented in Figure 1 for zero exit
swirl and for a range of inducer blade numbers. At the expected
maximum power work levels, tip speeds between 671 and 762 m/sec
(2200 and 2500 fps) would be required to achieve peak radial
3
m/see ftlsee
I T I I
OPTIMUM TIP SPEED BASED ON:

SLIP FACTOR CRITERIA
/ /
)
NO INDUCER INCIDENCE LOSS
.il' 'OW
ZERO EXIT SWIRL
)

793 (2600
762 (2500
732 (2400 )
701 (2300 )
...
:>
c. 671 (2200
)
w
w
3;
0.. 640 (2100 )
i=
0:
w
U
:> 610 (2000 )
c
z
0:
579 (1900)
o
a::
549 (1800)
518 (1700)
488 (1600)
457 (1500)
/

I W
I/jV
if

9
':y'?:

I
(I)
w
Vff
I
z
al
I
cO:C
u
I

II!
11
!
3; -' :i!
-(g ....

I-wofi'l
wO:zl-
N
/ W
T
I

I
ijwo:tl-
s: alit)

VJ'l

-'uo:-
I
I
:t:u.
0:
0
I
u.
rt I
I
I
I
,
'RESEARCH TURBINE I
DESIGN POINT
I_
- ,- .. --,
SINGLE-STAGE
RADIAL TURBINE
STAGE WORK RANGE
FOR SINGLE-SHAFT
AUTOMOTIVE ENGINE
427 (1400)
Btu/lb (80)
kJ/kg 0
(100)
233
(120) (140) (160) (180) (200) (220) (240)
4
279 326 372 419 465 512
SPECIFIC STAGE WORK, dHT
Figure 1. Optimum Tip Speed as a Function of
Stage Work and Blade Number.
558
turbine performance. Cycle studies indicate that high turbine
inlet temperatures [1370
0
C (2500
0
F)] also will be required to
achieve the fuel reduction goals. Ceramic mater ials have
received considerable attention for this application because
they eliminate the need for turbine cooling and the cr i tical
ma ter ials required for metal turbines, as well as having the
potential for low cost. However, at the start of the program,
the mechanical properties of ceramic materials were in a period
of such rapid improvement that an accurate estimate of maximum
allowable tip speed of a radial turbine was not possible.
If tip speeds from 549 to 610 m/sec (1800 to 2000 fps) were
required to satisfy the material properties, severe tip speed
limitations would exist relative to the optimum tip speed
required to achieve maximum performance (Figure 1). The impact
of tip speed limitations on radial turbine performance is illus-
tra ted in Figure 2. Resul ts from a research turbine tes t
program (designed with optimum tip speed) indicate that peak
efficiency was achieved at the design point. However, increas-
ing or decreasing tip speed from the design point results in a
severe performance penalty.
The radial turbine performance characteristics, shown in
Figure 2, raise a number of questions:
o What are the physical phenomena that cause the rapid
decrease in performance at higher or lower than opti-
mum rotor inducer tip speeds?
o Once the physical phenomena are established, what
design technique would tend to minimize the effect and
reestablish maximum performance at nonoptimum tip
speeds?
o What technique would tend to optimize turbine perform-
ance not only at maximum power but throughout the
entire duty cycle range?
These underlying questions formed the basis for arriving at
a radial turbine design technique that would optimize radial
turbine performance for the entire duty cycle of a single-shaft
automotive gas turbine engine.
First, assume that the effects of lower or higher than
optimum tip speed are attributed to rotor inlet incidence loss,
and then, that the incidence penalty is directly proportional to
the difference in kinetic energy between the optimum inlet work
coefficient and that actually implied by the rotor inlet vector
diagram. Figure 2 shows that, wi th these assumptions, the
pred icted and measured decrement in turbine eff iciency are in
good agreement (considering that the effect of stator loss and
rotor reaction are also present under these conditions). Given
this correlation between nonoptimum rotor tip speed and rotor
5
1 0
099

w
098 c..
fj"
l-
I=:"
--
097
fj"
l-
I=:"
096
095
......
//

DESIGN TOTAL-TO-TOTAL PRESSURE RATIO = 273

V-

t"
I'--.. "-.... ........
=090
7
L
V
.h
120
II
0.6
OPTIMUM WORK COEFF

'-.......
............
07
--D- TEST DATA AT DESIGN PRESSURE RATIO
OVER A RANGE OF CORRECTED SPEEDS
- - - - PERFORMANCE CHARACTERISTICS
PREDICTED FROM INCIDENCE EQUATIONS
08 09 10 1.1
STAGE WORK COEFFICIENT, ASTG'
3

"

1.2 13
Flgure 2. Effect of Rotor Tlp Speed on Attainable Turbine Efficiency.
1 4
inlet incidence loss, what design technique would minimize these
effects? Two approaches appear to be feasib+e: First, the
rotor inducer-to-exducer work split could be adjusted to the
point that optimum rotor inlet conditions would be reestablished
(i.e., high negative rotor exit swirl). Second, the rotor inlet
blade angle could be adjusted to reestablish optimum rotor inlet
conditions with a gO-degree radial inlet flow configuration.
Adjusting the inducer-to-exducer work split should be bene-
ficial for all types of rotor blading as long as a limitation on
tip speed or blade angle exists. Unfortunately, numerous exper-
imental studies indicate that the turbine downstream ducting is
relatively sensitive to rotor exit swirl, and that the magnitude
of duct loss is highly dependent on the type of geometry uti-
lized. For example, Figure 3 illustrates the loss characteris-
tics as a function of inlet swirl for an inter stage duct and
exhaust diffuser. Although Dovzhik's data (ref. 2) are for uni-
form inlet conditions, rig test results with the actual rotor
inlet conditions show the same basic characteristics. Unpub-
lished AiResearch results for radial diffusers indicate that
this type of exhaust duct could be the least sensitive to inlet
swirl; however, additional analyses and experimental studies are
required to fur ther def ine the design techniques. Wi th the
assumption that the turbine exit ducting loss is a function of
swirl, the duct loss characteristics must be included in the
optimization process for tip speed-limited radial rotor designs.
Even if optimum tip speed and zero exit swirl are achieved for a
given design point, the wide range of oper.ating conditions
experienced by the single-shaft turbine would still result in
relatively high rotor exit swirl during the normal duty cycle.
A turbine system optimization technique based on combining
the performance characteristics of the turbine stage and exhaust
duct is illustrated in Figure 4. At the corrected work level
shown, a tip speed of 781 m/sec (2562.0 fps) would be required
to achieve optimum rotor inlet conditions and zero exit swirl.
However, with an allowable tip speed of 610 m/sec (2000 fps),
zero exit swirl would occur at a rotor inlet work coefficient of
1.37. Although the exhaust duct loss would be minimized under
these conditions, Figure 4 shows that stage efficiency rapidly
decreases as this level of inlet work coefficient is approached.
By combining the performance characteristics of the turbine
stage and exhaust duct, the rotor inlet incidence effects are
minimized, and system efficiency is improved by 4.0 points.
Figure 4 was based on conventional radial element rotor
blades. Further improvement in system efficiency is expected
with three-dimensional (nonradial) blades. AiResearch has
7
...J
a:

5.01-----CHARACTERISTICS FOR INTERTURBINE
DUCT OR ANNULAR EXHAUST DIFFUSER,
DUCT INLET TOTAL TO EXIT TOTAL
PRESSURE LOSS COEFFICIENT RATIO
(CONFIGURATION 00-2 IN REFERENCE 2)

o

a:
I-
CHARACTERISTICS FOR A COMBINED
3.0
AXIAL-RADIAL ANNULAR EXHAUST 01 FFUSER.
DUCT INLET TOTAL TO EXIT STATIC
PRESSURE LOSS COEFFICIENT RATIO
(CONFIGURATION OR-1 IN REFERENCE 2)
(.)
u.
u.
w
o
(.)
en 2.0
en
o
..J
8
1.0

o 10.0 20.0 30.0 40.0
ROTOR EXIT SWIRL ANGLE, DEGREES
Figure 3. Illustration of Swirl Effect on
Rotor Downstream Ducting (ref. 2).
50.0
0.20
I I I I
TIP SPEED = 610 m/see (2000 f/see)
' ~
12 RADIAL BLADES
I I
dHTI8 = 38.3
I
V'"
ROTOR INLET WORK COEFFICIENT -
REQUIRED FOR ZERO EXIT SWIRL
~
.. 1.37
/
L ROTOR EXHAUST D : ~ ,
TOTAL PRESSURE LOSS
I'.....
(INCLUDES THE EFFECTS
OF BOTH SWIRL AND ............
I--- VELOCITY LEVEL)
0.15
0.10
005
0
f4-0PTIMUM ROTOR INLET
>
u
0.90
WORK COEFFICIENT I---
-
IT t'......
/
.............
1'1
,
L-STAGE TOTAL
~
EFFICIENCY
~
DUCT LOSSES ..
~
~ INCIDENCE LOSS
PREDOMINATE PREDOMINATES
z
w
U
u..
0.88
u..
w
..J
<
I-
0.86
0
I-
>
..J
Z
0 0.84
w
e"
<
I-
eI)
0.82
i--_
-.
">-.
y
~ ....
ROTOR EXIT/
......
~ ' l ~
SWIRL
./"'" --.....
V
....... ......
~ /
......
,
V
",
f\
j'.... SYSTEM TOTAL EFFICIENCY ......
(STAGE PLUS DUCT) ... to-
/
0.83
082
0.81
080
-50
w
I-
:lei)
..Jw
-40
Ow
~ a :
ctffi
ZC
< .
-30
Ww
:!:..J
e"
I-Z
x<
-20
W..J
a:!!:
O;!:
l-eI)
0
a:
-10
1.0 1.05 1.10 1.15 1.20 1.25 1.30
ROTOR INLET WORK COEFFICIENT
(ROTOR INLET ABSOLUTE TANGENTIAL VELOCITY/INLET TIP SPEED)
Figure 4. Turbine System optimization Technique.
9
investigated the aerodynamic and mechanical feasibility of non-
radial rotors, and the results indicate the following:
o With careful control of the rotor blade turning dis-
tributed through the rotor, rig test results show that
the increase in performance is in good agreement with
that predicted from a decrease in inducer incidence
loss
o Minimized rotor blade bending stress requires exten-
sive rotor blade thickness optimization and blade
restacking. At the present time, 20 degrees of rotor
inlet blade angle appears to be the upper limi t in
terms of rotor mechanical design
The recent study conducted on deswirl vanes by Mitchell and
Soileau (ref. 3) indicates that duct losses could be signifi-
cantly reduced for high exit swirl and Mach number turbine
designs. Therefore, deswirl vanes will be included in the
system optimization.
2.3 Scope/Purpose
The program objective was to investigate the performance
potential of a high-work, ceramic, single-stage, radial turbine
with radial and nonradial rotor blading, with and without
deswirl vanes, based on a system optimization approach applic-
able to a single-shaft automotive engine. This was accomplished
with a parametric study by examination of the aerodynamic and
mechanical effects of the following parameters:
o A range of shaft speeds to evaluate rotor size effects
o A range of inducer tip speeds to determine the effects
on efficiency and mechanical integrity
o A range of rotor inlet blade angles to determine the
effects on inducer blade loading and the impact on
rotor inducer blade stresses
o A range of rotor inducer-to-exducer work splits to
determine the effects of rotor exit Mach number and
exit swirl on exhaust diffuser losses
o A range of rotor inducer tip-to-exducer tip radius
ratios for each value of shaft and tip speed
o An examination of blade number effects
To evaluate turbine performance effects only, all cycle
parameters, with the exception of radial turbine efficiency and
exhaust duct pressure loss, were held constant for the analysis.
10
3.0 DISCUSSION OF RESULTS
3.1 Single-Shaft Engine Cycle Selection
The maximum power point cycle selected for the advanced
radial turbine is representative of operating conditions pro-
jected for a gas turbine automotive engine produced in the
mid-1980s for a l588-kg (3500-lb) automobile.
The engine specific fuel consumption (SFC) is 0.1965 kg/kW-
hr (0.323 lb/hp-hr), and the net power is 101 kW (135 hp). The
turbine system design requirements and performance goals at this
condition are presented in Table I. Station nomenclature used
in Table I is defined in Figure 5.
3.2 Configuration Definition
The selected system configuration consists of a turbine
inlet scroll, radial stage and downstream axial exhaust diffuser
(with and without deswirl vanes). Alternate turbine inlet and
exhaust configurations are indeed, further automobile
engine analysis may indicate these are required. However, the
primary objective of this program is to evaluate the performance
potential of a ceramic radial turbine through use of a system
optimization procedur,e that could be used to evaluate modif ica-
tions that may be required by further automobile engine
analysis.
The effect of selecting either a scroll or an annular inlet
is expected to be small for the present program since, in either
case, the objective is to maintain a relatively low Mach number.
Therefore, pressure loss wi th ei ther type of inlet would be
expected on the order of 1/2 to 1 percent. The type of exhaust
diffuser configuration selected is more significant due to the
difference in the levels of losses expected from an axial or a
radial diffuser. Since the type of exhaust diffuser selected
for the analysis will limit the generality of the results, both
at the selected maximum power design point and off-design condi-
tions, the selection will be based on available diffuser data
with actual radial rotor exit conditions imposed. The selected
exhaust duct configuration shown schematically in Figure 5 is a
conical diffuser and is the type normally used for small auxili-
ary power units (APUs). Detailed test results are available for
this type of diffuser.
3.3 Optimization Study Parameter Range
The design parameters evaluated in the optimization study,
and the associated range, are presented in Table II. The lower
limi t of turbine shaft speed (80,000 rpm) corresponds to the
lowest specific speed at which peak efficiency could be main-
tained, based on available data. The upper limit of shaft speed
11
TABLE I. CERAMIC RADIAL INFLOW TURBINE REFERENCE
CYCLE REQUIREMENTS
Parameters
Turbine Operating Life
Turbine Design Reliability
Turbine Scroll Inlet Total
Temperature
Cycle Pressure Ratio
Scroll Inlet Total Pressure
Rotor Exit Total Pressure
Mass Flow, w
Specific Stage Work,
llH !
T 1-4
Stage Total-to-Total Pressure
Ratio, P
R
!
T-T 1-4
Stage Inlet Corrected Flow,
w/e/o 11
Units Requirements
hr 3500
0.9999
1370 {2500}
5:1
kN/m
2
{psia} 478 {69.33l}
kN/m2
{psia} 116 {16.8}
kg/sec {lb/sec} 373 {0.823}
kJ/kg {Btu/lb} 509 {2l8.73}
4.087
kg/sec {lb/sec} 0.191 {0.42l}
Stage Corrected Work, llHT/e! kJ/kg {Btu/lb}
1-4
89.162 {38.333}
Stage Total-to-Total Efficiency,
n
T
-
T
!1_4
Scroll Total Pressure Loss,
llP/plo-l
Exhaust Diffuser Total Pressure
Loss, llP/PI4-6 {with or with-
out deswiri vanes
System Total-to-Total Efficiency
n
T
-
T
!
0-6
*Varies in parametric study
**Goal from parametric study
12
0.897*
0.01*
0.04*
0.870**
2
3
COMBUSTION
DISCHARGE
, - - - - - - - - - - - - - - - - - - - - - - - - - ~ I \
I I
: I
I REGENERATOR I
I OR RECUPERATOR I
I I
I I
I I
, I
\ I
.... _-------', ;'------...
I I
/ .
/ I
'" I
6 _ ~ / I
~ - ~ I
~ "
,
I
I
I
'4
/
//
//
90 BEND TO
REGENERATOR
OR RECUPERATOR
- - - - - ~ - - - - - - - - ~ - - -
_c ..... ____ _
NOTE:
TURBINE SYSTEM PERFORMANCE FOR PARAMETRIC STURY
IS DEFINED FROM SCROLL INLET (COMBUSTION DISCHARGE)
TO EXHAUST DIFFUSER EXIT PLANES.
CENTERLINE
Figure 5. Definition of Radial Turbine System for Parametric Study.
TABLE II. PARAMETER RANGE FOR OPTIMIZATION STUDY
Parameters
Rotational Speed, N
Inducer Tip Speed, U
t
3
Inducer Blade Angle,
Rotor Exit Swirl, a
Rotor Exit Absolute
Critical Velocity
Ratio, V/Acr
Rotor Blade Number, NB
Rotor Exducer-to-Inducer
Radius Ratio, /RT
4 3
Axial Shroud Clearance
(Ca)
Radial Shroud Clearance
(Cr)
Rotor Backface Clearance
(Cb)
Units Range
rpm 80,000 to 110,000
m/sec (fps) 792 (2600)
deg
deg
cm (in.)
(optimum with 10 blades)
753 (2470)
with 20 blades)
549 (1800)
(lower limit)
o to 20
o to -50*
0.30 to 0.60*
10 to 16
Minimum based on 50%
rotor exit hub blockage,
maximum of 0.80
0.0254 (0.010)
(constant)
0.0254 (0.010)
(constant)
0.0508 (0.020)
(constant)
*Estimated; absolute range will result from the rotor exit work
coefficient and annular area imposed.
14
(110,000 rpm) is based on maintaining acceptable single-stage
centr ifuga1 compressor design character istics. Higher speeds
are possible; however, this would involve a compressor perfor-
mance trade-off (due to the increasing inducer tip Mach number,
which is considered beyond the scope of the current program).
The range of selected shaft speeds is indicated on the specific
speed curve in Figure 6.
3.4 Aerodynamic Performance Correlations for Parametric Study
The initial parametric study was based on a one-dimensional
aerodynamic performance evaluation model for the turbine scroll,
stage and axial exhaust diffuser system. Correlations for the
scroll and exhaust diffuser were incorporated into an existing
radial turbine design program to arrive at a system performance
for the range of parameters presented in Table II and based on
the selected cycle conditions shown in Table I. The correla-
tions and methods utilized for the parametric study are:
o Scroll Loss: The pressure loss in a turbine scroll is
primarily a result of turning the flow and wall fric-
tion. Rogo (ref. 4) presented a method for predicting
the turning loss for a turbine scroll based on a mass
averaged dynamic head wi thin the scroll. Eckert
(ref. 5) developed a technique for predicting the
fr iction loss for a compressor scroll. A technique
for predicting total turbine scroll loss has been
derived from the work of Eckert and Rogo in terms of
a scroll inlet Mach number, scroll inlet diameter
(a) at the tongue, scroll inside radius (ro)' and fric-
tion resistance coefficient (X), as follows:
llP
q ( ~ ) 0 . 2 6 6
+ 21TA (r + ~ )
=
P
TOTAL
PINLET roc
a 0 3
where:
(r +
a)
c =
1n 0
ro
AR
1
(Rough) =
1n(:) +
1.74]2
[0.8686
AS
=
0.316
(Smooth)
- 1/4
Re
and Re
DHVU
=
--
\I
15
>
u
z
w
U
U-
U-
w
...I

t-
O
t-
0.95
0.90
0.85
NASA TJ 0-6605 I. J J I
DATA CORRECTED TO SPECIFIC SPEED RANGE
BASED ON SELECTED
ZERO CLEARANCE AND
Re = 3.0 X 105 '[\ II

/ --........ I "-
---\.----- ----I-
/ ,

/ FROM NASA TN 0-6605 \
/ \
/
I RPM
/
0 90,000---1
/
/
/
o 100,000
6 110,000
\l 120,000
I 0.80 ..... __ -.l. ___ ..... ___ ""--___ L..-__ --L ___ -'-___ -'-__ ......I
RPM (ft
3/4
/sec 1/2) - 30
0.233
40
0.310
50
0.388
60
0.465
70
0.543
80
0.620
90
0.698
100
0.775
110
0.853
SPECIFIC SPEED
Figure 6. Specific Speed Correlation for Radial Turbines.
The loss predicted from the preceding equation would
be the minimum loss that would be expected for a
scroll of circular cross section and design, using the
continuity and radial equilibrium scroll design equa-
tions.
o Maximum attainable (base) efficiency is established
from specific speed correlations based on NASA
(ref. 6) and AiResearch data. The NASA correlation is
presented in Figure 6.
o Base efficiency is then corrected for Reynolds number
effects based on results from NASA (ref. 7) and
AiResearch data. The correlation is presented in Fig-
ure 7.
o Rotor inlet incidence loss is calculated for a speci-
fied rotor inlet work coefficient from the following
relationships:
In general,
"'3 .
,l.
(
U
t
+ V
R
tan
3 3
= S.F. U
t3
For radial blades,
2
"'3 . = S.F. = 1 - -- (Stanitz Slip Factor)
,l. NB
"'3,act
"'STG = "'3,act -
=
1
17
I-'
CD
0
i=

a:
>
(J
2!
w
(J
-
u.
u.
w
-I

0
I-
w
C!'

en
..
w
en

m
!:'"
r.::
1.02
1.01
1.00
0.99
0.98
0.97
0.96
0.95
0.94
-
__
T
i----BASE REYNOLDS NUMBER IN
J---+
V
-+---I
I
I----+----1-----4---4--- SPECI FIC SPEED
I I
I
t---I
r
-.;: AUTOMOTIVE RADIAL TURBINE
REYNOLDS NUMBER AT MAXIMUM
j POWER
T-'
o AIRESEARCH RADIAL TURBINE
TEST DATA (R
T3
= 4.995 IN) -
t-t--+--+-----+---+----lf-----1----t----+-O NASA TN D3835 TEST DATA
(REF. ROTOR, RT3 = 3.01 IN)
-
J J I I I

o 1.0 2.0 3.0 4.0 5.0 6.0 7.0
Re, REYNOLDS NUMBER X 10-
5
Figure 7. Reynolds Number Correlation for Radial Turbines.
where:
U
t
= Rotor inducer tip speed
3
V
R
= Rotor inlet absolute radial velocity
3
VU
3
= Rotor inlet absolute tangential velocity
A3,act
A
STG
A
4
,M
YJ.
~
(3base
fiB
V
R
S.F.
NB
= Rotor inlet tip radius
= Rotor exit mean radius
= Stage total specific work
= Rotor iniet work
incidence loss
coefficient for zero
= Rotor inlet work coefficient from vector
diagram for a specific tip speed and
inducer-to-exducer work split
=
Stage work coefficient
=
Rotor exit mean work coefficient
=
Stage total efficiency with incidence
loss
=
Stage total efficiency before incidence
loss
=
=
Rotor inlet radial velocity
=
Slip factor for radial blades,
1 - (2/N
B
)
=
Number of rotor blades
o The effect of rotor shroud clearance is based on cor-
relations derived from references 8 and 9 and on
AiResearch test results. The performance penalties
are a function of both axial (Ca) and radial (Cr)
clearance. The experimental data from references 8
and 9 are presented in Figures 8a and 8b. Additional
rotor clearance effects are present with rotor scal-
lops. The performance effects as a function of back-
face clearance recently have been evaluated for the
Model GTP305-2 turbine (ref. 10) and are presented in
Figure 8. The size and scallop depth of this turbine
are similar to the turbine considered in this study.
19
o

a:
LL
> W
U a:
Zj::=
W I
-I-

u.1:-
u. I
W I-

I-
o
I-
1.00 __ - __ _
AXIAL CLEARANCE. PERCENT
I
0.98
I 7.1
0.96 __ 12.5
I 17.9
0.94
0.92
0.90 -. ...... ---'------'-_ ..... _.L.---li-.....
012345678
RADIAL CLEARANCE.
PERCENT OF EXDUCER PASSAGE HEIGHT
(A) Summary of Radial and Axial Clearance Effects on Total Stage Efficiency (Ref. 8)
20
o 1.02 ---.--....,..----.!"'-"--...------
RADIAL CLEARANCE. PERCENT

