Turbopump Systems For Liquid Rocket Engines
This document was uploaded by user and they confirmed that they have the permission to share it. If you are author or own the copyright of this book, please report to us by using this DMCA report form. Report DMCA
Overview
Download & View Turbopump Systems For Liquid Rocket Engines as PDF for free.
More details
- Words: 47,069
- Pages: 168
r-.,. I,,,,,, V_m
I
Z 0 m
i..._m ,m[
I< n,,
i[
,mc
m
I-(n m
Z m
5
L,J L,J
C3 < UJ
LLJ
(,1 < a. u)
Q Z < (n 0 I
A
Z 0
I... immmB
< Z 0
m
....I
I..I.I ---' C.3 •-_ m
•_
a..
"" U.J _.=
0 I=_ a.. ...j
m
m...--,,,w
m
r.n
r._
.._
Z
r._
_
r,.-3 v
m
._.J
n.J
< ,.I
L,_.
< Z 0 m
I-< Z
FOREWORD
NASA
experience
Accordingly,
has indicated
criteria
are being
a need
for uniform
developed
criteria
in the following
for the design
of space
vehicles.
areas of technology:
Environment Structures Guidance and Control Chemical Individual
components
of this
work
Propulsion
will be issued
as separate
monographs
as soon
as they
are completed. This document, part of the series on Chemical Propulsion, is one such monograph. A list of all monographs issued prior to this one can be found on the final pages of this document. These
monographs
are
to be regarded
as guides
to design
and
not
as NASA
requirements,
except as may be specified in formal project specifications. It is expected, however, that these documents, revised as experience may indicate to be desirable, eventually will provide uniform design practices for NASA space vehicles. This monograph, "Turbopump Systems for the direction of Howard W. Douglass, Chief, project written
management by A. J.
Corporation
and
Liquid Design
Rocket Criteria
Engines," was prepared under Office, Lewis Research Center;
was by Harold W. Schmidt and M. Murray Bailey. The monograph was Sobin and W. R_ Bissell, Rocketdyne Division, Rockwell International was
edited
by Russell
B. Keller,
Jr. of Lewis.
To assure
technical
accuracy
of this document, scientists and engineers throughout the technical community participated in interviews, consultations, and critical review of the text. In particular, W. W. Heath of Aerojet Liquid Rocket Company; H. M. Gibson of Pratt & Whitney Aircraft Division, United Aircraft Corporation; and D. D. Scheer of the Lewis Research Center reviewed the text in detail. Comments
concerning
the
technical
National Office),
Aeronautics and Space Cleveland, Ohio 44135.
August
1974
content Administration,
of this
monograph
Lewis
Research
will be welcomed Center
(Design
by the Criteria
For saie by the National Springfield,
Virginia
Technical
22 i 51
$6.25
Information
Service
GUIDE
TO
THE
USE OF THIS
MONOGRAPH
The purpose of this monograph is to organize and present, significant experience and knowledge accumulated in programs to date. It reviews firm guidance for achieving
and assesses current greater consistency
product, and greater efficiency in the design major sections that are preceded by a brief references. The State of the Art, section identifies which design elements
use in design, the and operational
design practices, and from them in design, increased reliability effort. The introduction
2, reviews are involved
for effective development
establishes in the end
monograph is organized into two and complemented by a set of
and discusses the total design in successful design. It describes
problem, succinctly
and the
current tecnnoiogy pertaining to these elements. When detailed best available references are cited. This section serves as a survey
information is required, the of the subject that provides
background Recommended
for the Design
The
material and Practices.
Design
limitation, successful project The
Criteria,
shown
a proper
in italics
or standard must design. The Design
to use in guiding
Recommended
Practices,
a design
also
clearly
and briefly
Criteria
what
rule,
sections
have
been
positive
or in assessing
guidance
organized
can be followed
through
3, state
criteria monograph or a design manual.
to
the
into decimally
both
correspond of subject sections
and
guide,
on each essential design element to assure serve effectively as a checklist of rules for the
in section
within similarly numbered subsections the Contents displays this continuity
The design specifications,
3, state
base
its adequacy. how
to satisfy
each
of the
possible, the best procedure is described; when this cannot be done references are provided. The Recommended Practices, in conjunction
Design Criteria, provide successful design.
design
technological
in section
be imposed Criteria can
manager
Whenever appropriate
Both
prepares
practicing
numbered
designer
subsections
on
how
so that
to achieve
the subjects
from section to section. The in such a way that a particular
as a discrete
criteria.
concisely, with the
format for aspect of
subject.
is not intended to be a design handbook, a set of It is a summary and a systematic ordering of the large and
loosely organized body of existing successful its merit should be judged on how effectively to the designer.
design techniques and it makes that material
iii
practices. available
Its value and to and useful
CONTENTS
Page 1
INTRODUCTION
2.
STATE
3.
DESIGN
.............................
OF THE ART CRITERIA A - Conversion
APPENDIX
B - Glossary
REFERENCES NASA
.........
2
..................
and Recommended
APPENDIX
1
Practices
.................
Units
to SI Units
of U.S. Customary
99 137
............
139
............................
................................
Space Vehicle
Design Criteria
149 Monographs
SUBJECT
Issued
STATE
to Date
153
.............
OF THE ART
DESIGN
CRITERIA
2.1
12
3.1
99
2.1.1
12
3.1.1
99
Pump Headrise and Flowrate Net Positive Suction Head
2.1.1.1 2.1.1.2
13 15
3.1.1.1 3.1.1.2
Propellant Properties Turbine Drive Cycle
2.1.1.3 2.1.1.4
18 20
3.1.1.3 3.1.1.4
2.1.1.5 2.1.1.6
26 28
2.1.1.7
30
3.1.1.5 3.1.1.6 3.1.1.7
and Cost
2.1.1.8 2.1.1.9
30 31
3.1.1.8 3.1.1.9
Type
2.1.2
32
3.1.2
108
2.1.2.1 2.1.2.2 2.1.2.3
33 33 34
3.1.2.1 3.1.2.2
108 108
2.1.2.4
35
3.1.2.3 3.1.Z4
2.1.2.5 2.1.2.6
39 44
3.1.2.5 3.1.2.6
109 109 109
PRELIMINARY System
DESIGN
Requirements
Throttling Efficiency Weight
Range
and Size
Conditioning Life, Reliability, Selection
of System
Number Turbopump Rotational
of Units Equivalent-Weight Speed
Turbopump Arrangement Pump Configuration Turbine Configuration
Factor
99 102 102 104 105 105 i07 107 107
III
SUBJECT
STATE
OF THE ART
DESIGN
CRITERIA
2.2
48
3.2
112
2.2.1
48
3.2.1
112'
2.2.1.1 2.2.1.2
48 57
3.2.1.1 3.2.1.2
112 114
2.2.1.3 2.2.1.4 2.2.1.5
59 61 63
3.2.1.3 3.2.1.4 3.2.1.5
114 116 116
2.2.2
63
3.2.2
118
2.2.2.1
63
Stability
2.2.2.2
64
3.2.2.1 3.2.2.2
118 118
Tip Speed
2.2.2.3
64
3_2.2.3
119
2.2.3
65
3.2.3
119
2.2.3.1
65
2.2.3.2
66
3.2.3.1 3.2.3.2
119 121
2.2.4
67
3.2.4
121
2.2.4.1 2.2.4.2
67 68
3.2.4.1
121
2.2.4.3 2.2.4.4
72 74
3.2.4.2 3.2.4.3
122 123
2.2.4.5 2.2.4.6 2.2L4. 7
75 77 78
3.2.4.4 3.2.4.5 3.2.4.6
124 124 125
2.2.5
79
3.2.5
126
2.2.5.1 2.2.5.2
79 80
3.2.5.1
126
3.2.5.2
127
2.2.5.3 2.2.5.4
80 81
3.2.5.3 3.2.5.4
127 127
2.2.5.5 2.2.5.6
81 82
3.2.5.5 3.2.5.6
128 128
2.2.6
82
3.2.6
128
2.2.6.1
83
2.2.6.2 2.2.6.3 2.2.6.4
84 85 86
3.2.6.1 3.2.6.2 3.2.6.3 3.2.6.4
129 129 130 130
DETAIL
DESIGN
Limits
AND INTEGRATION
to Rotational
Inducer
Cavitation
Bearing DN Seal Rubbing Turbine-Blade Gear Pitchline Pump
Speed Centrifugal Velocity
Stress
Design
Inducer
Turbine
Speed
Inlet
Flow
Coefficient
Design
Performance Optimization Exhaust Pressure Turbopump Bearing Turbine
Mechanical
Integration
Placement Rotor Assembly
and Attachment
Turbopump Housing Bearings and Seals Axial Thrust Balance Thermal
Barriers
Assembly System
Interfaces
Pump Inlet Pump Discharge Turbopump Mounting Gas-Generator Connection Turbopump
Service
Turbopump
Overhaul
and Mounting
on the Engine
Start Systems Main-Propellant-Tank Head Pressurized-Gas Start Tanks Liquid-Propellant
Start Tanks
Solid-Propellant
Start Cartridge
vi
3.2.4.
7
126
SUBJECT
DESIGN
STATE
EVALUATION
Engine-System
Characteristics
Design-Point System Balance Off-Design System Balance Control Constraints System
Dynamic
Analysis
Start Throttling Shutdown
DESIGN
CRITERIA
2.3
87
3.3
131
2.3.1
87
3.3.1
131
2.3.1.1 2.3.1.2 2.3.1.3
87 88 89
3.3.1.1 3.3.1.2 3.3.1.3
131 132 133
2.3.2
90
3.3.2
133
2..3.2.1 2.3.2.2
90 93
3.3.2.1 3.3.2.2
133 134
Instability
(Pogo)
2.3.2.3 2.3.2.4
94 95
3.3.2.3 3.3.2,4
134 135
Development
Testing
2.3.3
96
3.3.3
135
2.3.3.1 2.3.3.2
96 97
3.3.3.1 3.3.3.2
135 136
System System
OF THE ART
Turbopump System Engine System
vii
LIST OF FIGURES
Title
Figure 1
J-2 centrifugal-flow
2
J-2 axial-flow
3
RL10A-3,3
4
MA-5 booster
5
YLR87-AJ-7
6
F-1 turbopump
7
Typical
8
Speed
limits
for 5500
psi LOX turbopump
9
Speed
limits
for 4300
gpm LOX turbopump
oxidizer
turbopump
fuel turbopump turbopump turbopump
pump
13
Effect
of stage design
centrifugal 14
Effect
of turbopump
pressure 15
Effect
pumps
of centrifugal
tip diameter
to pump
specific
speed
with sweptback efficiency
16
Effects
17
Schematics
18
Efficiency
pump
of rotational
21
on pump discharge = 6) .................
throttling
on pump
stage specific
pressure 25 26
.................
on throttling impeller
characteristics
of basic turbopump speed
of 27
blades ................
discharge cycle)
pressure
vs chamber 28
...............
speed on pump
efficiency
..................
speed on required
vs stage specific
17
..................
(staged-combustion
> 10 in.)
16
drive cycles ....................
limits
requirements
14
...................
Effect of drive cycle and chamber pressure (oxygen/hydrogen engine, mixture ratio
turbine
and flowrate
il
..................
map .......................
11
Headrise
10
,
Schematics
12
9
......................
10
of basic
8
..........
.......
performance
7
....................
......................
assembly
assembly
6
...............
.............
assembly
turbopump
assembly
assembly
assembly
Page
(impeller 29
........ configuration
arrangements
for a LOX turbopump
36
......
37
..................
for various
types
..°
V111
of pumps ........
....
39
Title
Figure 19
Estimated
effect
for a pump
of pump
specific
Ns-D s diagram
for various
21
Typical
point
design
22
on efficiency
speed of 1000 ......
20
relative
configuration
Page
kinds
for 1-,2-,
as a function
of pressure
Influence
25
Influence of inducer design inlet flow coefficient efficiency ..............................
27
Thermodynamic vapor
of application design
of empirical
data
28
Schematics
29
Vapor
on suction
30
Schematics
31
Typical effects turbine-blade
capability
of bearing
for driving
support
of turbine centrifugal
Typical effect of type of mixture-ratio (staged-combustion cycle) .....
37
Zero-NPSH
engine
turbines
performance
LH 2 pump.
assembly
of incremental
of variations pumping
method
capability
50
......
hydraulic
of various
pumps
propellants
and inducers
as a function
.
51
of
inducers
.
.
55
................
56
................
57
.................. at the limit of
62 69
....................
76
..........................
in pump
47
.....
5O
pressure ratio on design stress .......................
34
Effects
on inducer
for various
arrangements
Balance
36
of rocket
on cavitation
an inducer
for two-phase
33
Illustration
temperature
54
Mark 29 experimental
35
kinds
performance
head factor
32
piston
ratio (inlet
.............................
of four methods
pumping
for various
inlet flow coefficient
suppression
pressure
(supersonic
46
24
Summary
turbines
45
Approximate
26
43
3-, and 4-stage
23
of inducer
41
..........................
Typical turbine velocity ratios of 1500 ° F) ............................. regions
throttling
of pumps ...................
efficiencies
Mach number)
and engine
................
control on pump .................... for determining
NPSH on various requirements
isentropic
design
factors
for hydrogen
ix
discharge
pressure 90
headrise
.......
............ and for oxygen ......
101 103 115
Figure 38 39 40
Title
Page
Flowchartsfordetermining turbopump speed andweightattheturbine-blade centrifugal-stress limit..........................
117
Effectof turbineinlettemperature onvariousdesign factorsin turbine optimization (gas-generator cycle)......................
120
Probability ellipses forturbopumpoperation ...................
132
LIST OF TABLES Title
Table Features
of Operational
Turbopump
II
Chief Features
of Operational
Turbopumps
III
Chief Features
of Operational
Turbines
IV
Effect
!
Chief
of Propellant
Density
Page Assemblies
on Turbopump
Design
.......................
V
Advantages
and Disadvantages
of Major Turbine
VI
Comparison
VII VIII IX
Guide
for Screening
Comparison Fluids
Candidate
of Typical
Suitable
for Pump
Testing
5
(Comparison
of Oxidizer
and 19
Limits
Turbopump
Drive Systems
,4
...................
for J-2 Engine)
and Design
3
...............
Fuel Pumps
of Operational
.............
Drive Cycles on Turbine
Arrangements
for Boost .....................
xi
Pumps
..........
Drive Cycles
22 ......
........... and Preinducers
23 110
......
113 136
_mnlmm
TURBOPUMP FOR
LIQUID
SYSTEMS
ROCKET
ENGINES
1. INTRODUCTION
The turbopump assembly for a modern liquid propellant rocket engine is a complete system in itself. It consists of many components, some of which are themselves subsystems (e.g., the
pump
and the turbine).
selection of the proper into a working system. pumps, turbines, monographs (refs. Rocket
engine
This monograph
system Details
the turbopump
as a system,
type for each application and integration on the design of the various components
gears, and bearings 1 through 10).
turbopumps
deals with
have
may
be
demonstrated
found
in
excellent
the
covering
of the components including inducers,
appropriate
reliability
component
in service.
However,
because of the strong emphasis on light weight and high performance, many of the. turbopump components are designed near the limits of the state of the art. Therefore, many problem areas must be avoided if the turbopump is to be reliable and compatible with the vehicle. For must operate
example, at high
to meet the performance speed and, consequently,
and weight requirements, the turbopump must have bearings and seals that will satisfy
life requirements at high speed; in addition, the pump must have a high-suction-performance inducer so that tank pressure and weight are minimized. At the same time, the turbopump housing must accommodate the wide variations in temperature between the pumps and the turbine without affecting alignment or imposing excessive radial or axial loads on the bearings. The design problem is made more complex cycles and by a constantly advancing state of the art. The
monograph
a turbopump
is organized system
from
to follow
the logical
preliminary
design
by
succession through
the
wide
of events
testing
on
range
in the the
of possible
duty
development
rocket
engine.
of This
process normally begins with a preliminary design phase in which the turbopump size, the component types, and the component arrangement are selected to meet the system requirements. The next phase is detail design and integration, in which the final rotational speeds limits,
are selected within the the pump and the turbine
constraints of the various mechanical and hydrodynamic are optimized within the constraints of the mechanical and
fluid dynamic limits, and the components assembly. The final phase is design evaluation, by both computer simulation and experimental
are integrated in which the testing.
into an turbopump
overall design
turbopump is evaluated
2. STATE OF THE ART
The chief features of operational in tables I, II, and Ill, respectively. To minimize
inert
for a liquid machinery. even that
weight
turbopump
and thereby
increase
pumps,
the delivered
and
payload,
turbines
are displayed :,
the turbopump
system
rocket engine has the highest power-to-weight ratio in the entire field of rotating Specific horsepower (hp/lbm)*, which was high in the mid-1950's, has increased
further in more recent engines. Table have been used on flight vehicles
illustrates
assemblies,
this progress.
In approximately
I, in which rocket engine are listed in approximate 10 years,
the
specific
turbopump assemblies chronological order,
horsepower
increased
from
2.22 for the Redstone turbopump assembly to values greater than 10 for the Saturn V (F-1 and J-2 engines) turbopump assemblies. The turbopumps for Space Shuttle Main Engine (SSME), which will undergo initial development testing in 1974, will increase the specific horsepower
by another
order
of magnitude
to a value
greater
than
100.
r
This
increase
systems
in specific
that
can
horsepower
operate
reliably
has been
achieved
at
speed.
high
developed and utilized in systems that inducers then were developed to enable pump inlet pressurization requirements,
primarily
by developing
High-speed
bearings
turbopump
and
seals
were
exert low axial thrusts on the bearings. High-speed high-speed pump operation without any increase in a very important condition because of the increase
in inert weight involved in raising the pressure level in large propellant tanks. As shown for the F-1 and J-2 turbopump assemblies, the high-speed technology has made it no longer necessary to use gearing to couple As a result, the latest designs have In general, centrifugal design
are
a high-volume pump the pump connected
the pumps for all propellants other than hydrogen designs (table II). Simplicity, light weight, and good the
reasons
for
this
selection;
the J-2
figure 1 is a good example of the current state requirement has made it necessary to employ centrifugal-stress limits stage) axial design was was
used
with an efficient, high-speed directly to the turbine.
for
the
centrifugal-flow
of the art. For multiple stages
have been performance oxidizer
hydrogen, in order
pump
turbopump
with the fuel and oxidizer has both pumps and the
pumps turbine
LR81-BA-11
had
for
converting
U.S.
shown
in
the high headrise to stay within the
(fig. 3). The
Atlas
MA-5
booster
turbopump
(fig. 4) has the fuel and oxidizer pumps on the same shaft and uses a double from the drive turbine. The Titan II YLR87-AJ-7 (fig. 5) has a geared-turbine
*Factors
single-stage, on and off
for pump impellers. As a result, an eight-stage (including the inducer used for the J-2 fuel pump (fig. 2) and a two-stage centrifugal design
RL10
engine
turbine.
have
customary
two
units
on separate shafts. The F-1 turbopump on the same shaft. The turbines for rotors,
to the
the
International
2
design
System
being
of
Units
either
(SI units)
assembly
gear reduction arrangement
assembly (fig. 6) all but the Agena
velocity
are given
compounded
in Appendix
or
A.
Table
I. -
Chief
Features
of Operational
Turbopump
Assemblies
Turbopump
Engine Chamber Application
Designation A-7
Redstone
MB-3
Thor
LR87-AJ-3
Titan
Thrust,
I,
lbf
Assembly Specific
Efficiency,
pressure, psia
Arrangement
78 000
318
170 000
594
i Geared
turbine
150 000
585
Geared
turbine
I st stage
t
Weight, Ibm
percent
Single shaft, turbine in middle
and
horsepower, hp/lbm
Start
system
26.4
332
2.22
Liquid tank
46.0
562
5.40
Solid propellant
45.8
720
5.11
Liquid propellant
start tanks
start
monopropellant
start
start
cartridge
pumps 80 000
682
Geared
O 2 pump
34.0
204
7.25
Liquid propellant
205 000
702
Geared
turbine
47.0
520
7.98
Solid propellant
start cartridge
Atlas
57 000
706
Geared
turbine
35.0
229
7.27
Solid propellant
start cartridge
MA-5 booster
Atlas
330 0002
577
Geared
turbine
48.0
875
3.59
Solid propellant
start cartridge
F-1
Saturn
Single shaft, turbine on end
44.6
3150
YLR8 I-BA-11
Agena
Geared
LR91-A J-3
Titan l, 2nd stage
H-1
Saturn
MA-5 sustainer
IB
L_ IC
1 522 000
1122
16 000
506
turbine
and
20.0
turbine
and
38.1
tanks
16.6
Tank head
5.81
Solid propellant
start cartridge
484
10.70
Solid propellant
start
256
8.30
Solid
start cartridge
9.03
Tank head Pressurized-gas
60.5
pumps YLR87-AJ-7
Gemini-Titan
215 000
784
1st stage
cartridge
pumps
Gemini-Titan
YLR9 I-A J-7
Geared
1O0 000
804
Geared
N204
15 000
400
Geared
02 pump
42.0
787
Dual turbopump, series turbines
37.4
305
7.73
44.9
369
21.60
56.5
555
50D
58.5
701
108.9
pump
propellant
2 nd stage RLIOA-3-3
Centaur
J-2
Saturn S-II and S-1VB
SSME (EPL*) 3
230
Space Shuttle
000
512 300
3237
high pressure
IBased 2Two
on
the
engines,
best each
available developing
data
as of 165
000
mid-t973.
Numbers
presented
Dual turbopump, parallel turbines
are
those
for
a turbopump
hp.
3Not operational, but presented for comparative purposes.
*Emergency po_,er level.
operational
76.1
system.
Tank
head
start tank
Table II. - Chief Features of Operational Turbopumps* Rated Number
Propellant Engine designation
Propellant
A-7
MB-3
LR87-AJ-3
density, lbm/fta( i )
Oxygen Alcohol (41
71.4 56.6
Oxygen RJ-I
71.4 53.2
Oxygen RP-I
71.4 50.5
Pump type
of stages
Centrifugal
I
Discharge
inlet
Head
Weight
Volume
Rotational
pressure, psia
pre_ure, psia
rise, ft
flowrate, lbm/sec
flowrate, gpm
speed, rpm
SPSHmin, ft 12)
NPSHc rit, riO)
Efficiency,
Power,
percent
hp
356
49.8
616
205
1 290
4 718
18
11
72.0
32O
464
42.5
1 139
150
1 190
4 718
40
35
70.0
418
867
53.0
1 651
456
2 870
6 303
55
79.0
1830
913
48.0
2 337
202
1 700
6 303
34
72.0
1 210
798
53.0
t 510
412.7
2 600
7 949
40
1034
22.0
2 881
183.3
1 630
8 780
30
i LR91 -A J-3
Oxygen RP-I
71.4 50.5
Oxygen RP-I
70.8 50.5
Oxygen RP-I
i H-1
MA-5 sustainer
MA-5 booster
F-I
YLR81-BA-I
YLR87-AJ-7
YLRgI-AJ-7
RL I0A-3-3
1
35.0
1613
175.6
1 100
8 945
31
42.0
3 024
74.1
659
25 207
100
980
65.0
1 851
537
3 410
6 680
35
25
77.8
2 340
1020
57.0
2 719
240
2 130
6 680
35
28
71.8
1 670
71.4 50.5
982
53.0
1 879
193.2
1 200
10 160
30
14
64.2
1 018
996
77.0
2 616
91.6
745
10 160
85
60
64.5
Oxygen RP-I
71.4 50.5
877
50.0
1 679
839
73.0
2 184
Oxygen
71.4
t600
65.0
RP-I
50.5
1856
45.0
IRFNA (s)
98.2
949
24.0
1 360
UDMH (6)
49.4
749
24.0
2 110
N204 A-50 (7) r
90.3 56.1
1182
84.0
1 740
1363
33.5
N204 A-50
90.3 56.1
Mixed Flow Mixed Flow
1112 1201
Oxygen
68.8
Cent rifugal
_r
597
60.5
t120
Centrifugal
2
990
30.0
31 800
1
1114
39.0
2 185
460.4
2 920
8 753
25
1238
30.0
38 000
83.6
8 530
27 130
130
31 000
(9)
37 400
(9)
Hydrogen J-2
I I
819 1097
Oxygen Hydrogen
4.35 70.8 4.4
Centrifugal Axial
7+ inducer
458
62O
2 862
6 314
40
211
1 867
6314
33
74.3
1800
73.6
I 151
3097
4070
25 200
5 488
65
5 168
1715
15 250
5 488
70
,60
74.6
30 200
55
72.6
22 I00
39.3
180
25 389
12
15.3
139
14 410
34
550
2 700
8 382
44
68.0
2 560
3 381
274
2 180
9 209
43
68.0
2 480
41.0
1 713
207
1 010
8 405
30
67.4
960
44.5
2 981
115
904
23 685
100
57.1
1090
I
352 352
28.2
184
12 100
17
62.9
5.6
581
30 250
132
55.0
592
94
18
80.0
2 358
75
73.0
7 977
(9)
78.1/ 69.6
27 400
(9)
74.1
76 400
stage SSME (EPL)**
Oxygen
70.4
Centrifugal
1/2 Is)
high pressure Hydrogen
4.38
Centrifugal
3
5174/ 8491
379/ 4940
9640/ 7100
6981
188
193 900
1137/ 120
7250/ 633
160.5
16 450
*Bared on the best data as of mid-1973. Numbers presented are those for a turbopump operational system. (l)At temperature specified by the application.
(6)Unsymmetrical dimetbylhydmzine,
(2)Contraentatly specified pump NPSI{, maximum acceptable.
(7)50 percent hydrazine and 50 percent UDMH,
(3)NPSII at a given drop in punlp discharge pressure, generally 2 percent.
(8) I0 percent of the flow goes through a second stage; numbers below slash are _r _eond stage alone_
(4)75 percent alcohol, 25 percent water.
(9)Boost pump upstream.
(5)Inhibited red fuming nitric acid.
(CH3) 2 NNH 2.
**Not operational, bat presented for comperative purpo-_.
Table
Number Engine
Working
designation A-7
fluid H202
I*
-
Chief
temperature, Type 2-row velocity
Features
of Operational
Turbines
_
Inlet
Inlet
of stages
Ill.
pressure, psia
°F
Pitchline Pressure
Flowrate,
Rotational
ratio
Ibm/see
speed,
740
37I
21.2
1204
523
17.6
1334
413
17.8
1240
560
28.2
622
17.7
6.68
rpm
Efficiency, percent
Power, hp
veloc!ty, ft/sec
4 718
37.2
758
412
15.4
30 540
-
3 162
1200
11.7
25 172
63.6
3 678
1035
25 207
63.1
1480
1020
32 700
70.2
4 141
1290
compounded MB-3
O2/RJ-I
2
pressure compounded
LR87-AJ-3
O2/RP-I
pressure
2
conrpounded LR91-AJ-3
O2/RP-I
2
pressure
6.31
compounded 18.0
H-1
O2/RP-1
2
pressure compounded
1200
MA-5 sustainer
O2/RP-I
2
pressure
1075
692
25.0
9.73
38 000
46.3
1 663
995
1240
497
15.7
14.07
30 986
66.9
3 140
1205
1450
920
16.4
5 488
60.5
52 900
840
compounded MA-5 booster
O2/RP-I
2
pressure compounded
F-1
02/RP-I
1*
2-row velocity
171.8
compounded
_q YLR81-BA-I
1
YLR87-AJ-7
IRFNA/UDMH
I
impulse, partial admission
14130
480
37.7
1.58
24 800
41.0
352
855
N204/A-50
2
pressure
1667
443
17.8
12.85
23 992
56.0
5 180
980
1650
420
29.0
5.46
23 685
53.0
2 122
964
106
698
1.42
5.35
30 250
74.0
687
782
2.5
5.0
8 753
48.4
2 358
590
7.3
7.0
27 130
60.1
7 977
1480
31000
72.9**
27 400
1363
37 400
79.0**
76 400
1661
compounded YLR9 I-A J-7
N204/A-50
2
pressure compounded
RL10A-3-3
Hydrogen
2
pressure compounded
J-2 - Oxidizer
O2/H 2
1"
2-row
velocity
760
89.5
compounded Fuel
O2/H 2
1"
2-row
velocity
1200
657
compounded SSME (EPL) (21 High Pressure Oxidizer Fuel
O2/H 2
2
reaction
1101
5880
159"*
63.3
O2/H 2
2
react ion
1391
5820
1.56"*
159.0
IBased on the best available data as of mid-1973.
Numbers presented are those for a turbopump
,xhal from the information given above. 2Not operational, but presented for comparative purposes.
system. Similar tables in monographs on system components (e.g., ref. 4) present design values that may differ some*l-stage because it has 1 nozzle, 2-row because it has 2 rotor blade rows. **Total-to-total fluid condition basis.
Turbine nozzles
Pump volu'
inlet -- Ind ivid ua I
turbine
rotors
Thrust balance cavit' i Impeller ! 1
package
/ bearing
Inducer inlet
/ ;i
am---
coupling
Vent bearing
for lube Forward
bearing Radial pins for thermal
Turbine
Figure
1. --J-2 centrifugal-flow
oxidizer turbopump
6
assembly.
inlet
volute Pump stator
Individual turbine rotors
Aft bearing
carrying clampingbolt
_or aft bearing lube
Forward r for
bearing
growth Inducer stator and forward bearing support
Pump rotoI
turbine inlet manifold
Thrust balance piston
Figure 2. --J-2 axial-flow
fuel turbopump
i
assembly.
pins thermal
Oxidizer
inducer
inducer
Hydrogen
Fuel pump impellers
pu_p bearing
Oxidizer gear
inlet manifold
drive
Turbine bearing
Turbine
Figure 3. - RL10A-3-3
turbopump
rotoz
assembly.
Imp • 11 er backface :ribs
Oxidizer impeller
pump Fuel
pump impeller
inducer Oxidizer
Turbine inlet manifold •
Gear reduction
Turbine quill shagt
Figure 4. - MA-5 booster turbopump
assembly.
rotors
1
Oxidizer
pump impeller
pump impeller
o
Oxidizer pump drive
gear cooler
Turbine inlet manifold
Oil re
Turbine rotors Fuel gear
pump drive Oil :pump Figure 5. - YLR87-AJ.
7 turbopump
assembly.
Turbine inlet mmifoid pump
Turbine rotors
.... '
impeller • i
Forward
Fuel inducer
bearings Oxidizer impeller
Oxidizer inducer
coupling
Purge¸ seal package
Radial thermal
Figure
6. -
pins for growth
F-1 turbopump
assembly.
pressure variety To
compounded;
the
of approaches
meet
future
LR81-BA-11
for turbopump requirements
had
design for
high
a single-stage
impulse
and arrangement chamber
turbine.
has been
pressure
and
Thus,
a wide
used. wide
throttling
range,
turbopump systems will continue to use single-stage centrifugal pumps for propellants tither than hydrogen and multistage centrifugal pumps for most hydrogen applications. Two-phase pumping
capability
will
become
increasingly
pressurization system by permitting the Future attitude control and auxiliary accumulators
charged
by
small
turbopump will have to start seconds. As a result, rapid
important
cryogenic* propulsion
turbopumps.
as a means
to simplify
the
tank
propellants in the tank to be saturated. engines may be fed through gaseous
To
minimize
the
accumulator
size,
each
thousands of times and, after each start, operate for only a few and efficient pump preconditioning will become increasingly
important.
2.1
PRELIMINARY
2.1.1
The
System
turbopump
will be used.
Requirements
system The
DESIGN
basic
requirements requirement
stem
from
is that
the
the engine turbopump
system
in which
operate
engine system while delivering propellants at the conditions chamber; more specifically, the various sytem requirements
the vehicle
Because
pump
to deliver speed
is
a proportionately a
performance, pump rotational important design parameters
strong
higher
influence
in
the turbopump the limits
of the
required by the engine thrust are evaluated individually and
collectively for their impact on the turbopump design. In turbopump a strong emphasis on attaining the lowest weight, because a relatively enable
within
system design, there is low system weight will
payload. all
phases
of turbopump
design
and
speed N and pump specific speed Ns evidence themselves as in the evaluation process. These parameters are related by the
expression
Ns -
*Terms
and
symbols,
materials,
and
vehicles
and engines
NQ y2
(1)
H3/_
are defined
12
or identified
in Appendix
B.
where Ns= specificspeed,rpm-(gpm)v2/ft N = rotational Q -- volume
speed,
An
increase
generally increase
reduces in speed and,
gpm
ft
in the
efficiently
rpm
flowrate,
H = headrise,
3/_
rotational
the head can reduce if
the
speed
increases
the
pump
specific
coefficient and, therefore, increases the number of pump stages required
specific
speed
is below
2000
to
3000,
speed,
which,
in turn,
the throttling range. An toJ provide a given head can
increase
the
pump
efficiency. In addition, increasing the speed will often increase the turbine efficiency and will almost always reduce the turbopump weight. On the negative side, increasing the speed increases the pump inlet pressurization requirements, decreases pump life, and increases the cost required to attain a given reliability.
2.1.1.1 Pump
PUMP HEADRISE headrise
and
AND
flowrate
FLOWRATE
are the
basic
requirements
imposed
on the turbopump
system,
because the specified pressure and quantity of propellant must be delivered to the engine thrust chamber if it is to develop its design thrust. Pump designers generally prefer to work with h6adrise and volume flowrate because for a given pump design these terms are •independent consequence
of propellant density, of this independence,
whereas a pump
flowrate (fig. 7) applies to all propellant as in high-pressure hydrogen pumps.
pressure rise and performance map
densities
unless
Pump,developed head at the design value for propellant the difference between the discharge head and the (discharge pressure) must be sufficient to overcome propellant example,
system
and
deliver
in a regeneratively
pressure must cooling jacket,
propellant cooled
engine
with
required density
in a pump or, more
chamber
frequently accurately,
compressibility
pressure
chamber
are present,
at the required cycle,
the
pressure. pump
will be less than
by dividing propellant
the
design
For
discharge
line losses, the pressure drop and the engine chamber pressure.
is approximated by the average
13
effects
flowrate (engine design thrust) is suction head. The discharge head the hydraulic resistances in the
a gas-generator
equal the sum of the propellant the pressure drop in the injector,
pressure is insufficient, the engine design thrust will not be reached. The headrise the propellant
to the thrust
weight flowrate are notl As a based on headrise and volume
in the If the
value,
and
the pressure rise by density. The most
l =-
,-r,
I
I Volume
I
flowrate
Q,
I
I
gpm
Figure 7. - Typical pump performance map.
accurate
method
that will headrise.
propellant
produce the required For situations in
high-pressure in increments adding
is to use the
hydrogen with the
the
actual
to determine
the isentropic
pressure rise. This enthalpy rise which propellant compressibility
pumps), enthalpy
increment
properties
increment
divided
by the
rise
can then be converted is significant (e.g.,
the procedure involving isentropic enthalpy at the beginning of a given increment being
(isentropic
enthalpy
pump
to in
rise is applied determined by
efficiency)
to the
previous actual enthalpy. This procedure is necessary because the propellant heating caused by the inefficiency in one increment decreases the flow density at the beginning of the next increment. As a result, the headrise based on isentropic enthalpy rise as the sum of the isentropic increments is roughly half-way between the headrise determined from th e inlet density
and
the
headrise
heating effect had discharge pressures propellant
heating
based
on
a single-step
little influence on were approximately is a very
significant
the
factor
RL10 1000
isentropic
for 6000-psi
14
enthalpy
rise.
This
and J-2 hydrogen pumps and 1250 psia, respectively. hydrogen
pumps.
propellant because the However,
The volume flowrates required from the pumps are establishedfrom the engine thrust, engine specific impulse, engine mixture ratio, and the density of the propellant being considered.Thereis often somequestionasto whether to usethe inlet, the discharge,or the averagedensity. The inlet density is easierto determine,but the dischargeor the average density is more accurateif the headriseis to be calculatedfrom the flow-passage geomet_ and the volume flowrate. A generalpolicy is to use inlet density for preliminary design where accuracyis not critical, and local densityfor the detail designof bladeflow passages. During off-design operation, the requirementsgenerally are determined over the entire operatingrange, becausea turbopump system often encountersits most severeoperating conditionsduring off-designoperation. Figures8 and 9, which presentvariouskinds of pumpspeedlimits for specifiedconditions, illustrate the effect of pump flowrate and headrise on the turbopump design speed. Regardlessof the mechanicalor hydrodynamic limit (sec.2.2.1) at which the turbopump is designed,increasingthe flowrate decreasesthe allowabledesignspeed.If the turbopump is designedat the bearing DN or the seal speed limits, increasingthe headrise(discharge pressure)also decreases the allowabledesignspeed.However,if the pump is designedat a givenvalueof specificSpeedto obtain the characteristicsof a giventype of pump, increasing the headrisewill increasethe speed.As far as the other limits are concerned,discharge pressurehaslittle effect on speed.
2.1.1.2
NET POSITIVE
By definition, the head due
SUCTION
HEAD
net positive suction head NPSH is the difference, at the pump inlet, between to total fluid pressure and the head due to propellant vapor pressure; it is
expressed in feet of the propellant being pumped. In vehicle, the NPSH is determined from an optimization
the preliminary that considers
design phase the weights
of the of the
vehicle tank, the tank pressurization system, the pressurization gas, and the feed line and, in addition, the system cost, the pump efficiency, and the turbopump weight. The trade often is made without the last two items because, in most cases, vehicle considerations far outweigh engine considerations. If the
NPSH
is less than
a certain
critical
value,
cavitation
will occur
in the
pump
inlet,
and
the pump headrise will be less than the design value. This critical NPSH is usually the value where the headrise is 2 percent less than the noncavitating value. Three methods have been used to correct the problem of an NPSH insufficient for the pump to meet design requirements:
(1)
increasing
the
tank
pressure,
which
increases
the
NPSH
supplied
but also
increases the required tank wall thickness and the tank weight; (2) decreasing the pump design speed, which decreases the NPSH required but also decreases the pump efficiency and increases the turbopump weight; and (3) redesigning the pump inlet by increasing the diameter and lowering the flow coefficient, a step that decreases the NPSH required but can decrease
the
pump
efficiency.
(Method
(3) was used
15
to meet
the
engine
requirements
with
Propellant = liquid oxygen Pump discharge pressure = 5500 psi Turbine centrifugal stress = 35 000 Seal DN
= 1.5
Shaft 100
rubbing
speed
= 325
psi
ft/sec
X 106
stress
= 40 000
psi
N s = 2000
OOC
= 63
000
_/o)
10 000
ill I
1000 !000
I
I
I
¸
I
I
I
I
I I Itl
I I 10 000 Flowrate,
Figure
8. -
Speed
limits
for
5500
i00 gpm
psi LOX
16
turbopump
(ref.
2).
000
100
000
Turbine
SO
000
20
000
(low
Turbine
NPSH
NPSH
O O
10
stress
=
SO
pressure
ratio_
stres_e
ratio)
ft
= 30
ft
000
NPSH
=
I0
I
ft
SO00 Propellant = liquid Pump flowrate - 4500 Turbine centrifugal 55 000 psi Seal rubbing DN= 1.5 xl0 Shaft stress 2000
Ns =
200
400
s_eed=525 =
I000
9.
-- Speed
limits
40
000
=
ft/sec psi
2000
2000
4000
Headrise,
Figure
oxygen gpm stress
for
4300
17
ft
gpm
LOX
turbopump.
10
000
20
0@0
the
J-2 hydrogen
pump
speed,
2.1.1.3
pump.)
and pump
A more design
PROPELLANT
detailed
discussion
of critical
NPSH,
great influence on turbopump system design-point requirements (i.e., headrise, the bearings and seals, the corrosive,
design. NPSH, cooling,
is contained
of the
in section
interrelation
2.2.1.1.
PROPERTIES
Propellant physical properties have a Propellant density affects the turbopump volume flowrate, and horsepower). In
lubricating, and viscosity characteristics of the propellant affect the speed limits (secs. 2.221.2 and 2.2.1.3), the method of lubrication, bearing-and-seal package. The size of the turbopump arrangement (sec. 2.1.2.4). material selection and thermal-conditioning
bearing-and-seal The propellant requirements
specific heat, s_ecific heat ratio, and molecular the blade heights and pitchline velocities, the (secs.
2.1.1.6
Density Densities
material selection, the and the size of the
package can also influence the saturation temperature affects for the pump. In the turbine, the
weight of the turbine working fluid affect type of staging, and the number of stages
and 2.1.2.6).
is the propellant of rocket engine
acid (IRFNA) to 4.4 lbm/ft illustrated by a comparison
property propellants
that has the greatest range from 98 lbm/ft
3 for liquid hydrogen of the J-2 liquid-oxygen
effect on turbopump design. 3 for inhibited red fuming nitric
(LH2). The effects and liquid-hydrogen
of this wide range are turbopumps (table
IV). As shown, the headrise whole LO2 pump, and
in the LH2 inducer alone is more than twice the headrise of the overall headrise for the LH 2 pump is almost 20 times the LO2 value,
even though the pressure difference in headrise, the
rises for tip speed
the two pumps are very similar. As a result of the of the LH 2 pump is more than twice as great as the tip
speed of the LO2 pump ; if the LH2 pump were a single-stage the LO2 pump), the tip speed would have to be four times addition, the LH2 much horsepower, times
pump weighs 64 pounds even though the weight
centrifugal type (i.e., similar the value for the LO2 pump.
to In
more and requires more than three times as flowrate of the LO2 pump is more than five
greater.
The
differences
the
fact
that
in size between hydrogen
LH2
turbopumps
turbopumps
and
can be rotated
other
faster
turbopumps
because
the
are minimized bearing
DN and
by seal
speed limits are higher. Because of the relatively high density of LO2, the LO2 pump usually can be designed to deliver the required pressure rise in a single stage at the optimum rotational speed without encountering any limitations on component speeds that cannot be alleviated. However, the cavitation characteristics and the density of LO2 are such that a preinducer would be required in order to avoid high the low density of LH2, a LH2 turbopump usually would
produce
the highest
payload,
because
tank pressures. Conversely, cannot be designed at the
the centrifugal
18
stress
on the turbine
because of speed that blade
limits
Table
IV.
-
Effect
of Propellant
(Comparison J-2 Engine)
Density
of Oxidizer
Pump characteristic
centrifugal
axial
Number
1
7 + inducer
lbm/sec
460.4
83.6
gpm
2920
8530
rpm
8753
27 130
1075
1208
of stages flowrate, flowrate, speed,
Pressure
rise, psi
Inducer
headrise,
Pump
headrise,
NPSH,
Power,
density,
RP-1
are
assumed discussion
-For
rocket
rotational same
shaft,
LO2/LH2
2.2.1.4).
Since
not
widely
for
such
on
engines
Ibm
However,
cavitation
known,
as
in the pumps
F-I
not
until
the and engine
generally
(no inducer)
5050
2185
38 000
18
75
390
865 - -
305
369
2358
7977
has
excellent
require
test
of
data
sufficient consult
two the
fuel
system. are
are
lacking, data
When
mounted
other
applications, NPSH
than
at
LH2,
like
obtained.
and
that
For
a
the the
tank
L02,
and
of water
a more
is
detailed
2.2.1.1.
oxidizer
are
similar
pumps
propellants on separate
19
zero
performance are
densities
and
some
at
propellants
test
characteristics
For
feasible
section
propellant
cavitation
a preinducer.
operation
characteristics
because
similar,
LH2
make
performance,
in which
), the
may
hydrogen
propellants
are
since
pump of
the
cavitation
speeds
ft/sec
hp
a hydrogen
2.2.1.1).
tip speed,
weight,
characteristics
(sec.
ft
discharge
Turbopump
cavitation
ft
ft
Impeller
low
for
Pump type
Rotational
(sec.
Design
Pumps
Liquid hydrogen, 4.4 lbm/ft 3
Volume
speed
Turbopump
Fuel
Liquid oxygen, 70.8 lbm/ft 3
Weight
the
on
and
(e.g.,
generally differ
shafts
widely so that
LO z/RP-1), are
pump
mounted
on
in densities each
can
operate
the
(e.g., at
its optimum rotational speed, as in the J-2 engine. The RL10 engine used only one turbine by.gearing the LO2 pump to the LH2 turbopump shaft (fig. 3). This arrangement avoided the low efficiency of a small, low-speed oxidizer turbine. Specific heat on isentropic
Cp and specific heat ratio "y of the turbine spouting velocity Co and, therefore, on
two-stage (LO2/RP-1
pressure-compounded and N204/A-50),
and
high-energy exceptions
hydrogen-fuel to this rule are
combinations the Redstone
working turbine
fluid have a large influence design (ref. 4). As a result,
turbines are used for low-energy-fuel two-row velocity-compounded turbines (table III). Low-energy-fuel A-7 and the F-1 turbines,
combinations are used for
turbines which were
that are two-row
velocity-compounded turbines because they had low pitchline velocities. The RL10 and the planned SSME turbines are high-energy-fuel exceptions because the turbine drive cycles for these engines call for pressure ratios of only 1.4 to 1.6 and, therefore, lower isentropic spouting velocities. Section 2.1.2.6 provides additional information on turbine relationships.
2.1.1.4 The
TURBINE
method
used
DRIVE
CYCLE
to drive
the
turbine
has
a direct
effect
on the
pump
headrise
and
power
requirements and on the pressure ratio and flowrate available to the turbine for supplying the necessary power. Thus, the turbine drive cycle used on a liquid rocket engine affects the design requirements of the turbine as well as those of the propellant pumps. Figure 10 (ref. 4) shows
typical
flow
schematics
for the
basic
types
of turbine
drive
cycles.
If the turbine
flow is in parallel with the combustion chamber (gas-generator and thrust-chamber-tapoff cycles), the pump head and power requirements are relatively low. However, if the turbine is in series with the combustion chamber (expander and staged-combustion cycles), the pressure
drop
across
the
therefore, discharge
the head and pressure is more
turbine
must
be
added
power requirements sensitive to changes
to
below
the turbine
pump
discharge
are high; in addition, in the pump and turbine
To date, most rocket engines have used the gas-generator which the turbine working fluid is derived by combustion at a temperature
the
temperature
limits.
pressure
the required efficiencies.
(GG) cycle (figs. 10(a) of the main propellants If the turbine
exhaust
the turbine a relatively
pump
and (b)), in in the GG
is afterburned
by the introduction of additional oxidizer, higher performance can be obtained GG cycle. The J-2S development engine utilizes a variation of the GG cycle called cycle (fig. 10(c)), in which injector at a location where
and,
from the the tapoff
working fluid is tapped off near the face of the cool gas is available. In the expander (or hot-fuel
tapoff) cycle (fig. 10(d)), which is used for the RL10 engine, the hydrogen that is evaporated and heated in the thrust-chamber regenerative jacket is used to drive the turbines. The turbine exhaust gas is then fed to the main chamber for combustion. In the staged-combustion then discharges
into
cycle the
(fig. 10(e)), a preburner generates main combustion chamber, where
Occurs.