a:u.
>w
ua:
Zi7

u.1-
u. I
wI-

.J

1.00 f"C1I::::;::x----t----t----t---t----.-t
0.98
0.96 1--.........
0.94 t----I--_+_-____ll----+_--+--__I
0.92
0 10 20 30
AXIAL CLEARANCE.
PERCENT OF INDUCER PASSAGE HEIGHT
<> 0.25
1::1. 0.92
o 3.0
o 7.0
(B) Summary of Radial and Axial Clearance Effects on Total Stage Efficiency
at Design Equivalent Values of Speed and Pressure Ratio (Ref. 9)
>
1.02
u
'''0...
Z
u.
r----__.
w
w
1.00
u .. a:
.....
ro-
-0 i='
r--- u.- I
r--.. 10...
u.1- I-
0.98
w
SCALLOP RADIUS! --
a:-
.J .. INDUCER TIP RADIUS = 0.769
I- 0.96
o 10 20 30
BACKFACE CLEARANCE.
PERCENT OF INDUCER PASSAGE HEIGHT
(C) Effect of Rotor Backface Clearance on Stage Total Efficiency
at Design Conditions (Ref. 10)
Figure 8.
summary of Radial Turbine Rotor Clearance Effects
on Total Stage Efficiency.
In addition, the results of a 1978 AiResearch-
sponsored radial turbine research program investi-
gated the effects of scallop depth on radial turbine
performance. These data will be used to supplement
the correlation derived from the Model GTP30S-2 tur-
bine.
o The performance effects of blade number are based on
the Models GTCP305-1, GTCP36-4, and published Pratt
and Whitney data (ref. 11). The resultant change in
turbine efficiency represents the effect of increased
blade loading as the blade number is reduced for a
given work requirement.
o Disk fr iction losses along the rotor backface are
estimated based on the results of the investigation
conducted in reference 12. This work is fully
described in reference 13. Note, however, that disk
friction is based on a full rotor backface disk. The
effect of rotor scallops on rotor disk fr iction is
currently not available.
o The exhaust diffuser loss correlation was based on the
Model GTC36-200 configuration shown in Figure ga. The
annular inlet to the diffuser is followed by a rotor
exit hub dump, a constant area section, and a conical
diffuser section. The constant area section down-
stream from the rotor exit dump allows the flow to mix
and stabilize, prior to the rapid diffusion near the
diffuser exit. The net result is a linear static
pressure rise along the diffuser shroud from rotor
exit to diffuser exit.
The non-dimensionalized diffuser geometry and perfor-
mance potential of this system is presented in
Table III. The loss character is tics of the Model
GTC36-200 exhaust diffuser as a function of inlet
swirl are presented in Figure 10. Although a number
of data points at high corrected speed or low-pressure
ratios result in relatively low loss levels, the pre-
dominant characteristic is at a high loss level, as
shown.
The lower loss levels would be equivalent to lower
diffuser inlet blockage levels, which is consistent
with diffuser test data. However, for the high work
and pressure ratio levels of this program, the higher
loss level is considered realistic. The minimum loss
coefficient (W . ) derived from Figure 10 is 0.30.
mln
21
COMBUSTION
DISCHARGE
~ I __ - - - L - - - - - - - - ~
EXHAUST
/DIFFUSER
.1R
____ 1-
AS = 1.865
A4
A ~ = 5.35
COMBUSTION
DISCHARGE
~ I __ - - - - L - - - - - - - ~
EXHAUST
/DIFFUSER
~ 1
--Cl.-4----+---m----- - Cl.-- --Cl.+--4---llHJ.-+-- -------1--Cl.--
~
A4 ~
ENGINE FLANGE PLANE ENGINE FLANGE PLANE
MEASURING STATION I MEASURING STATION I
FOR TURBINE EFFICIENCY FOR TURBINE EFFICIENCY
(A) (B)
RADIAL TURBINE SYSTEM PERFORMANCE FOR PARAMETRIC STUDY
IS DEFINED FROM SCROLL INLET (COMBUSTION DISCHARGE) TO
EXHAUST DIFFUSER EXIT PLANES (A) WITHOUT DESWIRL VANES'
(B) WITH DESWIRL VANES.
Figure 9. Exhaust Diffuser Configuration used in Parametric Study.
')
J
TABLE III. PERFORMANCE AND NONDlMENSIONAL GEOMETRY
FOR SELECTED DIFFUSER CONDIFURATION
Parameters
Exit Area/Inlet Area
L/flR
Diffuser Length
Inlet Height
Static Pressure Recovery at
Design Conditions, ~
Maximum Static Pressure Recovery
Diffuser Inlet Critical Velocity
Ratio at Design-Point, V/Acr
Average Diffuser Inlet Swirl at
Design-Point (a), deg
*lOO-percent corrected speed,
**llO-percent corrected speed,
PRT- T I1_4 =
PRT- TI1_4 =
Value
1.865
5.35
0.468*
0.60**
0.463
-3.0
4.07
3.85
23
4.0
13
z
1 3 ~ 3.0
0
i=
<t
a:
t-
2.0
z
w
CJ
LL
LL
W
0
CJ
1.0
~
0
..J
o
24
0
0
0
6
0
~
~
o
I I
80%
N/JO
90
100
110
120
P
RT
-DE = 2.0 TO 5.3
W
MIN
= 0.30
l ~
r
'/
;a/
y
A
~
~
~ (
0
0
Ofj
~
V
~ U
10 20 30 40
DUCT INLET SWIRL - DEGREES
Figure 10. Diffuser Duct Loss Data.
/
V
....
50
tv
VI
0.08
SERIES
Il)

I
I
S I
NON-SERIES

o
.J
w
a:
:J

w
a:
Q.
.J
<t
b
I-
I
I

__
o 0.2 0.4 0.6 0.8
MN - AVERAGE INLET MACH NUMBER
.
Figure 11. Deswirl Vane Loss Characteristics (ref. 3)
Using Arbitrary Air-Foil Shapes.
o When a rotor exit deswirl vane is used, the diffuser
non-dimensionalized parameters shown in Table III
will be applied at the deswirl vane exit. The losses
required for the deswir 1 vane to rees tablish ax ial
flow are based on the performance potential achievable
with arbitrary airfoil shapes indicated in Figure 11.
The test data shown are based on the work of Pratt and
Whitney (ref. 3). The deswirl vane geometry (solid-
ity, aspect ratio) and diffusion limits will be
derived from current axial compressor vane tech-
nology. A detailed description of the model used in
the parametric study is presented in Appendix B.
The above correlations represent the majority of the design
data available to the radial turbine designer. The current
sta te-of-the-art for radial turbine design is a result of two
major factors. First, the number of design variables associated
with the radial turbine are increased significantly compared to
the axial turbine. For example, rotor clearance effects are a
function of rotor inducer blade width, exducer blade heigh t,
backface seal clearance, scallop depth and scallop saddle con-
figuration. As a result, extensive (and expensive) rig test
programs are required to arr ive at meaningful correlations,
which tend to restr ict the dissemination of radial turbine
data. Secondly, the flow field in the radial turbine is highly
three-dimensional, radial stators are usually in the very low
aspect ratio region, and the relative vorticity complicates
rotor inlet and exit flow predictions.
Nevertheless, experience gained from axial turbine designs
has shown that the effects of rotor exit blockage and stage
reaction should not be ignored dur ing the turbine des ign pro-
cess. The effects of rotor blockage on overall radial turbine
performance do not appear to be available in the open litera-
ture. The rotor ex it hub blockage of AiResearch turbines used for
specific speed correlation is on the order of 30 percent, and
under this condition, excellent performance levels are still
achieved. However, for the automotive engine, turbine rotor
weight and inertia will have a significant influence on engine
acceleration and subsequent dr iveabili tYi for this reason, a
minimized rotor exit hub radius (lower rotor exducer hub-to-
inducer radius ratio) is desirable. The resultant increase in
rotor exit blade height would reduce the sensitivity to rotor
radial clearance.
Under these conditions, an order of magni tude effect for
rotor trailing edge blockage is necessary to arrive at a
rational tradeoff. Analytical analysis in terms of evaluating
boundary layer characteristics along the rotor meridional
streamlines is considered ineffective at the present time, due
to highly three-dimensional flow characteristics in the rotor.
The problem is further compounded by the local inducer flow sep-
aration associated with the highly loaded designs. Even if this
26
approach were pursued, the time required to analyze a large num-
ber of solutions on this basis would be prohibitive. Therefore,
a simplified approach is taken, whereby the relative effects of
rotor exit blockage are calculated, based on the compressible
mixing loss due to the rotor trailing edge thickness and local
relative velocity. The total pressure loss calculated in this
manner is compared with the loss calculated for an existing ref-
erence design, and the difference in calculated loss is used to
penalize the higher blockage designs. The model will be based
on modifying Stewart's (ref. 14) compressible mixing loss equa-
tions to account for trailing edge thickness alone.
The basis for this model is a result of numerous observa-
tions of test result that indicate the boundary layer thickness
in the rotor exit hub region is small. For example, Figure 12
shows the local eff iciency of the selected reference turbine
derived from rotor exit survey data. As indicated, the effi-
ciency is relatively constant at 95.5 percent up to approxi-
mately 40 percent of the blade height and then decreases rapidly
above this point. The implication from rotor powder traces and
predicted hub velocity distributions is that high secondary flow
generated at the rotor inducer region propagates to the rotor
exit shroud region. The flow along the hub line then expe-
riences high acceleration downstream from the rotor inducer
region at a relatively low loading level. The net result is
that the losses in the hub region are extremely low and account
for the high efficiencies measured.
On this basis, then, a calculated mixing loss due to the
trailing edge thickness alone would be expected to provide a
realistic estimate of the effects of increased blockage in this
region. However, even with this simplified approach, the calcu-
lation procedure becomes fairly complex in order to arrive at an
estimated performance decrement for rotor exit blockage. For
example, the rotor exit flow path dimensions, specified exit
work coefficient, and continuity must be solved to satisfy the
radial equilibrium equation inside the rotor trailing edge.
However, a radial thickness distribution is required that satis-
fies both the maximum tip thickness and maximum allowable blade
stress with a radial distribution of blade angle (usually radial
blades in this region). Then, the calculated radial relative
velocity in conjunction with the established radial thickness
distribution can be utilized to calculate a relative total pres-
sure loss from hub to tip based on blade blockage alone.
Since this type of analysis could not be achieved with the
one-dimensional performance model available for the parametric
study and the effect of rotor blockage should not be ignored,
the following procedure was adopted:
o First, the one-dimensional model was used to establish
peak performance character istics for a range of tip
speeds and rotational speeds.
27
29
I-
80
z
w
CJ
a:
w
a..
I-
60
::I:
e"
w
::I:
w
e"
<
100
(
ROTOR EXIT SHROUD
"
'"
~
ROTOR EXIT
~ V E Y D A T A
'\

40
<C
a..
l-
X
w
J
a:
e
20
....
e
a:
0.0 ROTOR EXIT HUB
I)
0.75 0.80 0.85 0.90 0.95 1.00
l1T-T' STAGE LOCAL TOTAL EFFICIENCY
Figure 12. Representative Rotor Exit Efficiency Characteristics
for Radial Turbines.
o Viable configurations were examined mechanically with
two-dimensional finite element methods that estab-
lished rotor blade thickness satisfying the maximum
allowable blade stress. A separate computer program
being developed was utilized to solve the radial equi-
librium equation with the specified blade thickness,
and an estimated rotor exit blockage performance
decrement was calculated.
The effect of stage reaction is expected to increase
directly with increasing work coefficient. However, isolating
this effect with highly loaded radial turbines is complicated by
rotor inducer incidence and stator loss, which will change as
inducer-to-exducer work split is adjusted to increase stage
reaction. If a systematic procedure is used to extract known
losses at tested off-design points, a relative rotor reaction
effect should evolve. This approach has been applied to the
Model GTP36-5l and 1976 research radial turbine test data to
derive a preliminary correlation of stage reaction on radial
turbine performance. Figure 13 shows results as desirable mean-
line reactions of 50 to 60 percent. The stage reaction shown in
Figure 13 is defined in the usual manner by the ratio of static
enthalpy across the rotor, relative to total stage work:
In terms of a stage work coefficient and relative velocity, the
stage reaction can be expressed as:
The effect of stage reaction was incorporated in the performance
model and the results of the parametric study will be presented
with and without these effects.
An additional effect that cannot be accounted for is rotor
shroud curvature. The shroud curvature, as a function of rotor
size and exducer tip-to-exducer tip radius ratio, increased with
increasing radius ratio and resulted in higher velocity peaks
and subsequent diffusion on the suction surface shroud stream-
line, as the radius ratio approached 1.0. Therefore, if the
parametric study indicates that high radius ratio designs are
more favorable, this effect would not become evident until the
detail rotor flow solution was in progress. To minimize this
occurrence, the radius ratio was limited to 0.80.
29
w
0
>
(.)
2:
w
(.)
1.01
u.
u.
w
~
::;)
~
1.00
X
<t
~
w
C!J
i:!
0.99
CI,)
0
~
>
(.)
0.98 2:
w
U
u.
u.
W
...J
0.97
i:!
v---
......
~ 0 .......
y
/
~ y
V

0
GTP36-51
0
1976 RESEARCH
/'
RADIAL TURBINE
0
I-
w
C!J
0.96
i:!
CI) 0.30 0.50 0.60 0.40
R
STG
- ROTOR MEANLINE REACTION
Figure 13. Preliminary Correlation for Radial Turbine Reaction.
3.5 Projected 1983-1985 Performance Improvements
The previously discussed performance model is based on cur-
rent 1979 technology. However, as indicated in Table I, the
high-stage work and performance level required by the turbine to
achieve a 50-percent reduction in fuel consumption is aggressive
and will require significant improvements in turbine efficiency.
With the smaller size associated with the automotive applica-
tion, both stator aspect ratio and rotor tip clearance penalties
will tend to restrict the attainable efficiency level, even if
optimum tip speed is achieved.
Historically, radial nozzle designs have been based on con-
stant section profiles with parallel end-walls. These condi-
tions, coupled with the favorable nozzle inlet-to-exit area
ratio, result in a theoretically high-acceleration two-
dimensional nozzle. However, in practice, the nozzle aspect
ratio is extremely low compared with an axial turbine, and the
rotor shroud curvature imposes cross-passage static pressure
gradients at the stator exit and transforms the highly acceler-
ating two-dimensional flow to a complex three-dimensional flow
field.
Indications that the stator is not the low-loss two-
dimensional device originally expected are implied from flow
visualization techniques (powder traces) and from an inability
to match radial turbine off-design test data utilizing tech-
niques similar to reference 15, unless stator loss coefficients
are increased significantly. Powder traces from a 0.30 aspect
ratio radial nozzle showed that secondary flow accumulations
covered approximately 90 percent of the passage in the trailing
edge suction surface region. Even with constant section pro-
files and parallel end-walls, secondary flows would be expected
due to the end-wall pressure-to-suction static pressure grad-
ients that result from the high turning. In addition, studies
conducted by Langston (ref. 16) and Gaugler (ref. 17) show that
blade-to-blade secondary flows are further compounded by the so-
called horseshoe vortex generated at the nozzle leading edge.
The concept of stator end-wall contour ing and non-free
vortex work distributions has been widely utilized in axial tur-
bine des igns over the past few years wi th excellent success.
Since the effect of end-wall contouring is to reduce the blade-
to-blade static pressure gradient (loading) at the nozzle end-
walls, this concept will apply to radial nozzles also. Indeed,
the benefits may be more pronounced with radial nozzles since
the aspect ratios are significantly lower compared with axial
nozzles. Therefore, in 1978 the potential benefits of end-wall
contouring were investigated for radial stators with a compress-
ible 3-D viscous flow analysis computer program developed by
AiResearch under USAF Contract (F33615-76-C-2110). The study
was conducted as part of the TARADCOM advanced radial turbine
program (ref. 18). Five different end-wall configurations were
31
examined (including a parallel wall baseline case) and compared
on a predicted total loss basis. The analytical results showed
that total pressure loss is reduced by 50 percent with end-wall
contouring. Additional analytical and experimental effort is
required to identify the true potential of the concept and also
evaluate the performance potential of combining both the end-
wall contouring and three-dimensional vane profile (non-free
vortex type) concepts. However, on the basis of preliminary
analysis, a 1.0-point improvement in turbine performance is con-
sidered feasible.
Although the sensitivity to rotor shroud clearance is
reduced compared with axial turbines, scaling the allowable run-
ning clearance with the reduced size of the automotive turbine
may not be possible. In addition, the rotor scallop introduces
a third clearance effect that is a function of both backface
seal clearance scallop depth and scallop saddle configuration.
Under these conditions, the estimated clearances shown in
Table II will result in relatively high performance decrements.
If higher engine running clearances are required due to shaft
dynamics or differential thermo growth, then attainable radial
turbine performance will be appreciably effected. Therefore,
the concept of clearance treatment to minimize clearance effects
will be considered for this application.
Dejc (ref. 19) has shown reductions of approximately
50 percent in clearance loss on axial turbine stages through the
use of circumferential grooves in the rotor shroud. Although
this concept does not appear to have been applied to a radial
turbine configuration, the clearance phenomena are basically the
same. The shroud and backface seal surface treatment is consid-
ered viable for this application. Based on a 35-percent shroud
and backface clearance loss reductions, a 1.35-point increase in
performance is considered feasible. The total improvement pre-
dicted for both stator loss reduction and rotor clearance
effects is 2.35 points for the 1983-1985 time frame.
Performance levels of the turbine inlet scroll, exhaust
diffuser and deswirl vanes will be maintained at current 1979
predicted levels. This does not imply that the performance
potential of these components cannot be improved, but except for
the deswirl vane, less research effort has been devoted to these
components. In addition, the configurations ultimately adapted
for this size shaft engine may differ from the representative
configurations used in the parametric study.
3.6 Preliminary Stator and Rotor Geometry
The radial turbine stator preliminary geometry was estab-
lished with the one-dimensional performance model. For a given
stator trailing edge thickness of 0.030 in. (from ceramic con-
siderations) and a trailing edge blockage of 10.0 percent (which
32
is consistent with the geometries used in the specific speed
correlation), the stator vane number and radial chord were
established from a radial nozzle loading coefficient. The load-
ing coefficient was based on the Zweiffel loading concept
applied specifically to the radial nozzle geometry.
The rotor inlet and exit radii and blade height were estab-
lished with the one-dimensional performance model. The rotor
flow path and axial length were evaluated from a meanline load-
ing distribution through the rotor over a range of meridional
flow shapes and axial lengths. The entire blade height through
the rotor was treated as a stream tube and the continuity equa-
tions, with an estimated rotor blade angle distribution, deter-
mine the rotor mean-line relative velocity distribution. The
condition of zero absolute vorticity was used to arrive at a
blade-to-blade loading.
3.7 Parametric Study Results
A one-dimensional performance evaluation model for the tur-
bine scroll, stage, and exhaust diffuser system was first estab-
lished by modifying an existing radial turbine design program to
account for the scroll and duct losses and projected 1985 per-
formance improvements. The computer program calculates geo-
metric and vector diagram quantities at the stator inlet, stator
exit, rotor inlet and rotor exit mean-line stations. Due to the
number of var iables required to def ine the turbine f low path,
the input to the program consisted of both free and fixed param-
eters. The free parameters were:
o
o
o
o
o
o
o
o
Rotational speed (N) = 80,000, 95,000 and 110,000 rpm
Rotor inducer tip speed (
Ut
3) = 549, 610, 701, and 782
m/sec (1800, 2000, 2300, and 2565 fps)
Rotor inlet blade angle (#B) = 0, 10, and 20 degrees
Inducer tip-to-exducer tip radius ratio, 0.80 maximum
Rotor exit annular area = 45.2 to 58.1 cm
2
(7.0 to 9.0
in.
2
)
Rotor inducer work coefficient (which establishes an
inducer-to-exducer work split, rotor exit swirl, and
Mach number)
With and without deswirl vane
With and without predicted reaction effects
33
Fixed parameters used in the basic parametric study were:
o Rotor blade number, set at 12 (The combined effects
of rotor blade number and rotor exit blockage were
examined for specific cases but are not included in
the basic parametric curves)
o Stator exit vane angle (defined at the rotor inlet
radii) was set at 72 degrees
o A stator loading parameter was fixed
o The vane trailing edge thickness was fixed at 0.076 cm
(0.030 in.)
o Rotor clearances were set at 0.020, 0.020, and 0.010
for rotor backface, axial, and radial, respectively
o The vane trailing edge blockage was set at 10 percent,
which was considered the maximum allowable for the
specific speed correlation to be valid
o The conditions of stator blockage, trailing edge
thickness and loading criteria then set the stator
vane number
o The cycle conditions, except for efficiency and duct
loss, were fixed
Resul ts of the basic parametr ic study are presented in Fig-
ures 14 through 22. The bas ic character istics of the aero-
dynamic study are summarized below:
34
o For a specified tip speed, the parametric study shows
that an optimum system efficiency is obtained for all
rotational speeds, both types of blading, wi th and
without deswirl vanes
o Rotor reaction significantly affects both the effi-
ciency level and optimum inlet work coeff icient at
lower tip speed levels
o Peak eff iciency increases uniformly wi th increasing
rotor inlet blade angle except at high tip speeds when
rotor inlet incidence is eliminated
o The level of tip speed required to achieve the system
efficiency goal is relatively high [579 to 701 m/sec
(1900 to 2300 fps)]
o The turbine efficiency level increases with rota-
tional s p e e d ~ however, when the effect of rotor exit
hub blockage is considered, this effect is reduced
088
086
084
08 2
080
z
UJ
C3
u::
u..
UJ
....I

....I

:!E
UJ
Iii

078
076
074
072
07
."...-
- ........
L

"
)/
782 m/sec
Ut3 = 2565 fps
".
N = 80,000 RPM
I3s = 0
NB = 12
_ T _ 701 m/.!c
Ut3 - 2300 fps

"""- I' "-

SYSTEM EFFICIENCY GOAL-
rl
NO DESWIRL VANE, BASE
---NO DESWIRL VANE,
WITH REACTION
-----REACTION,DESWIRLVANE
U - 610 m/sec 7
t3 - 2000 fps

.....
"
>'"

/
/"
''\
.... ,\
,
/
/
,,/
\
7
I
\

\

"
Ut3 = 1800 fps
V"
'"
S'
--
I
/
........
'\
I
I

I
"r.

-
08 09 10 11 12 13 14 15 16
ROTOR INLET WORK COEFFICIENT, X3
Figure 14. Turbine System Parametric Study, N = 80,000 rpm,
fiB = 0
35
36
N = 80,000 RPM
/36 -= 10
0
NO DESWIRL VANE
NS = 12
0.90
CD
I
0

0.88
I-

0.86 CJ
z
w
CJ
::tE
GOAL
u
t
_
782 m/sec
3
"V

1-- - ,-
--- - ------- -
/
v
V
U
t 701 m/sec \
3 .. 2300 fps
11.
0.84
u..
w
-I

0.82
0
l-
I
0
L..
"-
V
--,
"-
/"'3 .
610 m/sec
i',
2000 fps
I-
0.80
I
-I SASE

I-
0
0.78 I-

w
t-
(J)
0.76
>
(J)
/V::
-.........
[\

u
t

.. 549 m/sec
3
1800 fps /0-

0.74
0.8 0.9 1.0 1.1 1.2 1.3 1.4 1.5 1.6
ROTOR INLET WORK COEFFICIENT - A3
f3
B
= 10.
N = 80,000 RPM
PB = 20
NO DESWIRL VANE
NB = 12
0.90
cc
I
SYST1EM EFF1ICIENCY GOAL
0
"'i
0.88
I-

u
z 0.86
w
-
U
-
LL
LL
0.84
w
-I
<t
I-
0
0.82

0

0.80
-I
<t
I-
0
I-
0.78
:E
w
I-
en
>
0.76
en
_1
1--
n
V-

--- --- 1----
---- -
\
V Uta 701 m/S1!c ,
.. 2300 fps

/
U
t
= 782 m1S1!C
.--.........
3 2565 fps
Vu
ta
\
/
_ 610 m1S1!C -\
2000 fps
I
/
BASE' ..
I. -""TH -I'e
:

U 549 m1S1!C
I t3 - 1800 fps
I
-II 1\
0.8 0.9 1.0 1.5 1.6
1.1 1.2 1.3 1.4
ROTOR INLET WORK COEFFICIENT, A3
f3
B
= 20
0

37
09 0
088
08 6
'f084
o

"'"
>'
" 082
z
UJ
U
u:
II.
UJ
...J
08

0
...J
078
::E
UJ

07 6
07 4
072
U _ 782 m/sec
t3 - 2565 fps
.
N = 95,000 RPM
f3S= 0
NB = 12
u = 701
t3 2300 fps
DIFFUSER ONLY
-
NO DESWIRL VANE.

tIT
- ...... --........
WITH REACTION
./' "'
- - - - - REACTION. DESWIRL VANE
//;'
V

SYSTE"; EFFICIENCY GOAL 7
-
I

t-- _ ...
."
\
V
I'
610 m/sec
U
t3
= 2000 fps - "

..... .......

V /'
I' .....
\
'.