20
the turbine the second
working fluid, which stage of combustion
% (a)
(b)
Bipropellant gas generator
Honopropel 1ant gas generator
(c)
Thrust tapoff
chamber
0 o Oxidizer pump J F Fuel pump T Turbine GG Gas generator P Preburner
I
(d)
Figure
The V;
general this
advantages
table
compares
the
turbine
For
any
cycle,
and
the
turbine by
pressure
turbine. pump the
must
drive
the
discharge staged-combustion
figure cycle
pressure,
the
also,
because cycle,
In
1500
the
chamber
chamber the
GG
is in series pressure
21
can
of
tapoff
cycles,
fluid
with
the
as high
the
pump
penalties
is relatively expander
be
potential
imposed
working the
growth
on
low
on
cycle
chamber
combustion as 3000
turbine
account
to entering requires
of the
a higher
chamber. psi
the
discharge
the on
on
engine,
of high
prior
VI
pressure,
limitations
the
in table Table
requirements
and
pressure
listed
cycle.
performance,
pressure
to avoid
are
each
and
chamber
turbine
pressure,
turbine
flowrate
with
in order the
level,
pressureCratio and
cycles of
control,
and
the
engines
for
drive
thrust
of cycle
the
chamber
turbine
characteristics
pressure
psi
available Same
basic the
drive cycles (ref. 4).
mixture-ratio
within
effects 11.
below
of heating for
the
pump The
kept
expander
amount However,
in are
and
meet
the
system.
illustrated
In
limited
pump
of combination;
throttling cycles.
engine
of the
propellant
drive
must
generally
flowrates. the
the
the
are
pressures
of
combustion
of basic turbine
comparison
temperature,
of the
turbine
disadvantages
concise
effect
working-fluid each
a
Staged
10. - Schematics
and
gives
(e)
Expander (fue 1)
because
For the
Table
Cycle General
Monopropellant (a)
advantages
Extensive
GG
experience
Turbine
-
Advantages
Bipropellant (a)
Extensive
and
power
ent of main
independ-
engine
Disadvantages
experience;
(a)
response
multiple (d)
(e) bo bo
for
Cycles
Thrust
expander
(a)
performance
(e.g., Thor, Jupiter, Atlas, Titan II, S-IC,
(b)
S-IB, S-IVB)
(c)
Component insensitive
Minimum
(b)
performance to other com-
Simple,
reliable
(d)
Clean,
throttling
Good uprating
(d)
Control and operation over a wide thrust and
components/
Staged
combustion
(a)
Maximum
(b)
Allows
performance
working
Component performance insensitive to
pressure without
other some
penalty
components; interaction
high chamber and throttling performance
be-
tween T/C and turbine drive
noncorrosive
turbine
(c)
Minimum
tapoff
components
fluid
(c)
Good uprating capability
chamber
high reliability
ponents, i.e., little interaction
restarts
Simple control
Drive
system (b)
Quick
High
Turbine
used
operation (c)
of Major
Propellant
GG
most frequently
with H202 (b)
V.
Good
uprating
capability
capability
(d)
Simple control
(a)
More complex thrust chamber design
(a)
Interaction
(b)
Limited uprating capability
(c)
Engine performance sensitive to component
throttling
mixture ratio range relatively easy General
(a)
Low
(a)
performance
Low performance when
disadvantages
(b)
Additional system
propellant
required
one of main
(except
afterburning
(a)
is
(b)
1000
Carbon
buildup
on tur-
bine nozzles may occur with some fuel-rich
(c)
Heavy
propellant
combinations
(d)
Limited
Additional
combustor
(b)
pressures
because
performance
loss
of
(c)
feed system over tapoff cycles
psia because available fuel
and
High systempressures Poor uprating capability
required and expander
(d)
Engine sensitive design
Requires
higher
discharge
pressures
pump
of from
(b)
tapoff thrust
performance to component
between
system and chamber assem-
bly increases opment
(c) to low chamber
to approximately
power heated
is a monopropellant and complex
in chamber
pressure
used)
unless
propellants
Limited
risk
devel-
design
Table
__
Turbine
_
VI.
-
Comparison
of Operational
and
Design
Limits
on
Turbine
Drive
Cycles
Drive
Cycle
Monopropellant
GG
Bipropellant
Propellant
GG
expander
Thrust
chamber
tapoff
Staged
combustion
Constraint Propellant combination
(a)
H 2 02,
(a)
N 2H4, or similar
monopropellant used with pellant
can be
any main
Can be used propellant
ance
(h)
Fuel-rich
cycle usually is not considered unless one of
blades
main
major
level
(a)
Becomes
passage
(b)
problem
(a)
le_s attractive
increasing
level
of bipropellant
Equally
corrosive, blades
chamber is a
pumps
(a)
for all turbo-
are practical
as performance
and weight
penalties
but carbon
and thrust
walls
some
where
System
not feasible
(a)
at
high thrust
level,
transferred
per pound
propellant
pumped
used with
is a major
Equally thrust
as heat
(h)
cham-
problem
propellants
suitable levels
turbopumps
of
engine
there
is little
be.
perform-
ance gain
passage
some
an all-
and non-
on turbine
ber coolant with
levels
Not
monopropellant
gases are high-
deposits
with
suitable
thrust
thrust
Fuel.rich
est performing
walls
(a)
cause
carbon
and thrust
any
combination
fuel
propellants
with system tO
engine
Can be used with propellant
on turbine
coolant
is a
(a)
with
and nonbut
deposits
monopropellant Thrust
systems
I-I2 or light hydrocarbons
GG is highest
corrosive,
of low perform-
and high weight,
propellants
Limited'to as the main
combination
Because
(a)
any
properforming
(b)
with
combination
for all
(a)
Fuel-rich
is highest
and noncorrosive,
but carbon
deposits
on turbine
blades
and thrust
coolant
passage
problem
with
some
propellants
At low thrust
levels
and high
where
pressure,
are practical
ciencies
decreases
preburner
performing
chamber walls is a major
decreasing drive
pump
up pump
effi-
discharge
pressure_
in-
crease Performance
(a)
Lowest-performance when
used with
main
engine.
family
system
(a)
order
Hydrazine
gives much
performance
than
Low performance
at lO00-psia
peroxide
sure.
Loss
chamber Pressure
(a)
Pump
discharge
approximately chamber
pressure
is
(a)
I s on the
of 1/3 to
higher
Pump
pressure
rage over
to
creases
other
with
chamber
pressure
discharge
chamber
pres.
proportional
No loss in engine
pressure
perform-
(a)
is
(a)
1.5 times
Pump
same
as bi-
(a)
GG cycle
No engine advantage creases
in-
System
applicable
chamber
(c)
at all
pressures
(b)
Cycle
is equally
at all chamber
performance with
Performance
loss very
increases
large
chamber
pressure
at high
pressures
although
increasing
(b)
System
pressure
is
(a)
All comments propellant
2.5 times
for bi-
(a)
GG cycle
limited
to
(b)
chamber
sure by power ments
other
systems
increasing
in-
chamber
Pump
discharge
pressure
linear
function
of chamber
sure and may
pressure
1000-psia loss
over
loss due Performance
pressure
increasing
discharge
chamber
applicable
bleed.
with
chamber
(b)
performance
to propellant
pressure
approxilnately
pressure
systems
Performance propellant
ance due to propellant bleed. Performance advan.
1 percent
chamber
approximately
1.5 times
(a)
system.
Loss in engine
bipropellant
pres.
require-
and power
Turbine
pressure
available lower
may
than
pres-
2.0 times
pressure
ratio
be slightly
with
exceed
is a non-
GG
(h)
Cycle
has an upper
ber pressure
limit
of cham-
operation
available
pressure (continued)
Table
_
VI. -
Comparison
of Operational
and Design
Limits
on Turbine
Drive Cycles
(concluded)
cycTUrbine Drive Bipropellant
Monopropeltant GG
le
GG
Staged
Thrust chamber tapoff
Propellant expander
combustion
Constraint Working-fluid temperature
(a)
Temperature should be as high as practical to max. imize performance. Complex, cooled turbines increase performance and reduce weight
(a)
increase performance and complexity and reduce weight
(b)
t,_
Throttling and mixture.ratio
(a)
(b)
(c)
Growth potential
(a)
Controlof quired
four valves re-
Turbine bypass used for deep throttling to minimize GG injector throttling and GG control complexity Performance loss with throttling highest of cycles
Easily uprated ifmonopropellant storage is sufficient
Temperature should be as practical (_ 1500 ° F with uncooled turbine and GG) to maximize performance, Cooled turbines and GG
(a)
(a)
(b) Power balance and system pressures strongly affected by temperature
(c)
Temperature may be dictated by coking charac, teristics of certain propel. lants Simultaneous
control of
(a)
(0
(a)
Excess turbine flow in deep.throttled condition and additional injector pressure drop at nominal condition increase turbine flow and performance loss Easily uprated, flexible system
Allcomments for bi. propellant GG system are applicable.
(b)
Temperature may be dictated by coking characteristics of certain propellants
Temperature may be dictated by coking characteristics of certain propellants
Simultaneous control of turbine i(a) flow and oxidizer valves required
(b) Turbine bypass used for deep throttling to minimize gas generator injector throttling and GG control complexity
(a)
(a)
Performance not affected by working-fluid temperature
(b)
Power balance, system pressures, and maximum chamber pressure are strongly affected by temperature. Temperature should be as high as practical to minimize system pressures, maximize chamber pressure
(c) • Temperature may be dictated by coking characteristics of certain propellants
main propellant gas generator valves may be required (b)
Performance not affected by working-fluid temperature
(0
Low PR unchoked turbines provide coupling for possible feed system and combustion instability
Cb)
No performance throttling
(c)
loss with
(d) Very high fuel system AP and fuel pump discharge pressure required for deep throttling
(a)
Limited uprating capability. Limited by heat-transfer rate
(a)
Simultaneous control of main propellant valves required; no bypass Thrust chamber must provide acceptable tapoff gas over wide throttle range for deep throttling ! Performance loss(for throttling capability) at nominal conditions (on the order of 0.5 to 1.0 sec) same as GG cycle, lower in throttled condition (no excess flow)
Easily uprated, flexible system
(a)
Control of five valves required
(b)
Low.PR unchoked turbines provide coupling for possible feed system and combustion instability
(c)
Very high preburner injector AP or complex injector design required for deep throttling
(d)
No performance
(a)
Limited uprating capability. Cycle generally operated near maximum chamber pressure for maximum per. formance
loss with throttling
s000
40oo
_'
2oo0 _
Gas-genez'ator
cycle
i
i
i
t
I
500
1000
1500
2000
2500
Chamber pressure,
Figure
1 1. -- Effect
of drive
discharge ratio
preburner
provides
for this efficiency, Allowing injector,
the resultant
and
psia
chamber
pressure
(oxygen/hydrogen
engine,
on pump mixture
= 6).
high-energy
cycle because, the required for pressure
cycle
pressure
fluid
for the turbine;
3000
psi is an approximate
required
discharge
pressures
can be as high as 7000
and the
tapoff cycles, turbine pressure ratios of approximately flowrate of turbine working fluid and thereby maximize
impulse.
For
the
engine
illustrated cycle. For pressure Relative
expander
weight
by comparing these engines,
pump
discharge
ratios
and
staged-combustion
cycles,
the
to 8000
psia.
20 are required to the engine specific
optimum
turbine
pressure
cycles,
(sec.
2.1.1.7).
Some
of the
differences
between
these
cycles
are
the J-2 engine, a GG cycle, and the RL10 engine, an expander the chamber pressures are 787 and 400 psia, respectively; the ratios pressure
are 19 and
to GG
for expander result from efficiency
limit
generally less than 1.5, because of the large quantities of turbine working fluid In .design, expander and staged-combustion cycle turbine pressure ratios are in order to maximize chamber pressure and minimize turbopump weight, thereby
minimizing
of fuel
upper
for given values of turbine inlet temperature and turbopump pump discharge pressure rises steeply at higher values (fig. 11). drops in the lines, regenerative jacket, preburner, turbine, and
For GG minimize ratios are available. minimized
3000
to chamber
pressure
are
1.6 and
2.5; and the overall
turbine
1.4. turbopump
weight
and
horsepower
generally
are
somewhat
greater
cycles and are, much greater for staged-combustion cycles; these differences differences in the pump discharge pressure requirements. To meet the high requirements,
pumps
in expander
and
25
staged-combustion
cycles
must
either
operate at higher speedor havemore stages-thanthosein GG cycles.To meet the turbine pressure-ratio requirements, GG and tapoff cycles generally incorporate two-row velocity-compounded turbines, and expander and staged-combustioncycles generally incorporateeither two-stagepressure-compounded or reaction turbines.
2.1.1.5 In some
THROTTLING future
RANGE
applications
such
auxiliary propulsion systems, As a result, the turbopumps (fig.
as main
engines
for reusable
vehicles
and accumulator-fed
the engine thrust may be varied (throttled) during the mission. will be required to operate over a range of head and flow values
12).
Throttling
range
==2 120
/ 100 G
/
80 o
60
o
40
"r.
20
0
20
I
I
I
I
40
60
80
100
Flovtate
Figure
12.--Headrise
fztction,
and
flowrate
26
q/qde$,pet,
limits
to
cent
pump
throttling.
120
Pump throttleability unstable if the slope Therefore, zero-slope
is limited, because of the constant speed
for centrifugal pumps, which point is generally considered
a rocket-engine/pump combination can become lines on the pump H-Q map are positive (ref. 11). generally have discontinuity-free the stability limit (fig. 12). For
speed lines, axial pumps,
the the
stability limit occurs when the head drops abruptly or, in other words, when the pump stalls. Relative to the engine, the throttling limit occurs when the engine operating line, which is primarily linear, crosses the pump stability line. Consequently, pump throttleability improves
as the
stability-limit
flowrate
line moves
fraction
at the
stability
to the left relative
For centrifugal pumps, the decreases and, for sweptback head coefficient decreases
throttling impellers as stage
limit
to the design
decreases
or, in other
words,
as the
point.
cap_/bility increases as stage design head coefficient with a constant discharge blade angle, stage design specific speed increases (ref. 12). Consequently,
centrifugal pump throttling capability generally increases with and, therefore, with pump design rotational speed. As shown, down to zero flow if the specific speed is above approximately
stage specific speed (fig. 13) stable throttling is possible 2500. However, heating of
the trapped propellant makes operation at complete shutoff inadvisable except for very short durations (< 10 sec.). For axial pumps, rotational speed has little effect on throttleability. As shown in figure 12, centrifugal pumps generally have approximately twice the
throttling
range
of axial
pumps
and
therefore
are primary
candidates
engines.
In el
3.0_
IN
lqs
'
"_-
3400
=0
2.0
--
0
4.1 o_1 U .,'4 ID O U "CI gt
Figure
13. -
I
I
I
0.5
1.0
1.5
Flow
coefficient
of
stage design
Effect
characteristics impeller
fraction,
specific
of centrifugal
blades.
27
_/_
des
speed on throttling pumps
with
sweptback
for throttleable
2.1.1.6 High
EFFICIENCY turbopump
efficiency
(the
product
of
the
pump
and
the
turbine
efficiencies)
is
important to all types of pump-fed liquid rocket engines. It is important for engines with GG or tapoff cycles because engine specific impulse increases with decreasing flowrate of turbine working fluid. For engines with expander or staged-combustion cycles, increasing the turbopump efficiency increases the chamber pressure attainable discharge pressure (fig. 14). This increase in chamber pressure makes the size and weight of the thrust chamber and, therefore, of the engine.
(_mmhel"
Figure
14. -
Effect
In current
engines,
overall
(table I). As shown, the approaching 60 percent. meet this goal. Figure efficiency -tip flow velocity), maximum specific
15,
a plot and
of
stage
coefficient centrifugal value, then speed
efficiency
pressure
turbopump
efficiency
efficiencies
for a centrifugal
efficiency
in general
vs specific pump.
(ratio of fluid axial pump efficiency increases decreases as stage specific ranges
28
discharge
pressure
(staged-combustion
development pumps and
_Itl
at maximum
on pump
requirements
SSME now under High-specific-speed
pump
design
pressure
of turbopump
vs chamber
for a given pump it possible to reduce
speed,
cycle).
range
from
35
to 48 percent
will have turbopump reaction turbines are
illustrates
For a given
value
the
relation
efficiencies required to
of inducer
between design
inlet
velocity at the inlet to blade tangential with increasing speed until it reaches a speed increases (refs. 2, 12, and 13). Stage,
from
1300
at q_itl
= 0.05
to 2500
at
_Itl
=
9O 0.2
!.°
.IS
|
_o
_-
60
/
40
I
30
cn
,
300
I ,,,,I 5_
I lO00
Stage
Figure
15.
-
Effect
0.20. Consequently, maximum efficiency increase
(impeller
-
2000
specific
of centrifugal
efficiency
,
10
I,
000
speed
pump tip
SO00
'
stage
specific
diameter>
10
speed
on
pump
in.).
if a pump has a design stage specific speed that is less than that at the and has a constant number of stages, an increase in design speed will
efficiency.
With axial pumps, the efficiency will decrease 2000. At these low specific speeds, the blade rotor diameter, and consequently the blade height. Above a stage
as the stage specific speed is decreased heights become very small in relation
the blade tip clearance becomes specific speed of 2000, axial-pump
below to the
a significant fraction of stage efficiencies remain
at high levels up to specific speeds of 10 000 (refs. 12 and 13) (fig. 15 for _itl > 0.20). Consequently, if an axial pump has a stage specific speed less than 2000 and has a constant number of stages, the efficiency increases with increasing design rotational speed. Above 2000,
changes
in design
For high-flowrate efficiency, because
speed
do not have
high-horsepower turbines, the turbine can be full
given values is primarily
of working a. function
are designed
with
For
requiring
engines
rotational
rotational admission
fluid inlet temperature of pitchline velocity.
pitchline
velocities
turbines
effect
on axial-pump
speed has little with a reasonable
effect blade
efficiency. on turbine height. For
and turbine pressure ratio, turbine efficiency For this reason, many rocket engine turbines
close to the rotor
of low
much
horsepower,
because they have such small blade heights that of the blade height. In this case, partial-admission
29
stress
limit.
full-admission
turbines
are inefficient,
the tip clearances are a significant fraction turbines are used (refs. 14, 15, and 16).
For
example,
a partial-admission'turbine
The efficiency increases with
is used
of this kind of turbine decreasing diameter, and for More
low-thrust-level engines, detailed information on
2.1.1.7
AND
SIZE
Light
weight
payload.
is an important
Size is minimized
turbopump so that
and
YLR81-BA-11
turbine turbine
design
handling
the
engine
with arc of admission, in turn decreases with
speed. Therefore, rotational speed. 2.1.2.6.
WEIGHT
on
increases diameter
efficiency efficiency
requirement mounting
(table
III).
arc of admission increasing rotation
increases with design is contained in section
because
of the direct
are simplified.
Weight
effect
on
and size are
directly related to the turbopump design speed. An increase in the design speed decreases the diameters of both the pump and the turbine, because the allowable tip speeds, which are set by stress-limiting
factors
and
the discharge-pressure
requirements
of each
component,
essentially constant for a given application. To maintain proper geometrical the turbopump length also will decrease. Consequently, both the turbopump the turbopump weight decrease with increased rotational speed. Turbine turbines added added
are
relationships, envelope and
efficiency can be increased by using a large number of stages. However, multistage are heavier and longer because of the added rotors and nozzles and because the rotating mass must supporting structure.
2.1.1.8
be supported
by
an outboard
bearing
that,
in turn,
must
have
CONDITIONING
If a pump
for cryogenic
propellants
is not properly
flash into vapor on entering the pump, and consequence is that the turbopump bearings be in danger some way requirements
of failure. of
Therefore,
chilled
prior
to start,
the pump will become will not be lubricated
applications
with
cryogenic
the propellants
vapor locked. or cooled and
propellants
will
Another thus will
require
either
prechilling the pump or a pump surface that chills very rapidly. also exist for restart applications with cryogenic propellants because,
These during
the engine shutdown period, heat soaks back from the hot turbine to the cold pump and, if the shutdown time is long enough, the pump can reach temperatures as high as 0 ° to 100 ° F. The development of upper stages employing cryogenic propellants (e.g., the Saturn S-II stage, the Saturn S-IVB stage, and the Centaur stage) has provided solutions to the problem of turbopump conditioning. The solutions were particularly important for the Saturn S-IVB stage (J-2 restart. In the loops
engine)
Saturn for
and
S-IVB
turbopump
for
stage,
some
both
chilldown.
Centaur
the
oxygen
Electrically
stages
(RL10
engine)
and the hydrogen driven
3O
secondary
because
feed systems pumps
force
the
engines
include propellants
must
coolant from
the tanks through the inlet lines and the pumps. Return lines connectedfrom the pump dischargesthen permit the heated propellants to return to the tanks. This system has performed satisfactorily. However,propellant can be consumedduring chilldown (because the returned heated propellant can causeadditional tank venting), the restartsare by no meansinstantaneous,and the addedcomplexity of the secondarypumps,lines,andvalvesis undesirable. The loss of chilldown propellant and the chilldown time can be reduced by coating the surfaceswetted by pumped fluid with an insulating material and, with hydrogen, by increasingthe two-phasepumping capability (ref. 17). Analytical and experimentalstudies indicate that a thin layer of low conductivity material applied to a metal surfacewill make possiblea rapid surfacechill and will reduce the heat rejection rate from the mainbody of the metal (ref. 17).Two-phasepumpingcapability canbe improvedby designingthe inducer with a high ratio of fluid incidenceangleto blade angle,minimum blade blockage,and a large inlet annulusarea.Boost pumps,having low tip speedsand large inlet diameters,can further increasethe vapor pumpingcapacity(sec.2.2.1.1). The chilldown problem can be made less severeby minimizing the rate of heat soakback (ref. 17). To this end, the number and size of the contact points betweenthe turbine and the pump are minimized, and the remainingcontact points areinsulated(sec.2.2.4.6); in addition, the turbine massrelative to the pump massis minimized by using single-stage turbines rather than multistage turbines and by using smallerdiametersand lower turbine inlet temperature.However,reducingthe diameterandthe temperaturewill adverselyaffect performance. Gearedand single-shaftturbopumpsaresuitedto restartapplicationsbecause,in both cases, one turbine rejects heat to two pumps, thereby producing lower pump temperatures.In addition, for single-shaftturbopumps, one pumpinsulatesthe turbine from the other pump and, for gearedturbopumps, the gearbox insulates the turbine from both pumps and providesan additionalheat sink.
2.1.1.9 The
LIFE,
prime
RELIABILITY,
objective
in the
AND
COST
optimization
and
design
of a new
rocket
engine
system
is to
minimize the cost per pound of payload while meeting the mission reliability requirements. The turbopump plays an important part in this optimization process. The turbopump cost can be reduced by designing for low rotational speed, state-of-the-art components, and the minimum number of pump and turbine stages. The manufacturing, and maintenance costs. In addition,
reduced complexity minimizes assembly, the amount of development required to
attain a given reliability is reduced. Low-speed designs also may have a built-in uprating capability so that redesign is not necessary each time the thrust requirement increases. In
31
general,however,modifications aimed at low cost will increaseturbopump systemweight, decreaseperformance, and can produce a net increasein cost per pound of payload. Theretbre,an optimum turbopump systemexistsfor eachapplication.The designrotational speedat this optimum is a function of the mission,the number of enginesproduced, and other parameterssuchasthe reliability goal andenginereuse: A compromiseon rotational speedwas made for the J-2 hydrogenturbopump (table II). The number of pump stagesand the weight could have been reduced by designingat a higher speed.However,the resultinghigher valuesof bearingDN and sealrubbingvelocity would haverequiredgreaterdevelopmentcoststo meetthe life and reliability requirements. Therefore, the speedwas limited to 27 000 rpm. Another exampleof a designcompromise to obtain greaterreliability is the F-1 turbopump, in which both pumpsaremounted on the sameshaft as the turbine. This arrangementincreasedreliability by eliminating the gearbox. However,relativeto the gearedturbine arrangement(e.g.,the MB-3 andMA-5 turbopumps), the fuel pump inlet pressurerequirementsarehigherbecause,with the turbine mounted on oneend, the fuel pump inlet could no longerbe axial. In
general,
designing
for
increased
life
will
increase
the
weight.