,/
",
/

(\L

8..qs.
\

Lt

V
u 549 mI_',
1-' t3 = 1800 fps "1"

01-
1'-
08 09 10 1 1 12 13 14 15 16
ROTOR INLET WORK COEFFICIENT. X3
/3 - 0
B -
38
N = 95,000 RPM
PB = 10
0
NB = 12

, I ! !
= 701 m/sec
I
....-.'" t-'-- 2300 fps I
088 ....
.,\)', \ \ SYSTEM EFFICIENCY GOAL Ii
I' ,......;,,--'1--- -+---- -L.....-4- -
U = 782 m/sec ",...... ........
t3 2565 fps ' ..
I
, 610 m/sec \/ -
I /'
Ut3 = 2000 fps ./

/

I I '

g / \
U ='549
I'
080 f---jf-------I--------+--------l-------l-- t3 1800 BASE _ t-...
!
/ ...

w

---- NO DESWIRL VANE, BASE
.----REACTION, DESWIRL VANE
---NO DESWIRL VANE, WITH REACTION

08
Figure 18.
09 10 11 12 13 14 15 16
Turbine System Parametric Study, N = 95,000 rpm,
fiB = 10.
39
'1'
0

I

.,.
>'
u
z
w
U
u::
...
W
...I
<I:

0

0

...I
<I:

0

:E
w

II)
>
II)
40
N = 95,000 RPM
fJa= 20
0
NB = 12
090
.-7
r- U
t
= 701 m/sec

3 2300 fps

0< r--/
k1
t/'
'/
.<\

..
SYSTEM EFFICIENCY GOAL
-
088
086
084
082
080

1"
" = 782 m/sec
L' t3 2565 fps
-
...........
/
610 m/sec
''i
I'"
U
t3
= 2000 fps
-""

/ /
\
I '
L
U - 549 m/sec <:
t3 - 1800 fps
/

i'
NO DESWIRL VANE, BASE

----- REACTION, DESWIRL VANE
I
... iz'l)';
---NO OESWIRL VANE, WITH REACTION

"1eo)';
078
076
08 09
Figure 19.
10 1 1 12 13 14 15
ROTOR INLET WORK COEFFICIENT, ).3
Turbine System Parametric Study, N = 95,000
f3
B
= 20.
{01-
"
16
N = 110,000
I3B = 0
NB = 12
..
1
/ NO DESWIRL VANE, BASE
U
t3
= 701 M SEC _____ REACTION, DESWIRL VANE
2300 FPS
---NO DESWIRL VANE, WITH REACTION
o

..!.

>'
(,,)
z
w

u.
u.
W
-J
g

o

-J

w
li;

SYSTEM EFFICIENCY GOAL -----,.
0.86
U
t3
= 549 M/SEC
1800 f
PS
---+----1
U
t3
= 610 M/SEC ...... __ NO SOLUTIONS AT MAXIMUM
2000 FPS RADIUS RATIO (0.80) OR
MAXIMUM HUB BLOCKAGE
(50 PCT) FOR ROTOR EXIT

AREAS OF 45.15, 51.6,58.1
--......:...-+- cm
2
(7, 8, 9 m.
2
)

0.8 0.9 1.0 1.1 1.2 1.3 1.4 1.5 1.6
Figure 20. Turbine System Parametric Study, N = 110,000 rpm, P
B
= 0.
42
N = 110,000 RPM
{3s = 10
0
NO DESWIRL VANE
NS = 12
0.90
U)
I
0
..
0.88
I-
~
> ~
-- --
c.J
Z
w
0.86
~
SYSTEM EFFICIENCY SASE
LL
GOAL
LL
w
...I
<2:
0.84
I-
0
..
0
..
...I
0.82

I-
U 549 m/sec
0 t3 = 1800 fps
I-
NO SOLUTION
:2:
.
W
I-
0.80
en
>
en
0.78 a...----I. ___ .....L.... __ ---L ___ .....I-__ ---L ___ .....I-__ --'
0.8
Figure 21.
0.9 1.0 1.1 1.2 1.3 1.4
Turbine System Parametr ic Study, N = 110,000 rpm,
f3
B
= 10.
0.90
co
I
0

0.88

(.)
z
0.86
w
(.)
u.
u.
W
...J
0.84

0
'7
0
'7
0.82 ..J

0

w
0.80
lii
>
CI)
0.78
Figure 22.
N = 110,000 RPM
/3s = 20
0
NO DESWIRL VANES
NS = 12
U' LR\
.
hfpo) 1--,
I u
t3
- 701 mise!}
L V (23DOrp,' \
.-i----. ....,.-
tVfITEM EFFICIENCV
.----
\

I
/ Wrnt
..--
GOAL

',\011
V
\
u
t
_
549 mlsec
/
3
1..1800 fpsJ-
j
NO SOl!JTIONS
0.8 0.9
1.0 1.1 1.2 1.3 1.4
ROTOR INLET WORK COEFFICIENT - A3
Turbine System Parametric Study, N = 110,000 rpm,
.BB = 20
0

43
o At lower rotor tip speeds [549 to 610 m/sec (1800 to
2000 ft/sec)], system efficiency is increased from 2
to 6 points with the addition of deswirl vanes
In Figures 14 through 22, the peak efficiency for each
value of inlet work coefficient analyzed is based on the maximum
attainable efficiency for the range of rotor exi t areas and
radius ratios examined. Rotor exit areas of 45.2, 51.6 and 58.1
cm
2
(7, 8, and 9 in.
2
) were specified in the parametric study
and, in general, peak efficiency occurred at or near the maximum
value specif ied. The values of rotor exi t area selected were
established from preliminary rotor exit (zero swirl) continuity
calculations. However, as shown in Figures 20 through 22, cer-
tain combinations of rotation and tip speed will result in
either negative or unacceptably low exit hub radii and no solu-
tions will exist. Although high rotor exit area is desirable
from a reduced exducer radial clearance loss standpoint, the
higher exit area also requires increased hub blade thickness for
a glven blade number and blade stress level. The resultant
increase in hub blockage could therefore negate the gains in
clearance loss.
For the radius ratio, peak system efficiency occurred at
0.80 (the maximum allowable) for tip speeds up to 610 m/sec
(2000 fps). However, as optimum tip speed is approached, peak
system efficiency occurs at lower radius ratios (0.65). These
trends are associated with the rotor exit mean work coefficient.
As tip speed is reduced from the optimum value, the level of
rotor exit work coefficient increases rapidly. Therefore, the
rotor exit condi tions and duct loss are more sensi tive to
changes in the rotor exit mean radius. Under these conditions,
the reduction in duct loss at higher radlUS ratios is greater
than the increase in loss due to increased exducer clearance
effects. It should be noted that the optimum radius ratio of
0.80 for the lower tip speed cases also results in the highest
rotor shroud curvature, and the performance effects for this
parameter are currently not known. Therefore, if detailed mech-
anical analysis indicates this level of tip speed is required to
satisfy the rotor life goals, then a rotor internal flow solu-
tion at both 0.80 and at lower values must be compared and the
impact of the shroud loading evaluated. Figure 23 illustrates
the effects of rotor exit area and radius ratio on peak system
efficiency for 95,000 rpm and tip speeds of 610 m/sec (2000 fps)
and 701 m/sec (2300 fps).
Specific vector diagrams and system parameters along the
system efficiency envelope are presented in Figures 24 and 25
for 80,000 rpm and the range of tip speeds investigated. Of
particular note is the magni tude of the rotor exi t relative
critical velocity ratio associated with the deswirl vane cases
at peak system eff iciency. These higher values result in
increased rotor exit blockage losses. In addition, since the
44
RADIAL
MODE 3 MODE 4
FREQUENCY = 24,837 Hz FREQUENCY = 38,916 Hz
SCALE FACTOR = 2.00
RPM = 95,000
(!) MAXIMUM ~ MINIMUM
Figure 46. Normalized Deflections (Modes 3 and 4).
RADIAL
MODE 1
MODE 2
FREQUENCY = 12,295 Hz
RPM = 95,000
SCALE FACTOR = 2.00
(!) MAXIMUM 6 MINIMUM
Figure 45. Normalized Deflections (Modes land 2).
o
o

o
o

o
o
o
~ ~ - - - - - - - - - - - - - - ~ ~ - - - - - - - - - - - - - - ~
>-
u
z
Wo
::JO
o ~ - - - - - - - - - - - - - - ~ ~ ~ ~ - - - - - - - - ~ ~
W ~
~
lJ...
o
o

~ - - - - - - - - - - - - ~ - - ~ ~ ~ - - - - - - - - - - ~ ~
o
o
CO
o
o
o
0.00 25.00 50.00 75.00 100.00
SPEEO(PER CENT)
NASA ART CERAMIC TURBINE CASE NO. 2C.AE=7.0
VIBRATION FREQUENCIES 100"1. RPM= 95000
CAMPBELL DIAGRAM 50 i. RPM= 47500
125.00
Figure 44.
Campbell Diagram for 95,OOO-rpm Case with 45.2 cm
2
(7 in.
2
) Exit Area.
70
0'1
1.0
ISOPLETH INTERVAL = 0.051 em (20 mils)
TANGENTIAL THICKNESS
SCALE FACTOR = 1.50
C) MAXIMUM 6 MINIMUM
RADIAL
ISOPLETH INTERVAL = 3.447 kN/em
2
(5 ksj)
RPM = 95,000
TIP SPEED = 664 m/see (2180 fps)
NS = 12
A
EXIT
= 45.2 em
2
(7 in.
2
)
Figure 43. Tangential Thicknesses and Maximum Principal Stresses.
68
o
o

CO
IJ)
o
o

o
o

~ - - - - - - - - - - - - - - ~ ~ - - - - - - - - - - - - - - ~
>-
U
Z
Wo
:::JO
0'
W ~
a:::
l.J...
o
~ ~ - - - - - - - - - - - - ~ ~ ~ ~ ~ - - - - - - - - - - ~
co
-
o
o
co
o
o

o
0.00 2S.00 SO.OO 7S.00 100.00
SPEEO(PER CENT)
NASA ART CERAMIC TURBINE CASE NO.2
VIBRATION FREQUENCIES 100r. RPM= 9S000
CAMPBELL DIAGRAM 50 r. RPM= 47500
25E
17E
OE
E
E
E
E
E
E
E
E
E
12S.00
Figure 42. Campbell Diagram for 95,000-rpm Case with
58.1 cm
2
(9.0 in.
2
) Exit Area.
SCALE FACTOR = 1.50
RADIAL
2
(5 ksj)
RPM = 95,000
NB = 12
AEXIT = 58.1 em
2
(9 in.
2
)
Figure 41. Tangential Thicknesses and Maximum
Principal Stresses.
0'1
0'1
RADIAL
MODE 3 MODE 4
FREQUENCY = 26,445 Hz FREQUENCY = 35,951 Hz
SCALE FACTOR = 2.00 RPM = 80,000
o MAXIMUM 6 MINIMUM
Figure 40. Normalized Deflections (Modes 3 and 4).
MODE 1
SCALE FACTOR = 2.00
RADIAL
MODE 2
RPM = 80,000
Figure 39. Normalized Deflections (Modes land 2).
64
>-
u
Z
o
e

(0
UJ
o
o

<Xl
V
o
o

We ~ - - - - - - - - - - - - - - ~ ~ - - - - - - ~ ~ - - - - ~
:::Jo
O
W ~
0:::
l.L.
o
o
(0
- ~ - - - - - - - - - - - - ~ ~ ~ - - ~ ~ - - - - - - - - ~
o
o

<Xl
o
o
o-fM-
0'.00' 25.00 50.00 75.00
SPEEO(PER CENT)
NASA ART CERAMIC TURBINE CASE NO.1
VIBRATION FREQUENCIES tooY. RPM= 80008
CAMPBELL DIAGRAM 50 Y. RPM= 40000
Figure 38. Campbell Diagram for 80 , 000-rpm Design.
125.00
0'1
W
ISOPLETH INTERVAL = 0.051 CM (20 MILS)
SCALE FACTOR = 1.50
RADIAL
ISOPLETH INTERVAL = 3.447 kN/cm
2
(5 ksi)
RPM = 80,000
TIP SPEED = 701 m/sec (2300 fps)
NS = 12
Figure 37. Tangential Thickness and Maximum Principal Stresses.
en
:J
6.78 2.67
5.08 2.00
~ 3.38 1.33
a:
1.70 0.67
0.00 0.00
in. -.
em ...
I
-0.67
-1.70
I I I I
0.00 0.67 1.33 2.00
0.00 1.70 3.38 5.08
AXIAL LENGTH
Figure 36. Two-Dimensional Finite Element Model for ISOPDQ.
62
The basic flow paths depicted in Figure 31 were examined
for mechanical considerations using the 2-D finite element pro-
gram, ISOPDQ (ref. 20), illustrated in Figure 36. Each flow
path configuration was analyzed by an iteration on blade thick-
ness and wheel/disk backface contour to achieve 2-D maximum
principal blade stresses of 13.8 kN/cm
2
(20 ksi).
Figure 37 illustrates the tangential blade thicknesses and
maximum principal stresses for 80,000 rpm with 12 blades. This
configuration exceeds the 24.1 kN/cm
2
(35 ksi) stress limit in
the disk. However, the blade vibration character istics (Fig-
ure 38) are acceptable. The most critical blade vibration con-
dition expected for this engine application is the stator vane
passing frequency. This size engine will be designed with a
range of stator counts from 17 to 25. With a stator count of 23,
this design has only one blade-natural frequency in the engine
operating range. This poss ible frequency interference is an
exducer torsional third mode (Figures 39 and 40), which is not a
problem with any existing radial turbines.
Figure 41 illustrates the blade tangential thicknesses and
the maximum principal stresses for 95,000 rpm, with 58.1 cm
2
(9 in.2) of rotor exit area and 12 blades. Both blade and disk
stresses are within established limits. Based on the Campbell
diagram of Figure 42, a stator count selection between 17 and 25
would not reduce the number of blade natural frequencies below
two in the operating range. To alter the blade for this speed to
a shape similar to the 80,000-rpm case, which exhibits better
vibration characteristics, an exit area of 45.2 cm
2
(7 in.2) was
examined. Figure 43 illustrates the blade tangential thick-
nesses and maximum principal stresses for this exit area. Fig-
ure 44 shows that 23 stator vanes will reduce the number of
blade-natural frequencies in the operating range to one. Fig-
ures 45 and 46 illustrate the mode shapes for this blade config-
uration.
Figure 47 illustrates the blade tangential stresses, and
Figure 48 depicts the blade-natural frequencies for 110,000 rpm.
The maximum principal stresses in the disk are somewhat lower
than for 95,000 rpm, but the blade vibration characteristics
appear to add considerable risk for this wheel configuration.
Although maximum principal disk stresses continue to decrease as
physical speed increases (Figure 49), blade vibration charac-
teristics set a practical upward limit on physical speed to the
95,000- to 100,000-rpm range.
After the physical speed range was established, other mech-
anical parameters were examined. The first of these variables
was the number of blades. When the blade count was varied from
10 to 16, very little change was observed in maximum stress at
the center of the wheel (Figure 50). These slight changes in
bore stress wi th changes in blade count indicate that bore
stress is more dependent on disk size than blade configuration.
61
with an axial blade length of 4.1 cm (1.6 in.). However, the
final axial length and blade number was established during the
detail rotor design.
3.9 Parametric study for Two-Dimensional Mechanical Analysis
The mechanical analysis was initiated with the character-
i.stic flow path configurations that resulted from the parametric
study for 80,000, 95,000, and 110,000 rpm.
The life objective of the program was to achieve a fast
fracture modulus of rupture (MOR) probability of success of
0.9999, using a selected ceramic material with a characteristic
strength of 62.1 kN/cm
2
(90 ksi) and a Weibull modulus of 15.
These material property values are projections of selected
ceramic material capability for the 1983 time period and were
derived from a review of current ceramic matrials and associated
processing capabilities. The following projected 1983 proper-
ties for near-theoretical density Si3N4 and SiC were used as a
basis for this design study:
Projected Strength (KSI) Weibull m
Material RT 1100C RT 11000C
(20l2P)
Si3N4
100 90 15 15
SiC 90 90 15 15
To provide meaningful probability of success calculations,
stresses must be integrated over both the surface area and vol-
ume. Three-dimensional stress analysis is required to make the
integrations. Initially, however, to examine a large number of
configurations in the parametric study, a 2-D stress analysis at
room temperature was used.
Based on previous experience with ceramic radial turbines
(utilizing similar material properties), stress limits of 13.8
kN/cm
2
(20 ksi) in the blades and 24.1 kN/cm
2
(35 ksi) in the
disk were eftablished. The blade limi t for the 2-D analysis
[13.8 kN/cm (20 ksi)] was fixed at a level providing adequate
stress margin to accommodate stress concentrations in the com-
plex geometry of the blade/disk interface region. The 2-D fin-
ite element program does not compute concentrations in this
transition region.
60
m/see fps
N = 95,000 RPM
U
t
= 701 m/see (2300 fps)
3
= 0
A4 = 45.2 em
2
(7 in.
2
)
610
3: 457 1500 ........ =_
>
!:::
u
e
...J

w
>

Z = 0
....I
W
c:: 152
(1.5 in.)
I-Z=3.
8em
l
a:
e
l-
e
a:
o
-152 -500
in .... 0
em ... 0
0.5
1.27
1.0
2.54
1.5
3.81
2.0
5.08
MEANLINE MERIDIONAL LENGTH
2.5
6.35
3.0
7.62
Figure 35. Rotor Solidity Study, Axial Length = 3.8 em
(1.5 in.).
59
N = 95,000 RPM
U
t3
= 701 m/sec (2300 in. )
{JB = 0
A4 = 45.2 em
2
(7 in. 2)
m/see fps
610
3:
>.
l-
0305
o
...J
W
>
W
>

-I
w
a:
a:
o
bOO
a:
-152 -500
0
em 0
0.5
1.27
1.0
2.54
1.5
3.81
2.0
5.08
MEANLINE MERIDIONAL LENGTH
Z = 0
I-z = 4.3 em
I (1.7 in.)
---ct
2.5
6.35
3.0
7.62
Figure 34. Rotor Solidity Study, Axial Length = 4.3 crnCl.7 in.).
58
3:

I-
0
0
..I
W
>
W
>
i=
<C
...J
w
a:
a:
0
I-
0
a:
W
..I
(!)
Z 60
<X:
W co.. 50
c
<X:z
40
I-
30
-'a:
In 20
W-
10
a:
o 0
I-
o
a:
m/see fps
610 2000
I

sa
II
N
I
N = 95,000 RPM
U
t
= 701 m/see (2300 fps)
3
{3s = 0
A4 = 45.2 em
2
(7 in.
2
)
V
/
-
---
/
Z = 0
V
Z = 4.8 em (1.9 in.)

/
Z = "b"-- .t-

\
STREAMLINE
-.

I
----........
/VELOCITY 1
1500 457
305 1000
152 500
0 0
r--

/fJ
... ::-........

... -
1/1
' ..

/
MC.-

W'ROTOR EXIT-.J
,
RELATIVE VELOCITY
h9
FROM 1-0 VECTOR
DIAGRAM

PS
/ '\2
,
",._.../ .. J
./
'/
V
-152 -500
in ....... 0
em ..... 0
0.5
1.27
1.0
2.54
1.5
3.81
2.0
5.08
M, MEANLINE MERIDIONAL LENGTH
2.5
6.35
3.0
7.62
Figure 33. Rotor Solidity Study, Axial Length = 4.8 ern (1.9 in.).
57
56
!'I = 95,000 rpm
U
t3
= 701 m!sec(2300 fps)
iJe = 0
C-I!l !!.1..1..
203 080
a:
<1
-I )----
ci
a:
178 070 0
:I;
u
-'

C AR

a:
152 06
a:
-I
0

'"
1 02 040
i=
w
-'

076 030
a: :1;"

0-
a::i:
;:w

051 020
wgj
a:
0"-
-<

!;t
ti
025
010 --
-
C!I
060

'"
a:
z"
........ [
p-
........
........
0
t=
u 050

w
a:
w ---
-[
C!I

'"
040
If
089
..2
-'
,

0"-

--(
1>----
::!:u 088
w
z

"'U
>-
"' ... ...
w
--C
087
A3 = 1.0
A4 = 58.1 cm
2
(9 in.
2
)
NO DESWIRL VANES
R
t
i
Rt
3 = 0.650
---
---l
- -
--
-
-
0'
-{'

//

"..,
,.'!.---
,,"'"
".,
.--"
>-

-
I---
V BASE FOR PARAMETRIC
STUDY
RSTG .J
-
-
-
_I u-

--
---
:-- _ ViA
cr
)3
--
--
-
l.----C
y-

---
---=

....:;..>-<-( P ........ ____

I- - --c

-
400

O.l"",
!;t<! w
300 tid
-,z C!I
z w
c
O:J

200
060
050 0

a:
t;
w
040
a:

030
1 o
095
09
300 Z
cr."
w
III
::!:
:J
Z
200 w

a:
1
S
00 II>
66 68 70 72 74 76 78
STATOR EXIT ANGLE, a40 DEGREES
Figure 32. Stator Solidity Study, N = 95,000 rpm, P
B
= 0.
3.8 Stator and Rotor Solidity Studies
Since rotor inlet incidence loss is a function of the inlet
stage work coefficient, the effect of stator exit flow angle can
be examined independently at a constant work coefficient. The
parametric study was based on a constant exit angle of
72 degrees, which has proven to be a reasonable compromise
between the sensitivity of manufacturing tolerances and rotor
shroud and backface clearance effects. The effect of varying
the stator exit angle is illustrated in Figure 32. For the
study, stator trailing edge thickness and blockage were held at
0.030 in. and 10 percent, respectively. The following charac-
teristics are observed:
o From continuity, increasing the stator exit angle
increases the stator exit "b" width (also the rotor
inlet "b" width)
o For a constant stator loading, exit blockage, and
trailing edge thickness, increasing stator exit angle
increases the stator aspect ratio (b/AR)
o Increasing the stator exit angle reduces the rotor
inlet relative velocity and increases rotor reaction
o For an increase in stator exit angle from 66 to
78 degrees, system efficiency increases by 1.3 points
due to reduced shroud and backface clearance effects
and increased reaction effects
The rotor axial chord was evaluated on the basis of a mean
line loading. The loading was established by treating the
entire flow path as a stream tube and by solving the equations
of continuity and energy from inlet to exit with the estimated
flow path and rotor blade angle distribution. The condition of
zero absolute vorticity was used to establish the blade suction
and pressure surface velocities.
The results of this analysis for the 701-m/sec (2300-fps),
95,000-rpm configuration are presented in Figures 33 to 35 for
axial lengths of 4.8, 4.3, and 3.8 cm (1.9, 1.7, and 1.5 in.),
respectively. For each axial length examined, a range of rotor
blade numbers from 12 to 16 was evaluated. The results show
that inducer loading was not affected by axial length, since the
blade angle distribution in this region was fixed with radial
blades. However, in the exducer and trailing edge region, the
blade loading was significantly affected by axial length due to
the rate of change in rotor blade angle. Conversely, the rotor
blade number resulted in a change in rotor blade loading
throughout the entire flow path, as expected. On the basis of
this study, an axial length between 4.1 and 4.3 cm (1.6 and 1.7
in.) resulted in satisfactory rotor blade velocity distr i-
butions. The majority of the mechanical analysis was performed
55
em in.
7.62 3.0
5.08 2.0
2.54 1.0
o
o
710 m/see
- U
t3
= (2330 fps)
625 m/see
- Ut3 = (2050 fps)
N = 80,000
A4 = 58.1 em
2
(9.0 in.
2
)
A4 = 51.6 em
2
(8.0 in.
2
)
A4 = 45.2 em
2
(7.0 in.
2
)
U - 674 m/see
- t3 - (2210 fps)
U - 591 mIse
- t3 -(1940fps)
N = 95,000
664 m/see
r-----, --U
t3
= (2180 fps)
_ 588 m/see
--Ut3 - (1930 fps)
N = 110,000
Figure 31. Characteristic Rotor Flow Path Configurations for
Selected Parametric Study Rotational Speeds.
111
W
c
w
W
Q.
en
Q.
t-
o::
w
CJ
=>
C
Z
NB = 12
.fin.
TIP SPEED ENVELOPE
WITH REACTION AND
DESWIRL VANE EFFECTS
TO ACH I EVE SYSTEM
EFFICIENCY OF 0.870
ROTOR EXIT AREA = 58.1 cm
2
(9 in.
2
)
c3 = 72.0
732
671
610
- ""--'7--).'
.y'-...p.---
WITH DESWIRL VANES
540
80,000 90,000 100,000
ROTATIONAL SPEED (RPM)
Figure 30. Tip Speed Envelope with Reaction and Deswirl Vane Effects,
but No Rotor Exit Blockage Effects.
110,000
52
0.90
A
0.88
co
0.86
I
0