Design
/ procedures
for
increasing life include designing at relatively low speeds, avoiding resonant frequencies, using low stress levels to avoid fatigue limits, using low turbine temperature, and using low values of inducer inlet incidence angle and inducer tip speed to reduce cavitation erosion. Information on designing to obtain the presented in references 1 through 10.
2.1.2
In the
Selection
selection
turbopump
of System
of system
components
type, and
the various
system.
Headrise
NUMBER and
life
in turbopump
system
components
is
Type
system
requirements
Optimization
requirements are conflicting. Even after and design judgment must be exercised.
2.1.2.1
desired
are used
is often
necessary
many
decisions
optimization,
to select
because
the best
some
of the
are still not obvious,
OF UNITS
flowrate
requirements
can be met
more turbopumps, each delivering a fraction to be considered, then both R&D (research considered in the selection of turbopump
either
by a single
turbopump
or by two
or
of the total flowrate. If more than one pump is and development) and production costs must be size. Research and development costs increase
with size because of the increased hardware costs, increased propellant costs, and higher test facility costs. Production costs for an individual unit will also increase with size; however, for a constant than one pump
flow, two pumps, each pumping that is pumping all the flow.
32
one-half
the
flow,
usually
will cost
more
Researchand developmentcostsandunit production costsare determinedasa function of size. A production learning curve is used to obtain averageproduction costs for various numbers of production units. Typically, an equivalent93-percentlearning curveis used(a 93-percentlearning curve reducesthe averagecost per unit by 7 percent for eachdoubling of the number of units). Total production costsarethen determinedfor constanttotal flow for various size turbopumps by multiplying the averageunit cost by the number of units required to give the total flow. Total costsareobtainedby addingthe R&D andproduction costs.Sizecan then be selectedon the basisof minimum costs. For low production rates,total costswill be minimal at smallpump sizes,becausethe R&D costswill be predominantfor this type of requirement.As the production rate increasesand production costs become the predominant cost factor, the minimum costswill occur at largerand largerpump sizes.
2.1.2.2 The
TURBOPUMP
EQUIVALENT-WEIGHT
best-performing
with
given
turbopump
values
of
is the
thrust
level,
one
FACTOR that
mission
results
velocity
in the heaviest
increment,
gross
payload
for a vehicle
weight,
and
specific
impulse. Because the turbopump weight is part of the stage burnout weight, a reduction in turbopump weight allows an increase in the vehicle payload. In addition, a decrease in turbine flowrate for a GG cycle increases the vehicle payload by increasing the engine overall
specific
specific The
impulse
equivalent
expressed
impulse. lower weight
in pounds
This than
effect
that
factor
of stage
occurs
because
of the thrust EWF
is used
payload
the turbine
chamber to convert
per pound
exhaust
for a GG cycle
has a
exhaust. turbine
per second
flowrate
of turbine
to payload
and
is
flowrate:
(2) (EWF)T
-
v - O(-_s)E
(IS)E .]
F
where (EWF)T
e = turbopump PL = stage
(Is)F. = engine
equivalent
payload, specific
weight
factor,
Ibm impulse,
lbf-sec/lbm
33
lbm/(lbm/sec)
(Is)x2 = turbine exhaustspecificimpulse,lbf-sec/lbm F = enginethrust, lbf 0PL O(Is)E
= f (missionvelocity increment,systemgrossweight,systemspecific impulse,andstagepropellant fraction)
For GG cycles,EWF rangesfrom5 lbm/(lbm/sec)fOr boosterenginesto 200 lbm/(lbm/sec) for upper-stageengines(ref. 18). For staged-combustioncycles,(Is)w2 = (Is)E, andtherefore the correspondingEWF is essentiallyzero. The total effect of the turbopump on stagepayloadis determinedby addingthe turbopump weight to the product of EWF and the turbine weight flowrate.The sumis the turbopump equivalentweight: (EW)TI' = WTv +
(EWF)T
p(WT )
(3)
where (EW) T v = turbopump
This
W T p
_-
turbopump
WT
=
turbine
equivalent
weight
equivalent weight,
weight
can
weight,
Ibm
Ibm
flowrate,
be reduced
lbm/sec
by
decreasing
either
the
turbopump
weight
or the
turbine turbine
flowrate. Turbine inlet temperature,
flowrate, in turn, is decreased by increasing turbopump efficiency, and turbine pressure ratio. For the staged-combustion cycle,
turbine
inlet
and turbopump
temperature
efficiency
have
a direct
effect
upon
engine
weight
and delivered payload. High turbine and pump efficiencies will reduce the required pump discharge pressure and reduce the engine system weight. High turbine inlet temperature (within the capabilities of the turbine materials) also will decrease the required pump discharge pressure and reduce engine weight.
2.1.2.3
ROTATIONAL
SPEED
Design speed has more influence on turbopump parameter. It influences all the turbopump design selection of the turbine, the pump, for speed that is the best compromise complex decision facing a turbopump
system design than any other single requirements and in addition affects the
and the turbopump configuration. among all the important factors system designer.
34
Selecting is perhaps
the value the most
To select the speed,its effect on eachof the designrequirementsmust be known. These various effects can then be weighed, and the decisioncan be made.The effects must be consideredcarefully because,if the speedis too low, the weight will be too high, and the turbopump efficiency and consequentlythe enginespecificimpulse will be too low. As a result, the vehiclemight not meet the payloadrequirements.On the other hand,if the speed is too high, the resultingturbopump may not meet the reliability and life requirements.This failure would result in a costly and time-consumingdevelopmenteffort to modify the hardwareand then provethat the modification waseffective. In the process of increasing rotational speed to improve vehicle performance, several mechanicaland hydrodynamic speedlimitations may be encountered(sec.2.2.1). Figures8 and 9 illustrate these limits for a given propellant as a function of pump flowrate and headrise.As shown in figure 16, the turbopump configuration must be modified in order to exceedeach of thesespeedlimits and achievesuccessfuloperation at a higher speed.For example,if the speedis limited to 13000 rpm, the turbopump will not require any special components,will weigh 800 lbm, and will have a pump efficiency of 67 percent.If the speed is increasedto 27 000 rpm, the weight decreasesto 500 lbm and the efficiency increasesto the maximum value of 77 percent. However,this designrequiresa preinducer and an outboard turbine bearingthat, in turn, requiresanadditional sealandprovisionsfor cooling; it might be better to designat 18000 rpm and settle for 600 lbm and74 percent. On the other hand, if weight is critical, it may be desirableto designat speedsabovethe maximum-efficiencyvalue of 27 000 rpm. The upper limit is the speedat which the benefit of the decreasingweight is balancedby the penalty Of the decreasingefficiency or, in other words,it is the speedat the optimum payload(sec.2.1.2.2). In some cases,a turbopump with a different type of component is also a potential candidate.For example, a two-stagepump for the application on figure 16 would have a different stage specific speed and, therefore, from figure 15, would have an efficiency characteristicdifferent from that in figure 16.Consequently,a thorough speedoptimization often requiresanalysisof severalturbopump configurations,eachwith its own pump type and turbine type.
2.1.2.4 The
TURBOPUIVlP
term
turbines
"turbopump for both
the
ARRANGEMIENT arrangement" fuel
refers
and the oxidizer
to the (fig.
physical
17). The
relation
arrangement
of the pumps has a strong
and
the
effect
on
vehicle payload because it influences the weight and the speeds (and, consequently, the efficiencies) at which the various components can operate. However, it is rather difficult to generalize on the best arrangement because, as shown in figure 17; there are many options. In addition, and current
the selection is influenced technology. For example,
by a number of factors, including prior experience a gear drive might have been necessary 10 years ago
35
Speed
limiting
_/'/_//////_/'_/_/
r////////Pre_'n'ducer Pump
= centrifugal
Stages
= I
Flowrate
= 4200
Headrlse
= II
////i
factors
Outboard
t.urbin_
_e:a_i_n_
i////_,
components I
required Special
}
stress Turbine limits
///High- PR////
gpm )0 ft
100
O
80-
U .,4 U
60-
/
40 O
20-
1000
M ,-4
\
80O
d ,z= O_ .e4 ¢)
600
400 0
200
0
1
I
1
10
20
30 Design
Figure
16. -
Effects
of rotational
a LOX
turbopump.
40 rotational
I
I
50
60
speed,
speed on required
36
70
80xl 03
rpm
configuration
for
_
Single
(a) Pumps
back
Dual
Geared
shaft
to back
(b)
Pancake
(c)
shaft
Turbines
in
series
_3
(e (d) Turbine
between
.ne_
pumps
SYMBOLS
--VTurbine
Pump (f)
Single
geared
(g) Turbines
pump
Gears
Figure 17. - Schematics
of basic turbopump
arrangements.
in parallel
because of a lack technology would
of technology for dealing permit an entirely different
If the propellant 17(a) and 17(d)) only one turbine.
densities are similar offers the advantages Within this category,
as in the F-1 engine In some
cases,
the
a disproportionately
large
between
arrangement
diameter
the same shaft and the turbine (fig. 17(e)). This arrangement
permits
optimum
(fig.
in a turbine
In this event,
17(d))
more
advanced
that
as in the A-7 engine.
has very
the two pumps
short
blades
or
may be placed
on
height and diameter. The MA-5 booster (fig. 4) engine utilize turbopumps of this type. The
(fig. 5) engine and the YLR81-BA-11 (fig. 17(b)) in which each pump is geared
configuration components.
today
on a separate shaft that is geared to the pump shaft the turbine to operate at a higher rotational speed,
thereby alleviating the problems of blade and sustainer engines and the YLR-91-AJ-7 YLR87-AJ-7 arrangement
the pumps
results
or both.
placed permits
whereas
(e.g., LOX/RP-1),the single-shaft arrangement (figs. of less complexity and less weight because it requires the turbine may be mounted on one end (fig. 17(a))
(fig. 6) or mounted single-shaft
with cavitation, selection.
rotational
speeds
engine separately for
each
utilize the "pancake" to a single turbine. This of
the
three
turbopump
For small hydrogen-fueled upper-stage engines, the single geared turbopump (fig. 17(f)) can be used to avoid the efficiency penalty of having too small an arc of admission in the oxidizer In
the
turbine.
The
dual-shaft
optimum particularly
speed true
RL10
engine
arrangement and, for
utilizes
(figs.
this arrangement.
17(c)
therefore, the hydrogen/oxygen
and
17(g)),
each
pump
can
overall pump efficiency will engines where the propellant
be designed be higher. densities
at its This is are very
different. Within this dual-shaft category are two turbine drive arrangements: series turbines (fig. 17(c)), and parallel turbines (fig. 17(g)). Series turbines permit the initial turbopump to rotate faster and reduce the turbine flowrate requirements by permitting a larger overall turbine
pressure
The dual-shaft such as starting, the problem development.
ratio.
The
J-2 and J-2S engines
utilize
this arrangement.
parallel-turbine arrangement has great flexibility throttling, and making mixture-ratio excursions; of separating the liquid This arrangement is also
propellants and efficient because
for off-design operation in addition, it eliminates
simplifies advances
turbopump in bearing,
system seal, and
cavitation technology in combination with increases in headrise requirements have made it possible to run the pump at the same speed as the turbine. The experimental X-8 engine (ref. 19) utilized parallel turbines, and this arrangement is used in the SSME now under development.
38
2.1.2.5
PUMP CONFIGURATION
A rocket lightweight
engine and
pump receives propellant delivers the propellant
at low pressure at high pressure
so that propellant so that it can
tanks can be flow into the
high-pressure combustion chamber. Pumps are divided into two basic categories: [1) positive-displacement pumps, in which fluid is forced into a high-pressure region by reducing the volume of a chamber that is momentarily sealed off from a low-pressure region, and (2) nonpositive-displacement
pumps,
and diffusing the kinetic nonpositive-displacement generally the axial
in which
the
fluid
pumps rocket
is raised
by alternately
energy. As shown in figure 18, the design specific pumps (axial, centrifugal, Barske, Tesla, Pitot, and
higher than those for positive-displacement and centrifugal pumps having the highest
rocket engine (eq. (1))for
pressure
have been either engines are high.
adding speeds drag)*
for are
pumps (Rootes, vane, and piston)*, values. Up to the present time, almost all
axial or centrifugal The high speeds
because the pump specific speeds are necessary because of the high
I00 _Inducer
(_-
0.3)
8O
g
O
/
,7-,
/
/
,,.o.,,
4C
2O
Tesla
l
I ] til[ll
2
S
I 10
i i llZJtl
20
SO
I I00
Stage
---o
I i Itlilt
200
SO0
specific
speed,
i 1000
I I litlli
2000
5000
I 10
000
20
I I
000
Ns
Figure 18. - Efficiency vs stage specific speed for various types of pumps.
*Pumps these
named pumps
are is given
described in
reference
briefly
in
Appendix
B;
sketches
of
20.
39
the
pumps
appear
on
figure
20.
Detailed
information
on
flowrate requirementsand the high rotational speedsrequired to minimize turbopump weight. Pumpsother than axial or centrifugalbecomemore competitivewhen flowratesare of the order required for enginethrust levelslessthan 5000 lbf. Sincealmostall pump-fed rocket engineshave been larger than this (table I), the pump selection for enginethrust greaterthan 5000 lbf will be discussedfirst. For large engines(F > 5000 lbf), centrifugalpumpshavebeenusedfor all propellantsother than
hydrogen
able
to meet
because the head
they
are simple
requirements
and
flexible
efficiently
and, in relatively
in a single
stage.
dense
However,
propellants,
are
the low density
of
hydrogen requires overall pump headrise so high that multistaging is necessary to obtain a stage specific speed sufficient for high efficiency (fig. 15). For this reason, the RL10 engine uses a two-stage centrifugal hydrogen pump and the J-2 engine uses a seven-stage (plus one inducer
stage)
that, for single-stage three-stage order In
axial
both
the efficiency
preliminary
hydrogen, the best coefficient
Another
reason
for multistaging
hydrogen
pumps
design
of
requirement a single-stage
and
the tip-speed
centrifugal
pump
is for A in
limitation. for
propellants
other
than
the inducer inlet tip flow coefficient _bit 1 must be selected such that it provides compromise between efficiency and NPSH requirements. Increasing the flow increases the efficiency (fig. 15) but decreases the suction specific speed (sec.
2.2.1.1) and therefore the added complexity operation (above NPSH penalties,
increases the NPSH requirement. This problem of a preinducer (boost pump) can be tolerated.
can be eliminated if To permit high-speed
the peak-efficiency specific speed, fig. 15) without suffering efficiency or a double-entry impeller can be used. Impellers in liquid-oxygen pumps
usually are shrouded avoid rubbing against For and
pump.
pump discharge pressures greater than 2000 to 2500 psi, the tip speeds pumps would exceed the allowable stress limits for existing materials. pump has been designed for the SSME high-pressure-hydrogen application
to meet
the
hydrogen
in order to obtain the housing.
hydrogen, the preliminary the number of stages and
high efficiency
design involves selecting resolving the axial versus
with the large
the inducer centrifugal
clearances
required
to
inlet flow coefficient question. For a pump
specific speed (based on pump head rather than stage head) of 1000, the effects of these variables on engine throttling ratio and pump efficiency are illustrated in figure 19. If throttling to low thrust levels (high throttling ratio) is required, a centrifugal pump is the primary candidate because axials cannot be throttled much beyond 2 to 1. As discussed above, the more efficient high-flow-coefficient designs require more NPSH. The use of inducers with low inlet flow coefficient has less effect on the efficiency of axial pumps than on that axial
of centrifugal
pumps
Another
have more
factor
approximate 2000 ft/sec
that
pumps
because
than
one stage.
influences
the inducer
hydrogen-pump
tip-speed limits are 2800 ft/sec for shrouded titanium centrifugal
flow coefficient
design
is the
affects
impeller
only one stage
tip-speed
limit.
and
The
for unshrouded titanium centrifugal impellers, impellers (1700 to 2300 ft/sec, depending on
40
1oo o_ .
J _
(engine line between design point and origin
_ I __/
_Itl O.1S
0 IS O*
_
IOI_
assumed
to
be
!.
linear)
"
0.05
Axial
--_
C_triNgal i
4_
1
3
5
7
9
3
tl Number of stages inducer stage for
(includes
S
7
9
U
axial)
Figure 19. - Estimated effect of pump configuration on efficiency and engine throttling for a pump specific speed of 1000.
design
specific
attachment,
speed,
hole
in the
amount
of
center,
etc.),
blade and
sweepback, 1500
These tip-speed limits and parametric analyses the basis for hydrogen-pump selection. For (approximately
7000
psi in liquid
ft/sec
blade for titanium
height, axial
method rotors
of (sec.
shroud 2.2.2.3).
similar to those illustrated in figure 19 form example, assume a headrise of 200 000 ft
hydrogen),
an overall
pump
specific
speed
of 1000
to
correspond to figure 19 (which means a constant rotational speed set by one of the rotational speed limits), a pump inlet flow coefficient of 0.15, and a centrifugal pump to meet a wide throttling range (> 2); then 1-, 2-, and 3-stage pumps would have tip speeds of 3300, 2300, and 1900 ft/sec and efficiencies of 78, 84, and 86 percent, respectively. A comparison with the impeller because the tip speed would impeller. three-stage
limits indicates that a single-stage pump would be impossible far exceed the tip-speed limit for an unshrouded titanium
The tradeoff then becomes one shrouded pump, The unshrouded
simple crossover inlet. However,
tubes to carry the it would be heavier
between a two-stage unshrouded pump and a two-stage machine would have fewer parts and
flow from the first-stage discharge to the second-stage due to the larger diameter and to the heavier housing
required to minimize housing deflections that would cause excessive leakage losses at the impeller blade tip; reference 21 shows the effect of blade tip clearance on the efficiency of an unshrouded centrifugal pump. The shrouded three-stage machine would be more efficient and easier to balance for axial thrust. However, the three stages would make a nose-to-back arrangement between the probably capabilities.
necessary discharge
would
be
Analytical
to avoid sealing problems and to avoid extremely of one stage and the inlet of the next. In this based
on
studies
engine-system for
the
SSME
41
performance high-pressure
analyses
complex case, the and
liquid-hydrogen
ducting selection
fabrication pump
demonstratedthat the three-stageconfiguration had an efficiency advantageof 6 to 7 percentagepoints over the two-stage,becausethe pump specific speedswereabout half of thoseshown in figure 19.This madethe selectionof a three-stagemoreobviousthan that in the aboveillustrative examplewherethe advantagewasonly 2 percentagepoints. For low-flow applications(F < 5000 lbf), the pump clearances,tolerances,andflow-passage heights become smaller, and manufacturing becomesmore difficult. To stay within manufacturinglimits, the rotational speedmust be held essentiallyconstant as flowrate is decreasedbelow a given minimum level. As a result, the pump design specific speeds decreaseand low-specific-speeddesignssuch as Barske, Pitot, vane, and piston pumps become more attractive. The regionsof application for some of these low-specific-speed candidatesare shownin figure 18. Also, to obtain reasonablebladeheightsandefficiencies for the turbines, partial-admissionor geared turbines or both becomenecessaryat low flowrates. These turbopump configurations are lessefficient than higher flowrate designs, andconsequentlythe turbine flowrate penalty is.greaterfor smallengines. The subjectof low-flowrate limits for pump-fedsystemsis well exploredfor rotating pumps (ref. 13). The pumping action of the rotor and the lossesin the pump flow path are interrelated with the similarity parametersNs (defined in equation (1)) and Ds, specific diameter,definedby DH¼
(4)
D s -
where Ds. = sPecific D = impeller
diameter,
fts/4/gpm
diameter,
1/2
ft
Thus, the pump efficiency becomes a unique function of Ns and D_ ; its maximum value for given N_ and D_ can be determined by differentiating the equations for the interrelations between losses and geometry and solving for the geometry that yields the minimum losses for
given
operating
conditions
(Ns
D_ values).
This
procedure
is bes_t performed
by
high-speed digital computers (ref. 13). The resulting data can then be presented in the form of Ns - D_ diagrams (fig. 20) in which lines of constant efficiency and lines of constant geometrical parameters (representing the pump geometry) are plotted as functions of specific speed and specific diameter. The validity of the calculated data depends only upon the accuracy of the functions used to interrelate losses with pump geometry. The data presented for high Ns values at the right-hand side of figure 20 show that axial pumps attain highest Dhl/Dtl
efficiencies at high specific speeds and that decreasing values are desired for increasing specific speeds. Centrifugal
medium-specific-speed
regime
efficiently;
increasing
42
values
of the
rotor
of the hub ratio pumps cover diameter
ratio
u = the 6-
(_[
"_aJ) sduand
_o spu!_
sno!JeA Jo_ uJeJ6e!p
s(] -- sN -- "OZ aJn6!-4
sN 000
OOZ
000
001
I
000
OZ
I
000
0I
I
000Z
I
O001
OOZ
I
I
OOI
I
OZ
I
OI
I
tO'O I_n_=_uoD
Dtl/Dt2 are partial-emission
desired with pumps and
pump type, the Tesla pump, speed regime. In single-stage and is directly interrelated application The
of rotating
wear-ring
are scaled are likely
pumps
centrifugal
is indicated
by the dashed
can become
TURBINE
The
turbine
the
fluid
(open
the latter
regime, Another
offers a fair efficiency potential in the medium-to-low specific rotating pumps, the head coefficient has certain limiting v_tlues with specific speed and specific diameter. The limit for the
and tip clearances
Tesla (although
pressure. drive the
speeds. In the low-specific-speed the highest efficiency potential.
or shrouded
two
types
line in figure
disproportionately
down in size. Thus, losses resulting from to be large for small pumps. Pump
conventional
2.1.2.6
decreasing specific Pitot pumps offer
high as conventional
wear-ring and blade-tip types that share this
impeller
do not have
20.
or partial
emission),
the wear-ring-clearance
pumps
clearance problem Barske,
leakage are the drag,
and
problem).
CONFIGURATION
receives
working
to mechanical
fluid
at high
energy,
The mechanical energy pump. The turbine must
and
temperature
exhausts
the
and pressure, spent
fluid
is delivered to the turbopump remove the maximum amount
converts
at lower
the energy
in
temperature
and
shaft where it is used to Of energy from each pound
of working fluid; thus for GG and tapoff drive cycles the turbine flowrate is a minimum, and for topping and expander drive cycles the turbine pressure ratio is a minimum. At the same
time,
the
turbine
must
not
impose
an unacceptable
weight
penalty
on the
turbopump
system. As shown being
in table
two-stage
III, nearly
all rocket
engine
pressure-compounded
turbines
for
have had
low-energy
two axial
propellants
rotors, (fuels
the designs other
hydrogen), two-row velocity-compounded for high-energy propellants (hydrogen GG and tapoff cycles, and two-stage reaction for expander and staged-combustion Exceptions to these rules are the A-7 and the F-1 turbines, which have two-row compounded low. Another an expander Turbine
turbines
propellants
because
the turbine
pitchline
is a function
of the
turbine
pitchline velocity U to the isentropic shown in figure 21. Note that these
because include
they are for subsonic relative clearance losses. Since radial
turbine
U/Co's
are
generally
21(a)) for rocket engine same U/Co, it is heavier.
less than
applications. Therefore,
type,
the
number
(fig.
22),
axial
Another problem with rocket engines have used
44
are
turbine turbine.
and the ratio
in
of
spouting velocity Co. These design-point curves are for illustrative purposes only,
Mach numbers (Mach turbines are difficult 0.4
of stages,
fuel) in cycles. velocity
velocities
exception is the RL10, which has a two-stage pressure,compounded cycle because the turbine pitchline velocity is too low for a reaction
efficiency
the turbine trends are
for low-energy
than
number losses = 0) and do not to multistage and since rocket turbines
are more
efficient
the radial turbine is that, axial turbines exclusively.