I
0.84

>
(J
0.82
z
w
(J
/V

-
"V

.""
-- ---
---
1--'-
--
:---
---.-
-.
/
/.
" " ,,aB = 20.0
0

/
[
,aB = 10.0
0
W'
,aB = 0.0
0
I
u.
0.80 u.
w

w
0.78

en
BASE WITH REACTION
-
- ViNES
>
en
NB = 12
I I
0.76 I-- - ROTOR EXIT AREA 51.6 to 58.1 cm
2
(8 to 9 in.
2
) - -
cx3 = 72.0
0
I
0.74
fps ...... 1800 1900 2000 2100 2200 2300 2400 2500 2600
m/sec...... 549 579 610 640 671 701 732 762 792
U
t3
, INDUCER TIP SPEED
Figure 29. Peak System Efficiency for 95,000 rpm with
Deswirl Vanes.
N = 110,000
NB = 12
REACTION EFFECTS
NO DESWIRL VANES
ROTOR EXIT AREA '" 58.1 cm
2
(9 in.
2
)
= 72.0
0.90
SYSTEM EFFICIENCY
0.88
GOAL
co
- -
I
-2..
'7
0.86
I-

(J
0.84
z
w
(J
u.
u.
0.82 w
:!:
w
I-
en
>
0.80
en

0.76
fps .... 1800 1900 2000 2100 2200 2300 2400 2500 2600
m/sec .... 549 579 610 640 671 701 732 762 792
U
t3
, INDUCER TIP SPEED
Figure 28. Peak System Efficiency for N
=
110,000 rpm.
51
U)
I
0
l-

{::"
> ..
(.)
:2:
w
(.)
u..
u..
w
==
w
I-
CI)
>
CI)
50
N = 95,000 RPM
NS = 12
REACTION EFFECTS
NO DESWIRL VANE
ROTOR EXIT AREA 58.1 cm
2
(9 in.
2
)
a3 = 72.0
0.90
EFFICIENCY
GOAL
0.88
--- ----
0.86
0.84
0.82
0.80
0.78
0.76
fps .... 1800
1900 2000 2100 2200 2300 2400 2500 2600
m/sec ... 549 579 610 640 671 701 732 762 792
U
t3
, INDUCER TIP SPEED
Figure 27. Peak System Efficiency for N = 95,000 rpm.
CD
I
0
~
~
> ~
u
z
w
U
u.
u.
w
:is
w
I-
CI)
>
CI)
N = 80,000 RPM
NB = 12
REACTION EFFECTS
NO DESWIRL VANE
ROTOR EXIT AREA ~ 58.1 cm
2
(9 in.
2
)
x
3
= 72.0
0.90 r - - - r - - - . . . . . , . . - - - , . . - - - , . . - - - - - r - - - - - r - - - - - , - - - - - , ~ - __
SYSTEM EFFICIENCY
GOAL
0.86
0.84
0.82
0.80
0.78 Ir-""T"'t----:lr--t----+---+---+---+----t----t-----f
0.76 t--+t----+---+---+----t----t---1-----.,f-------4
fps ... 1800 1900
2000 2100 2200 2300 2400 2500 2600
m/sec ... 549 579 610 640 671 701 732 762 792
U
t3
, INDUCER TIP SPEED
Figure 26. Peak System Efficiency for N = 80,000 rpm.
49
characteristic shown is at the rotor exit mean-line, signif-
icantly higher relative critical velocity ratios would occur in
the rotor exit tip region and shock losses should be evaluated.
To define a tip speed envelope that will achieve the cycle
performance goals, the peak system performance from Figures 14
through 22 was first plotted as a function of tip speed and
range of inducer blade angles, as shown in Figures 26 through 29.
The results showed the magnitude of tip speed required for each
rotational speed and also the effect of rotor inlet blade angle
on required tip speed. Figure 29 shows these characteristics
with deswirl vanes for 95,000 rpm. These results were
cross-plotted in Figure 30 in terms of tip speed versus shaft
speed, and a tip speed envelope was obtained. The envelope
shows that without rotor exit blockage effects, tip speed
requirements range between 579 and 716 m/sec (1900 and 2350 fps)
depending on inlet blade angle, shaft speed and deswirl vane
specification.
Characteristic rotor flow paths for the three shaft speeds
are presented in Figure 31. At 80,000 rpm, the basic flow path
resulted in a relatively small blade, large disk configuration.
Conversely, at 110,000 rpm the flow path resulted in a rela-
tively large blade, small disk configuration. Although the
radius ratio at 80,000 rpm could be decreased to match the exit
hub radius at 110,000 rpm, this was not shown for two reasons.
First, the parametric study showed increasing performance as
radius ratio increased. Second, the inducer blade height
increased by a factor of approximately 2.0, requiring a signif-
icant increase in blade thickness to maintain equivalent inducer
blade stress levels. In addition, the increased blade height
was expected to result in undesirable blade vibration charac-
teristics.
48
Additional observations from Figure 31 are:
o In add i tion to the smaller blade he igh t d imens ions
associated with 80,000 rpm, the large disk also would
increase rotor weight and inertia, which penalizes
vehicle response time; therefore, driveability may
suffer. However, rotor ex it hub blockage would be
minimized, and the number of blades could be
increased.
o At 110,000 rpm, rotor weight and inertia would be min-
imized, and rotor blade height would be maximized, but
the reduced rotor exit hub radius would significantly
increase exit blockage or severely limit the number of
blades. Mechanically, the relatively tall exducer
blades could result in unacceptable blade vibration
character is tics.
1.4

53:
W -' 12
.
1-0:1-
-0
X W 0: 1.01
w
0:-1-
aI--
I- 0 0.81
0-'0
o:
w..J
o::w
>
...
u
w
1->
=> ,
..JO
0-
(1)1-
m
<CO::
:z>
<c!::
Wo

1--'
0.9
0.8'
0.7
0.61
N = 80,000 RPM

NB = 12
_W
x>
w..J

0::
au
1--
a!::
0::0::
U
-'
0.3'
+101
0:: 0,

----NO DESWIRL VANE, WITH REACTION
----REACTION, DESWIRL VANE

:z 0:: _10,1----jfJ---+--{:J---f---+--t-/--+----t---hr--1
t!'
Ww

1-' -201
-
x '

o::t!'
02
b
0::

0.7 0.8 0.9 1.0 1.1 1.2 1.3 1.4 1.5 1.6 1.7
Figure 25. Turbine Vector Diagram and System Parameters
N = 80,000 rpm, fiB = 0 (Cont).
47
46
(!)
I-
eI)
a:

z
o
I-

w
a:
w
(!)

..J CD
I
.-- o:t
0-

a:Q.
w<1

u.o
u. ..J
C W

w
J: a:
X Q.
W
...J

I-
o
'I 'I
o I-
'I
..J

o w
I- (.)
w u.
(!J LL
W
t;
0.9
N = 80,000 RPM
0.8 i3S = 00
0.7
NS = 12
0.5
0.4
0.3
0.25
0.8 0.9 1.0 1.1 1.2 1.3 1.4 1.5 1.6 1.7
Figure 24. Turbine Vector Diagram and System Parameters
N = 80,000 rpm, = 0.
>"
(.)
2!
w
(.) 0.S4
u.
u.
W
..J

o c
.... - .....
O'
........
J (:" 0.S2

0.S1
t;

'>"
(.)
2!
w
(.) 0.S9
u.
u.
W
..J
,O.SS
o 0
.... -
.....
o'
J 0.S7

:E 0.S6
w
t;
X3 = 1.25
!
A4 = 5S.1 cm
2
(9.0 in.
2
) t
A4 = 51.6 cm
2
(S.O in.
2
)_I, \
J
A4 = 45.2
1
cm
2
(7.0 \
.I)
.0
6-: \
"'t .A
J,F
RADIUS

Cr 1
RATIO USED IN
J
PARAMETIC STUDY
.1
0.65 0.70 0.75 O.SO
RTiRT3 - EXDUCER TIP-TO-INDUCER TIP RADIUS RATIO
(A) U
t3
= 610 m/sec (2000 fps)
x; = 1'.0
MAXIMUM RADIUS-
--I
RATIO USED IN
I STUDY

'\.
A
9- I .r
-6 LIl
I
-r;r
I 1 -,
?-
>--J1.4 - 58.1 cm
2
(9.0 in.
2
,JI-/
A4 = 51.6 cm
2
(S.O

f--A4 = 45.2 cm
2
(7.0 in.2) I
J I I I I I I
>
en 0.65 0.70 0.75 O.SO
RTiRT3 - EXDUCER TIP-TO-INDUCER TIP RADIUS RATIO
(B) U
t3
= 701 m/sec (2300 fps)
Figure 23. Effects of Rotor Exit Area and Radius Ratio on Turbine
System Efficiency, N = 95,000 rpm, #B = 0, No Deswirl
Vanes.
45
RADIAL
ISOPLETH INTERVAL = 0.127 em (50 mils) ISOPLETH INTERVAL = 3.447 kN/em
2
(5 ksi)
SCALE FACTOR = 1.50
RPM = 110,000
(!) MAXIMUM 6 MINIMUM NB = 12
74
o
o
CD
lJ)
o
o

o
o

~ - - - - - - - - - - - - - - - - 4 - - - - - - - - - - - ~ ~ ~

...-
>-
u
z
Wo
o ~ _________________ ~ ______ ____
c' r
W ~
~
lJ...
o
o

CD
-
o
o
CD
g
o
0.00 25.00 50.00 75.00
SPEEO(PER CENT)
NASA ART TURBINE CASE 3.
VIBRATION FREQUENCIES
CAMPBELL DIAGRAM
100i. RPM= 11 0000
50 i. RPM= 55000
100.00
Figure 48. Campbell Diagram for IIO,OOO-rpm Case.
17E
DE
E
E
E
E
E
E
E
E
E
125.00
o"J
U1
34.45- 50
27.56- 40
20.67- 30
13.78- 20
"

...............
~ -.
80 85 90 95 100 105 110
PHYSICAL SPEED (RPM X 10-
3
)
Figure 49. Effect of Speed on Maximum Principal Disk Stress.
-....J
0'1
kN/cm
2
ksi
N = 95,000 RPM
27.56 - 40 .... - ..... --------T---------,----------r--....

24.12 - 35 ....

en
C

:J 20.67 - 30
:E

17.23 - 25 ..... -..&...--------"---------'----------..... - ....
10 12 14 16
NUMBER OF BLADES
Figure 50. Maximum Disk Stress Versus Number of Blades.
Another parameter to be varied was tip speed. This vari-
ation was accomplished by keeping the flowpath constant for
95,000 rpm, with an e x ~ t area of 45.2 cm
2
(7 in
2
.) and varying
the lnducer tip radius. Figures 51 and 52 show the tangential
thicknesses and maximum princlpal stresses for tip speeds of 634
m/sec (2080 fps), and 70 m/sec (2300 fps), respectively. The
maximum dlSk stresses for the 701 m/sec (2300 fps) are somewhat
higher than for the 634 m/sec (2080 fps) but are wi thin the
24.132 kN/cm
2
(35 kSl) stress limit. Figure 53 summarizes the
maximum disk stresses as tip speed is varied.
3.10 Parametric Study of Detail Aerodynamic Analysis
Since mechanical analysis indicated that disk stress and
blade thickness would increase with the higher rotor exit area,
and the rotor blade vibration characteristics were more favor-
able at the lower 45.2 cm
2
(7 in. 2) rotor exit area, typical
vector diagrams with and without deswirl vanes were selected for
45.2 cm
2
(7 in.
2
) cases. Two-dimensional vector diagrams were
calculated as a function of rotor exit blockage.
For the case without deswirl vanes, the rotor exit average
relative critical Mach number (W/Acr) varies between 0.70 and
0.80, and with a deswirl vane, the critical Mach number varies
between 1.0 and 1.1. The general results in terms of a rotor
relati ve total pressure loss ratio and hub blockage are pre-
sented in Figure 54. When the effects of rotor blockage and
blade number are combined, a minimum loss wi th and without
deswirl vanes is defined as in Figure 55. The high penalty
associated with the deswirl vane case is attributed to the
higher rotor exit Mach number and exit angle due to the
increased rotor exit swirl.
To illustrate the combined effects of stage reaction and
rotor exit blockage on system efficiency, the 95,000 rpm base
solutions from Figure 17 were utilized. The results are
presented in Figure 56. Since the parametr ic study indicated
system efficiency increased with increasing rotor exit area, the
overall effect of exit area, with reaction and blockage also was
included. The radius ratio was fixed at 0.80. The charac-
teristics derived from the basic parametrlc study are shown at
the bottom of Figure 56 as a function of inducer tip speed and
exit area. The change in efficiency as a function of stage
reaction alone is next shown (derived from Figure 13). As indi-
cated, the benefits of higher exit area are already partially
offset due to lower reaction. In addition, the higher incidence
loss assoclated wlth low tip speeds is further compounded by
relatively high losses due to reduced reaction.
The rotor exducer blade thickness and blockage correlation
derived from the 2-D mechanical analyses, together with the cal-
culated 2-D rotor exit vector diagrams then were used to obtain
a efficiency penalty for rotor exit blockage. Since the rotor
77
SCALE FACTOR = 1.50
RADIAL
2
(5 ksi)
RPM = 95,000
NB = 12
RADIAL
ISOPLETH INTERVAL = 0.127 em (50 mils) ISOPLETH INTERVAL = 3.447 kN/em
2
(5 ksi)
RPM = 95,000
SCALE FACTOR = 1.50
Figure 52. Tangential Thicknesses and Maximum Principal
Stresses.
00
o
27.56- 40
~
w
~ 24.12-35
(I)
~
(I)
c
:E
i 20.67-30
~
:E
17.23-25
fps
m/sec
~
~
",-
2050
625
2100
640
2150
655
TIP SPEED
2200
671
2250
606
2300
701
Figure 53. Maximum Disk Stress Versus Tip Speed, A4 = 45.2 cm
2
(7 in.2),
N = 95,000 rpm, NB = 12.
-
2350
716
o

c::
CI)
CI)
REFERENCE TURBINE FROM
SPECIFIC SPEED CURVE
1.0
o
-' 0.99 1-----;----1----+----4..,.....
w
c::
:l
CI)
CI)
w
0.98
-'

w
>
I-

-'
w
c::
a:
o
l-
e
c::
% BLOCKAGE = TANGENTIAL THICKNESS X NB
2
1T
RHUB
0.97 1-----;----1----+----+-----+----.-1----1
0.96 t-----!r-----t----+----t----+----f---t
MAXIMUM ALLOWABLE -;-----s
HUB BLOCKAGE FOR
PARAMETRIC STUDY
0.95 t----!f------+----+-----r----+----t---f
o 10 20 30 40 50
ROTOR EXIT HUB BLOCKAGE, PERCENT
Figure 54. Effect of Rotor Exit Hub Blockage on Rotor
Loss.
81
82
60
w
~ 40
<t
~
U
o
..J
III 30
~
~
~
~ \ ~
~ ;,
~ ~ L..---
V
v ~
I--"'"
~
~
.fl
~
~
~
~
..............
.,
50
20
-3.0
V
I
"
\
~
Q
~ ~
-2.0
-1.0
~
V
v
~ N O O ~
to-.
0.0
10 11 12 13 14 15 16
NB - NUMBER OF BLADES
Figure 55. Effect of Rotor Exit Hub Blockage and Blade
Number on System Efficiency.
f/FINAL
N = 95,000
PB = 0
0
NB = 12
RT iRT 3 = 0.80
E:CIEN: GOAL
2 2 - - 4,---1
Z A4= 45.2 em . (7 in. ),
A4= 51.6 em
2
(8 in.
2
)

58.1 em
2
(9 in.
2
)
0.80
+0.76 L---L. __ -'---__ .l..-_----lL-_---L __ --L __ ...L-__ ..l.-_..-l
+0.005 df/BLOCKAGE SYSTEM --r--""""'--
0.00
(.J
o
-0.005

<J -0.01 0
-0.020
z
0- 0.015 1-:-, 2
I- A4 = 45.2 em (7 if!. )
= 51.6 em
2
(8 in.
2
)
W -0.005 A4 = 58.1 em
2
(9 in.
2
)
<J
0.000
0.90 SYSTEM EFFICIENCY, f/T-TIO-6
o
"'i
:E 0.86
w .
.... >
cnu
ffi A4 = 45.2 em
2
(7 in.
2
)
(J 51.6 em2 (8 in.
2
)
tt A4 = 58.1 em
2
(9 in.
2
)
w
0.78 L--'-__ ..J-__ .l..-_---I __ --L. __ --L. __ ..J.... __ ...L..-_---I
fps -. 1800 1900 2000 2100 2200 2300 2400 2500 2600
em/sec... 549 579 610 640 671 701 732 762 792
INDUCER TIP SPEED
Figure 56. Effects of Reaction, Rotor Exit Area, and
Blockage on System Efficiency.
03
radius ratio was set at 0.80, the higher tip speeds result in a
blockage penalty that is less than the reference turbine (i.e.,
positive AYJ blockage). The combined effects of reaction and
blockage are indicated by YJ final. The advantages of the higher
rotor exit area designs are reduced significantly and are lim-
ited to higher tip speeds due to the maximum allowable blockage
of 50 percent.
up to this point, the aerodynamic and mechanical analyses
indicate the following trends:
o The effects of reaction, exit area and blockage
increased the required tip speed of all solutions by
37 to 76 m/sec (120 to 250 fps)
o The effect of rotor exit hub blockage increased the
tip speed level of the 95, OOO-rpm solutions to the
level required at 80,000 rpm. Since the disk stress,
weight, and inertia are higher at 80,000 rpm, these
solutions were eliminated
o Comparison between the solutions at 95,000 and 110,000
rpm show that disk stress is slightly lower at 110,000
rpm, but the rotor exit hub blockage is significantly
higher. To reduce the thickness requirement would
require a reduction in rotor ex it annular area to
approximately 36.1 cm
2
(5.6 in.2), which decreases
stage efficiency and increases the required tip speed.
For this reason, the optimum solutions would be
between 90,000 and 100,000 rpm.
The tip speed characteristics for 95,000 rpm are presented
in Figure 57 as a function of imposed effects of reaction,
blockage, and exit annular area. The relative merit between
radial and nonradial blading was examined next wi th a three-
dimensional stress analysis. The rotor blade angle distr ibu-
tions for a 10- and 20-degree inlet angle were estimated from
previous nonradial rotor optimization.
84
(XI
U1
---- NO DESWIRL VANE
---WITH DESWIRL VANE
m/sec fps
732 2400
701 2300
c
w
W
Q.
671 2200 (I) 13B = 0
Q.
i=
a:
13B = 10
w
640 2100
(J
:J
C
13B = 20
Z
a:
610 2000
0
13B = O.t--;B = 10
--
I-
0
a:
13B = 20
~ - - - -
579 1900
BASE
549 1800
BASE:
1983 TECHNOLOGY
N = 95,000 RPM
NB = 12
STATOR EXIT ANGLE = 72
13B = 0
13B = 0
-'
13B = 10
13B = 10//
13B = 20 // 13B = 0 /
f3B = 2U
~ / //
13B = 10
~ /
~ , , , ~ /
13B = 0 V' ~ /
13B - 20
~ ~ / .J'
13B = 10
1// \) //
--13B = 0._
~ ~ /
------
13B = 10
~ ~ ;/
13B = 20
/'
~ - - - - -
1/
13B = 20
-----
COMBINED
COMBINED
REACTION
RADIUS RATIO
BLOCKAGE AND
AND EXIT-AREA BLADE NUMBER
PARAMETER EFFECTS
Figure 57. Effect of Reaction, Blockage, and Exit Annular Area on Tip Speed
Characteristics for Design System Efficiency of 0.87.
4.0 DETAIL ROTOR AERODYNAMIC AND MECHANICAL DESIGN
The three rotor configurations (without deswirl vanes) cor-
responding to inlet blade angles of 0, 10, and 20 degrees were
selected for detail aerodynamic and mechanical analysis. The
tip speeds for these designs were based on the data presented in
Figure 57 and represent the predicted values required to achieve
the cycle system efficiency goal of 0.87 with all known losses
and mechanical considerations accounted for. The design param-
eters for the detail design configurations are presented in
Table IV. Rotor flow path and blade geometries are presented in
Appendix C.
4.1 Aerodynamic Detail Design
The rotor inlet and exit flow path dimensions (derived from
Figure 57) were established from the optimized one-dimensional
vector diagrams. The rotor solidity study, described in Section
3.8, indicated that an axial length of 4.06 cm (1.60 in.) should
result in satisfactory rotor blade loadings for the radial blade
case. For the nonradial rotors, previous designs have demon-
strated that additional axial length is required, due to the
increased rotor turning.
The detail rotor design was initiated wi th preliminary
estimates of hub and shroud contours, blade thicknesses, and
angle distributions. These quantities were input to a radial
turbine geometry program along specified station lines (quasi-
orthogonals) for the internal flow solution. The resultant data
matr ix was curve-f i t to def ine the total blade geometry. The
computer program is a modification of the current centrifugal
impeller geometry program that allowed an arbitrary blade to be
defined.
To perform the internal flow analysis, the level and dis-
tribution of losses in the rotor must be specified in addition
to the rotor blade geometry and thermodynamic conditions. The
magnitude of the total stage loss is determined from the effi-
ciency analysis described in paragraph 3.4. The stator loss is
estimated from stator reaction rig tests performed at
AiResearch; the remaining loss is assigned to the rotor.
The computer program for the rotor internal flow analysis
solves the radial-equilibrium equation by satisfying the conti-
nuity, momentum, and energy equations in the meridional plane in
a manner similar to references 21 and 22. Blade surface veloci-
ties are computed from the local rate of change in moment of
momentum, the condition of zero absolute vorticity, and linear
velocities between suction and pressure surfaces. Stanitz
(ref. 23) has shown that this method produces satisfactory
results compared with relaxation solutions of the potential flow
equation. Numerous iterations between the geometry program and
internal flow analysis program are required to achieve satisfac-
tory blade loading for each thickness distribution examined.
86
TABLE IV. SELECTED CONFIGURATIONS FOR DETAIL AERODYNAMIC
AND MECHANICAL DESIGN
Configurations
Design Parameters Units f3. =
B
0 !1 = 10
B
!1 =
B
20
Rotational Speed N, rpm 95,000 95,000 95,000
Inducer Tip speed'.U
t
' cm/sec (fps) 701 686 671
3 (2300) (2250) (2200)
Stage Work Coefficient, 1.033 1.091 1.142
A
STG
Rotor Inlet Work 1.000 1.030 1.085
Coefficient,
1..3
Specific Speed, Ns rpm 0.537 0.537 0.537
(f t
3
/
4
/sec
1
/
2
)
(69.20) (69.20) (69.20)
Stage Reaction, R
STG
0.541 0.545 0.511
Reynold's Number, 1.028 1.050 1.074
Re X 10-
5
Exducer Tip-to-Inducer 0.632 0.646 0.661
Tip Radius Ratio,
R
t
/Tt
5 4
Rotor Exit Annular cm
2
(in.
2
) 45.2 45.2 45.2
Area,A
4
(7) (7) (7)
Rotor Exit Absolute, degrees -0.76 -13.40 -11.50
Swirl, a 4
Number of Rotor Blades, 12.0 12.0 12.0
NB
Turbine Inlet Scroll 0.0057 0.0056 0.0056
Loss, .!lP/P
Rotor Exit Diffuser 0.0303 0.0322 0.0311
Loss, .!lP/P
Stage Tota1-to-Tota1 0.889 0.892 0.893
Efficiency, ~ T - T
1-4
System Tota1-to-Tota1 0.870 0.872 0.873
Efficiency, ~ T - T 0-6
87
The criteria for satisfactory blade loadings are monotoni-
cally increasing velocities and minimum blade surface diffu-
sions. For highly loaded rotors (that normally imply increased
loading in the rotor inducer region), these criteria are not
satisfied. In addition, inlet flow deviations based on
Stanitz's criteria are not expected to apply; therefore, the
loadings in the inducer region become ill-defined. Under these
condi tions, more rigorous stream-function methods for evalu-
ating the blade surface velocities are not warranted. The rotor
internal flow analysis is further compounded by rotor exit flow
deviations. Deviations are expected even with isentropic flows
due to the rotor relative vorticity. Approximate methods for
predicting these deviations were proposed by Mizumachi (ref. 24)
in 1971, then Khalil (ref. 25) der i ved a rigorous method in
1977. However, these deviations have been modified signifi-
cantly by the transport of secondary flows from the inducer to
the exducer tip region.
The final rotor flow path configurations are presented in
Figure 58. As indicated, the hub and shroud contours were main-
tained for all three rotor inlet blade angle solutions. A con-
stant flow path configuration allowed a more meaningful mechani-
cal comparison, and previous nonradial rotor designs have pre-
sented no particular advantage from modifying the rotor meridi-
onal flow path under these conditions. Conversely, rotor axial
length was fairly sensitive (to achieve satisfactory loading)
when a rotor inlet blade angle was introduced.
The change shown in Figure 58 is based on iterating axial
length until approximately equal rotor inducer loadings for all
three blade angles were achieved. An alternate approach would
be to evaluate a rotor loading parameter ~ l o n g the hub, mean on
shroud lines and iterate axial length until equal loading param-
eters were achieved. The final rotor blade angle distributions
for inlet blade angles of 0, 10, and 20 degrees are presented in
Figure 59 as a function of percent meridional distance. Also
shown are the locations at inlet and exit where flow deviations
from the blade were assumed.
The rotor inlet deviations are apparently due to rotor
relative vortici ty effects, and the location where the flow
starts to follow the blade was based on Stanitz (ref. 1) corre-
lation and was applied to radial turbines by Katsanis (ref. 22).
The deviations at the rotor exit are even more complex, since
the rotor relative vorticity effects are compounded by large
secondary flow migration to the rotor exi t tip region. The
approach currently utilized at AiResearch is to assume the rotor
exit flow is turned to the rotor mid-throat angle, and the addi-
tional uncovered downstream turning is the total deviation at
each radial location. The net effect with this procedure is to
assume deviation increases from approximately zero at the hub to
88
en
:>
c
<C
c.r::

10.1 6
8.8 9
7.6 2
6.3 5
5.0 8
2.54
2.03

4.0
- --------
-r- j' ,1
-------
- "
3.5
55-_ 5
\
,... _ en 1IIIt,...
en!:!
;Ln c:n
M RPM95oo0
!:! -
A 4 45 15 cm
2
(70 In 2,
U
t3
701 m/sec (2300 FPSI. f3 = 00
1 I
- - - - - U
t3
686 m/sec (2250 FPSI. f3 = 10 0
I
- - - - - U
t3
671 m/sec (2200 FPSI. f3 200
3.0
------

------
- E"
-----;
E 5
--1-
" E ....
.... ---_.-1
:!E
f--.-- -
E 5
!:i NN
- ;
\
.... -
2.5
E 5 g:::
I'N
M .... -
;t8 ..