(fig. at the
100 50%
reaction
50%
reaction
80 y
Ires sure compounded
60
(zero
reaction)
40
20
/_
t m'bine
//_
0
I
,
I
'
I
,
I_
I
,
I
,
I
,
I
[-,(a)
-j o
Single
stage
(b)
.°.F
100
50%
Two
stage
(two
row
for
VC)
reaction
/_ction Pressure
60
_
/y_
Velocity
.
compounded
/ v.,ooit f
/!
0
0.2
(C)
Three
0.4
stage
0.6
(three
row
compounded
,
l 0.2
,
(d)
Four
stage
0.8
for
Turbine
VC)
velocity
ratio,
I 0.4
J
(four
,
I 0.6
row
for
I 0.8
VC)
U/C °
Figure 21. --Typical design point efficiencies for 1-, 2-, 3-, and 4-stage turbines (supersonic relative Mach number).
To maximize the energy available per pound GG and tapoff cycles, pressure ratio generally
of flow, the turbine inlet temperature are maximized. Since this procedure
and, for produces
a high turbine isentropic spouting velocity Co and since a high U/Co is required to obtain a high efficiency (fig. 21), turbine pitchline velocities are generally maximized. However, because of centrifugal stress, these velocities are limited to a range of 1500 to 1800 ft/sec. Additional limitations result from the fact that, to obtain proper proportions for the turbopump, the pitch diameter is limited to approximately diameter and, to avoid excessive leakage losses, the approximately 0.15 in. Two-stage (two-row for velocity compounded) turbines offer higher efficiency than single-stage turbines (cf.
45
three times the pump impeller lower limit on blade height is
generally figs. 21(a)
are selected and 21(b))
because without
they the
1.6
1.4
1.2
O
.5 r-,
0.8
°_
0.6
U,
ft/sec
0.4 1500
1000
I O2/RP-I
1000 02/H 1500
2I_______
o
lo I
Turb£ne
4;
I
pressure
ratio,
(Po)T1/PT2
Figure 22. - Typical turbine velocity ratios as a function of pressure ratio (inlet temperature of 1500 ° F).
added
complexity
and
weight
of turbines
with
three
or more
stages
(three
compounded). As shown in figure 21(b), two-row velocity-compounded as efficient or nearly as efficient as two-stage pressure-compounded ratio is less than 0.2. In addition, velocity-compounded turbines
rows
because they have a turning vane, rather than a nozzle, between the rotors. nearly zero) pressure drop across the turning vane also results in a lower axial velocity,compounded design. Therefore, are used for rocket engines with turbine between 0.2 and 0.34, two-stage pressure
is little
or no pressure
drop
across
The zero (or thrust for the
two-row velocity-compounded turbines generally velocity ratios less than 0.2. For velocity ratios compounded (impulse staging) turbines generally
are used; they are more efficient up to 0.3 (fig. 21(b)) efficient between 0.3 and 0,34, they have less axial thrust there
for velocity
turbines are either turbines if the velocity are shorter and higher
the
rotor.
46
Above
and, than velocity
even though slightly less reaction turbines because ratios
of 0.34,
reaction
TURBINE
HARDWARE
SYMBOL
o
2
row,
compounded
velocity
Pressure Reaction
compounded
Velocity
I
3
2 Number
compounded
of
rotor
4
blade
rows
Figure 23. --Approximate regions of application for various kinds of rocket engine turbines.
turbines generally are used because, these ranges agree reasonably well
as shown, they are more with existing hardware.
efficient. It should
Figure23 be noted
shows that that these
ranges are not efficient than
absolute. For example, for two stages, a 25-percent-reaction turbine is more either a pressure-compounded or a 50-percent-reaction turbine at a velocity
ratio
therefore,
thrust
of
0.31; relatively
less
it would important.
be selected On
the
other
47
if efficiency hand,
were
a two-stage
very
important
and
pressure-compounded
axial
turbine could be selectedat a velocity ratio of 0.36 if axial thrust were relatively more important than efficiency.This overlapis shownby the designdatapoints on figure23. Turbine efficiency can be increasedby using more than two stages.However,as shown in, figure 21, the rate of increasedecreasesas stagesareadded.For example,at a U/Co of 0.3, typical 1-, 2-, 3-, and 4-stageturbines (50-percentreaction) haveefficiencies(for relative Mach numberslessthan 1.0) of 68, 80, 85, and88 percent;respectively.In addition, these multistage (more than two stages)turbines are more complex, are heavier, and require outboard bearingsthat, in turn, require an extra sealand specialprovisionsfor cooling and lubricating. Figure 23 alsoshowsthe approximateregionsof application for theseturbines. An additional factor consideredin turbine designis that, for a given rotational speed,a diameter tradeoff between weight and efficiency exists. In addition, a tradeoff among turbopump weight, turbine pressureratio, and turbine inlet temperatureexistsif the speed is limited by turbine-bladecentrifugalstress.Theseandother tradeoffsarediscussedin more detail in sections2.2.1.4and 2.2.3.1. For low-horsepowerapplications, partial admissionor gearedturbines often are usedto obtain bladesthat are high enough to avoid excessiveclearancelosses.Someearly RL10 turbines were partial admission and the YLR81-BA-11 turbopump uses both partial admission and gearing. Geared turbines also eliminatesome of the controls that are necessaryfor dual-shaftarrangements.This reduction is desirablefor smallengines,because control componentweightsgenerally do not scaledown to smallsizeaswell asturbopumps do. For staged-combustioncycles, reductions in system pressuredrops and interconnection complexity may be achievedby reversingthe turbine so that the high-pressurestageis outside of the low-pressurestage.The hot gas from the precombustion can then enter through an annularaxial inlet and,with the aid of an annular180-degreeelbow downstream of the low-pressurestage,the turbine dischargegascan flow back to the maincombustion chamberthrough anannularaxial discharge. 2.2
DETAIL
2.2.1
2.2.1.1 The pump
DESIGN
AND
Limits to Rotational
INDUCER
inducer impeller
is the
INTEGRATION Speed
CAVITATION low-pressure
on (usually)
pumping
the main
element
turbopump
48
shaft.
located
immediately
Its purpose
upstream
is to permit
of the
the pump
to
operate
at a lower
inlet
NPSH
(sec.
2.1.1.2)
or, for a given
pump
inlet NPSH,
to operate
at
higher speed. If the inducer is operated above its suction specific speed limit, excessive cavitation will occur and the pump will not deliver the design headrise. This limit to inducer suction specific speed encountered in design.
generally
Suction
Ss is a useful
speed,
specific
speed
flowrate,
and
is the
net positive
first
and
suction
(lowest)
limit
significant
design
to turbopump
parameter
rotational
that
relates
speed
rotational
head: NQ v2
ss =
(NPSH)
(5) _
where Ss = suction NPSH Corrected
specific
= net positive suction
specific
speed,
suction speed
with zero inlet hub diameter speed, rotational speed, and made by numerically at the inlet:
rpm-gpm
head,
'_/ft _
ft
S's is the
suction
specific
speed
of a hypothetical
inducer
that operates with the same inlet fluid axial velocity, inlet minimum required NPSH as the test inducer. The correction
increasing
the flowrate
to compensate
for the area blocked
tip is
by the hub
Ss
S's -
(6) (1 - v2) _
where inlet hub v = inlet
diam.
tip diam.
Dh Dt 1
Selection of inducer design is based on the suction performance and efficiency at the design point (figs. 24 and 25, resp.). The inducer design-point suction-performance correlations in figure inducer
24 were obtained by drawing curves designs that were tested in hydrogen,
through oxygen,
the peaks of the test curves for 20 and water; these basic test curves are
shown in figure 26 (adapted from ref. 2). The efficiency correlation in figure 25 was obtained by drawing curves through the experimental best-efficiency points of 20 inducer designs, some of which were the same as those listed in figure 26. The
allowable
suction
at a lower inlet pumps presented this
action
specific
speed
of an inducer
flow coefficient, as shown in figure 24. (Also shown
reduces
the inducer
efficiency
can be increased
by designing
by the correlation of test is that the data peak at q_It (fig.
49
25)
and,
particularly
data
the inducer on high-speed However,
1 "_ 0.05.)
at high
pump
specific
-----------200
correlation deviation from
correlation
X 103 !_,_
Range
Test
-Ji NPSH
[
Ss '
Propellsnt
Cm2/2g "--'___
I
hydr°gen
8160/_Ttl
I
OXY' en
4850/',t I
2(cm212g)
0 .,-I
-
_
-
100 --
0 o1-1 4J
3(Cm2/2g)(.O721/_Itl)
I
u_ _:1
2,
II
50-
U ¢) I-I 0 CJ
I
2O 0.03
I 0.05
Design
I
I
inlet
I
tip
I
I 0.I0
flow
I 0.20
coefficient,
_Itl
Figure 24. -- Influence of inducer design inlet flow coefficient cavitation performance.
on
100 Q) .,.4 fJ .PI q4 q_ U
-_Itl
h£dS-_'_S
(high
)
_
_O@itl
-
0.1
G)
50
Cflat plate> m
0
30 0.03
I
Design
Figure 25. -
I 0.05
I
inlet
I
1
I
I 0.10
tip flow
coefficient,
I 0.20 _Itl
Influence of inducer design inlet flow coefficient inducer hydraulic efficiency.
50
on
4t9
0 co _>
;'.7
E 0
O_
0
¢L
J
°
iI
il iIII
--
--_¢
._-
-o
=.>o
¢_
-
=.
x
•
•
•
.....
r
.............. - _o. _ ._'_"_= .......
N
o_ °_,.,,, _
/
,,
0
//N
. u
I1_ 4.J
_o -0
O_ v
D-
O_ CJ
.c_ O
_-_-_
_
._-
J_':,'TZ/" w._'_
U
E E E bO
I
--"
t'N
P 03 LL
I
1
I _/l_(_a-l)sS
I
I
I = s)S
I _paads
I
I
o!._!oads
I
I
uo!_,ons
1
I
pal:)aJJO::)
"0 cco
CL
speeds, relation.
the overall pump efficiency The original pump inducer,
(fig. which
15). The J-2 hydrogen pump had an inlet diameter of 7.25
illustrates this in. and an inlet
fl0w coefficient of 0.095, did not meet the NPSH requirements. The inducer was then redesigned for an inlet diameter of 7.8 in., the result being a reduction of the inlet flow coefficient to 0.073 and an increase in the suction specific speed capability (fig. 24). This change
decreased
the pump optimized
the
NPSH
requirement
efficiency. Therefore, in terms of turbopump
suction
specific
become
an important
speeds
are often design
(TSH). suction
derated
to reduce
cavitation
limit of suction performance, thereby reducing the vapor
This effect permits the The difference between
effects and is approximately above-mentioned reduced
level,
but
damage,
and
pump to operate the basic NPSH,
equal to the value of NPSH
vaporization pressure and
decreased
therefore
life can
within the inducer increasing the local
satisfactorily at a reduced value of inlet which is independent of thermodynamic
NPSH in water at room temperature, and the is called the thermodynamic suppression head
This parameter varies with propellant type and is the reason performance shown in figure 24; at low inducer tip speeds,
higher suction were correlated
also slightly
factor.
During operation close to the chills the surrounding liquid, NPSH. NPSH.
to an acceptable
for a given application and mission, the configuration, is weight, pump efficiency, and tank weight. For long life,
for the variation TSH would result
in in
specific speeds than those shown. Theoretical relations from reference 22 with data on the J-2 oxidizer pump, J-2 fuel pump, Atlas-sustainer oxidizer
pump and Thor oxidizer pump that include these thermodynamic
to produce the following empirical expressions for NPSH effects (all parameters referred to inducer inlet):
114.7 tc NP-SH m 2/2g
/
_ 0.931
/
_)t 4/9
(Ot/Z)
0"16
/_hydrogen
(7) _t 2
Ut 115
hydrogen
6.35
(Dt/Z)
0"16
/3 other
(8) _NPSH) m2/2g
_ 0.931 other
_t 2 Ut
_t4/9
where cm
=
fluid axial velocity
g = acceleration
at inducer
due to gravity,
q_t = inducer
inlet
tip flow
Dt = inducer
inlet
tip diameter,
inlet, ft/sec 2
coefficient in.
52
ft/sec
1"15
Z = numberof inducerbladesat inlet /3 = thermodynamic ut = inducer
The
first
term
suppression
inlet
on
the
tip speed,
right
side
head
factor
(fig. 27)
ft/sec
in each
of these
equations
(room-temperature water) to the inlet fluid axial velocity entering fluid expressed in feet of head) and is derived from correlation (fig. 24) and the suction specific speed equatiQn each head
equation is the ratio of the thermodynamic of the pumped fluid. The thermodynamic
propellant vapor pressure and is shown NPSH to inlet velocity head is greater ratio and the inlet axial velocity head. two-phase-flow must inlet
Mach
number
possible, principles of NPSH
ratio
of the
basic
suppression head to the inlet axial velocity factor _3is a function of propellant type and
than
(refs. Mach
1.0,
the
inlet
line is choked,
and
23, 24, and 25). If the ratio is less than number is less than 1.0, two-phase
the NPSH can be less than the inlet velocity apply. (These principles are discussed at the end to inlet velocity head is greater than 1, equations
diameter
and
blade
is contained
number
are slight.
in references
1 and
Additional
discussion
the
NPSH
1,0 and the pumping is
head, and the two-phase-flow of this section.) When the ratio (7) and (8) show that the pump
NPSH requirements may be adjusted by changing tip speed and vapor pressure inlet flow coefficient. Because of the small exponent on the (D/Z) term, the head
NPSI_
head (kinetic energy of the the water suction-performance (eq. (5)). The second term in
for various propellants on figure 27. If the ratio of than 1.0, the NPSH is equal to the product of the If the ratio is less than 1.0 but the inlet equilibrium
is greater
equal the inlet velocity head equilibrium two-phase-flow
is the
of thermodynamic
as well as effects of suppression
22.
The inducer speed limit due to cavitation preinducer upstream of the main inducer
can be essentially removed by placing to provide the required inlet pressure
a low-speed to the main
inducer. A preinducer can be driven by several methods, four of which are illustrated schematically in figure 28. The gear drive (fig. 28(a)) is the most positive and efficient but, because of the gear train, may be the most complex and least reliable. The through-flow hydraulic
turbine
drive
(fig.
28(b))
is both
simple
hydraulic turbine drive (fig. 28(c)) can be installed the pump inlet, whereas the gear and the full-flow efficient than the other two drive systems. The installed at a location remote from the pump
and
efficient.
The
recirculated-flow
in the engine at a location remote from drives cannot; however, this drive is less electric drive (fig. 28(d)) can also be inlet. The electric motors often are
propellant-cooled to save weight and reduce the number of rotating seals. This drive is attractive for low-thrust applications. Another configuration that can be installed at a remote location is the gas-turbine drive, the type used for the RL10 engine in the Centaur i!/.
stage.
!!;i!
each is gear driven by its own hydrogen-peroxide gas turbine. The gasrturbine boost pump system is complex and usually requires either a separate gas generator or a catalyst bed, but has the advantage (as used on the Centaur) that it can be started before the main pumps.
The
Centaur
boost
pumps
(fuel
and oxidizer)
t
53
are located
in the propellant
tanks,
and
i00
I
2 gJ ( A hv) = _hydrogen
PL
Pv _other
o
=PL
k_
| (adapted from ref. 22)
_h v (Cp)L
(Thor
data
correlation)
I0
o
,= o .,-_ ul u_
k
HYDROGEN (LH,
u
N204 o
1.0
METHANE (CH
WATER
(H20) (LF 2)
(CH6N 2 OXYGEN
(h02)
I i0
0.i I
Propellant
vapor
Figure 27. --Thermodynamic propellants
i00
pressure,
suppression
as a function
54
Pv' psia
head factor for various
of vapor pressure.
___Preinducer
/-
Gear
drive Hydraulic
Inlet-_
i
__L-_"_ turbine__ _t
____-_ {
///
¢-
!_
_=__j/
___L_.A. 4_I_= (a)
Gear
"°w
shaft
(b)
drive
..... .ydrau_ic
ta
_Discharge
//# l
Inlet flow
-"
_
Throughflow
Electric
motor
hydraulic
turbine
stator-
7
drive
Elec.%ric r°tor r
--
/
Inle&
.,o. =>r v
/__i
II Irr (c)
// _Discharge
//___A_---_
Recirculated-flow
flow
__.____=--=_-_
_ _hydraulic
Figure
L_ turbine
28.
"_I-L
Inducer
(d)
drive
-- Schematics
of
four
methods
for
driving
an
____LS__ v_[J _
Electric
inducer.
motor
drive
motor
Recent developments (refs. propellants can be pumped
23, and,
24, and therefore,
25) have indicated that two-phase (vapor-liquid) that cryogenic propellants that are saturated in
the tank (zero tank NPSH) can be pumped. from figure 29 or from the expression
Vapor-pumping
o_= 1-
1-B
capacity
can
be determined
J
(9)
where o_ = vapor
volume
fraction
at the
(i//3)L = 1 -- arc tan 4_L = ratio (for a pure liquid) _L = flow coefficient B = blockage boundary The
values
for pumping
expansions from for the increased
of incidence
at inducer
fraction layer
inlet,
vapor-volume/mixture-volume
angle
to blade
inlet (for a pure
(normal
capacity
pump
to flow
with
a saturated liquid are used to size size, two-phase inducers are similar
at the inducer
of the inducer
the thermodynamic
tJ
60 4a
c%j
so 40
_'_
o_
20 -
lo -
_o
0 Blade
Figure
":
,0×NNN
4a
29. -
10 plus
Vapor
20
30
40
bound&ry-layer
pumping
capability
56
blade
plus blade
relationships
for
the pump inlets for zero NPSH. Except to other inducers. In oxygen, low-speed
80
f-t
inlet
liquid)
direction)
in combination
angle
50
60
blockage,
70
80
B, percent
for two-phase
inducers.
,
boost pumps accommodate
generally are the extremely
to avoid
choking
2.2.1.2
BEARING
The
second
required in order to obtain the large inlet areas necessary to low acoustic velocity of two-phase oxygen in equilibrium (i.e.,
in the inlet line).
DN
turbopump
speed
bearing. This parameter, (rev/min), is proportional the inner race. If the DN
limit
reached
the product of to the tangential limit is exceeded,
is usually
the
the bearing may fail because of overheating and contact contains the information on bearing capabilities, including engine Lower
applications in which DN values are necessary
in figure
of the
wear and fatigue. Reference 6 the DN limits for typical rocket
to shaft diameter, bearing location.
it becomes a function of everything The most common bearing locations
30.
(a)
Inboard
(c)
Outboard
bearings
(b)
(d)
bearings
Outboard
Outboard inducer
Inducer (e)
Outboard
rolling-contact
the life requirement and the radial and axial loads are low. if long life and high load-carrying capacity are required.
Since bearing DN is proportional affects shaft diameter, including shown
DN limit
bearing bore D (mm) and rotational speed N velocity of the bearing at the inside diameter of lubrication and cooling will be insufficient, and
ptmp
turbine
bearing
bearings
without
_Turbine
bearing l_lmp
B
Bearing
Figure 30. - Schematics of bearingsupport arrangements.
57
that are
The
overhung-turbine
rocket
engine
bearing
turbopumps
arrangement
in which
(fig.
the
pump
30(a))
has been
is mounted
used
almost
on the same
shaft
exclusively
for
as the turbine;
the turbopumps for the J-2, RL10, and F-1 engines (figs. 1, 2, 3, and 6) are examples of this arrangement. This configuration avoids the additional supporting structure and the separate lubrication and sealing systems required for the outboard turbine bearing (figs. 30(b), 3_(c), and 30(d)). For the overhung arrangement, the shaft is often sized so that the turbopump design
speed
is below
the
lowest
simple
shaft
bending.
This
value
sizing
of shaft
critical
is particularly
speed,
important
which if the
in general turbopump
over a wide speed range or, in other words, if the turbopump application The following equation is used to estimate this limiting speed:
is that must
requires
due
to
operate
throttling.
(DN)4 / 3 N -
(10) KDN (HP) 1/3
where DN = bearing HP = shaft
bore
× bearing
horsepower,
KDN = empirically
speed,
mm-
hp
derived
coefficient
for bearings,
Number of pumps on shaft
The final design speed equation (10), because for existing spring rate,
the speed limit pump impeller "educes length, diameter
and, therefore,
Propellant
and and
KDN
t
LH2
325
1
Dense
374
2
Dense
478
because weight
not agree with the value from correlation of the shaft sizes
critical speed is influenced of the components. If this
by factors such as bearing deviation causes difficulty,
such as bearing spring rate, may be adjusted without having to redesign the whole turbopump.
can be obtained by placing (fig. 30(e)) and using the
the overhang increases the
as follows
based on a critical-speed analysis may the equation is based on an empirical
turbopumps shape, size,
some of these factors, critical-speed problem
rpm
the pump bearing between inducer stator to support
(relative to that shown in fig. 30(a)) and, bearing span. Both of these factors would would
increase
the speed
58
limit.
so as to avoid the Some increase in the inducer and the it. This arrangement
for a given turbopump permit a smaller shaft
Figure 8 showsthat the bearing limit for this arrangement(labeled critical speed)differs from the other limits in that turbopump geometricsimilarity (constantspecificspeed)is not maintainedwhen scalingwith flowrate. The reasonfor this is that bearingsandsealsdo not scale as the other turbopump components do. Consequently, turbopumps for low-thrust-level engineshave proportionally longer shafts and more overhangthan db turbopumps for high-thrust-levelenginesand therefore have a more severecritical-speed problem. By placing the turbine bearingoutboard (fig. 30(b)), the shaft diameter can be sizedby torsional stress; since this practice generally produces a smaller shaft diameter, the turbopump therefore can be designedfor a higher speed.The following equationexpresses this speedas a function of the torsional stressthat would occur in a solid shaft equalin diameter to the shaft diameter at the bearing(an equivalent-shaftstressof 25 000 psi has producedgood correlationwith final design): N=
1.378×10-
5
Seq 1/2 (DN) 3/2
(11)
(Hp) 1/2
where Seq = torsional stress.in a solid shaft at the bearing, psi
of the same
outside
diameter
as the pump
shaft
Even higher speeds may be obtained by moving the bearing on the pump end to a location between the inducer and the impeller (fig. 30(c)). This placement minimizes the required bearing
diameter
by
minimizing,
the
torque
transmitted
through
the
boost pump, a main pump inducer may not be needed, because provide sufficient NPSH for the main pump impeller. In this case, placed The
outboard arrangement
options
pads that act leakage losses,
expensive and contamination
bore.
With
a
would can be
(fig. 30(d)). in figure
30
have
thus
far
rolling-contact bearings for nearly all applications. become restrictive, hydrostatic bearings (refs. 26 pressurized minimize
bearing
the boost pump the pump bearing
against these
permitted
sufficient
If rolling-contact and 27), which
latitude
bearings function
to use
eventually by having
the shaft to keep it centered, can be used. ttowever, bearings have very close clearances and therefore
are subject to the various and differences in thermal
problems growth
to are
associated among the
with close clearances (e.g., shaft, the bearing, and the
speed limit that generally occurs at approximately DN limit is the seal rubbing speed. Seal rubbing
the same rotational speed speed is the speed at which
housing).
2.2.1.3
SEAL
A turbopump as the bearing
RUBBING
SPEED
59
the rotating mating ring on the shaft rubs againstthe stationary sealnosepiece.If the seal speedlimit is exceeded,the nosepiecewill weardown too rapidly or the sealmay fail due to overheating.The sealspeedlimits in commonpropellantsmay be found in reference7. The turbopump rotational speedat the sealspeedlimit dependson the shaftdiameter.For a turbopump with an overhung turbine (fig. 30(a)) and a shaft sizedon critical speed,the following equationis often usedto estimatethe limiting rotational speed: Kss(SS)4/3
N=
(12)
(HP)1/ 3
where SS = seal rubbing Kss = empirically
speed, derived
ft/sec coefficient
for seals,
Number of pumps on shaft
as follows
Propellant
SS
LH2
240
Dense
208
Dense
164
As with equation (10), equation (12) is based on an empirical correlation of shaft diameters for various turbopump configurations that have overhung turbines and it does not scale along a constant specific speed line (fig. 8). The seal speed limit, like the bearing DN limit, can be alleviated by utilizing a turbopump configuration with outboard turbine bearings (fig. 30(b)) instead of an overhung-turbine configuration. The limit then is expressed by
N = 4.37
Seq 1/2 (as) 3/2
(13)
(HP) _!2
As before, Seq is the (equivalent) usually 25 000 psi.
torsional
If the turbopump is to be designed configuration in figure 30(b), special
stress
in a solid
shaft
of the same
diameter
and is
at a rotational speed in excess of that allowed by the seals are required. The seal function can be broken into
6O
two parts: static, anddynamic.The liftoff seal,which sealsduring static conditionsandlifts off when pressurizedduring operation, can perform the static function. The dynamic or operatingfunction canthen be performedby a secondsealthat is in serieswith the first and that doesnot haveasseverea speedlimitation asthe rubbing type of seal.This arrangement is being used in the SSME high-pressurehydrogen turbopump in which the operating function is performedby a stepped-labyrinthseal. A type of sealthat performs both functions is the hydrodynamic face seal,which is being developedfor the SSMEhigh-pressureoxygen turbopump. During static conditions,sealing is accomplishedconventionally by having the spring-loadednosepressagainstthe mating ring on the shaft. However,during operation, groovesin another part of the noseassembly trap pump fluid andforce the noseawayfrom the matingring. As a result, the noserideson a film of liquid duringoperation andthereforeis not subjectto the rubbing-sealspeedlimit. More information on the various kinds of seals,including the hydrostatic seal,canbe found in references7, 26, and 28.
2.2.1.4
TURBINE-BLADE
Turbine-blade centrifugal
CENTRIFUGAL
centrifugal force
on the
stress
refers
blades
during
to the
STRESS stress
turbine
at the
operation.
turbine-blade This
stress
roots
caused
is proportional
by the to the
square of the turbine rotational speed times the turbine annulus area (N 2 Aa). When this stress exceeds the stress limit of the turbine blade material, the turbine blades will pull off the turbine rotor. The N 2 Aa limits for candidate materials as a function of temperature may
be found
in reference
4.
The previously mentioned methods for alleviating allow the turbopumps for most propellants to maximize payload. However, for liquid-hydrogen
component speed limits are sufficient operate at rotational speeds needed turbopumps, the rotational speed
to to for
maximum payload generally exceeds the turbine-blade centrifugal-stress limit. Because there is no simple way of alleviating this limit, hydrogen turbopumps usually are designed close to the turbine stress limit; this practice generally results in a speed less than the maximum payload value. The J-2 hydrogen turbopump, cavitation limits, is an example of such a design.
which
also
operates
To optimize the turbopump design at the centrifugal-stress limit, tradeoffs among turbine inlet temperature, turbine pressure ratio, blade materials, turbopump blade materials, turbopump weight, and speed are required. For example, raising the turbine inlet temperature turbine required flowrate but will also decrease the allowable rotational the allowable N 2 Aa. This speed decrease, in turn, will increase the weight.
In
addition,
it will
tend
to decrease
the
61
efficiency
close
to the
inducer
some rather involved turbine type, turbine turbopump rotational will decrease the speed by decreasing turbopump size and
of centrifugal
pumps
and
to
increasethe number of stagesfor axial pumps.The sametype of optimization is necessary to determinethe optimum turbine pressureratio (fig. 31) because,asshown,the maximum speedand consequentlythe minimum turbopump weightboth occur at a low pressureratio, where the dischargeannulus areaAa is minimum, whereasthe minimum turbine flowrate occursat a much higher pressureratio and consequentlyat a larger dischargeannulusarea (assuminga constantinlet pressure).
(T°)TI
o2/n2l
-
Propellants '
2000°R
l
hp - 1o ooo
I
°OO00[
i_
g z5
40 000
30
000
;
2°°°I
i
4a
°
10
S
.2
i0000[
.o
O" 0
• i0
I 30
i 20
I 40
_
0
|
0
(Po)T1/PT2
Figure
31.
-
Typical limit
•
30
40
(Po)T1/PT2
effects of
|
20
I0
of
turbine
turbine-blade
pressure
centrifugal
ratio
on
design
at the
stress.
q
Cooling
the
turbine
blades
speeds above the stress determine the optimum cooling Turbine
with
the propellant
limit for uncooled amount of cooling
being
pumped
blades. A payload and to determine
to allow an increase in turbine cooling is common in jet-engine
inlet temperature turbines and may
turbopumps. One problem with turbine outside of the blade and cold cryogenic
cooling is thermal propellant on the
stress resulting from the severe thermal and therefore can adversely affect life.
gradient
can cause
In some cases when multiple restart is required, even if the blade is not cooled. The temperature blade blades
and with
the hot
cold gas
center ducted
of the to the
blade hollow
makes
may
62
optimization the proper
to operate
at
is necessary to tradeoff between
and cooling to raise the speed. be used in future rocket engine stress inside.
caused by hot gas on the For repeated rapid starts,
cracking
due
cracking due to thermal gradient between the
be sufficient
center
it possible
to cause
can alleviate
the
to thermal
fatigue
fatigue can occur hot surface of the problem.
this problem.
An
Hollow additional
benefit of hollow bladesis that the blade thicknesscanbe taperedto reducethe masswith radial distance from the base without affecting the outer contour of the blade. This techniquecanpermit a higherlimit on centrifugalstresswithout penalizingperformance.
2.2.1.5
GEAR
PITCHLINE
VELOCITY
For geared turbopumps, the limit limits results in an upper limit turbopump.
The
pitch diameter. rubbing velocity, at the
contact
For a given be increased
gear
on gear pitchline velocity in combination on rotational speed for the high-speed
pitchline
velocity
is the
If the upper limit on po0r lubrication (high
points),
and heat
tangential
can cause
rapid
2.2.2
There
on limits
limit, the rotational the limit on tooth
speed at that point is the of the turbopump. Reference
to gear pitchline
teeth
at the
velocity
and gear-tooth
speed bending
can or
upper limit on 5 contains the
stress.
Pump Design
are
several
performance, inducer inlet the maximum
2.2.2.1
pump
design
decreasing stability, flow coefficient, the pump
INDUCER
impeller
INLET
generally
To meet the requirements have helical blades with
mounted
Qit i (ratio
of
flow coefficient pump efficiency
on the
fluid
turbopump
axial velocity
decreases inducer (fig. 19). However,
flow
until
performance of 0.05 are
decreases. The extremely difficult to manufacture
inducer
design
oxidizer
and fuel pumps
cannot
of liquid very gradual
performance
inlet
that
be
exceeded
without
penalizing
are the minimum pump stages, and
FLOW COEFFICIENT
impeller. inducers coefficient
are
limits
or risking failure. Among these limits stability limits for axial and centrifugal
tip speeds.
Inducers
design overall
gear
a combination of high the amount of lubricant
wear or failure.
horsepower transmitted at this pitchline-velocity by decreasing the gear pitch diameter until
information
of the
this velocity is exceeded, centrifugal forces decrease
compressive stress is reached. The rotational rotational speed for the high-speed component design
velocity
with tooth stress component of the
coefficient
flow
value
coefficients
have
inducer
shaft
upstream
of the
to blade
tangential
velocity).
efficiency (fig. 25), which, a decreasing flow coefficient
of 0.05
first-stage
pump
rocket engines for low pump inlet pressure, curvature and low values for the inlet flow
is reached
(fig. 24); below
Decreasing
the
in turn, decreases increases suction that,
the suction
thin blades required to operate at a flow coefficient and are often inadequate structurally. Consequently, are inlet
seldom
less
flow coefficients
63
than
0.07.
of 0.11
For
example,
and 0.073,
the
respectively.
J-2
2.2.2.2
STABI LITY
If an axial-pump stage is operated at a flow coefficient, the pump will stall, thereby causing performance to become poor and unpredictable. the
original
problem
existed
with
problem, addition
it was allowed
necessary to add a seventh stage identical to the original the pump to deliver the same flow at a lower rotational
resulting margin.
higher
operating
Stall may be predicted 29). In the J-2 engine
flow
J-2
coefficient much less than the design flow the developed head to drop (fig. 12) and the In addition, damage can occur. This .stall
hydrogen
coefficient
pump,
unloaded
which
had
the blades,
six stages.
thereby
To solve stages. speed,
increasing
centrifugal
pumps,
operation
in the positive-slope
region
of the head/flow
(figs. 12 and 13) has been known to cause instability during operation Although the effect is not nearly as drastic as it is in axial pumps, that region avoided. Therefore, centrifugal pumps generally are designed such that the coefficient
2.2.2.3
during
This The
the stall
by using a blade loading parameter called the diffusion factor (ref. program, it was found that axial-pump stall occurred when the blade
diffusion factor at the mean flow-passage diameter exceeded 0.70. Therefore, are designed such that during operation the maximum blade diffusion factor diameter is not larger than this value for either the rotors or the stators. For
the
operation
falls to the right
of the zero-slope
point
axial pumps at the mean
characteristic in the engine. of operation is minimum flow
on the head/flow
curve.
TIP SPEED
For a given stage in a centrifugal pump, and the square of the impeller discharge
headrise is proportional tip speed:
to the stage head
coefficient
J(ut:) 2 H =
(14)
g
where = stage head ut2
----
coefficient
stage impeller
g = acceleration
discharge due
tip speed,
to gravity,
ft/sec
ft/sec 2
Because stage head coefficient for a centrifugal pump generally falls between 0.4 and 0.6, the primary component of stage headrise is tip speed, and consequently high-headrise stages require high tip speeds. However, if the tip speed is too high, the centrifugal forces will exceed
the
material
strength
of the
impeller,
and
64
it will burst.
Therefore,
although
high tip
speedsare often desired to minimize the number of stages,pump impellers are always designedwith an adequatetip-speedmargin betweenthe maximumoperatingvalue andthe burst value. For titanium, which has the highest tip-speed capability of the candidate impeller materials, the approximate tip-speed limits are 2800 ft/sec for unshrouded centrifugalpump impellers,2000 ft/sec for shroudedcentrifugalpump impellers,and 1500 ft/sec for inducers and axial pump rotors. (Thesevaluesare difficult to generalizebecause the blade,sweepback,the designspecificspeed,the blade height, and the method of power transmission- all affect the allowabletip speed.)Multistagingis sometimesrequiredto stay within theselimits. It must be noted that hydrogen pumps generally are the only pumps that approachthe tip-speedlimits becausehydrogen with its low density is the only propellant that has headriserequirementsgreater than the capability of a singlecentrifugal impeller. It must alsobe noted that multistagingis usedto obtain a higherstagespecificspeedandthereforea higher efficiency (fig. 15). As a result, somemultistage hydrogen pumpshavetip speeds much less than the limits. For example, the J-2 and the RL10 (the only flight-proven hydrogen pumps) havetip speedsof approximately 900 and 800 ft/sec, respectively;the impellersweremadeout of "K" monel andaluminum,respectively.
2.2.3
Turbine Design
There
are
several
performance, and turbine tradeoffs
design
limits
that
decreasing stability, or risking blade centrifugal stress limits, to
requirements
2.2.3.1
turbine
obtain
optimum
imposed
PERFORMANCE
Among form
are
conducted,
the interrelated
moment
start
without
penalizing
these limits are the rotor disk the boundaries within which and
the
exhaust
pressure
OPTIMIZATION
effects
of inertia,
exceeded
configuration.
Tip speed is a function of rotor disk diameter and the centrifugal stress on the rotor will exceed the fail. Therefore, turbines are optimized within the best combination of turbine flowrate, turbopump treat
be
failure. which
performance
by the engine
cannot
of tip speed,
time,
inlet
rotational speed. If tip speed is too high, material strength of the rotor, and it will rotor disk stress limitations to obtain the weight, and start time. The optimizations
tip clearance,
temperature,
and
efficiency, material
flowrate,
weight,
strength.These
rotating
interrelations
may be illustrated by considering the effects of an increase in tip speed for a stress-limited candidate with constant rotational speed. In this case, an increase in the tip speed (1) increases
the
and
therefore
the
diagram
turbine the
diameter
turbopump
efficiency
(ref.
and therefore start 4);
time; (4)
the weight;
(3)increases
decreases
65
the
(2) increases the velocity
blade
height,
the moment ratio
U/Co
which,
for
of inertia
and therefore a given
tip
clearance,increasesthe tip leakagelosses;(5) increasesthe materialstresslevel,which makes it necessaryto reduce the inlet temperature to obtain a higher allowable stresS;and (6) causesa net changein turbine flowrate as a result of the changesin diagramefficiency, tip leakagelosses,and inlet temperature.Turbopump equivalentweight (sec.2.1.2.2) is often usedin suchan optimization to convertthe turbine flowrate to equivalentvehicleweight. The turbine blade centrifugal-stresslimit, discussedin section2.2.1.4,generallyappliesonly to hydrogen turbopumps andinvolvesan evenmore complex interrelation becauseit affects rotational speedand therefore alsoaffects the pump design.As with the tip-speedanalysis, optimizations involving turbopump equivalent weight are usedto determine the optimum combination of factors. Another tradeoff is that of weight and efficiency to determine the optimum number of stages.Multiple stagesincreasethe efficiency (fig. 21), but also increasethe weight and require an outboard turbine bearing with its added supporting structure, seals, and lubrication system.Again, turbopump equivalentweight is usedin determiningthe optimum numberof stages.All existing flight turbopumps(table I) haveusedno more than two stages becausethe relatively short flight times havemadeweight relatively more important than turbine efficiency.
2.2.3.2 The
EXHAUST
turbine
PRESSURE
exhaust
pressure
is the
static
pressure
of the
turbine
gas as it leaves
the
last
turbine blade row. For GG cycles, this pressure is low so that a high turbine pressure ratio and therefore a low turbine flowrate are achieved. However, the exhaust pressure must be high enough to permit an efficient and stable exhaust. For example, the flareback of turbine exhaust gases into the boattail of early ballistic missiles occurred because of the subsonic discharge problem. On
engines
turbine turbine. have
of the
turbine
gas. A higher
with
conical
or bell nozzles
gas disposal can affect If the turbine exhaust this
nozzle
"choked"
turbine
that
discharge
employ
the available is discharged
(operated
could
or tapoff
turbine pressure ratio into a separate nozzle,
above
discharge pressure will not be affected flight. This practice will increase the
a GG
pressure
the
critical
pressure
by changes in atmospheric turbine discharge pressure
have
cycle,
avoided
the
this
method
of
and the design of the it may be desirable to so that
turbine
pressure during and affect the
ratio),
vehicle turbine
pressure ratio. If the turbine exhaust is discharged into the main nozzle, as on the J-2 engine, the point at which the turbine exhaust is put into the nozzle will determine the turbine back pressure.
When
the
turbine
exhaust
is used
to cool
the expansion
nozzle
(flowing the turbine exhaust down the coolant passages and discharging expansion at the exit being used to increase performance), the pressure must be considered in establishing the turbine discharge pressure.
66
by "dump
cooling"
the gas at the. exit, loss in this circuit
For staged-combustionand expandercycles (sec. 2.1.1.4), the turbine dischargepressure must be sufficient to permit the flow to passthrough the downstreamducting and the injector into the thrust chamber; therefore, the turbine dischargepressureexceedsthe enginechamberpressureinsteadof being a smallfraction of the chamberpressure,asin the GG cyclel As a result, the turbine pressureratios for expanderandstaged-combustion cycles are minimized to minimize the pump dischargepressurerequirements;however, for _G cycles, which show little effect of turbine pressureratio on pump dischargepressure,the pressureratios aremaximized within the previouslydiscussedlimits. The RL10 engineis an example of the expander cycle arrangement. No operational engines employ the staged-combustioncycle,but this cycle is usedin the SSMEnow under development. 2.2.4
Turbopump
Mechanical
In turbopump mechanical attachment of the various design
objectives.
problems are pointed out. interdependent considered
2.2.4.1 From 30)
In the
integration, the detailed construction, arrangement, and turbopump components are planned to best meet the turbopump following
section,
the
major
options
in solving
the
basic
design
discussed, and the options that in general provide the best solutions are However, because there are many possible requirements and many components, solutions other than the best general one often must be
and used.
BEARING the design
are
PLACEMENT standpoint,
primarily
the turbopump
a function
of the
single-shaft turbopump design, shaft that usually is supported and one aft toward the turbine. is within usually
Integration
limits
and
is overhung
overhung simplify housing. overhung this case
arrangement
if the
of the
the
has been
bearing lubrication and However, if the turbine
has,
in general,
aft bearing
used
almost
minimize has more
turbine generally is too large and also for very-high-speed
housing.
In
and the placement and
turbine
of bearings
components.
For
(fig. the
selecting
the
no
inboard exclusively
the than
more (fig.
than 30(a)).
two
rotors,
At the present
for nongeared
turbines
the
turbine time,
the
in order
to
structural requirements of the turbopump two rotors, the shaft required to support an
to permit the bearing to stay within DN limits. For designs, the aft bearing usually is placed outboard.
This arrangement is more complex because seal between it and the turbine, provisions turbopump
pump
the pump impellers and the turbine disks are mounted on a by two bearing sets, one located forward within the pump If the rotational speeds are low enough that the bearing DN
turbine
by placing
arrangement
sizes
an outboard aft bearing requires an additional for lubrication, and structural support by the
bearing
arrangement,
the
complexity
required
operate at a higher speed often must be weighed against the potential benefits of increased pump efficiency or decreased weight. In short, the added complexity must be justified.
67
to
On the forward end, the bearing for a single-stagepump usually is placedinboard of the pump impeller, as in the J-2 oxidizer turbopump (fig. 1). If the pump has two or more stages,the forward bearinggenerallyis placedbetweenthe first and the secondstagesasin the RL10 fuel pump (fig. 3); this location makespossiblea reduction in the shaftdiameter and consequentlythe bearing DN by reducingthe amount of torque transmitted thrmagh the bearing.If the pump has a separateinducer stage,the bearingoften is placedbetween the inducer and the remainder of the pump, with the inducer statorsactingas a bearing support, asin the J-2 fuel (fig. 2) andthe Mark-29fuel pumps(fig. 32). /
For
cases in which
and
the
turbine
both can
the
be
turbine
designed
as separate
seals. Under these circumstances, and later attached to one another The
turbopump
configuration
and the pump
units,
the pump and by a coupling. may
have
each
the
incorporate
more
than
with
three
its own
impellers,
the pump
sets of bearings
turbine
are developed
as separate
a reduction
gear between
the
and units
turbine
and
the pump, thereby permitting each to be designed at its optimum rotational speed. In general, geared turbopumps are restricted to small sizes in which the turbine speed must be high enough to obtain reasonable blade heights and in which multiple turbines would require
excessive
important
2.2.4.2
control
because
TURBINE
In general,
the
elements
controls
ROTOR
ASSEMBLY
considerations
involved
turbopump depend on whether separate unit. For a single-shaft the turbine considerable combination
in terms
do not scale
of size and
down
weight.
in size as rapidly
AND
The
latter
in the
relation
the turbopump configuration,
of the
turbine
to the
whole or a of
rotor to the main drive shaft, in which case the drive coupling receives attention. This drive coupling (ref. 8) may consist of a curvic coupling in with an involute spline as in the Mark-29 fuel turbopump (fig. 32); a curvic as in the
involute (fig. 2).
spline Here,
alone, or just a bolt and torque-pin joint as in the later the main considerations are normality and concentricity
proper
operation
F-1 (fig.
of the
6), J-2 oxidizer
turbine
rotor
(fig.
relative
MA-5 sustainer turbopump or through a quill bearings as in the turbopumps for H-l, MB-3, some
gear is mounted separation, into
misalignment
between
directly on the turbine the turbine drive shaft.
the
1), and early
to the pump
If the turbine is geared to the pump, the turbine rotor this case, power may be transmitted by a gear attached
allows
rotor
configuration is single shaft, geared, the main consideration is the attachment
alone
shaft
is
ATTACHMENT
coupling
ensure
consideration
as turbopumps.
gear
68
and
torque
turbopumps;
the
turbine
deflections
an
J-2 fuel turbopumps limits in order to
rotor.
is considered to be a separate directly to the turbine shaft
shaft into a power and MA-5 booster
shaft,
J-2 fuel
unit. In as in the
gear mounted on separate (fig. 4). The use of a quill drive
shaft.
If the power
may be introduced
by gear
Cylindrical for thermal
growth
bearing
I_ducer star and forward
Turbine inlet manifold
support
Inducer Curvic couplings
Intermediate drive coupling
Involute spline Ch
Forward bearing
.Centrifugal impeller
Turbine rotors Thrust balance piston
Figure 32. - Mark 29 experimental
LH 2 pump.
tf the speed,
turbine is a separate the drive coupling
unit and is mounted in line with is of major importance. Such
transmitting the torque from the turbine to deflection in regard to alignment and to duration thrust in the turbine and the pump.
the pump. It must also be adaptable to in axial distance that may result from axial
tip speed low, the
Redstone A-7, the turbopump
Centaur RL10 (fig. 3), and MA-5 booster (fig. 4) turbopumps, and therefore shaft can pass through them. However, if the tip speeds are high, the rotor
may
have
to be solid
influences rotor disks
operating at the same must be capable of
Turbine rotor tip speeds are
disks
strongly turbine
the pumps a coupling
at the center
turbine rotor assembly can have holes through
in order
to withstand
and attachment. If the their centers, as in the
the high centrifugal
in the J-2 (figs. 1 and 2) and F-1 (fig. 6) turbopumps. Under these partway out on the disks are required for attachment to the turbopump. Extreme with a
thermal cryogenic
difference rotors range
conditions may exist within a turbopump assembly fluid and the turbine with an extremely hot
between
the
are mounted occurs in the
the centrifugal demand utmost
turbine
and
the
pump
may
be as high
on the same drive shaft as the pump rotor assembly. These severe temperature
forces,
circumstances,
as
bolts
if the pumps operate gas. The temperature
as 2000 ° F. If the
turbine
impeller, the total temperature gradients in combination with
stresses introduced by the high tip speeds of the turbine attention and care during the design of the rotor assembly.
and
pump
rotors
Thermal growth is an important factor in the rotor assembly. Pilots designed
that must be considered in maintaining concentricity to maintain concentricity among the various rotor
components components
if different materials are used for those centrifugal force act together on the pilots.
designed neglected. the not
may not be dependable or if thermal growth and to maintain concentricity may The material of the mounting
properly if thermal considerations are bolts, which fasten the turbine disks to
main drive shaft, is carefully selected so that the difference in thermal expansion result in loss of the clamping torque. Such loss of torque occurred during
development adequately
of the J-2 hydrogen pump. tightened. As a result, the
coupling
teeth
improvements Differential are
not operate (or clamp)
used,
cracked
in growth
because
and
The original bolts were first-row turbine disks
fretted.