2.0
I \ '-...

--
T

I
" -
I

...

I
,
I
\
1.0
0.8
in .... 0.0
em -. 0.0

r-.
0.5
1.27
1.0
2.54
AXIAL LENGTH
-----
1.5
3.81
\ \
\'
\\
'-', .........
Figure 58. Final Rotor Flow Path Configurations.
!i 5

Ne
2.0
5.08
89
CI)
w
w
a:

w
c

z
0
i=
:J
al
a:
I-
CI)
c
W
...J

2

w
C

...J
al
a:
0
I-
0
a:
70
60
50
40
30
20
10
0
-10
-20
-30
o
90
SHROUD
,

;;;-
ROTOR EXIT
DEVIATIONS
K
,
N
{//
47
.,
/;
/ , HUB
'/

//

jj
If
/

/

'/ ,;
/'/
/
ROTOR INLET ...........
-,/k/
/
DEVIATIONS /
,,/

&
/ ;,'//
..--'/
-,

V';
"
" ,8B = 0
0
-"
-"

.,..,.
-
1-:::--
,8B = -10
0
,;
lP'"
----- ---

-----
,8B = -20
0

10 20 30 40 50 60 70 80 90
PERCENT MERIDIONAL DISTANCE
Figure 59. Rotor Blade Angle Distribution Versus Percent
Meridional Distance.
100
a relatively high value at the tip and accounts for the devia-
tions expected in the tip region from the secondary flows. How-
ever, the parametric study results are based on precise
inducer-to-exducer work splits. Achieving the exact vector
diagrams desired would require either a significant increase in
rotor internal flow knowledge or experimental iterations of
rotor exit blade angles.
The rotor suction and pressure surface velocity distribu-
tions for the three inlet blade angle cases are presented in
Figure 60 as a function of meridional distance. With 12 rotor
blades and the adjustments in axial length, the loading through-
out the blades is similar and is expected to achieve equal per-
formance levels.
4.2 Detailed Mechanical Design
Ceramic materials possess large variations in strength
properties, due to brittle characteristics. Consequently, prob-
abilistic techniques must be utilized to accurately evaluate
ceramic components.
The probabilistic component evaluation technique employed
by AiResearch is based on the Weibull statistical strength model
(ref. 26). The working relationships permit the calculation to
be made from a combination of local stress, component size, and
Weibull parameters for both area- and volume-controlled frac-
ture.
CPF = 1 - e-
R
where R is defined as:
and At = Test-bar surface area
a
V
a
A
()ov
()oa
()uv
()ua
Ba
= Test-bar volume
= Elemental volumetric stress
= Elemental surface stress
= Volumetric zero failure stress
= Surface zero failure stress
= Volumetric characteristic stress
= Surface characteristic stress
= Surface Weibull slope
91
a:
1.0
w
In

::;)
0.8
2:
:J:
CJ
ct 06

w
> 0.4
j::
ct
-I
w
0.2
a:
-I
w
a: 00
2:

SHROUD STREAMLINE

",'
",.--....
.......... -

j....oo"" ....

.--

-

/
' .
-,,'
V
SUCTION SURFACe/

" .
..
PRESSURE SURFACE,
...... ;
-).
... '"
V

f3s .. 0
---
.....

t--....
... '"
---==
------ f3s .. 10
" ..
1--- - ---f3s "20
-0.2 -
a: 1.0
w
In

::;)
".8
2:
:J:
CJ
0.6
ct

w
0.4
>
j::
ct
-I
02
w
a:
-I
w
00
a:
z

-0.2
--...
MEAN STREAMLINE
,r

"",

-
--
if'"

--
-..-::
.. V
....
..... A
f7

V
....

f3s .. 0

------ f3B = 10
.... -
--"
---f3B - 20
a:
1.0
w
co

::J
08
2:
:J:
CJ
ct
06

w
>
04
j::
ct
-I
w
0.2
a:
-I
w
a: 0.0
2:

-0.2
,
1/
HUS STREAMLINE
........
..
..
V/
.
'\..

,
\

r'.

" ,-r?

1,.. ..

---

...
V

:,....-
1'"

L-_-;:-:::
.;,..".-'
f3s .. 0
t:t
.... ..;. ..
-----. f3B .. 10
'0: .... ... '"
","
_0- f3S" 20
...
o 10 20 30 40 50 60 70 80 90 100
PERCENT MERIDIONAL DISTANCE
Figure 60. Velocity Distributions.
92
Bv = Volumetric Weibull slope
da = Differential surface area
dv = Differential volume
This probabilistic computer code distinguishes between surface
and volume strength relationships and permi ts assessment of
overall component risk in a method compatible with finite
element techniques. All three principal stresses are considered
and the codes can assess effect of multiaxial stress states and
change in reliability due to proof testing (ref. 27).
As previously mentioned, cumulative probabilities of suc-
cess computations require 3-D finite element stress results.
The 3-D model for the optimized radial-blade configuration is
shown in Figure 61. Detailed stress analyses for a similar
radial inflow turbine rotor configuration in the AFTlOl program,
under DOE/NASA contract DEN3-l67, have shown that stresses and
cumulative probabilities of success are quite similar for both
room-temperature and maximum power steady-state operating con-
di tions. This AGT study also showed that minimum cumulative
probabilities of success are computed during transient opera-
tion. Unfortunately, transient thermal studies could not be
completed for the radial and non-radial blade configurations of
this study because of financial restrictions. Consequently, all
comparative stress analyses were carried out for room-
temperature operation. However, probabilities of success were
computed using anticipated 1983 material properties at elevated
temperature (characteristic strength of 62.053 KN/cm
2
(90.0 KSI)
and a Weibull modulus of l5.0}. The parametric study indicated
that the 95,000-rpm wheel, with an inlet tip speed of 701 m/sec
(2300 fps), was the preferred radial-blade configuration and is
summarized on Table V. When this configuration was first exam-
ined mechanically (Figure 52), disk stresses were marginally
high. To reduce disk stresses, the saddle height of the wheel
was reduced by removing disk material above the self-sustaining
radius of the disk. The resulting wheel stress field at room
temperature is illustrated in Figure 62. The peak bore stress
was reduced from 24.132 kN/cm
2
(35.0 ksi) to 22.339 kN/cm
2
(32.4 ksi). The resulting probability of survival for the rad-
ial wheel using anticipated 1983 ceramic mater ial properties
[character is tic strength of 62.053 kN/cm
2
(90.0 ks i) and
Weibull modulus of 15.0] is 0.9999. This cumulative probability
of survival (CPS) meets the objective of the program.
Once the baseline radial blade design was analyzed, the
nonradial inlet design feature was studied using the 3-D stress
and probabilistic evaluation computer code. Initially, an inlet
angle of 10 degrees was examined with the blade thickness dis-
tribution for the radial blade configuration. To achieve the
same efficiency as the radial blade design, the required tip
speed for this nonradial configuration was reduced to 686 m/sec
93
94
Figure 61. Typical Three-Dimensional Finite Element
Model.
TABLE V. OPTIMUM CONFIGURATION DERIVED FROM PARAMETRIC
STUDY AND 3-D MECHANICAL ANALYSIS
Parameter
Shaft speed
Inducer tip speed
Rotor blade number
Rotor exit swirl
Rotor exit critical velocity
Rotor inducer-to-exducer tip
radius ratio
ratio
Optimum
95,000
701 em/sec
(2300 fps)
o degrees
12 to 14
-10 degrees
0.449
0.632
95
RADIAL
SUCTION SIDE
- - - : } . - ~ O
32.4 KSI
SCALE FACTOR = 1.50
2
(5 ksj)
(!) MAXIMUM 6MINIMUM
32.4 KSI
RPM = 95,000
PRESSURE SIDE
Figure 62. Maximum Principal stresses.
(2250 fps). Principal stress results for this configuration
(Figure 63) indicate high bending stresses in the wheel inlet.
These produce a low probability of survival of 0.9394. Experi-
ence with nonradial blades from other research programs indi-
cated that these bending stresses could be reduced by incorpo-
rating rake to the blade design. Figures 64 and 65 illustrate
configurations without and with rake (-30 degrees), respec-
tively. The principal stresses for the rake case ( F i ~ u r e 66)
improve the cumulative probability of success to 0.9986.
Although this CPS value is significantly better than the no-rake
design, the goal of 0.9999 was not reached.
Blade thickness distribution was the next parameter to be
studied. Increased blade thickness was incorporated in the
inducer portion of the -30 degree rake case (Figure 67). The
resulting principal stresses, illustrated in Figure 68, produce
a cumulative probability of survival of 0.9993.
Based on the improvements in blade stresses and CPS with
the modified blade thickness distribution, a new distribution
was generated that increased the blade thickness in both the
inducer and exducer. The resulting pr incipal stresses (Fig-
ure 69) created a cumulative survival probability of 0.9998.
The disk stresses were increased slightly due to thicker blades.
The thickness distribution was modified once again by
thickening the blades in the high stress regions. The resulting
blade stresses were significantly reduced (Figure 70). Disk
stresses were again increased slightly. The survival rate for
this configuration, 0.9997, was lower than the previous blade
thickness design because of the higher disk stresses.
The 20-degree inlet configuration was analyzed with two
different blade thicknesses: the radial blade and the thickest
distribution for the 10-degree inlet. To achieve an efficiency
equivalent to the radial design, tip speed was reduced to 671
m/sec (2200 fps). As expected, the radial blade thickness dis-
tribution produced extremely high blade stresses (Figure 71) and
an unacceptable cumulative probability of survival of 0.0001.
with the thick-blade distribution, the blade stresses were con-
siderably reduced (Figure 72), and resulting cumulative prob-
abili ty of success was 0.4865. The blade stresses could be
reduced by a further increase in blade thickness, but these
increases would increase disk stresses. Increased thickness
would prevent the 20-degree inlet case from ever achieving a
survival rate as high as the radial blade design.
The peak blade and disk stresses and cumulative probabili-
ties of survival for the different wheel geometries are tabu-
lated on Table VI. In summary, the radial blade is the only wheel
design that meets the 0.9999 CPS program goal; nonradial config-
urations do not possess survival rates as high. Blade thick-
nesses had to be increased significantly to keep the blade
97
NON RADIAL, BETA = 10
RAKE = 0
86.1
SUCTION SIDE PRESSURE SIDE
31.0 KSI 31.0 KSI
2
(5 ksi) ISOPLETH INTERVAL = 3.447 kN/cm
2
(5 ksi)
(!) MAXIMUM ~ MINIMUM TIP SPEED = 686 m/sec (2250 fps)
Figure 63. Maximum Principal Stresses.
0
w-
II
<t
....
0
W
II
al
..J-
W
~
<t <2:
C
a::
<t
a:
:2:
0
:2:
-
."
.e-
o
Ln
N
N
-
II
C
w
w
c..
en
a..
....
o
...
OJ
~
ro
..c:::
Ul
OJ
ro
ro
r-I
r:Q
99
I-'
o
o
NONRADIAL, BETA = 10
RAKE = -30
Figure 65. Blade Shape, Original Thickness (Rake = -30).
I-'
o
I-'
SUCTION SIDE
NONRADIAL. BETA = 10
RAKE = -30
64.5 KSI
PRESSURE SIDE
D
31.0 KSI
2
(5 ksj)
31.0 KSI
2
(5 ksi)
SCALE FACTOR = 1.50 RPM = 95.000
(!) MAXIMUM ~ MINIMUM TIP SPEED = 686 m/sec (2250 fps)
Figure 66. Maximum Principal Stresses, Original Thickness.
0
0-
Il
<
I-
W
m
...i
<{
C
<{
a:
z
0
z
102
0
M
I
II
w
~
<
a:
g
~
E
to
ClO
co
II
C
W
W
Q.
en
Q.
j::
I-'
o
w
SUCTION SIDE
RAKE = -30
PRESSURE SIDE
2
(5 ksj) ISOPLETH INTERVAL = 3.447 kN/cm
2
(5 ksj)
(!) MAXIMUM 6 MINIMUM TIP SPEED = 686 m/sec (2250 fps)
Figure 68. Maximum Principal Stresses, Modified Thickness(l).
......
o
~
SUCTION SIDE
RAKE = -30
PRESSURE SIDE
2
2
(5 ksi)
(!) MAXIMUM 6 MINIMUM TIP SPEED = 686 m/sec (2250 fps)
Figure 69. Maximum Principal Stresses, Modified Thickness (2).
I--'
o
111
SUCTION SIDE
RAKE = -30
- ~
PRESSURE SIDE
2
2
(5 ksi)
Figure 70. Maximum Principal stresses, Modified Thickness (3).
......
o
0'1
RAKE = -30
SUCTION SIDE
31.1 KSI
PRESSURE SIDE
2
(5 ksi) TIP SPEED = 671 m/sec (2200 fps)
(!) MAXIMUM ~ M I N I M U M
Figure 71. Maximum Principal Stresses, Original Thickness.
I-'
o
~
SUCTION SIDE
RAKE = -30
PRESSURE SIDE
2
(5 ksj)
2
(5 ksj)
Figure 72. Maximum Principal Stresses, Modified Thickness (3).
......
o
<XI
TABLE VI. STRESS AND CUMULATIVE PROB\BILITY OF SUCCESS SUMMARY AT 95,000 RPM
Cumulative Probability
of Success
Configuration Rake Blade Inlet Max Blade Max Disk
Tip Speed, Stress, Stress, Contribution
m/sec kN/cm
2
kN/cm
2
Total
(fps) (ksi) (ksi) from Blades from Disk
Radial Blade 701 16.8 22.3 1.00000 0.99989 0.99989
(2300) (24.4) (32.4)
10-Degree Inlet 0 686 59.4 21.4 0.93950 0.99992 0.93942
Radial Blade Thicknesses (2250) (86.1) (31. 0)
-30 44.5 21.4 0.99862 0.99992 0.99855
(64.5) (31. 0) I
Modified Thickness 33.8 22.0 0.99940 0.99987 0.99928
1
Distribution U (49.0) (31. 9)
12 31.4 22.5 0.99996 0.99981 0.99976
(45.6) (32.7)
13 28.1 23.0 0.99996 0.99972 0.99970
(40.7) (33.3)
20-Degree Inlet 671 91.5 21.4 0.00006 0.99992 0.00006
Radial Blade Thickness (2200 ) (132.7) (31.1)
Modified Thickness 671 52.4 23.0 0.71518 0.99972 0.71498
Distribution t3 (2200) (76.0) (33.3)
stresses at acceptable levels. Increased blade load increased
disk stresses and caused disk survival rates to be lower than
those for the radial blade design. With sufficient blade thick-
ness and rake adjustments, moderate amounts of backsweep (10
degrees or less) can be tolerated with nearly the same cps as a
radial bladed wheel.
The 10-degree inlet design that possessed the highest cumu-
lative probability of survival (modified thickness No.2) and
the radial blade configuration were compared, using ceramic
properties more representative of present day materials [a char-
acteristic strength of 44.816 kN/cm
2
(65 ksi) and a Weibull mod-
ulus of 12]. With these lower values, the radial wheel again
possessed a better survival rate--0.9686 versus 0.9359 for the
nonradial design.
109
5.0 OFF-DESIGN PERFORMANCE EVALUATION
For a single-shaft configuration, turbine operating condi-
tions will vary significantly from maximum power to idle.
Therefore, the turbine design der i ved from the maximum power
optimization study may not provide minimum SFC and maximum
driveability over the entire duty cycle. Although the turbine
design point could be selected at any power setting and then
examined off-design, this approach is not necessary. Evaluating
the off-design characteristics at different "simulated" power
settings can be accomplished merely by selecting a range of vec-
tor diagrams at maximum power setting. For example, since tur-
bine flow, speed, and pressure ratio will be reduced at idle,
high positive swirl would be expected and, with the duct loss
characteristics used in the parametric study, would result in
decreased performance. Of course, turbine performance at idle
is also dependent on a number of other parameters, such as work
coefficient and reaction.
The point is that if the vector diagram at maximum power is
changed (for example, from -10.0 to -40.0 degrees exit swirl),
all off-design vector diagrams will shift, and new efficiency
characteristics will be defined. In addition, the system effi-
ciency envelope already provides alternate vector diagrams that
would maintain maximum efficiency at the design-point.
The effect of design-point vector diagram selection on
overall duty cycle performance will be illustrated based on the
results of the current NASA-DOE Advanced Gas Turbine Power Train
System Development Program (Contract No. DEN3-l67). Since the
cycle and selected design point parameters are very similar to
the 701 m/sec
2
(2300 fps) tip speed radial-bladed configura-
tions, the results are directly applicable to the present study.
Figure 73 shows three selected inlet work coefficients (and
the corresponding rotor exit swirl levels) on the system effi-
ciency envelope for 95,000 rpm and a tip speed of 701 m/sec
(2300 fps). The three cases are designated low (-9.8 degrees)
swirl, high (-27.0 degrees) swirl, and maximum (-37.0 degrees),
and the low swirl case corresponds to peak system efficiency at
maximum power.
Figure 74 shows the results of each vector diagram over the
entire duty cycle. At idle, the maximum swirl design increased
efficiency by approximately 6.0 points compared to the low-swirl
design. At maximum throttle and cruise, the performance of the
high-swirl and maximum-swirl designs was almost equivalent. At
maximum power, the maximum swirl efficiency was reduced by
approximately 2.5 points. Overall, the high-swirl design
appeared to offer the best compromise between reduced maximum
power and increased part-power characteristics. However, cycle
studies 'will be required for each case before a final selection
can be made. Not only were the off-design character istics
110
(I)
w
w
0:
C!l
w
c
N = 95,000 RPM
U
t3
= 701 m/sec (2300 fps)
A4 = 45.2 cm
2
(7 in.
2
)
i +20 t----------I--
-i
RT iRT 3 = 0.65
0:

w
....
::>
...J
o
en
co

2

w
:2:
....
X
w
0:
o
....
o
0:

o
I-j
....

(,)
z
w
U
U.
LL
W
...J

....
o

o

...J

....

:E
w
t;

t;

0.0
0
(I) ....

2>
....

-w
X>
W...J
0:
0(')
.... -

: HIGH SWIRL
:..-.---! DESIG
1
POINT
0.4 0 !::
a: a:
(,)
I
0.88r-----+----+---+--I----+-
1
-----+------1
I
MAX
DESIGN POINT I
I

0.84
0.82 .... ___ ...... ___ .-. ____ -'-___ ...&. ___ -'"
0.7 0.8 0.9 - 1.0 1.1 1.2
ROTOR lNLET WORK COEFFICIENT, >"3
Figure 73. Maximum Power Design-Point for Off-Design
Performance Evaluation.
III
112
en
w
w
ce
C!I
w
0
.. ..i
ce

en
....
X
w
ce
0
....
0
ce
eO
>
u
z
w
+70
+60
+50
+40
+30
+20
+10
0
-10
-20
-30
-40
088
086

ettl:
Ww
:!!:..J
alet
ce ....
082
Oce
ww
zen

:::!:U.
080
80
II)
I
o

078
076
10
I
t'
.. C
<;
\
l\ \
1\ \ \
\\
\
\
f
1- __
1--
11 )
...
\

:>.
1-/
SEA LEVEL 294C (85F) DAY
'J..>V/A cr MAX
",'
I I
( --
./
V
V
.. 0 VIA cr HIGH
I I
I
t-'!.,l.OW
V"

-
V/Acr LOW
--
..
","
./
"\ '[
I ..........

r- ....

Vr-

V
....

1"- ....

V
r---....
..
;...:
r--.....
....

V
r--
1--
"'r--.....
-
r-......

V
..
r-
['--...
r---
......
1"<>

1'-;:..1

1/3)
DESIGN POINT
I
I
_

-
- 1JHIGH
-
-
--
p"":
'7
-
I-- -
-1--t-
-
t<v
j
I
I
I
'OJ
I
I
I,)
-
I-
--
OPERATING
PERCENT
IDENT CONDITION P/PIT-SYS N..[O
0)
IDLE 134 585
0
MAXIMUM 168 550
THROTTLE
0
CRUISE 178 673
0
MAXIMUM 426 1000
POWER
20 30
P/PIT-SYS COMBINED TURBINE AND DIFFUSER
TOTAL PRESSURE RATIO
NOMINAL
kW
(hp)
224
(30)
1342
(180)
1566
(210)
10067
(1350)
40
060
050
...I
ct
u
E
040 ce
u
w
....
::J
...I
030
00
la-

.... ce
x>
w ....
020
ceu
00
.... ...1
OW
ce>
I
010
!!:.tJ
>
000
Figure 74. Off-Design Performance Characteristics as a Function
of Desiqn-Point Exit Swirl (from NASA-DOE Advanced
Gas Turbine Power-Train System Development program)
strongly dependent on the exhaust duct swirl loss charac-
teristics (and, therefore, had to be known before a vector dia-
gram was selected), but the turbine Reynolds number was reduced
significantly at part-power and had to be accounted for in the
off-design analysis.
The off-design analysis also confirms that the use of
deswirl vanes is not viable for the single-shaft engine applica-
tion due to the large rotor exit swirl excursion during the nor-
mal duty cycle operation. The parametric study has shown that
deswirl vanes can significantly reduce downstream duct losses;
therefore, deswirl vanes should not be ignored during the tur-
bine optimization process.
113
6.0 CONCLUSIONS
The results of this program show that exceptionally high
radial turbine tip speeds are feasible with the projected 1983
ceramic material properties (62.1 KN/cm
2
characteristic
strength and a Weibull modulus of 15). These results indicate a
high-work ceramic radial turbine with 0.076 cm (0.030 inch)
thick blade tips could achieve both the cycle performance goal
of 0.87 7J T-T and the rotor mechanical integr i ty goal of 0.9999
cumulative probability of success for a single-shaft automotive
engine application.
Three-dimensional rotor analyses showed that viable high
tip speed designs were derived from a design technique that com-
bined the characteristic of low disk volume and relatively deep
scallops wi th a blending of the blades and disk in the rotor
back-face region. Wi th radial blade element rotors, three-
dimensional mechanical analysis showed that the cumulative prob-
ability of success (CPS) goal of 0.9999 was achieved at a tip
speed of 701 m/sec (2300 fps).
With a rotor inlet blade angle of 10 degrees, the rotor tip
speed requirement was reduced to 686 mysec (2250 fps), and
detailed mechanical optimization of this design showed that the
CPS was reduced to 0.9998. At a tip speed of 671 m/sec
(2200 fps) and a rotor inlet blade angle of 20 degrees, addi-
tional blade thickness was required to achieve tolerable blade
stress levels. The increased loading from this additional blade
material caused higher disk stresses that adversely effected the
CPS value. Therefore, the conclusion is that although nonradial
rotor blading allowed a reduction in tip speed, this advantage
was offset by an increase in rotor blade stress levels. This
conclusion also would apply with reduced ceramic material prop-
erties; that is, the unique ceramic characteristic of low duc-
tility and density favors the lower global stress field asso-
ciated with the radial bladed designs. Therefore, if the pro-
jected ceramic properties are not achieved, turbine performance
potential would be reduced with both radial and nonradial blad-
ing, and fuel economy would suffer.
Additional observations, resulting from the advanced
radial turbine program, are:
114
o The turbine system optimization technique more accu-
rately identified peak system efficiency configura-
tions compared to an analysis approach involving indi-
vidual stage and exhaust duct optimization.
o The estimated effects of reaction and rotor exit bloc-
kage were relatively important for a large number of
solutions examined. Additional work is desirable to
more accurately refine/define performance correla-
tions for these effects.
o For engine operation requiring only relatively small
variations from turbine design-point operating condi-
tions, the analysis showed that significant improve-
ments in performance were provided with the applica-
tion of deswirl vanes. However, for a single-shaft
automotive engine, the large excursion in rotor exit
swirl eliminated consideration of deswirl vanes for
this application.
o The off-design analysis showed that the integrated
duty cycle performance was highly dependent on the
selected design-point vector diagram. This implies
that ei ther the effect of rotor exit swirl on duct
performance must be accurately known in advance of the
study analysis, or that several design-point vector
diagram rotor configurations must be evaluated exper-
imentally to arrive at the optimum duty cycle config-
uration.
115
APPENDIX A
ABBREVIATIONS AND SYMBOLS
117
118
APPENDIX A
ABBREVIATIONS AND SYMBOLS
a - Scroll inlet diameter, cm (in.)
b - Stator exit and rotor inlet passage width, cm (in.)
b / ~ R - Stator aspect ratio
c - In{r
o
+ a)/r
o
g - Force-mass conversion constant, 32.174 lbm-ft/
lbf-sec
2
q - Dynamic pressure, kN/m2 (lb/in.
2
)
r - Scroll inside radius, cm (in.)
o
w - Stage mass flow, kg/sec (lb/sec)
wJ8/o - Stage inlet corrected flow, kg/sec (lb/sec)
A - Annular area, cm
2
(in.
2
)
J - Mechanical equivalent of heat, 778.0 ft-lbf/Btu
L - Rotor axial length, diffuser length, cm (in.)
M - Meridional length, cm (in.)/Rotor exit mean, cm
(in. )
N - Rotational or shaft speed, rpm
R - Radius, cm (in.)
U - Blade speed, m/sec (ft/sec)
V - Absolute velocity, m/sec (ft/sec)
W - Relative velocity, m/sec (ft/sec)
Z - Axial dimension, cm (in.)
ABBREVIATIONS AND SYMBOLS (Contd)
Acr - Critical velocity ( ~ l 9RgTT)1/2, m/sec (ft/sec)
AR - Deswirl vane aspect ratio
Ca - Axial rotor shroud clearance, cm (in.)
Cb - Rotor backface clearance, cm (in.)
Cr - Radial rotor shroud clearance, cm (in.)
DH - Hydraulic diameter, m (ft)
Max - Maximum
MC - Midchannel
Min - Minimum loss coefficient
MN - Absolute flow Mach number
M - Relative flow Mach number
NREL
N/}B - Corrected speed, rpm
P
NS - Specific speed
N
V
- Number of stator vanes
R
T
-
DE
- Total-to-diffuser exit static pressure ratio
P
R
- Pressure ratio
PS - Pressure surface
P
s
- Static pressure, kN/m2 (lb/in.
2
)
P
STD
- Standard atmospheric pressure, 1.01325 kN/m
2
(14.696 lb/in.
2
)
P
T
- Total pressure, kN/m2 (lb/in.
2
)
119
120
RD - Diffuser recovery ( PS6 - P
S4
)1Q4
Re - Reynolds number for turbine stage (W/R
T
Re - Reynolds number for turbine scroll
Rg - Gas constant, J/kg K (ft-1b/1b-OR)
R
STG
- Stage reaction
S.F. - Slip factor for radial blades
SS - Suction surface
TSTD - Standard atmospheric temperature, 288.2 K (518.7R)
TPLP - Total pressure loss parameter
T - Absolute total temperature, K (OR)
T
U
t
- Blade tip speed, m/sec (ft/sec)
V
R
- Absolute radial velocity, m/sec (ft/sec)
Vu - Absolute tangential velocity, m/sec (ft/sec)
Vx - Absolute axial velocity, m/sec (ft/sec)
W - Relative velocity, m/sec (ft/sec)
W/Acr - Relative critical velocity ratio
a - Rotor exit swirl angle, absolute flow angle,
degrees (measured from axial direction)
~ - Rotor blade angle, degrees
~ B - Rotor inlet blade angle, degrees
Y - Ratio of specific heats
o - Ratio of turbine inlet total pressure to standard
atmospheric pressure
Aa - Total stator turning, degrees
Ah - Static enthalpy change across rotor, kJ/kg (Btu/lb)
AHT - Specific stage work, kJ/kg (Btu/lb) (based on total
conditions)
AHT/e - Specific stage corrected work, kJ/kg (Btu/lb)
(based on total conditions)
AP/P - Total pressure loss
AR - Stator radial chord, rotor exit blade height,
cm (in.)
- Change in efficiency
- Efficiency
e- Ratio of turbine inlet total temperature to
standard atmospheric temperature
eeq - Equivalent cone angle defined in Appendix B
A - Work coefficient (Vu/U)
A - Friction resistance coefficient for scroll loss
STG - Stage work coefficient
A3 . - Ideal rotor inlet work coefficient
,1
A - Actual rotor inlet work coefficient from vector
3,act diagram
A4,M - Rotor exit mean work coefficient
v - Kinematic viscosity, m
2
/sec (ft
2
/sec)
M - Absolute viscosity, N sec/m
2
(lb-sec/ft
2
)
W - Loss coefficient (PT. - P
T
\/q
ln out)/
121
SUBSCRIPTS
base - Base turbine total efficiency from specific speed
curve
blockage - Rotor exit blockage
D - Diffusion efficiency
i-Turbine stage total efficiency with incidence loss
final - Total system efficiency with blockage and reaction
effects
ISEN - Isentropic
loss - Rotor relative total pressure loss
max - Maximum
min - Minimum
reaction - Stage reaction
T-DE - Total to diffuser exit static
T - Tip
T-sys - Inlet total to exhaust diffuser exit total
T-T - Stage or system total-to-total efficiency
Turbine System Station Nomenclature
o - Scroll inlet (combustor discharge)
1 - Stator inlet
2 - Stator exit
3 - Rotor inlet
4 - Rotor exit (deswirl vane inlet)
5 - Deswirl vane exit (exhaust diffuser inlet)
6 - Exhaust diffuser exit
122
APPENDIX B
DESWIRL VANE LOSS MODEL
123
APPENDIX B
DESWIRL VANE LOSS MODEL
For highly loaded radial turbine designs, Section 3.7
showed that system performance is significantly increased with
the application of rotor exit deswirl vanes. The concept is to
relieve the rotor inducer loading by increasing rotor exit work
coeff icient, which in turn increases rotor exit velocity and
swirl levels. The deswirl vane is then utilized to reestablish
axial flow and reduce velocity by the difference between the
absolute and axial components.
A comprehensive evaluation of deswirl vane losses was con-
ducted by Mitchell and Soileau (ref. 3) between 1974 and 1977
under Air Force sponsorship. From a preliminary design stand-
point, the major results of this program are:
o Non-ser ies - Airfoils demonstrated a reduction in
total pressure, higher diffusion eff iciencies, and
reduced tendencies for flow separation relative to
series airfoils.
o Conical angle - An equivalent conical angle correla-
tion (8
eg
) was extended for diffusing cascades into a
range of cone angles applicable to advanced engine
turbine exit guide vanes. The correlation is used to
define optimum diffusion efficiencies and was suc-
cessfully demonstrated in 3-D annular cascade test-
ing.
The correlation derived by Mitchell and Soileau (from dif-
fusing cascade data) is presented in Figure 75. The correlation
relates equivalent cone angle (the rate of diffusion) with dif-
fusion efficiency ( ~ D ) as a function of equivalent area ratio
(amount of diffusion).
The equivalent cone angle is defined with the nomenclature
of this report as:
l[ (AR 1/2( 1/2 )1/2J
8
eq
= tan- 0.564 or) cos as - cos a
4
With parallel end-walls, the equivalent area ratio is:
A /A - 1
5 4 - cos a
4
125
......
'"
0\
I-
Z
w
U
a:
w
0..
6

>
u
z
w
u
100 i ,
_/ MAXIMUM 110
I I / j '(, I=-- I EjUIVALr AREA RATIO = 1.26
80 7' 1 .......... 1.36 I I
60 I , III I / / I , =t----...........: -=---f coed:
H: 40 I I II
w
z
o
(I)
20 I I I CASCADE, DATA_+--__ -I
c
PHASE I FGV CASCADE DATA
o ' ,
o 2 4 6 8 10 12
EQUIVALENT CONE ANGLE - DEGREES
Figure 75. Equivalent Conical Angle Correlation of
Diffusing Cascade Data (ref. 3).
14 16
The diffusion efficiency relates the actua1-to-idea1
recovery in the following manner:
n = (PS5 - - PS4 (measured)
D P
S5
- P
S4
) [P
T4
- P
S4
(ideal)]
For the advanced radial turbine parametric study, the line
of optimum diffuser efficiency shown in Figure 75 was used to
define a relationship between equivalent area ratio and effi-
ciency (Figure 76). Since parallel end-walls were assumed for
the study (consistent with the selected diffuser geometry), the
rotor exit swirl angle is shown as a function of area ratio.
From the data of Figure 76, rotor exit swirl levels from 37.47
to 50 degrees can be evaluated. Fifty degrees appeared reason-
able for the upper limit and, therefore, the maximum rotor exit
angle considered with deswirl vanes was set at this value. How-
ever, at the lower swirl levels, the trade-off occurrence between
deswirl and no deswirl vanes is not clear. Therefore, the data
from Figure 76 was extrapolated to 30 degrees, and then conven-
tional axial compresson cascade correlations were used to define
the deswirl vane loss to O-degree swirl.
The loss model for 0- to 30-degree exit swirls was based on
an in-house correlation between a total loss parameter (TPLP)
and diffusion factor (Df) with corrections for end-wall effects.
With representative values of solidity (a) and diffusion factor
of 1.4 and 0.4, respectively, a loss coefficient (w) was calcu-
lated with the following relationship:
w =
2 (TPLP)
cos a
4
For exit swirl levels from 30 to 50 degrees, nonseries air-
foils to apply, and the total pressure loss for the
deswirl vanes were calculated.
From the one-dimensional rotor vector exit mean-line vector
diagram, the following quantities were known:
and
wIT
A P
T
4
The deswir1 vane ideal exit conditions can be obtained from
the area ratio and continuity relationships. The equation is
rearranged to solve for the deswirl vane exit static pressure
(since diffusion is known):
P )
S4 P P
P S4. [ T4 - S4 (measured)]
J.deal
127
en
w
w
a:
G
w
c
-50
i
w'
..J
G
-40
Z

..J
a:
:::
-30
en
w
I-
::J
..J
e
-20
en
co

l-
X
w
-10
a:
e
l-
e
a:
0
c 0.90

>'
(J
Z
w 0.80
u
u..
I.L.
W
0.70
en
::>
I.L.
I.L.
c 0.60
-

"if

./
iv/
-
II ,-
DESWIRL VANE
CORRELATION FROM REF 3
I I I
DESWIRL VANE CORRELATION
BELOW -30 DEGREES BASED ON
L-CURRENT AIRESEARCH AXIAL
COMPRESSOR TECHNOLOGY
n.-
l--Cr

----
k::::::

"-

UM
DESWI RL
VANE INLET
LEVEL
TIGATED
RAMETRIC
Y
SWIRL
INVES
IN PA
STUD
8 en
w
w
a:
a:
G
WW
6
enC
::> '
I.L. C"
u.. Q)

,
I-w
4
Z
..J
wG
..JZ

::JZ
2eO
wu
1.0 1.1 1.2 1.3 1.4 1.5 1.6
o
1.7
EaUIVALENT AREA RATIO, A5/A4)
eq
Figure 76. Deswirl Equivalent Area Ratio for Maximum
Diffusion Efficiency (ref. 3).
128
The deswirl vane exit measured static pressure is used to
solve for continuity, and an exit measured total pressure is
defined.
The downstream diffuser loss is based on the axial velocity
(zero swirl), total pressure, and loss coefficient established
at the deswirl vane exit. The summation of the deswirl vane and
diffuser losses are used with the stage efficiency to define the
system efficiency for the particular configuration being exam-
ined in the parametric study.
129
APPENDIX C
ROTOR FLOW PATH AND BLADE
GEOMETRY FOR ~ B = 0, 10 AND 20
0
131
APPENDIX C
ROTOR FLOW PATH AND BLADE
GEOMETRY FOR = 0, 10 AND 20
The rotor flow path and blade geometry for the three rotor
inlet blade angle configurations are presented in Tables I, II
and III for = 0, 10 and 20, respectively. The data array
included is for 7 equally-spaced modal points along a quasi-
orthogonal station line. The first station point is on the
shroud streamline and the last (seventh) point defines the hub
streamline. The blade normal thickness is defined as the base-
line distribution for = 0. The variables for each station
line are defined below:
R - Local radius on station line (inches)
Z - Local axial dimension on station line (inches)
(Z = 0 is defined at rotor trailing edge)
BETA - True blade angle (tan =
where:
= Polar angle about the axis of rotation
m = Meridional distance along a streamline
DEV - the difference between the rotor flow
and blade angle in degrees. Rotor inlet and exit
deviations were applied in the internal flow solu-
tion
THICK - Blade normal thickness (lnches)
The flnal thickness array used to minimize the blade bend-
ing stresses for = 10 and 20 is presented in Tables IV and V.
Station data in these two tables starts at the blade inlet,
instead of the exit, and contains tangential thickness in place
of normal thickness. Beta distr ibution for these thickness
distributions is similar to the distributions in Tables II and
III. Dimensions are in inches.
133
Rob Number 4 = Blade Traillng Edge
Ind Number 19 = Blade Leading Edge
S T
R
A 1.59300
DEV .70900
I
THICK 10.67000
1 __ -0.00000
, Sl .03803
IR
IZ
'IBETA 1.65300
DEV .83800
'THICK !4.18000
I -0.00000
'\ S1 .03816
'IR
1.73100
DEV .97800
, 'THICK 6.43000
\
I 0.00000
I _ .04046
I 'Z
1.50300
.79100
24.07000
-0.00000
.05136
1.57900
.9)000
1B.3BOOO
-0.00000
.05302
1.61700
1.07400
12.36000
-0.00000
.05711
1.81250
, ,BETA 1.85000
I DEV 1.11500 1.20300
I
THICK 8.49000 - 6.38000-
, 0.00000 -0.00000
!: S1 .04758 .06803
i R
II
"OEV 1.24300
I 'THICK 3.20000
\ , -0.00000
I' 51 .05690
I R
1.99300
1.31800
2.38000
-0.00000
.08034
1.H200
.87300
17.47000
-0.00000
.07159
1.50500
1.02200
12.58000
-0.00000
.07266
1.62200
1.16900
8.29000
-0.00000
.08627
1.77500
1.29000
-4.26000
-0.00000
.10223
1.97200
1.39200
1.56000
-0.00000
.10276
I
,
OEV 1.33300
I "'THICK 1.12000
2.18000 2.17000
1.39500 1.45700
.83000 .54000
I
I, I
-0.00000 -0.00000
- .06933 -- .08792
'R
I'" Z 2.39300
\
",'ggA 1.31HOO
'THICK .16300
: 0.00000
I: s .04252
2.39000
---1.44500
.13000
-0.00000
.04704
l
' iETA 300 - 2.59000
,DEV 1.42300 1.46800
THICKO.OOOOO -0.00000
0.00000 -0.00000
5 .03498 .03515
R
Z --- ------
BETA_2.77430 2.77430
DEV 1.43850 1.47890
!
THICKO.OOOOO 0.00000
0.00000 -0.00000
S .03013 .03010
IR
2.92390-- - 2.92390
1 DEV 1.43850 1.47890
-0.00000
, 0.00000 -0.00000
.03596 .03596
2.38700
1.49700
.09700
-0.00000
.05687
2.58700
1.51200
-0.00000
-0.00000
.03535
2.77430
1.51930
0.00000
-0.00000
.03008
- 2.92390
1.51930
-0.00000
-0.00000
.03596
1.32100
.95500
10.87000
-0.00000
.09816
1.43200
1.11300
6.78000
-0.00000
.10703
1.56800
1.26500
4.21000
-0.00000
.12351
1.73750
1.37800
2.14000
-0.00000
.13418
1.95100
1."6600
.74000
-0.00000
.13592
2.16000
1.51800
.25000
-0.00000
.09870
2.38300
1.54800
.06300
-0.00000
.06238
2.58300
1.55600
-0.00000
-0.00000
.03554
2.77430
1.55960
0.00000
-0.00000
.03008
2.92390
1.55960
-0.00000
-0.00000
.03596
1.23000
1.04000
4.27000
-0.00000
.15148
1
35800
.20500
.98500
-0.00000
.15625
1.51300
1.3'1780
.14000
-0.00000
.17630
1.70000
1."
65
8
0
.020 0
-0.00000
.16345
1.93000
1.54000
-.08000
-0.00000
.17514
2.15000
1.58000
-.04000
-0.00000
.11225
2.38000
1.60000
.03000
-0.00000
.06696
2.58000
1.60000
-0.00000
-0.00000
.03601
2.77430
1.60000
0.00000
-0.00000
.03009
2.92390
1.60000
-0.00000
-0.00000
.03596
135
TABLE II.
ROTOR FLOW PATH AND BLADE GEOMETRY
FOR = 10.0 DEGREES
Rotatlonal Speed = 95,000 rpm
Inducer Tip Speed = 2250 fps
, I
"p
, z
STATION
, BETA
,DEV
rHIC K
o S TAT I ON
1
'eETA
, 0 EV
,THICK
': --,-STATION
,p
.,1
BETA
JUOEV
nTHICK
,
NU"IBFR 1
1. B300
-.200UO
60.50000
-0.00000
.02781
NUI1R ER 2
1.75300
-.13330
60.45000
-0.00000
.02781
NIJI1BER 3
1.75300
-.06070
00.10000
-0.00000
.02..1JU
0'1 STATION 4
, R - --- - -- 1.75300
J Z O.OCOOO
, BETA 59.40000
PEV -0.00000
.02781
, STATION NUMBER 5
R 1.75300
"'Z ____ _ .06700 _
wBETA 58.45000
4,OEV -0.00000
"fHICK .02842
1.61417
-.20000
57.70000
-0.00000
.02781
1.61417
-.13330
57.53]00
-0.00000
.02781
1.61480
-.00670
57.13300
-0.00000
.02181
1.01530
0.00000
56.43300
-0.00000
.02781
1.01120
.01550
55.30000
-0.00000
.02A42
1.47533
-.20000
54.90000
-0.00000
.03532
1.47533
-.13'330
54.61700
-0.00000
03532
1.47670
-.C6670
54.16700
-0.00000
.03532
1.41710
O.OOOCO
53.46700
-0.00000
.03532
1.48130
.08400
52.1,000
-0.00000
.03532
.. STATION NUMBER 6
wP 1.75300 1.61920 1.48530
4 Z .13600 .15450 .17300
a ETA. _______ 56 .. 90000 __ 53.66700 _ 50.43300
"OEV -0.00000 -0.00000 -0.00000
4,rHICK .02861 .02901 .03573
.... 1 _ STATION NUMBER 7
4JR 1.75300 1.02250 1.49200
, Z .21000 .23800 .26700
'BETA 54.67000 51.28000 47.89000
"'OEV ___ -0.00000 -0.00000 -0.00000
o THICK .02875 .02905 .03580
, ,
STATION NUM8ER 8
1.33f,50
-.20000
52.10000
-O.OOCOO
.04771
1.33650
-.13330
51.1COOO
-O.OOOOC
.04771
1.33850
-.06610
51.20000
-0.00000
.04171
1.34000
o.oecoo
50.500CO
-0.00000
.04711
1.34550
.09250
49.00000
-0.00000
.04824
1.35150
.19150
41.20000 _
-0.00000
.04830
1.36150
.29500
44.49000
-0.00000
.04878
o-'R _ _ 1.75300 1.62750 1.50200 1.37650
"Z .28500 .32450 .36400 .40350
,BETA 52.00000 48.41000 44.83000 41.24000
I.;OEV -0.00000 -0.00000 -0.00000 -0.00000
____ .&..02.885 __ .& 03002 __ , .03b45 _ .04937 _
STATION NUMBER 9
1.197h7
-.70000
49.01700
-0.00000
.Ot44tl
1 .1 97h 7
-.13330
4!l.be300
-0.00000
.06448
1.2C030
-.06670
4R.20000
-o.oeooo
.06" 4A
1.2C230
0.00000
41.50000
-o.oceoo
.00448
1.2C910
.10100
46.00000
-O.OCCOO
.00513
1.21710
.21000
43.9t700
-O.OCCOO
.06592
1.23100
.32300
41.1CCCO
-0.00000
.00707
1.25100
.44300
31.65000
-0.00000
.Ot734
R 1.15400 1.03610 1.51900
_1 .36000 .41500 .4MOO
1.402CO 1.28500
BETA 4A.85000 44.90000 41.070CO
--0.00000---0.00000
.02901 .03051 .03619

STATION NUMBER 10
1.76000 1.05400 1.54800
, ,1 .44500 .51500 .5R4CC
r,-
'BETA --- 44.78000,40.00000 36.43000
'OEV -0.00000 -0.00000 -0.00000
.THICK .02921 .03080 .03798
.52400 .'>7900
37.17000 33.28000
-0.00000 - -0.00000
.04943 .Ot772
1.443CO
.05400
32.25000
-O.OOOCO
.05080
1.33100
.12300
28.07000
-0.00000
.01156

-.70000

-0.00000
.08341
1.05883
-.13300
45.h6100
-0.00000
.08341
1.06220
-.Cth70
45.20000
-0.00000
C8 341
1.0t410
o.cecco
44.50000
-O.CCOOO
.C8341
1.07380
.10950
43.00000
-0.00000
.08415
1.08380
.27850
40.13300
-0.00000
.086n
1.10050

31.11000
-0.00000
.OB810
1.12550
.48250
34.07000
-0.00000
.0BBbO
1.16100

29.390CO
'-0.00000
.OG415
1.23100
.79300
23.9COCO
-0.00000
.09593
.9;>000
-.20000
42.85000
-0.00000
.10101
.92000
-.13100
42.65000
-0.00000
.10101
.92400
-.06,.,10
42.20000
-o.oeoeo
.10101
.9?100
0.00000
41.5COOO
-0.00000
.OG865
.93800
.11 800
40.00000
-0.00000
.10754
.95000
.24100
37.5COOO
-0.00000
.10800
.91000
.3AOOO
34.32000
-0.00000
.11511
1.00000
.57200
30.48000
-0.00000
.12580