Stronger,
tighter
thermal
distribution
within
the
turbine
may
temperature downstream
on the upstream side of a turbine disk may side, deflection of the disk is also a consideration.
to high
a uniform
in
tip speed,
the
disk profile,
rotor which
disks
do not
means
that
70
the clamping between the
ha_ze a hole both
sides
be
cause
will the to be curvic
combination
curvic design and disk material processing eliminated the between turbine disks is also considered even though similar
hotter than the other. In this event, to this differential radial growth
If, due
not strong enough cracked, and the
bolts
considerably themselves
have
various A pilot
one
with problem. materials
disk to operate
bolts are designed to adapt turbine disks. Since the higher
in the center,
are symmetrical
than
that
on
the
it is desirable
to
in coutour.
For
thesecases,the disks are attachedto the shaft by an intermediatedrive coupling,asin the Mark 29 experimental hydrogen turbopump (fig. 32). Since the torque that is to be transmitted from the turbine to the pump usuallyis large,the diameterof the intermediate coupling at which the turbine disks are attachedusually is large. A large diameter also contributes to disk stability and reduces the unit loading. Such an intermediate drive coupling is attachedto the main driveshaft with an involute splineand concentricpilots. Curvic couplings(ref. 8) frequently areused to attach the turbine disks to eachother (fig. 32). This type of coupling providesmaximum torque-carryingcapability andconcentric and normality relationshipsamong the individual components.The curvic coupling is usedin connection with clamping bolts, usually through-bolts usedin combinationwith clamping nuts. These clamping bolts are generally positioned at the mean diameter of the curvic couplingin order to avoiddeflection of the turbine disk dueto clampingforces. If torque-Carryingrequirements are relatively small, a key or pin drive often is used. Frequently, the clampingbolts are modified in sucha way that they are able to carry the torque as well as clamp the disk to the main drive shaft, asin the later J-2 fuel turbopump (fig. 2). Sometimesadditional radial pilots areusedto maintain concentricity betweenthe maindrive shaftand the turbine disk. However,this function canalsobe performedby drive pins; if so, at least three pins are used.Another method for maintainingconcentricity is to cut the involute splineandradial pilots directly on the extendedhub of the turbine disk and attach the disk directly to the main drive shaft. Clamping bolts are usedto attach the turbine disk to the main shaft at a suitablebolt-circle diameter.If the radial stressesof the turbine disk are moderate,a hole in the centerof the ttirbine disk is often used.In this case, the disksare mounted on a shaftthat passesthrough their centersandareeither attachedto each other with clamping bolts or shoulderedagainstthe main shaft and clampedwith a centerclampingnut. Torque pinsoften areusedfor power transmissionin suchanassembly. Two distinctly different turbine rotor designshavebeenemployed: one in which the rotor disks are individual, as in the J-2 (figs. 1 and 2), MA-5 booster(fig. 4), YLR87-AJ-7 (fig. 5), and F-1 (fig. 6) turbopumps; and the other in which they are made from one piece or welded together into one unit, as in the RL10 (fig. 3) and A-7 turbopumps.If the rotor consistsof individual disks, the mounting of the disks to the drive shaft is a step-by-step procedure,Such a configuration is designedso that the diskscannot be mounted onto the drive shaft backwards.Backwardmounting is preventedby asymmetricalpositioningof the bolt holes or by the addition of a so-called"idiot pin" (a locating pin that matchestwo parts in correct orientation). Since turbine disks are usually heavy and very difficult to handlebecauseof their delicatenature, they areusuallyinstalledvertically. This procedure ensuresaferhandling and better control for positioning the disk centerrelativeto the shaft center. In an installation where the disks are mounted directly to the drive shaft of the 'pump and the bearing is inboard, the shaft is already mounted in the bearingsand is therefore positioned.However,if an outboardbearingis used,additional stackuptolerances, concentricity tolerances, runouts, and thermal deflections have to be considered.The
71
possibility the
of backwards or counter-rotating drive shaft usually does not exist
main
has
a
thrust
positioned the rotor
balance
interference
that
TURBOPUMP
For
single-shaft
allows
axial
travel
of the turbine rotor in relation to is made from one unit. If the pump of
the
to the main drive shaft because, may be reduced. This reduction
and consequent
2.2.4.3 a
piston
axially relative and the stator
installation if the rotor
loss of power
shaft,
the
turbine
rotor
is
if neglected, the spacing between would introduce the danger,of
in the turbine.
HOUSING turbopump,
joining
the
pump
housing,
which
is superchilled
for
cryogenics, to the superheated turbine casing constitutes a major design problem. The interface between those two elements must be designed to adapt readily to the temperature differential.
The
differential
turbopumps. Structural thermal growth of the most
existing
large
In the structural configuration on the heated along
shrinkage
turbopumps method,
have used
a relatively
results in a cone smaller side. Theoretically, the
this cone
and,
therefore,
The
lbf thrust)
mechanical
turbopumps
(figs.
in diameter temperature at that
pronounced
in
large
method.
is used
for the turbopump
housing.
on the chilled side and larger should be equal to ambient location
should
remain
This
in diameter somewhere
constant.
It is at
housing is flanged and bolted onto the pump casing. The is the key to solving the problem of differential thermal was used for the Mark 29 fuel (fig. 32) and the uprated F-1
for accommodating
1, 2, and
differential
thermal
to
is used, the turbine hot gas is hotter
that it may freely expand is attached to the manifold
the turbine discharge manifold within the
to this system,
be accommodated strains and deformation
growth
however,
by
the in the
anchored. Many mechanical problems have the pins were attached directly to the hot-gas
When a structural cylinder containing the high-pressure such cone
6). A drawback
turbine has may introduce
the radial pins are brackets containing
interstage
long cylinder
more
is an arrangement
or keys on which the hot turbine casing can slide unrestrained without losing relation to the pump; this method was used in the J-2 and standard F-1
torque of the accommodation
mounted structural
are
are available for ensuring unrestrained method presently is favored, although
turbopumps.
method
of radial pins its concentric
growth
the mechanical
the diameter
this point that the cylindrical ambient-temperature interface growth. This structural method ( 1.8 million
and
and mechanical methods turbine casing. The structural
side of the structural outside turbine casing, this structural
72
the
full reactive
arisen when the mounting manifold or torus.
manifold is attached to it; since the manifold than the surrounding shell, the manifold is
relative on one
stators.
is that
radial pins or keys; this mounting brackets in which
to the cylinder. In this mounting, end and, on its larger-diameter end, cylinder. In addition to suspending cone may also be used to support
a to the the
The structural cylinder has been usedalso to alleviatethermal-distortion problems. For example, the rear bearing support on liquid-hydrogen pumps is generally subjected to liquid-hydrogen temperaturesnear the bearingand to high temperaturesnearthe turbine mount. On one of the early versionsof the J-2 hydrogenpump, gradualdistortion of the bearingsupport occurred.The distortion wasdue to large,repeatedthermalstressesddring eachfiring that resulted from the radial temperaturegradientthrough the bearingsupport. The solution was to changeto a strongermaterial. However,as previously discussed,the solution for a similar designproblem on a later pump, the Mark-29 hydrogen,is a more fundamentalone. Here, the bearingsupport and turbine mount function areseparatedinto two assemblies.The turbine mount is arrangedin the form of a thin cylindrical part. The temperaturegradient:is then essentiallyaxial, andthermalstressesarevery low sincegrowth is unrestrained. If the turbine is a separateunit with bearingson both sidesof the rotor, the turbine rotor is fixed relative to the turbine casingby the shaftbearingmounts, while the manifold with the gasdischargenozzle is flexible relative to the turbine casing.The problem is not so much misalignmentbetween the manifold and the rotor, but the relative positions of the two bearingsupports,which are more or lessindependentmembersbecausethey arepositioned in housingsthat are connected to each other by the casing.Hot gasshould not contact either the casingor the bearingsupport structure because,if it does,the resulting thermal deflectionswill produce bearingmisalignment.The passageof the hot-gasdischargethrough the bearingsupport often is avoidedby using a dischargemanifold that exhauststhrough a hole in the casing.If the hot-gasdischargepassesthrough the bearingsupport, the bearing support struts are shieldedby havingthem passthrough vanesthat contact the hot gas.The spacebetweenthe vanesand the struts may be filled with insulation.Insulation is alsoused betweenthe hot-gaslnanifold andthe casing. For large multistage turbines in which some thermal deflection is unavoidable,a linkage arrangementfrequently connectsthe casingto the bearingcarrier at the turbine discharge. This linkage keepsthe bearingcenteredrelative to the casing,regardlessof the amount of thermalgrowth in the casing. Material cross sectionsare heId to a minimum wherever high temperature is expected. Flangeswith bolt holes or threadedholes are avoidedin hot-gas regions,becausethose flangesare usually of heavycrosssectionandaresubjectto thermal deflection andcracking. The reason for this is that the core of the heavy section will not heat as rapidly as the surfacematerial and therefore high thermal stresseswill be introduced. In addition, hot-gas flanges have to be sealed, a requirement that is often difficult to satisfy becauseof differential growth anddistortion of the sealsand the flanges.
73
2.2.4.4
BEARINGS
If the turbine
AND SEALS
and pump
are in a single-shaft
bearing considerations for arrangement shown in figure propellant
is tapped
off
at
a
and
the pump the pump
the
turbine
and the turbine
high-pressure
propellant passes through both venting to either the pump inlet cooling method is used for both bearing
unit
the turbine are similar 30(a) with propellant-cooled region
to
bearing
is inboard,
seal and
those for the pump. For the bearings, a portion of the puml_
such
as
the
pump
discharge.
This
bearings into a low-pressure region that is obtained by or a low-pressure region in the impeller flow passage. This J-2 turbopumps (figs. 1 and 2). A seal between the turbine
minimizes
leakage
from
the
pump
to the
turbine.
The
pressure
on
side exceeds the pressure on the turbine side so that turbine gas cannot leak into and affect the bearing cooling. For an oxidizer turbopump, such as the J-2 oxygen
pump (fig. 1), fuel-rich turbine hydrogen
pump
propellant
back
purge Seals generally are used to prevent any oxidizer leakage into the gas. If the support for the pump bearing is the inducer stator, as on the J-2 (fig. 2), the
If an outboard
static
pressure
through
the bearing.
turbine
bearing
(fig.
rise across
30(b))
is used,
the
stator
can be used
lubrication
may
to recirculate
be a problem.
pumping fluid is hypergolic, then a separate lubricant flow has to be source. A separate turbine unit does not have an internal source for Therefore, lubricant and cooling flow have to be introduced into the removed. Bearing races on the turbine shaft or on the main drive shaft
supplied bearing bearing usually
If the
by another lubrication. cavities and are installed
with a slight interference fit in order to prevent looseness and possible sliding of the bearing race on the beari_ag journal Of the drive shaft. Such sliding generates heat and contributes to an early selected;
failure of the bearing. If axial and radial loads are severe, a split-race bearing often in this design, additional balls can be inserted into the bearing, thereby making
is it
capable of absorbing higher loads. Selecting the proper material for sealing the lubricant flow from the hot gas in the turbine is critical, because a sealing material that is compatible with one lubricant may not be compatible with another. The cavity pressure of the lubricant flow
and
operate
the
and
DN radial
in the
this step
balance,
beating
axial
pressure
effectively;
on pressure The
cavity
turbine
are
is particularly
important
bearings
limit
depends
loads.
have
have no axial-load To
prevent
hot
bearing and shaft-tiding assembly
If
on operating
the
turbine
approximately
capacity turbine
gas from
rings
three
gnd are often
the hot-gas chambers. seal and the other
of carbon
established
when
so that
a dynamic
the
seal, which
seal
will
operates
is used. conditions
is part
made
entering
such
as temperature,
of a single-shaft
bearing may be a roller bearing. If the turbine likely to be ball bearings because they have Roller
carefully
lubricant,
configuration,
the
and
turbine
times
is a separate assembly, the bearings are most to absorb axial loads as well as radial loads.
subject
the
radial-load
the
to end wear bearing
cavity,
capacity and skewing
of ball bearings (ref.
a seal is installed
6). between
This seal arrangement usually consists of two seals, a face-contact type of seal. The shaft-tiding seal in segments
around
74
the
periphery
and
bound
but
together
the one a is an by
small coil springs.It acts like a labyrinth sealin that it causesa pressuredrop in the turbine gas.That portion of the gasthat may leak through the sealis collectedin the cavity between the two sealsand drained. The second seal,which is located close to the bearing, is a dynamic seal that necessitatesa mating ring. This sealis activated statically by built-in springs that press the seal ring against the mating ring. During operation, a carefully calculatedpressurebalanceensuresconstant contact of the seal-ringface with the mating ring. Sincethe sealring is alwaysin contact with the matingring, the peripheralvelocity is limited. This rubbing-velocitylimit is a function of the pressuredrop acrossthe sealandthe propertiesof the bearinglubricant. In some turbine configurations, the bearingand sealhousingsare alsopart of the turbine casing.In others, particularly in largeconfigurations,a flexible diaphragmis attachedto the turbine manifold and bolted to the bearinghousing.In that case,the bearinghousingis a separatecomponent.If shaft whirl or critical speedbecomesa problem, it canbe alleviated by changingthe bearingpreloadso asto changethe bearingspringrate.
2.2.4.5 For
AXIAL
sm_ill
However,
THRUST
turbopumps, for
the
high
BALANCE axial
thrust
bearing
DN's
may
be
in liquid
controlled rocket
by
antifriction
turbopumps,
the
bearings
alone.
axial-load-carrying
capacity of the bearings generally is too low to carry all the load without some other design provisions. For example, the H-l, the MA-5 booster (fig. 4), and the MA-5 sustainer turbopumps have ribs on the backface of the centrifugal-pump impellers. These ribs cause the propellant to spin with the impeller, thereby decreasing the pressure with radial distance from the impeller tips. The effect is to reduce the net axial thrust on the rotating assembly. For centrifugal pumps in which the axial load is too great to be controlled by balance ribs, it can be controlled by putting a shoulder on the impeller backface that mates with a labyrinth seal (fig.
1). The
cavity
surrounded
the turbopump, thereby reducing reducing the axial thrust. This sensitive solution.
than
balance
by that
the pressure acting wear-ring arrangement
ribs to impeller
Open-faced and shrouded acting against the housing,
seal is vented
axial position
centrifugal pump impellers and thus centrifugal pump
on
to a lower
pressure
region
the impeller backface has the advantage of
and therefore
produce shrouding
is generally
within
and being
thus less
the preferred
very different pressure forces strongly affects axial thrust.
In general, shrouded impellers have more predictable pressure distributions and are less likely to produce forces that will cause the turbopump shaft to flex: However, pumps with unshrouded impellers have been designed to accommodate the different forces and have been operated successfully. In high-speed, high-pressure pumps, the high bearing DN's reduce the bearing capacity at the same time that the high pressures increase the axial loads. If either
75
axial-load the thrust
variation betweenoperating points or a reasonablemargin for error in thrust prediction exceedsthe bearingaxial-loadcapacity,the bearingsaredesignedto movefreely in the axial direction, and a balancepiston (fig. 33) is usedto centerthe rotating assembly.The balance piston accomplishesthis by varying its axial force with axial position. With a series-type balancepiston as shown in figure 33, if movementis to the right, a high-pressureorifice (in
Highpressure seal Transducer Stator
assy
Support
assy
Low-pressure seal _ssure
orifice Spring assy
High-pressure orifice
Balance
piston
Rotor
assy
Labyrinth
Figure
Turbine seal
33.
-
end
Balance
piston
bearing
assembly.
hydrogen, rub rings) opens, and a low-pressure orifice closes such that the pressure acting against the piston increases and provides a restoring force in the opposite direction. If movement is in the other direction (to the left in fig. 33), the high-pressure orifice closes. and the direction position. balance
low-pressure decrease,
The J-2 hydrogen pistons, and current
-for balance The rub exhibited occasional
orifice opens such that the pressure and thereby permitting the net force to return (fig. 2) and Mark 29 hydrogen designs for the SSME turbopumps
the force the rotor
acting to the
in that neutral
(fig. 32) turbopumps utilize now under development call
pistons. rings low
of the friction
impact
J-2 hydrogen balance system originally were carbon, a material that and associated low wear. However, rotordynamic transients caused
against
the
carbon
rings
and, on a few occasions,
76
the carbon
cracked.
This
failure led to a materials testing program in liquid nitrogen. From it, (1) leaded bronze (BeariumB-10)ringswith an Inconel 718 piston and(2) glass-filledTeflon (Armalon) with a titanium piston wereselected.Thesetwo combinationshave.workedwell. Balancepistons are often made integral with the pump impellersso as to minimize axlal length and therefore weight. Also, by acting on fluid that is in an impeller clearancespace, the integral arrangementdoes not require an additional supply of fluid and, therefore, reducesleakagelosses.The integral balancepiston is sensitive to housing and impeller deflectionsand therefore requiresanalysisof thesedeflectionsduring design.To avoid the complication of deflection analysis,a separatebalancepiston may be mounted elsewhereon the turbopump shaft. However, as previously discussed,this arrangementincreasesboth axial length andleakagelosses. Ball-bearingsystemsfor turbopumps haveundergonemuch development.If the bearing DN's are not too high, the bearingaxial springsareshimmedin sucha way that the bearing loads are controlled within acceptablelimits as long as the rotor axial motion doesnot exceedthe "free motion" betweenthe rub rings. This practiceallowsthe rub ringsto carry large transient loads during turbopump starts and stops without bearing overload (for periods of 1 or 2 sec).On the other hand, if significant rub-ring wear occurs,the bearings pick up a larger shareof the load andprovide someoverloadcapacity.Duplex bearingpairs, which are free to float axially, arid rubbingstops,which carry the excessaxial loads,canbe usedto eliminateall axial,loadson the bearings. In pumps with a single spring-loadedangular-contactbearing at eachend in combination with a balance-piston system, a serious problem arises in designing and setting the axial-spring loads. Differential thermal shrinkage of rotor and stator assemblies,axial shrinkageof the rotor at high speedfrom Poisson'seffect, anddifferential axial growth from pressureeffects - all enter into the axial bearingloads under variousoperatingconditions. This combination of effects has often led to operation of unloaded ball bearings,an undesirablecondition for high-speedoperation. One solution to this problem that has worked well on two models of LH 2 pumps is to use duplex bearings at each end, the bearings
being
loaded
against
each
other
(1) the inlet differential-growth
bearing set problems
can be mentioned
The effect increased.
spring
can
2.2,4;6
on
THERMAL
rates
be
through
allowed above,
turbopump
springs.
a great
to float axially, and (2) the radial advantage
if shaft
Two
advantages
thus spring critical
bypassing rates are speeds
result: all the doubled.
need
to be
BARRIERS
When a hot-gas turbine is used in combination from the hot turbine to the cryogenic pump After
Belleville
shutoff,
the pump
temperature
77
with a cryogenic pump, thermal soakback becomes a problem if restarts are required. rises rapidly
if the turbine
is not thermally
isolated from the pump. This temperaturerise causesthe cryogenicfluid in the pump to evaporate,thereby making the restart of the pump very difficult. Consequently,specialcare is taken to prevent thermal paths betweenthe turbine and the pump. Various mechanical meansare availableto accomplishthis. Couplingshavebeenusedbetweenthe turbine and the pump rotor; e.g., an involute spline together with a quill shaft or, if centrifugal stress permits, a coupling from a dissimlarmaterial suchasplastic.Another useful couplingis the ball joint; this joint has the advantagethat, assoon as torque transmissionceases,the balls sink back into their sockets,thereby disengagingthemselvesfrom the pump drive coupling and essentiallydisengagingthe two shafts.Cooling the turbine bearingwith the cryogenic fluid is often helpful in preventingtemperaturecreepbackinto the pump through the drive shaft. The housingsmay be attachedto eachother with an arrangementof radial pins that act as thermal insulators; an insulating material may be clampedbetween the turbine and the pump housing;or a thermal barrier may be added.The thermalbarrier may consistof a manifolding devicethrough which cryogenicfluid is circulatedafter turbine shutoff so that the fluid may absorbsomeof the heat. This fluid may then be carried in the form of gas back to the cryogenictank andusedfor tank pressurization.
2.2.4.7
ASSEMBLY
To avoid
contamination
engine turbopumps manufacturers have the
clean
meet
room
critical
Several
that
is temperature
assembly
other
the turbopump the assembly checked and
wear rates
controlled
to permit
and possible
environment. In addition
verification
explosions,
rocket
For this purpose, most to being dust controlled,
of critical
dimensions
example,
seals
and
to
dimensions.
requirements
are also met
during
assembly.
For
positioned
in
are checked to verify that leakage will not exceed the amount specified in specification. Rotor axial and radial clearances and the bearing drag are verified. Since every component is subjected to considerable vibration, all
bolts, nuts, and other however, is never used could
can lead to excessive
are assembled in a dust-controlled an area designated a "clean room."
do considerable
fasteners are secured within the turbopump damage;
lockwire
with locking devices or lockwire. because pieces of wire that failed
usually
is used
externally
or in areas
Lockwire, in fatigue
where
there
is
no possibility that broken pieces can escape. As a safety measure,.locking devices are never used twice. After assembly, a torque check is performed on the rotor to ensure that there is no interference between the rotor and the housing and that the drag on the seals and bearings Some
is within rotor
Matched maintained.
the specification.
assemblies
assemblies For
are often
example,
particularly in a multiple-stage of the rotor after assembly components is the best
is difficult answer.
matched, are
and
necessary
close
individual if extremely
clearances
between
components
cannot
accurate the
be interchanged.
relationships
rotor
tip
rotor assembly, frequently require grinding to avoid tip clearance losses. Interchangeability
to achieve
in such
a case, and therefore
78
have
and
the matched
the
to
be
housing,
the tip diameter of individual assembly
often
The selection Of
materials
for
the
turbopump
as well as operational requirements. some period of operation, a process but
not
with
others.
Some
For that
materials
components
must
consider
fabrication
needs
example, a turbine may require weld repair after may be easily accomplished with some materials
(e.g.,
Hastelloy)
do not
require
annealing
after
weld
repair, whereas others (e.g., Ren6 41) must be annealed if weld repair has taken place bn a finished machined component. The annealing process, however, may produce deflection in such a component and therefore may make its reusability questionable. The selection of tolerances may either ease the higher rate of rejection. Often,
manufacturing process or make it more difficult by causing a the use of magnetic material eases the manufacturing process
by enabling for machining
to use magnetic
the manufacturer those parts.
If the turbopump
is designed
for long service
of considerable importance. Periodic procedure should be possible without components reassembly. selected assembly
are Easy
chucks
and
other
life and overhaul
inspection difficult
capability,
holding
devices
ease of assembly
of some components often disassembly and reassembly.
indexed against each other so that mistakes access without complicated tooling is provided
so that they are are made different
magnetic
cannot be to all points.
not interchangeable. For example, fasteners in thread size in order to prevent installation
is
is necessary; this Therefore, rotor made during Fasteners are
within a given of a given bolt in
an area that may require a longer or shorter bolt. Components often have to be separated by force as a result of thermal deflections that have taken place during high-temperature operation. Provisions are made for extracting those components without damanging them, usually by adding threaded the disassembly sequence
holes to the clamping and used as extricators.
flanges so that bolts can be inserted during Interference tolerances often are selected
for components that must maintain close relation to each other. In such a case, the assembly process calls for chilling down one component or heating up the other before assembly, so that they are easily joined. Provisions for disassembly of such joints are also necessary.
2.2.5
The
System
interfaces
between
vehicle and engine the fluid entering contraction
Interfaces
Interfaces desired
turbopump
performance the inducer
can cause
(3) maintenance costs serviced after use.
2.2.5.1
the
system
and
the
engine
and vehicle
systems
affect
and maintenance costs. For example, (1) the flow pattern of affects pump suction performance, (2) thermal expansion and
mechanical can become
connections
to fail if the connections
high if the
turbopump
system
is not
are not flexible, designed
and
to be easily
PUMP INLET
between fluid
flow
the
propellant
pattern
at the
feed
line and
pump
inlet.