, -0. COOOO
.12654
1.125CO
.80500
19.12000
-0.00000
.14224
Station Number 4 = Blade Traillng Edge
Statlon Number 19 = Blade Leading Edge
'I' STAT ION
111
'1
I
BETA
, 0 EV
, ;THICK
I
',
" 1
1\ BETA
DEV
'1 THICK
,J ,
NUI"BER 11
1.17500
.54500
39.34000
-0.00000
.02931
NUMBER 12
1.80000
.65500
32.3bOOO
-0.00000
- .02947
1.oA400
.62100
35.09000
-0.00000
.03107
1.12600
.74100
28.35000
-0.00000
.03124
,
1.59300
I .109 CO
30.8'>000
-0.00000
1.03803
1.65300
'1 3B CO
24.33000
-0.00000
.038 It:
,.! STATION NUMBEII 13
II 1.84000 1.78600 1.73100
"I -__ .78700 .88300,- .97800
',SfTA 23.48000 20.03000 16.58000
IIDEV -0.00000 -0.00000 -0.00000
l
it HI CK .02952 .03180 , .04046
\I I
",i STATION 'NUMBER 14
",P 1.92500 1.88750 1.850eo
Z .94000 1.02800 1.11500
"B ElA.-___ -_13.59000 - -11. 03000 __
"OEV -0.00000 -0.00000 -0.00000
, ,T\ HICK .02981 .03454 .04758
I', __ STATION NUMBER 15
i II 2.05600 2.03500
,,2 1.09500 1.10900
BETA 4.20000 2.50000
"OEV - ---- -_-0.00000 --0.00000
I "liJHICK .03299 .04162
i,J STATION NUMREII 10
2.01400
1.24300
.80000
.=0.00000
, .05690
1.50300
.19100
2t.6COCC
-0.00000
.05136
1.57900
.93000
20.320CO
-0.00000
.05302
1.67100
1.07400
13.130CO
-0.00000
.05111
1.81250
1.20300
5.91000
-0.00000
.06803
1.99300
1.31800
-.90000
-0.00000
.08034
1.412CO
.87300
22.35000
-O.CCCOO
.07159
1.5C500
1.02200
16.31000
-O.OCCOO
.07200
1.6<'200
1.16900
9.6!lCOO
-0.00000
.0 f621
1.17500
1.29000
3.35000
-O.OCOOO
.10223
1.97200
1.39200
-2.60000
-0.00000
.10276
,"11 -- 2.21000 2.20000 2.19000
, 'Z 1.21000 1.21200 1.33300
"BETA -2.50000 -3.48000 -4.45000
!,DEV -0.00000 -0.00000 -0.00000
2.18000 2.11000
1.39500 1.45100
-5.43000 -6.4COOO
-0.00000 -0.00000
I
THICK------ .03515--- .04363-'.05519 ,- .06933 - .08792
>11
" STATION NUMBER 17
I P 2.40000 2.39700 2.39300 2.39000
I
'" 2 -, 1.29000 1.34200 1.39300 1.44500
JIBETA -7.20000 -1.03000 -8.05000 -8.48000
-0.00000 -0.00000 -0.00000 -0.00000
K ________ , .0_3482 __ ,', ___ i_" 042 52 __ .04704
Iw, STATION NUMBER 18
I
"R 2.60000 2.59700 2.59300 2.59000
1 1.33500 1.31900 1.42300 1.46800
,1 BETA -9.10000 -9.75000 -9.80000 -9.85000
OEV -0.00000 -0.00000 -0.00000 -0.00000
1 THICK .03411 .03480 .03498 .03515
I
"0 __ ST AU ntL.HUKB ER 19_
-R 2.71400 2.71400
4,2 1.35120 1.392 10
BETA -10.00000 -10.00000
OEV _ -0.00000
"THICK .03038 .03025
.,: STATION NOl'te'H 20
"'R __ 2.86780 2.8b780
-2 1.35120- 1.39270
"BETA -10.00000 -10.00000
, DEV -0.00000 -0.00000
"THICK .03522 .03522
i.71400 2.11400
1.43410- 1.47500
-10.00000 -10.00000
-0.00000 -0.00000
1.03013 .03010
_ 2.86180 2.80780
1.43410 1.41560
-10.00000 -10.000CO
-0.00000 -0.00000
'.03522 .03522
2.38100
1.49700
-8.90000
-0.00000
.05681
2.58700
1.51200
-9.90000
-0.00000
.03535
2.11400
1.51110
-10.00000
-O.CCCOO
.03008
2.8t180
1.51710
-10.00000
-O.OCCOO
.03522
1.32100
.9
18.11000
-0.00000
.09Bl0
1.43?00
1.11300
12.29000
-o.oeooo
.lC103
1.5MOO
1.2t:5CO
6.23000
-c.OCOOO
.12351
1.73150
1.31800
.19000
-0.00000
.13418
1.95100
1.4bbOO
-4.30000
-o.COOOO
.13592
2.16000
1.51!l00
-1.3BOOO
-0.00000
.09810
2.38300
1.54800
-9.33000
-0.00000
.00236
2.58300
1.55600
-9.95000
-0.00000
.03554
2.71400
1.55850
-10.00000
-0.00000
C 3008
2.8t:1eo
1."i5B50
-10.00000
-C.CCOOO
.03522
1.230CO
1.04000
13.8t:000
-0.00000
.15148
1.35800
1.20500
8.28000
-0.00000
.1'j 825
1.51300
1.357eo
2.78000
-0.00000
.11030
1.10000
1.46500
-1.71000
-0.00000
.lB345
1.93000
1.54000
-6.00000
-0.00000
.17514
2.15000
1.58000
-R.35000
-0.00000
.11225
2.3BOOO
1.60000
-9.15000
-0.00000
.0069b
2.58000
1.bOOOO
-10.00000
-O.COOOO
.03001
2.11400
1.60000
-10.00000
-O.OOOCO
.03009
2.86180
1.60000
-10.COOOO
-0.00000
.03522
137
,
Rotational Speed
=
95,000 rpm
Inducer Tip Speed
=
2200 fps
STATION 1
R 1.75300 1.61417 1.47533
Z -.30000 -.'10000 -.30000
BETA 60.60000 58.54000 56.70000
OEV -0.00000 -0.00000 -0.00000
THICK .OZ781 .02781 .03532
STATION NUMBER 2
R 1.75300 l.b1417 1.47533
Z -.20noo -.20000 -.20000
BHA 60.50000 57.70000 54.90000
OEV -O.COCCO -0.00000 -0.00000
THICK .02781 .02781 .03532
STATION NUMBER 3
R 1.75300 1.61417 1.47533
1 -.13330 -.13330 -.13330
BFTA 59.80000 56.88000 53.91000
DEV -0.00000 -0.00000 -0.00000
THICK .02781 .02781 .03532
STATION 4
R
-
1.75300 1.61480 1.47670
Z -.C6670 -.06670
BETA 59.18000 56.77000 53.'15000
DEV -0.00000 -0.00000 -0.00000
THrCK .02781 .02781 .03532
STATION 5
R 1.75300 1.61530 1.47770
Z - - o.OOCOO 0.00000 0.00000
BETA 58.20000 55.30000 52.40000
DEV -0.00000 -0.00000 -0.00000
THICK .02781 .02781 .03532
STATION NUI!RFR 6
R 1.75300 1.61720 1.48130
l .Ob700 .07550 .09400
, BETA-,----,- 55.50000 5? 75000 50.00000
I DEV -0.00000 -0.00000 -0.00000
I THICK .02842 .02842 .03532
STATION HUI1BFR 1
R 1.75300 1.61920 1.48530
Z .13bOO .15450 .17300
AETA 53.50000 50.55000 41.60000
I DEY - -
- -0.00000 -0.00000 -0.00000
, THICK .02861 .02901 .03573
STATION HUMRfR B
R 1.75300 1.62250 1.49200
Z .21000 .23800 .26700
BETA 50.55000 47.46000 44.37000
, OEV -0.00000 -0.00000 -0.00000
TABLE III. ROTOR FLOW PATH AND BLADE GEOMETRY
FOR = 20.0 DEGREES
Station Number 5 = Blade Trailing Edge
Statlon Number 20
=
Blade Leading Edge
STATrON 11 I
R 1.76000 1.65400 1.54800 1.44300 1.33700 1.23100 1.12500
,Z .44500 .51500 .65400 .17300 .79300 .8b500
1.33b50 1.19767 1.05883 .92000 AETA 39.50000 35.27000 30.93000 26.b5000 22.37000 18.08000 13.80000
-.30000 -.30000 -.30000 -.30000 DEV -o.onooo -0.00001) -0.00000 -0.00000 -0.00000 -0.00000 -0.00000
53.5300a 50.48600 46.99000 42.96600 THICK .02927 .03080
I
.03798 .05086 .07156 .09593 .1'tZ2't
-0.00000 -0.00000 -0.00000 -0.00000
STATION 1Z .04711 .Ob448 .08341 .10701
R 1.77500 1.68400 11.59300 1.50300 1.41200 1.32100 1.23000
Z .54500 .62700 .70900 .79100 .67300 .95500 1.04000
1.33b50 1. H767 1.05963 .97000 RETA 33.00000 28.51000 24.02000 19.52000 15.03000 10.54000 b.05000
-.20000 -.20000 -.20000 -.20000 DEV -0.00000 -0.00000 10.00000 -0.00000 -0.00000 -0.00000 -0.00000
52.10000 49.01700 45.93300 42.85000 THICK .02931 03107 .03803 .0513b .07159 .0981b .Hllt8
-0.00000 -0.00000 -0.00000 -0.00000
STATrON NUMBER 13 .04771 .06448 .08341 .10701
R 1.60000 1.72600 1.b5300 1.57900 1.50500 1.43200 1.35800
Z .t-5500 .74700 .63800 .9"1000 1.02200 1.11300 1.20500
1.33650 1.197&7 1.05A"3 .92000 BETA 25.75000 21.33000 1&.90000 12.48000 8.05000 3. &3000 -.80000
-.13330 -.13330 -.13300 -.13300 OEV -0.00000 -0.00000 -0.00000 -0.00000 -0.00000 -0.00000 -0.00000
51.05000 48.13000 45.22000 42.30000 THICK .02947 .03124 .03816 .05302 .072b6 .10703 .15825
-0.00000 -0.00000 -0.00000 -0.00000
STATION 14 .04771 .06448 .08341 .10701
R I.B4000 1. 78t.OO 1.73100 1.67700 1.62200 1.5MOO 1.51300
Z .78700 .88000 I .97310 1.06610 1.15910 1.25220 1.34520
1.338'50 1.20030 1.06220 .92400 BETA
- - 16.20000 12.46000 8.73000 4.99000 1.25000 -2.48000 -b.22000
-.06670 -.06670 -.Obb70 -.Obb70
OEY -0.00000 -0.00000 -0.00000 -0.00000 -0.00000 -0.00000 -0.00000
50.44000 41.53000 44.b1000 41.70000 THICK .02952 .03180 1 .04046 .05711 .06627 .12351 .17b30
-0.00000 -0.00000 -0.00000 -0.00000
.04771 .Ob448 .08341 .10701
STATION HUMRFR 15
I
R 1.92500 1.8A750 1.85000 1.81250 1.77500 1.73150 1.70000
I .94COO 1.02800 11.11500 1.20300 1.29000 1.37800 1."b500
1.34000 1.20230 1.0b470 .92700
BETA 5.44000 -.19000 -3.01000 -5.83000 -8.64000 -11.46000
0.00000 0.00000 0.00000 0.00000 DEV - ---0.00000 -0. 00000 ' iO'ooooo -0.00000 -0.00000 -0.00000 -0.00000
49.50000 46.63000 43.7b700 "0.90000
THICK .02981 .03454 .04758 .06803 .10223 .13418 .183lt5
-0.00000 -0.00000 -0.00000 -0.00000
STATION 16 .04771 Ob4't6 .08Hl .10701
R 2.05600 2.03500 2.01400 1.99"300 1.97200 1.9'HOO 1.93000
Z 1.01)500 1.16900 1.24300 1.31800 1.39200 1.4b600 1.54000
1.34550 1.20970 1.07380 .93800 BETA -6.00000 -7.60000 -9.20000 -10.80000 -12.40000 -14.00000 -15.60000
.09250 .10100 .10950 .11 800
DEY -0.00000 -0.00000 -0.00000 -0.00000 -0.00000 -0.00000 -0.00000
41.25000 4".50000 41.75000 39.00000 ' THICK - -- .03299-- .04162 .05690 .08034 .10276 .13592 .17514
-0.00000 -0.00000 -0.00000 -0.00000
STATION NUMBER 17
.04624 .06513 .08475 .10751t
R
2.21000 2.20000 12.19000 2.18000 2.17000 2.16000 2.15000
Z 1.21000 11.33300 1.45700 1.51600 1.58000
1.35150 1.21770 1.08380
95
90
BETA -13.30000 -14.12000 -14.93000 -15.75000 -lb.57000 -17.38000 -18.20000
.19150 .21000 .22 A5 .24 00 DEV -0.00000 -0.00000 -0.00000 -0.00000 -0.00000 -0.00000 -0.00000
44.65000 H.70000 38.75000 35.80000
THICK .03515 .043b3 .05579 .06933 .08192 .09870 .11225
-0.00000 -0.00000 -0.00000 -0.00000
.04830 .06592 .08b91 .10800
lR
STATION NUMBER 18
2.40000 2.39700 2.39300 2.39000 2.38700 2.38300 2.38000
Z 1.29000 1.34200 1.39300 1.44500 1.49700 1.54800 l.bOOOO
1.36150 1.23100 1.10050 .97000 ' BETA -18.00000 -18.30000 -18.60000 -18.90000 -19.20000 -19.50000 -19.80000
.29500 .32300 .35170 .38000
DEV -0.00000 -0.00000 -0.00000 -0.00000 -0.00000 -0.00000 -0.00000
41.27000 38.18000 35.09000 32.00000 THICK .03482 .03794 .04252 .0410" .05667 .Ob238 .Ob696
-0.00000 -0.00000 -0.00000 -0.00000
19
'---'-'T H ( E K .0281S--.0296"5---.03566---.0"818 -, .06701- .08810 -.1151.,..
iR 1
2
'.59300
i60000 r59700 2.59000 2.'58700 2.58300 2.58000
STATION NUMlER
q :z .33500 .37900 1.42300 1."6800 1.51200 1.55600 1.bOOOO
R 1.75300 1.62750 1.50200 1.37b50 1.25100 1.12550 1.00000
' BETA
-20.00000 -20.00000 -20.00000 -20.00000 -ZO.OOOOO -20.00000 -20.00000
- - Z .28500 .32450 .36400 .40350 .44300 ."8250 .52200
OEV -0.00000 -0.00000 -0.00000 -0.00000 -0.00000 -0.00000 -0.00000
THICK
.03471 .03486 I .03498 .03515 .03535 .03554 .03bOl _AETA-- ---41.70000- - 44.28000 --40.87000--H.45000 - 34.03000 30.62000 - 21.20000
IOEV -0.00000 -0.00000 -0.00000 -0.00000 -0.00000 -0.00000 -0.00000
STATION NUI1AEP 20
I
THICK .02885 .03002 .03b"5 .0"937 .06734 .068bO .12580 I
R -'- - 2.65370 - 2.65370 l2.65370 2.65370 2.b5310 2.b5310 2.65310
STATION NUMBER 10 Z 1.34520 1.36770 11.43010 1.47260 1.51510 1.55750 l.bOOOO
I R 1.75400 1.63670 1.51900 1."0200 1.28500 1.16700 1.05000
I BfTA
-20.00000 -20.00000 -20.00000 -20.00000 -20.00000 -20.00000 -20.00000
I Z .36000 .41500 .46900 .52400 .57900 .b3300 .b6600
DEV -0.00000 -0.00000 -0.00000 -0.00000 -0.00000 -0.00000 -0.00000
,- RFTA "5.34000 41.25000 37.16000 33.07000 28.98000 24.89000 20.80000
THICK
.03038 .03025 I .03013 .03010 .03006 .03008 .03009
I DEV -0.00000 -0.00000 -0.00000 -0.00000 -0.00000 -0.00000 -0.00000
THICK .02901 .03057 .03b79 .OH43 .Ob772 .09475 .12b51t
STATION 21
l
R 2.81110 f.AII10 2.81110 2.81110 2.81110 2.811l0 2.61110
Z -- - ---- 1.34520 -- .36710 1.43010 1.47260 1.51510 1.55150 l.bOOOO
BETA -20.00000 -20.00000 -20.00000 -20.00000 -20.00000 -20.00000
DEV -0.00000 -0.00000 -0.00000 -0.00000 -0.00000 -0.00000 -0.00000
THICK
.03458 .03456 I .03458 .03459 .03458 .03458 .03456
139
Rotatlona1 Speed = 95,000 rpm TABLE IV. OPTIMIZED TANGENTIAL THICKNESSES
Inducer T1P Speed = 2250 fps FOR f3
B
= 10 DEGREES
NOTE:
X AXla1 dlstance (lnches)
1 .2.77 2.4435 .0621 8 .1326 2.1532 .1523 7 .6052 1.5560 .1091 3 1. H36 .07S1 STREAMLINE R T
T = Tangentla1 thlckness (lnches) 2 .2707 2.H15 .vb6S 'I .0'11'1 2.1483 .1bb'l STATION NO. 11
8 .53qz 1.5176 .1213 4 1.2829 1.)27b .0903 3 .2437 2.4397 .0746 10 .0552 2.1434 .1746
9 .H12 1.47'12 .1504 5 1.2522 1.452b .1027 1 1.6670 1.7530 .0554
STATION NO. 4 .21bb 2.4378 .0823 11 .01'>5 2.138b .177'1 STREAMLINE R T 10 .4072 1.4408 .1876 6 1.2215 1.3774 .1161 2 1.6722 1.665l .06'13
5 .18'16 2.435'1 .08'14 12 -.0222 2.133b
11 .3412 1.4025 .2376 7 1.1908 1.3024 01412 3 1.6778 1.5714 .0805
S TRE AHlI NE 6 .lb26 2.4340 .0'17'1 13 -.ObO'l 2.1287 .1673
,
, 1 .6bOO 1.9?4'1 .0535
8 1.1601 1.2273 .15'13 4 2 .58'13 1.90H .0614 1.66H 1.48'15 .0'138
7 .1356 2.4322 .1007
,
'I 1.12'14 101522 .1'180 5 1.6<122 1.4020 .1130
1 .248a 2.7140 .0417 8 .1086 2.4303 .10<12
I 1-
3 .5185 1.8'118 .06'13
STATIOI'! NO. 15 10 1.0'187 1.0771 .2231 6 1.700'1 1.3141 i-I 4 .4478 1.8752 .0832 I
.1439
2 .2261 2.7140 .0427 9 .0816 2.4284 .1106 STAtION NO.
" 11 1.0680 1.0020 .2572 7 1. 7l3e 1.2211 .11, 79
3 .2073 2.7140 .0433 10 .0546 2.4265 .1145 5 .3170 1.6586 .0'185 I ' STREAMLINE X T
,
8 1.7317 1.13'16
T " I 6 .3063 1.3421 .1264
.1'170
4 .18b6 2.7140 .04'12 11 .0216 2.42H .1148 S T RE HE
'I 1.7595 1.0556 .2153
5 .165'1 2.713'1 .04'14 12 .0006 .1134 7 .2355 1.6254 .1803 1 1.1057 1.7664 .0521 STATI ON NO. 1'1
10 1.8030 .97'18 .2316
6 .1451 2.7140 13 -.0264 2.4209 .1151 .4572 2.1116 .0607 8 .1646 1.S0d'l .2160 2 1.0452 1.7172 .0625 ,
11 1.8670 .<1200
7 .1244 2.7140 .0548 .4127 2.104'1 .0679 'I .0'141 1.7923 .24H
3 .'1847 1.6681 .0651 STREAMLINE .2517 I't
,
R
8 .1037 2.7140 .053'1
3 2.0'183 .0771,
"I
10 .0233 1.7757 .2743 4 .'1241 1.6186 .0718 I 'I .082'1 2.7140 .054?
I"
STATIOI'! NO. 5 4 .3237 2.0nb .086'1 I 11 -.0414 1.7591 .30)1 5 .8636 1.5697 .0176 1 1.4529 1.7530 .0542
" 10 .0622 2.7140 .0575
,
5 .2792 2.0850 .1064 I 12 -.1046 1.75'14 .3282 6 .8031 1.5206 .0'102 2 1.4287 1.6147 .0b77
11 .0415 2.7140 .0560 SHEAMLINE T 6 .2H6 2.0763 .134a 13 -.1600 1.7500 .3424 7 .H26 1.4714 .1054 3 1.4067 1.596'0 .0794
12 .0207 2.7140 .0560
7 .1'103 2.071 7 .1436 8 .6820 1.4222 .1283
"
1.3654 1.5162 .0'122
13 0.0000 2.7140 .0587 1 .3248 2.3552 .Ob27 8 .1458 2.0650 .1b72 9 .6215 1.3730 .1563 5 1.3645 1.4399 .1075 2 .2947 2.3525 .0741 9 .1013 2.0584 .1850 STATION NO. 12 10 .5010 1.3238 .1785 6 1.3446 1.31>19 .1275 3 .2646 2.34'18 .0775 10 .0566 2.0518 .1946 11 .5004 1.2747 .2226 7 1.3254 1.2639 .1543
STATION NO. 2 4 .2345 2.3472 .0'136 11 .0123 2.0451 .2048 STREAMLINE X R T
8 1.3083 1.2063 .1831 5 .2044 2.3445 .0'176 12 -.03Z2 2.0365 .2146
9 1.2907 1.127<1 .2140
X R T 6 .1743 2.3419 .1137 13 -.07b7 2.0318 .2223
1 "
1 .7730 1.8564 .051'1 STATION NO. 16
10 1.2766 1.0483 .2360 7 .1442 2.33'13 .1171 2 .7033 1.8390 .0573
1 "-
1.2668 .9585 .2601
i
' ' ,-
.1141 --2.33b6
I ' - - 3 .b337 1.81'15 .Ob18 S TRE AML INE R T
.2613 2.6235 .0484 8 .127<1
, I .0841 2.3340 .129'1 I '
"
.5b41 1.8001 .0716
2 2.b228 .0541 9 STATION NO. 9
, ' I, ,
1 3 .2163 2.6223 .05bl
, ,
10 .0540 2.3313 .1344 , , 5 .4945 1.7608 .0805 1 1.2025 1.7559 .0529
,
STATION NO. 20 4 .1938 2.6217 .0588 11 .0239 2.3287 .1350 StRE x T b .4248 1.7613 .1011 2 1.1501 1.6967 .0641 I
...
5 .1713 2.6211 .0b37 12 -.0062 2.32bO .1355 7 .3552 1.7419 .12'10 3 1.0</78 1.b375 .071l0 R T
6 .1489 2.6205 .01>5'1 13 -.0363 2.3233 .1357 1 .5187 2.0419 .05'11 8 .2856 1.7225 .1733
"
1.045" 1.5783 .074'1 1 .1264 .Ob99
2 .4673 2.0331 .0649 9 .2159 1.7031 .2149 5 .'1930 1.51'10 .0826
- --X-
--1.7531 .0549
e .103'1 2.6193 .0705 3 .4160 2.0243 .0734 10 .1463 1.6837 .2500 6 .'I40b 1.45'18 .1008 : 2 1.5119 1.67l6 .0687
'I .0814 2.6187 .0718 STA TION NO. 6 4 .3647 2.0154 .0865 11 .0767 1.0643 .2622 7 .8883 1.4006 .1156 1 3 1.4'167 1.5901 .0804
10 .0589 2.6181 .0746 5 .3133 2.0066 .1045
8 .635'1 1.3413 .1352
I'
4 1.48b6 1.5087 .0<127
11 .0364 2.6175 .0736 S TRE AI'ILINE R T 6 .2620 1.9977 .1238
9 .7835 1.2822 .1668 5 1.4756 1.4273 .1117
12 .0139 2.616'1 .0743
7 .2107 1.98'10 .1543 STATlOIl NO. 13 ___
10 .7311 1.2229 .189'1 6 1.4660 .1332
13 -.0086 2.blb3 1 .35'17 2.26QZ .0673 8 .15'13 1.9802 .1824
11 .6788 1.1637 .2245 7 1.4588 1.2654 .1585 2 .3258 2.2b57 .0701 9 .1080 1.'1714 .202'1 , STREAMLINE X R T
8 1.4536 .la58 3 .2918 2.2621 .0840 10 .0567 1.9625 .2157
9 1.453'1 1.10)3 .2157
STATtON NO. 3 4 .257'1 2.2584 .0'148 11 .0053 1.9537 .2285 1 .8690 1.6147 .0507 STATION NO. l? 10 1.4504 1.0225 .2472 5 .2240 .1067 12 -.0460 1.944'1 .2481 2 .8201 1.78bB .0569
11 1.4553 .9334 .2649
STREAI'ILINE R 6 2.2512 .11'17 13 -.0'173 1.9361 .2633 3 .7511 1.7588 .0591 STREAI'ILINE R T 7 .1561 2.2476 .132d 4 .6822 1.7308 .0656 1
1 .2771 2.5331 .0565 8 .1221 2.2440 .1459 5 .b133 1.702'1 .07b6 1 1.2920 1.7530 .0534 I': STATIOI'! NO. 21 2 2525 2.5318 .0628 9 2.2403 .1487 STATION NO 10 6 .5444 1.674'1 .085'1
, ,
2 1.H96 1.684'1 .0655 3 H80 2.5306 .0660 10 .0542 2.231>7 .1547 7 .4755 1.b470 .1005 3 1.2076 1.6166 .0160 I STREAMLINE X R
T
4 .2035 2.5294 .0704 11 .0203 2.l331 .1561 T 8 .4066 1.61<10 .1306 4 1.1654 1.5484 .0665 1 '
5 .17<10 2.5282 .077& 12 -.0136 2.22Y5 .1582
I
9 .3377 1.9340 .3527 5 1.1232 1.4602 .0941
I'" 1 1.5978 1.7530 .0565 6 .1544 2.5270 .0622 13 -.0476 2.225'1 .1589 1 .5869 1.97<17 .0577
,
10 .2687 1.5632 .2231 6 1.0810 1.4120 .1035 2--1. 5928--,-.6684 - --;0705 --- --- 2.5258 .0857
.5276 1.'1662 .0637 14 I 11 .1998 1.5352 .2018
,
7 1.0389 1.3438 .1262 3 1.5890 1.5837 .0605
7 .12'19
Z
8 .1054 2.5246 .0908
3 .4683 1.'1566 .0706 I', 8 .9'167 1.2756 .1520 I i 4 1.5864 1.49'11 .0937 'I .080'1 2.5234 .0'117 STATION NO. 7 4 .40'11 1.9453 .0781
I '1 .'1545 1.2072 .1754 5 1 5847 1."146 .1125 10 .0563 2.5221 .0'125 5 .34'18 1.'1338 .0978 STATION NO. 14 10 .9123 1.13n .1972
I"' 6 1.3303 .1358 11 2.5210 .0950 X T 6 .2905 1.'1223 .1275
,
11 .8701 1.0710 .2386
I 7 1r882 1.2461 .1607
' ' 12 .0073 2.51'16 7 .2312 1.9108 .1641 I 1 STREAMLINE X R T
;- - 8 1 5965 1.1634
.1872 13 -.0172 2.5185 .0'158 1 .4035 2.1876 .Ob30 8 .171'1 1.8'1'13 .1982
I , 9 1 60H 1.0817 .2134 2 .3649 201 82 7 .0740 'I .1l2b 1.8878 .2205
i- 1 1.0012 1.7863 .0527 STATION NO. 19
1 10 1,633'1 1.0036 .2403 3 .3261 2.1776 .078i! 10 .0533 1.'764 .?42'1
1
2 .9352 1.7478 .0554
, 11 1
1
6344 .9221 .2631
NO. 4 4 .2874 20172'1 .0'156 11 -.0060 1.864'1 .ZI>31
L
3 .8692 1.7095 .0bB STREAMLINE X T
LL '
,
5 2.168,) .lJQ4 12 -.0653 1.a534 .Zij67
"
.8032 1.6711 .0654
STRE
'--
.2100 2.1b31 .1253 13 -.1246 1.8420 .29'17 5-- .7372 1.6327 .0757 -
1 1.3750 1.752Q .0550 STATION 1'10.- 22 - ---- - -- -----
b
7 .1713 2.15d2 .14b'l 6 .b71Z 1.5'144 .0866 2 1.3443 1. 77A .0679
, I
141
'.
,I
I

,

..
' ..
, 1
RotatIonal Speed = 95,000 rpm
Inducer TIP Speed = 2200 fps
NOTE:
X AXlal dlstance (lnches)
T Tangentlal thlckness (lnches 1 .307'
.27'17
.2517
.2238
.195A
.H7'1
.13'1'1
l.41JR1
7.405'1
2.4034
2.400'1
2.3QB4
2.Hbl

2.3'111
2.3881>
2.3863

2.3B13
2.37P6
STATION NO. l !
r
;y -
iSTRFAMLINF
1
_- 1-
2 .23H
I 3 .2123
I .1'111
I 5 .lb'lQ
I I>
1
--7- .1274
8 .1062
'I .OM'I
I - 10 .OP7
11 .0425
12 .Cll2
-13-- 0.0000
STUION NO.
,STREAMLItlf
\--
1
Z
3
I

i 5

I :
10
11
-12
13
.2703
.2471
.2139
.2e07
.1774
.1542
.1310
.lC7P
CP45
.Oft3

.OH9
-.0084
STATION NO. 3
STREAMLINE x
1
2
3
4
-- 5----
1
8
'I
10
--ll
12
13
.2en
.2618
.2364
.2110
.1
.H07
.134B
.1094
.0840
.058t
.0332
.C07B
-.0176
STUION NO.
STROMUNF
2.6537
2.65'7

2.65'7
2.6537
2.6537
2.1>537
2. 65
2.6537
7.6537

2.6537
- 2.6537
2.5720
2.5712
2.5703
2.5bQb
2.5686
2.5679
2.5670
2.5661
2.5653
2.5M5
2.5637
2.5628
2.5t20
-R

2.48R3
2.4867
2.4B50

2.4B02
2.47R5
2.476'1
2.4H3
2.47310
2.4710
7.4704
T
.044'1
.04'16
.0504

.05'12
.01,02
.0"4a

.066(1
.Ob'l6
.nOR8

.0725
T

.0"52
.0616
.0650
.0724
.074&
.0771
.0814

.0845
.087'
.OAQI
.0'116
T
.'16n
.'1640
.'1581
.'1561
.'1515
.'150'1
.'14'10
.'141>5
Q45'
.'1433
.'1414

.1150
T
, 7
i
'l -8
'r- 'I
I
-I 10
': 11
i
'l 12
, 13
':,1
---.1120
.OB40
.0561
.02B1
.0002
-.0276
": STATInN NO.
, I
5
:; IS TR E 4HLINf X

"I Z .3038
"I 3 .2727
lill _ It .2"15
.2104
o I> .17'12
"
\
"' 8 .1170
, 'I .0858
I' 10 --.05H
I 11 .0235
I 12 -.007t
- ,
I
, l STATIO_D. 6
'I
',STREAMLINE x
"f---
.... i 1 .3t'l4
I 2
,- 3 --.2993
I 4
" 5 .22'13
-6 ---.1'143
'I 7 .15'12
, , 8 .1242
I
I 9 .OA92
'I 10 .0541
I' 11 .01'11
I I
17 --.015'1
13 -.050'1
" I
STATION NO 7
p
2.3271>
2.3243
2.3208
2.3175
2.3142
2.31e9
2.3015
2.3041
2.3008
2.2'174
2.2'141
2.2'107
2.28H

2.24'10
2.2446
2.2402

2.2314
2.226'1
2. 2Z 26
2.2182
2.ZlH
2.20'13
2.204'1
2.2005
2.1Q61
-x R
-,I 1 .H24 2.1738
" 2 .3721> 2.16Rl
1 I 3 .332'1 2.1624
L, 4 .2'132 2015H
- 5 .25-14-- 2.150'1
I> .2137 2.1451
7 .1739 2.13'14

.0687
.()777

.0'122
.1015
.101>0
.1131>
.IIe'l

.1314
.1372
T
-- - .0601>
.075l
.0838
.0'lRI
.1040
.1161
.1227
.1303
.1372
.1484
.1510
.1612
.1!>38
T
.0662
.0731
.0'11'1
.0'1'1'1
.1150

.1367
.1471

.1737
.1780
01'104
.1'136
.0642
.0760
.0871
.107"
--.117'1
.1360

[
I
,
8
'I
10
11
12
13
.13'02
.0'14'

.0150
-.0747

.' STATInN NO. 8
" 'STRE AHL PIE
, i
1111
11\
1
2
3
4
5
I>
7

q
10
11
12
13

.3134
.3280
.lAn
.2373
.1'120

.1013
.055'1
.0101>
-.034 8
-.0801
, STATION Nn. 'I
X
1 .5232
2
3 .'1'13
4
5
I> .2635
7 .2115
8 .15'16
q .1076
10 .C556
11 .0037
12
13 -.1002
i
,
STATTON Nil. 10
:' sue AMllNE x
I
I
--1
4
5
" 7
8
9
10
11
17
13
.'8'1l
.52'15
-.
.4103
.35r7
.2'110
.2'14
.1716
.1172
.05?6
-.0070

-.1763
2.1317
2.1 77'1
2.1222
2.1164
2.1107
2.1050
2.1030
2.0'15"
2.08Rl
2.CAr"
2.0734
2.06"0
2.0587
2.0512

2.02'10
2.02H
Z.OIU
2.03H

2.0181>
2.0007
1.'1'1'18
1. '1'103
1.'180'1
1.'1715
1.'11>21
1.'1526
1.9433

1.9245
o
1.'l71Q
1.'1661

1.'1'07
1.'lIPS
1. '1070
1.6'152
1.B835
1. P717
1. R
I.R4AO
1.83P
.1"V
01"77
.2015

.2221
.?337
T
.OM5
.071>1>
.0'll8
.102'1
.1265
.1470
.16H
.187'1
.7074
.22'1B
.2H2
.25'10
.2785
.0"11
.0743
.0856

.17QZ

.lno
.202l
.7311
?5R 7
.21>'17
.2Qbb
.317'1
T

.0,,48
.O"ZO
.10P
.127Q

.1054
.215"
.7H7
.2 RR I)
.3104
.HRA
.'51>4
TABLE V.
i STlTl'lN Nil. 11
X
,
I' ,
, ,
I,i
i
, I
1 .6fOO
2 .58'11
3 .5182
.H73
5 .37M
I> .3055
7 .2H6
8 .1637
II .0'128
10 .021'1
11 -.C4'10
12---.1054
13 -.HOO

i:: If--S-T-A-T-I-O-N-N 0 12
I'l!
L:iS"TAEAIILINE

" 7008
" 3 .1>310
" 4 .5tt2
, 5-- .4'114
I> 'lib
,17 .35lA
, 2820
, 'I .21ll
., 10 .1424
----11-- .0726

l.J 13
X
" 1 .8855
..3, 2 .8163

4 .b77P
5 .1>085
-f>--.5392
7
'"
B .4007

10 .2622
11 .H2'1
STATION NO. H
"'ISTREAMllNf
-,
1 .11'184
2 .'1317
3 .e650
, I 4 __
5 .7317
b
OPTIMIZED TANGENTIAL THICKNESSES
FOR = 20 DEGREES
p
1.'1250
1.'1085
1.8'120
1. f7 54

1.8423

1.80'13
1.7'128
1.
1.75'16
1.75Ql
1.7500
1.B5'15
1.8403
1.8212
1.8021
1.7830
1.763'1
1.7448

1.7066
1.1>874
1.66(13
R
1. e157
1.788\
1.7611
1.7338
1.70M
1.1>7'10
1.6517

1.5'170
1.51>97

R
1.78M
1.74'13
1.7117
1.f>741
1.6365

T
.0547
.0624
.0756
.10'1'1
.1

.2465

.3258

.3RH
.4088
T
.0512
.05'11
.0674
.0786
.1019
.H7'1
.16'14

.2424

.12'17
T
.04'17
.0555

.0866
.10'12
.1414
.1735
,?PZ

.7'114
T
.0511>

.0612
_Ob81
.0741
.0'155
( 7
8
II
.5'18'

.33B
i
!
I
1"1
I:'
H
\:'i
I !
10
11
HATION NO. 15
HRE IHL TNE X
1
2
3 Q81A
4 .'1203
5 .e588
6 .7'173
7 .735"
e
'I ." 127
10 .5517
11
, I STUlnN NO. 16
"ISTRE !Nf
"I
x
" 1
2
3
4
5
6
7
8
'I
10
11
1.2038
1.1504
1.0'170
1.0436
.'1'107
.93"8
-.8R34
.83eo
.7766
.7232
.1>6'lB
STATION NO. 17
,
n'STAFllIlINf Y
1
..
M'
I::!
1
Z
3

5
6
7
e
'I
10
11
1.2'148
1.7518
1.208A
1.1658
1.1228
1. CHA
1.036P
Q'l38

'107"
A6"A
STl flON NO. lA
STRHMltNF X
1 1.37A2
7 1.3470
1.5H4
1.'737

1.4486
1.4110
1.7/01>6
1.71A'

1."?}4
1.5730
1.537'1
1.471>1
1." 276

1.3310
1.2A'5
1.755'1
1.I>Q7Z
l.t3A5
1.57'1'1
1.5212

1.40'A
1.3451
1.2Pt5
1.1'77
1.16'11
1.75 '1
1./oPH

1.54'll
1. loA 17
1.4113
1.3454
1.?7n
1.2CQ5
1.141"
1.
1.757'1

.1160
.15l3
.IR40
.2251

T
.OH7

.0702
.0753
.1043
.103'1
.1323
.11-08
.1'157
.ll.7
T _
.055.
.0624
.0"'10

.084'1
.0'1"2
.1080
.1376
.IHO
.I Al"
.22H
T
.0561>
.,!>5n
.0144
.0805
.0'112
.1055
.PQl
.14n
.1731 ,
.155'1
.2371 I
I
:1
3 1.315'1
1.2847
5 1.253"
I> 1.
7 1.19lZ

'I 1.128'1
IG---l.C'l78
11 1.0b61>
STATION NO. III
STII E AIIL-INf
1 1.4558
I --- 2 --
"H-l.4085
" 4 1.3868
,- 1.3657
'" I>
"I 7 1.325'1
<u 10 1.276'1
21 l--h-l682
"
"
"-SlA'ro_O. 20
- I
"STREAML INf x
'i------

1.5278

1.3776
1.3026
1.2275
1.1525
1.0774
1.0023
1.7530
1.6747

1.5183
- 1.4400
1.3620
1.2841
1.7065
1.1281
1.0484
.1I58l ---
'I 1 1.528'1 1.7530
" Z ',1.5131> 1.6716
"/--3---1.5000 - 1.5'101
" 4 1.4877 1.50Rb
1.4274
J.l 1.3461
"" 7 ' 1.45'12 1.2655
., 8 1.453'1 1.1850

"I 10 1.4515 1.0225
ST::ION .9'31
STREAMLINE x
- 1--1.IIP'I
2 1.5931>
3 11.58'17
4 ,1.5870
5 11.5852
I> 1.5857
7 , 1.5884
'I 1.1>101
R
T
I '
I
8 11.5Q67
10 1.6347
11 I 1.6369
1.7530
1.6t>n
1.5837

1.4146

1.2461
1.11034
1.0R17
1.0035
.9220

.(\6'11
I

STATION NO. 22
.OR03
.(\'124
.1071
.11
x
1
.1140
, 3
I '
.1'108
4
5
f>
7
.77"0 :'!
.2577
'i
,I
T
i'
8
'I
10
11
.0576
.0705
.OR17
.OQ46
.10'1;
.1257
.14A4
1::1-
.1755
?052
.2367
1.6670
1.1>722
1.6778
1.6843
1.1>922
1.7009
1.
1.7317
1.75'15
1.8030
1.a1>70
"
1
2
;r------
R
1.7530
1.6652
1.5773

1.4020
1.3141
1.2271
1.13G6
1.055h
.'17Q8
.9200
T 1"----
_ __ __ _ _ __ _
.0583
.071b '"
.0813 "'--
.OQ4b II
.112h "
.131>0 _
.1607 I"
.1877 ,
.7170 "I-
24 7q
.2655
T
.O'RO
.0724

.0'160
.1153
.13A5
.1!>7Q
.1873
.2140

.2600
I
1---
,I
"
!-- -
; 'I
I "I
T
.0587
.071A
.OR25
.OQ59
.1157

.16Q2
.1923
.2151>
.2322
.2517
143
REFERENCES
1. Stani tz, J. D., "Some Theoretical Aerodynamic Investi-
gations of Impellers in Radial and Mixed Flow Centrifugal
Compressors," Trans. ASME Vol. 74, No.4, May 1952.
2. Dovzhik, S. A. and V. M. Kartavenko, "Measurement of the
Effect of Flow Swirl on the Efficiency of Annular Ducts and
Exhaust Nozzles of Axial Turbomachines," Fluid Mechanics
Soviet Research. Vol. 4, No.4, July-August 1975.
3. Mitchell, W. S. and J. F. Soileau, "Turbine Exit Guide Vane
Program," AFAPL-TR-77-75, November 1977.
4. Rogo, C., "Development of a High Tip Speed Radial
System for a Small Turboal ternator , " SAE Paper
Presented at SAE Mid-Year Meeting, Montreal,
Canada, June 1971.
Turbine
710552,
Quebec,
5. Eckert, B., "Axial Kompressoren and Radial Kompressoren,"
Berlin/Bottingen/Heidelberg: Springer-Yerlag, 1953.
6. Kofskey, M. G. and W. M. Nusbaum, "Effects of Specific
Speed on ExperImental Performance of a Radial-Inflow
Turbine," NASA TN D-6605, February 1972.
7. Nusbaum, W. J. and C. A. Wassenhauer, "Experimental Perfor-
mance Evaluation of a 4.59-Inch Radial- Inflow Turbine Over
a Range of Reynolds Numbers," NASA TN D-3835, February
1967.
8. Penny, N., "Rover Case History of Small Gas Turbines," SAE
Paper No. 634A, January 1, 1963.
9. Futral, Jr., W. M. and D. E. Holeski, "Experimental Results
of Varying the Blade-Shroud Clearance in a 6.02-Inch
Radial-Inflow Turbine," NASA TN D-5513, 1970.
10. Kidwell, J. R. and G. D. Large, "Advanced Technology Compo-
nents for Model Aircraft Auxiliary Power System,"
AFAPL-TR-2106, February 1980.
11. Calvert, G. S. and U. Opapuu, "Design and Evaluation of a
High-Temperature Radial Turoine," USAAVLABS Technical
Report 68-69, Phase I - Final Report, January 1969.
12. Daily, J. W. and R. E. Nece, "Chamber Dimensions Effects on
Induced Flow and Frictional Resistance of Enclosed Rotating
Disks," Journal of Basic Eng ineer ing , Vol. 82, No.1,
March 1960, pp. 217-232.
145
REFERENCES (Contd)
26. Weibull, W. A., "Statistical Distribution Function of Wide
Applicabi li ty," Journal of Applied Mechanics, 18 [3] ,
1951, pp. 293-297.
27. Evans, A. G., "A General Approach for the Statistical
Analysis of Multiaxial Fracture," Journal of The American
Ceramic Society, Vol. 61, No. 7-8, July-August 1978.
147
NASA-LeW1s
21000 Brookpark Road
Cleveland, OH 44135
Kerry L. McLal11n, MS 77-2
ProJect Manager (10 cop1es)
George Virosteck, MS 501-11
Contracting Off1cer
Norman T. Musial, MS 500-318
Patent Counsel
Library, MS 60-3
(2 copies)
Robert Y. Wong, MS 77-2
Report Control Office, MS 5-5
Harold E. Rohlik, MS 77-2
Richard A. Rudey, MS 86-5
Morton H. Krasner, MS 500-210
Melvin J. Hartmann, MS 3-7
W1111am E. Goette, MS 500-210
David G. Evans, MS 500-210
Harry W. Davison, MS 500-210
Stanley M. Nosek, MS 501-11
Paul T. Kerwin, MS 500-210
Roger S. Palmer, MS 500-210
James N. Deyo, MS 500-210
W1111am E. B. Mason, MS 500-211
Harold H. Valentine, MS 500-125
Charles Wagner
CIMS 418-37-18
Chrysler Corporation
P.O. Box 1118
Detroit, MI 48288
DISTRIBUTION LIST
E. E. Bailey
AF APL/DO
Wright Patterson Air Force Base, OH 45433
NASA Scientific and Technical Information
Fac111ty
Attn: Accessioning Department
P.O. Box 8757
Baltimore/Washington International
Airport 21240
(50 copies)
NASA Headquarters
Attn: REC-l/J. F. Slomski
Washington, DC 20546
Jet Propulsion Lab.
4800 Oak Grove Drive
Pasadena, CA 91103
Attn: R. C. Heft
Robert B. Schulz
U. S. Department of Energy
MS 5H-039
20 Massachusetts Avenue
(6 cop1es)
USDOE TIC
Attn: T. Laughlin
Building 1916-T-l
Oak Ridge Turnpike at Athens Road
Oak Ridge, TN 37830
(154 cop1es for distribution under DOE
Category UC-96)
H. E. Helms T-15
General Motors Corporation
Detroit Diesel Al11son D1vision
P.O. Box 894
Ind1anapolis, IN 46206
Edw1n E. Strain
Garrett Turb1ne Engine Company
Division of Garrett Corporation
P.O. Box 5217
Phoen1x, AZ 85010
William I. Chapman
Williams Research Corporation
2280 West Maple Road
Walled Lake, MI 48088
149
1 Report No DOE/NASA/0106-l I 2 Government Accession No 3 RecIpient's Catalog No
NASA CR-165l70
4 Title and Subtitle 5 Report Date
February 19B1
ANALYTICAL DESIGN OF AN ADVANCED RADIAL TURBINE
6 Performmg Organization Code
7 Author(s) B Performmg Organization Report No
G. D. Large, D. G. Finger, C. G. Linder
31-3653
10 Work Unit No
9 Performmg Organization Name and Address
A Division of The Garrett Corporation
11 Contract or Grant No
111 S. 34th Street DEN3-l06
Phoenix, Arizona 85034
13 Type of Report and Period Covered
12 Sponsoring Agency Name and Address Contractor Report -
U.S. Department of Energy ?-l-'7Q ton R_,':'Dn
Office of Transportation Programs 14 Sponsoring Agency Code
DOE/NASA/0106-l
15 Supplementary Notes
project Manager, Kerry McLa11in Final Report prepared under
NASA-Lewis Research Center Interagency Agreement
Cleveland, OH 44135
No. DEAl Ol-77CS51040
16 Abstract
The aerodynamic and mechanical potential of a single-stage ceramic radial inflow
turbine was evaluated for a high-temperature single-stage automotive engine. The
aerodynamic analysis utilized a turbine system optimization technique to evaluate
both radial and non-radial rotor blading. Selected turbine rotor configurations
were evaluated mechanically with 3-D finite element techniques. The analysis
showed that exceptionally high rotor tip speeds (2300 ft/sec) and performance
potential are feasible with radial-bladed rotors if the projected ceramic material
properties are realized. Non-radial rotors reduced tip speed requirements (at
constant turbine efficiency) but resulted in a lower cumulative probability of
success due to higher blade and diSk stresses.
17 Key Words (Suggested by Author(s)) 1 B Distribution Statement
Advanced Radial Turbine Unclassified - Unlimited
CeramiC Engine STAR Category - 85
Automotive Gas Turbine DOE Category - UC-96
19 Security Classlf (of this report) 20 Secunty Classlf (of this page) 21 No of Pages 22 Pnce"
Unclassified Unclassified
For sale by the National Technical Information SerVice, Springfield. Virginia 22161
NASA-C-168 (Rev 10-75)
End of Document

Analytical Design of Radial Turbine Design-Nasa

Caricato da

Informazioni sul documento

Titolo originale

Copyright

Formati disponibili

Condividi questo documento

Condividi o incorpora il documento

Opzioni di condivisione

Hai trovato utile questo documento?

Questo contenuto è inappropriato?

Copyright:

Formati disponibili

Analytical Design of Radial Turbine Design-Nasa

Caricato da

Copyright:

Formati disponibili

11111111111 1111 II Ilil 111111 1111 1111 II 111111 II 1111111111 I 11111111 ,I

3 1176 00161 6656

.02901 .03051 .03619

Potrebbero piacerti anche