79
the The
pump flow
inlet
are designed
pattern
entering
to produce a pump
the
strongly
affects the pump suction performance, and a poor flow pattern can causeexcessive cavitation. Therefore, bends and changesin inlet ducting crosssection are minimized. If elbowstoo closeto the inlet (approximately 15 to 20 pump-inlet diametersupstream)are required, turning vanesare usually used.If there is sufficient NPSH,tangentialpump inlets often are usedto minimize ducting weight and complexity. Straighteningvanesat the talak exits are employed to minimize swirl of the pump inlet flow (ref. 1) and to prevent fluid vortexing in the tank.
2.2.5.2
PUMP DISCHARGE
For ease of assembly discharge
line.
To
pressure-actuated discharge lines
and disassembly, seal
this joint,
pump
O-rings
seals generally are are often welded
disproportionately
large
and heavy
discharges
are often
chamber The
vehicle thermal
ball-joint
propellants,
and
small pumps, the flanges become
discharge line and the downstream plumbing, of the pump volute tongue. The amount of because too little diffusion produces a large
and too much diffusion produces be large and therefore heavy.
a discharge
line
the connections between the turbopump assembly and the loads due to the weight of the turbopump assembly and react inertia, propellant inertia, engine gimbaling, fluid pressure
flange forces, expansion and
common end
types
and gyroscopic forces. The contraction of the turbopump
connection
of turbopump
arrangement ball-ended
is Close-coupled, struts at the other for the
J-2S hydrogen
This method
struts, tension
rigid pads at one end to accommodate pump
systems
dimensional
ball-ended in pure
shrinkage. The thick sections required for cast structure. If the pump is very flange.
mounting
to accommodate
growth or shrinkage. Theoretically, structure by loading the members
discharge
to the
mounts also adapt to the assembly and the thrust
assembly.
most
was used
for noncryogenic
bolted
MOUNTING
The turbopump mounts are engine. These mounts support to loads due to machinery differentials, differential
and
for small line sizes.
pressure drop due to high fluid velocity that, because of low fluid velocity, must
TURBOPUMP
used
are flanged
used for cryogenic propellants. For to the pump discharge, because
To minimize the pressure losses in the diffusers generally are used downstream diffusion is determined by optimization,
2.2.5.3
usually
with
tolerances
and
with
at least
differential
one
thermal
end of the turbopump and one or two the dimensional variations. One large pad
rectangular,
for designs
80
struts
arranged triangularly, yield the lightest or compression. However, a simpler
for this type small, it often
is best suited
utilize
integral
keys
to accommodate
of design make it particularly is mounted directly through with
cast volutes.
the
suitable the pump
2.2.5.4
GAS-GENERATOR
CONNECTION
Gas generators usually are influence on the turbopump
are
connections
often
interrelated,
of the
The major temperatures are used
and
discharge
better to relatively small support the GG weight.
to connect
the
gas manifold, and and contraction in some
duct
cases
also serve
turbines,
problems with and pressures
MOUNTING
close coupled to the turbopump design. The GG connections join
GG to the hot-gas inlet of the turbine loads and adapt to thermal expansion mounts
AND
because
(e.g.,
as the GG the
stiffer
and, therefore, have some the hot-gas discharge of the
the GG mounts (ref. 9). These the
J-2
mount.
turbopump) Such
manifold
support the 'GG connections and the
a system
structures
welded
lends
can more
itself easily
the connections are that they must seal gas at extremely high (e.g., 1550 ° F and 920 psia on the F-1 engine). If bolted flanges GG
to the
turbine,
pressure-actuating
metallic
seals
generally
are
used. Dual sealing lands increase reliability if a vent to low pressure is provided between the lands (thus providing a high pressure drop only across the pressure-actuating portion of the seal) in combination with a secondary low-pressure seal that precludes external leakage even if the primary seal develops slight leakage. Welded connections between the GG and the turbine
manifold
provide
a
more
reliable
joint
if
thorough
X-ray
crack-detection methods are employed. Welded connections are difficult the GG or turbine is damaged, and therefore may not be advanced-development systems. However, these connections are more expensive
for high production
rates
once
the system
and
penetrant
to repair in case best suited for reliable and less
is developed.
Gas-generator mounts must adapt to a high degree of thermal expansion and contraction. Solid-propellant GG's for start systems can produce a severe tendency to low-cycle fatigue. Maximum transient gas temperatures can be as high as 2400 ° F (above 2000 ° F for about 1 sec). On the J-2S liquid-hydrogen caused a Hastelloy-C manifold greater elongation configuration was gradients.
2.2.5.5 When
resulted _ in warped made of Inconel 625
TURBOPUMP a
reassembly,
turbopump may
turbopump, to crack after
have
SERVICE
to undergo
flanges with with a zirconia
ON THE
is disassembled,
the steep temperature ten starts. A change
it
gradients during starts to 347 CRES to obtain
consequent leakage. lining to reduce the
ENGINE must
a full-speed
be
removed
operational
from
test.
the
costly removal from the engine may tasks is performed. In the extreme, testing may be necessary.
81
engine
It is obviously
from a time standpoint to be able to replace seals, instrumentation, and without major disassembly. If provisions for performing these procedures the turbopump design, these relatively simple followed by operational
The final temperature
and,
perhaps are not
be necessary each major turbopump
after
advantageous bearings made in
time one of disassembly
it is highly desirableto designthe turbopump system such that main shaft sealsmay be replaced without disturbing the bearing assembliesand that leak checks following the installation of sealscan be madeto verify the correctnessof the installation. In addition, provision for hand turning the turbopump permits a simple torque checkfor early detection of damagedbearings,excessiverubbing, and worn seals.Boroscopeports are someti_nes providedfor inspectionof turbopump condition prior to reuse.
2.2.5.6 For
TURBOPUMP
major
overhaul
turbopumps (e.g., they are reinstalled
OVERHAUL or
disassembly,
the
turbopump
is removed
from
the
engine.
Most
those of the J-2 engine) then require a full-speed operational test before on the engine. If the turbopump overhaul involves other than a simple
change of bearings and seals, the engine should then be hot fired for recalibration. Therefore, simple turbopump overhaul is less costly than replacement, because a full-speed operational test of a turbopump is generally less costly than an engine recalibration. The cost of replacement parts is an important consideration in the design decisions as to whether to integrate certain parts for ease of original manufacture. For example, fir-tree-mounted turbine buckets can be replaced individually; if the turbine buckets are integral
with
bucket
the
disk,
is damaged.
weighed
against
The turbine individually.
however,
In deciding
entire
which
the probability
rotor This
the
turbine
design
wheel
to use,
the
may
have
cost
of original
assembly and the pump permits the replacement
2.2.6
Energy
the shaft
if one
manufacture
is
of failure. rotor assembly of either unit
are often designed without rebalancing
the case of the J-2 fuel turbopump (fig. 2), this approach replacement of the entire turbine assembly (manifold, stators, disturbing
to be replaced
seals or bearings
of the
to be balanced the other. In
is extended to permit the and rotor assembly) without
turbopump.
Start Systems
storage
systems
that
supply
propellant main tanks, pressurized-gas and solid-propellant start cartridges.
initial
drive
start tanks, All of these
power
for
the
turbine
include
pressurized-liquid-propellant systems have been used
on
the
liquid
start tanks, production
flight engines and are within the current state of the art. The type of start system on specific engines is shown in table I. The selection of an energy storage system, for starting an engine is dependent on several factors including maximum thrust), repeatability of starts, number of environmental conditions, weight, commonality valves,
and type
of turbine
drive
cycle.
82
allowable start time (time to 80-percent starts, tanked propellant conditions, of start and normal-operation control
The primary areasof impact of the start systemon turbopump designare provision of adequatestall marginduring the pump accelerationtransient,and possibleprovision of small boost pumpsupstreamof the main turbopumps.Adequatepump stall margin normally can be provided by adjustment of the energy level of the start system, although it may,be necessaryalsoto provide greatermargin through pump design.The J-2 enginesystem,which uses a pressurized-gas start system, was initially designedfor a storagepressurelevel of 800-+50 ° psia and a temperature of 200-+50 ° R. Thesevalueswere selectedto permit the start tank to be refilled, for restart capability, from the thrust-chamber fuel injector manifold. During the initial phase of the J-2 engine developmentprogram, there was a tendency for the fuel pump to stall when the oxidizer systemwasprimed under the above start-tank conditions. This problem was corrected by increasingthe fuel pump speedto move the operatingpoint during the start transientfarther from the pump stall curve.The additional stored energy required to increase the fuel pump speed was obtained by increasingthe start-tankpressurelevel to 1250 + 50 psia. Small
boost
pumps
are
considered
for
liquid-propellant
main-tank
start
systems
when
marginal start power is available. These small turbopumps have small rotating inertia low torque requirements and can be designed to operate with a wide range of propellant conditions to minimize thermal preconditioning requirements.
2,2.6.1 This
MAIN-PROPELLANT-TANK
start
system
uses
the head
for initial turbopump concept, is the simplest One of the principal results from start
HEAD available
in the main
rotation. This method of the start systems.
advantages repeatability
and inlet
propellant
is used
of this sytem requirements
in the
is minimization that make
separate start tanks, operating under blowdown components for refilling in order to accomplish
conditions, in-flight
tanks F-1 and
to provide RL10
the energy
engines
and,
in
of weight. The reduced weight it necessary for systems with to have restarts.
either These
multiple tanks weight-reduction
or
benefits are" amplified for engines that operate on a closed cycle with turbines that normally operate with a high flowrate and low pressure ratio; providing these conditions with propellants from separate start tanks would require large, heavy systems. Propellants a bipropellant
from
the
main
preburner
tanks
can be used
or gas generator
for engines
or the
thrust
that utilize chamber
either
cooling
the exhaust fluid
from
as a source
of
energy to drive the turbines. Although the exhaust from a bipropellant GG operating under tank head was used with the F- 1 engine, it should be noted that this kind of circuit normally is a low-flow circuit, and the GG should not be operated greatly off its design mixture-ratio conditions.
However,
for
engines
staged-combustion cycle (turbine main-propellant-tank start system is designed for liquid propellants,
with
low-pressure-ratio
turbines
operating
on
working fluid exhausting to thrust chamber injector), may develop combustion instability if the GG (preburner) since in effect it would be operating under deep throttled
83
a a
conditions. Solution of this problemrequiresa sophisticatedinjector or control system.Use of the thrust-chambercooling fluid as a sourceof energyto start the turbopumps,asin the RL10 engine, posesdifferent problems. The enginemust be conditioned within certain limits from what may be widely varying initial conditions to provide start repeatability. Under certain initial conditions, only marginal power may be availablefor turbopump breakawaytorque. Other problems associatedwith main-tank-headstarts for certain types of enginesand mission requirementsare exemplified by early developmentof the J-2 engine.The start method was eventually changedfrom tank-headstart to a pressurized-gas start systemfor the following reasons: (1) The start transient wastoo slowandunrepeatablefor the clusteredenginesof the SaturnS-II stage(5 to 6 secfrom start signalto mainstage). (2)
Because the engine was to be used in several different applications and had to meet orbital start requirements, engine inlet-pressure requirements were low and varied over a wide range (fuel pressure = 27 to 46 psia, oxidizer = 33 to 48 psia). As a result not
(3)
of the
wide
range
of start
conditions,
a common
start
sequence
was
feasible.
A common large
Development
start
difference and
main-tank-head
sequence in turbine
qualification
start
at sea level and altitude
systems
was not feasible
because
exit pressure. of
the
to other
SSME
types
will
of engine
advance cycles
the and
state wider
of
the
ranges
start requirements. This advance will be accomplished by placing a portion transient under closed-loop control and using a digital-computer controller monitor and command engine operation. The SSME also will incorporate
of fast-start
2.2.6.2
PRESSURIZED-GAS
art
for
of mission of the start package to low-pressure
turbopumps upstream of the main turbopumps. These low-power turbopumps with a wide range of inlet propellant conditions and, being low in rotating capable
of the
can operate inertia, are
transients.
START
TANKS
This system, employed on the J-2 engine system, utilizes high-pressure gas stored in a small engine-mounted tank to initially spin the turbines during engine start. This system typically includes a tank; fill, vent, relief, discharge, and check valves; and a duct connecting the start tank to the turbine. Both inert gases (e.g., nitrogen and helium) and fuels (e.g., hydrogen) have is that
been the
used start
successfully tank
as the turbine
can be replenished
developed on the J-2 engine; however, to refill the start tank. The alternatives
drive during
fluid. engine
The advantage operation.
of using This
the engine
procedure
has
fuel been
up to 60 seconds of mainstage operation is required are to use multiple tanks or a single large tank for
84
restart capability. The latter of thesealternativesresults in successivelylower start-tank pressuresfor eachrestart; this lower tank pressuremust be compensatedfor to achieve repeatablestarts. Pressurized-gas start tanks are alsoat a disadvantagewhen consideredfor closed-cycle engines because the high-flow, low-pressure-ratioturbines result in large fluid-storagerequirements. Tank volume is establishedon the basisof the energyrequiredto start the engine.Pressureis the most significant independentparameter for this type of energystoragesystem.It has been empirically determinedthat, for minimum weight, the gasstoragesystemshould be designedfor pressuresbetween 1500 and3000 psia.Becauseof thesehigh pressuresandthe largenumber of components(leakagepaths),the systemis susceptibleto leakage. A critical component for pressure/temperaturecontrol of cold storedgasis the relief valve. Relief valvesarenormally consideredas safety devices.Therefore,toleranceson relief and reseat pressuresare large,usually +5 percent on relief pressures, with the reseat pressure being 10 to 15 percent below relief pressure. control of start tanks where consistent start device
or a narrow-band
2.2.6.3
relief
These energy
tolerances is required.
are not adequate for pressure A pressure-regulator type of
valve is required.
LIQUID-PROPELLANT
START
TANKS
Liquid-propellant start tanks have been used on the Atlas MA-5 sustainer early versions of the Thor MB-3 engine. The start tanks can be ground based devices for single-start engines or they can be mounted within the vehicle.
engine and on with disconnect Very repeatable
starts
tanks.
this
(from
engine
type
of
pressurization, State-of-the-art
to
system
engine) is very
fill, vent
relief,
can
be provided
complex and tank
liquid-propellant
and
by liquid-propellant requires
discharge
start-tank
start
high-pressure
However,
propellant
tanks
and
valve systems.
systems
have
been
used
with
the
LOX/RP-1
propellant combination; however, many of the design elements are applicable to other propellant combinations. The tanks are filled with propellants and pressurized prior to engine start. Tank capacities and pressure levels are established to accelerate the turbine to a speed sufficient into mainstage This
method
to produce pump discharge pressures operation. The ratio of tank pressures of engine
start
has
not
been
used
where
that will allow the engine governs the mixture ratio engine
restart
to bootstrap during start.
is required.
Repeatable
refilling of start tanks with cryogenic propellants such as liquid oxygen and liquid hydrogen is difficult because of rapid vaporization due to warm tanks and unpredictable heat transfer rates. Since of the
the two liquid turbopumps,
propellants a separate
must small
be burned combustor
85
to provide
the energy
for initiating
is required
for engine
cycles
that
rotation do not use
a gas generator to provide a turbine drive fluid. As with pressurized-gas start systems, liquid-propellant start tanks areat a disadvantagewhen consideredfor closed-cycleengines becausethe high-flow, low-pressure-ratioturbines result in largefluid-storagerequirements. 2.2.6.4
SOLID-PROPELLANT
START
CARTRIDGE
Solid-propellant start cartridges have been developed into very reliable start systems for the YLR87-AJ-7, YLR81-BA-11, J-2S, and current versions of the MB-3 and MA-5 engine systems. Restart capability is available by using a separate cartridge for each start. The system
for the
of residue closed-cycle Design
J-2S
engine
provides
contamination, engines in which
considerations
the the
involved
three
starts
with
solid-propellant turbine exhausts in the integration
separate
cartridges.
However,
start technique cannot be to the thrust-chamber injector. of a solid-propellant
because used
with
GG into a turbopump
start system include determination of burn duration and flowrate, control of grain thermal environment, integration with the ignition system for the liquid-propellant GG, and selection of the solid-propellant start-cartridge method for multiple-start systems. A detailed analysis must be made of the engine start system solid-propellant burn duration and flowrate. Both parameters repeatability,
prevention
of thrust
cartridge on the J-2S engine 50 percent of their mainstage propellants in the pumps, propellants from the
overshoot,
and start
reliability.
starts the engine by accelerating speed. During this acceleration because the vehicle propellant
at burnout.
The performance because the grain
The
total
start-cartridge
propellant
effect engine
start
J-2S
weight
engine,
an
orbital
heat-transfer
after
analysis
was
temperature range of the start-cartridge engine with a three-start mission use
between
engine
The
is approximately
installation
maximum expected was a multiple-start starts.
recirculation to achieve The nominal grain burn rate from 2.3 lbm/sec
is strongly dependent on propellant is conditioned to 50 ° + 10 ° F for
installation on the engine and then is maintained, 75 ° F by the vehicle boattail environment. the
solid-propellant
the pumps to approximately period, there are mixed-phase ducts contain mixed-phase
engine time is to 9.9
13 lbm.
of the solid,propellant start cartridge is sensitive to grain temperature, burns faster and provides a higher flowrate as the grain temperature
increased. Grain-temperature start cartridge for the H-1
For
The
the required to start time,
that must pass through the pumps and be replaced with subcooled propellants propellant tanks prior to burnout of the solid-propellant start cartridge. The J-2S
engine does not require propellant bleeding or propellant start, since the pumps will start on mixed-phase propellants. 2.3 seconds and the flowrate increases at an exponential lbm/sec
to determine are critical
starts
satisfactorily
with
86
composition. The 24 hours prior to
on the engine,
conducted
is
at 40 ° to
to determine
the
grain. This experimental engine and up to 10 hours coast time
a grain-temperature
range
of-50
° F to
+140° F. However,start time varied from 3.2 to 4.2 sec.The J-2Sturbine drive systemwas made lesssensitiveto grain temperatureby (1) utilizing a checkvalvein the combustion chamber tapoff line, a step that allowed automatic changeof the power source,and (2) using a low-energysolid-propellantstart cartridgethat allowedpower overlapwithout thrust overshoot. , During engines,
the initial development contamination of the
Combustible start-cartridge
products forced into gas caused detonations
purge for the gas-generator H-1 engine gas generator. pressure
level and
The ignition solid-propellant RP-1
of the solid-propellant start cartridge for the MA-5 and H-1 injector for the liquid-propellant GG was a major problem.
flowrate
of GG start
enter
the
the LOX injector when LOX entered
LOX manifold A series of tests to prevent
combustor.
To
was developed was conducted
contamination
liquid propellants cartridge, which
manifold by the manifold.
the solid-propellant A gaseous nitrogen
to prevent contamination to determine the required
and maintain
reliable
reliable
ignition
and
minimize
overlap, two pyrotechnic igniters are used to ignite the gas generator. autoignited by the heat from the start cartridge and burn for approximately a redundant heat source to ignite the liquid propellants.
2.3 DESIGN 2.3.1 If
the
engine
start
cartridge
These igniters are 2 sec to provide
EVALUATION
system system
turbine power The turbopump meet a nominal system turbopump
and
is to
must
meet
be within
specifications,
the
on-
certain
The
turbopump
engine system
and
off-design system
operation must
of the
provide
the
component
tolerances
and
efficiency
are
considered
during
the
design.
DESIGN-POINT system mixture
SYSTEM
BALANCE
balance, known ratio, and thrust
as pump efficiencies, turbine efficiencies, turbine inlet temperature, and various pump
limits.
necessary to match the required pump power at all test and flight conditions. system must also be designed such that all delivered engine systems will performance specification. Therefore, the effects of variations in turbopump
In a design-point nozzle area ratio,
headrise,
the
by the LOX and
Engine-System Characteristics
turbopump
2.3.1.1
ignition.
in the H-1 engine system is accomplished burns for approximately 100 msec after
ensure
in the purge
flowrate,
and
turbine
factors such are combined turbine engine
flowrate
87
as engine type, chamber pressure, with initially assumed factors such
pressure ratio (for a GG or a tapoff cycle), performance factors to predict the pump requirements.
In making
this
balance,
the
turbine power is equatedto the sumof the oxidizer pump power,the fuel pump power, and any auxiliary power that might be in this system. A design-pointsystembalanceof the final design-pointcharacteristicsof the turbopump is also necessarybecauserefinements and changesthat occur during the detail designo,f the rocket engine systemin turn can affect the turbopump system operation. For example, during the developmentof the MA-5 sustainerengine, the turbine and pump efficiencies when tested were below the initial values usedin the engine-systemdesignbalance.This condition necessitatedchangesto increasethe turbine drive pressureavailablefrom thegas generatorso that the enginewould meet its ratedthrust value, i 2.3.1.2
OFF-DESIGN
SYSTEM
BALANCE
Frequently, an off-design operating for a particular component than operating range is not carefully at the design point may either
condition of the turbopump system may be more the nominal design condition. Therefore, if the
severe entire
investigated, a turbopump system that operates satisfactorily fail or perform unsatisfactorily during off-design operation.
An example of the effect of operating range on turbopump be found in the J-2 upper-stage engine. To meet flight 230 000 lbf thrust while operating at an oxidizer-to-fuel
system design and operation can requirements, this engine delivers mixture ratio of 5.5; the thrust
then
ratio
decreases
requirement requirement
to a value
of 170 000 pounds
at a mixture
of 4.5. The
for the fuel pump occurs at a mixture ratio of 4.5, while for the oxygen pump occurs at a mixture ratio of 5.5.
The first step in defining the operating range is to determine A system balance (sec. 2.3.1.1) that includes the predicted turbopump is used to determine engine throttling and mixture-ratio
the
(which are due to engine throttling, and tolerances and the
turbopump
range
must
power
the maximum
power
of planned operation. characteristics of the
entire range of turbopump operation over the planned excursions. In addition to the predictable excursions of
turbopump operation atmospheric pressure, due to manufacturing operating
the range operating
maximum
be
known variations of pump inlet pressure, mixture ratio), engine component variations effect of these variations on the required
allowed
for
in the
initial
design.
These
effects
are
treated by first obtaining known component tolerances from existing liquid rocket engines as a basis for estimating, for example, the hydraulic-resistance tolerances of lines, valves, and nozzle coolant passages; the tolerances in the pump head/flow and efficiency/flow relations; and
the
2.3.1.3), normal
tolerances
of the
etc. The (Gaussian)
tolerances of all of the components are assumed to have a statistically distribution. The effects of all of the component tolerances are not
algebraically probability taking the
added,
since
main-engine
this
would
nozzle,
the
be
of occurrence. The component square root of the sum of the
a
turbine-exhaust
condition
the controls
with
(sec.
a very
low
tolerance effects are summed statistically squares of the effect of each component.
by The
88
"worst-case"
nozzle,
individual component effects are determined either by performing an engine balance assumingthat the given component is at the extremeof its tolerancerangeor by using a linearizedversion of the engine-system equationsto determinethe influencecoefficientsfor eachof the enginecomponentsunderconsideration.
2.3.1.3 The
CONTROL
control
affect
points
the
majority
CONSTRAINTS used
required
for regulation
pump
of engines
discharge
to date
have
of engine pressure
used
thrust
and
a simple
and
thus
the
open-loop
mixture
ratio
turbopump system
can
system
markedly
design.
to control
The
turbopump
power. Either calibration orifices were used to control the flowrate to the turbine GG or a pressure regulator was used to control the oxidizer flow to the GG. The MA-5 and MB-3 rocket engines are examples of the latter. The J-2, F-l, and H-1 rocket engines all use calibration orifices in the turbine gas-generator lines to set power level. The RL10 engine used
fuel
(hydrogen)
heated
power the turbine; pressure and adjusts closed-loop
by being
to control a turbine
thrust-control
passed
through
the
thrust-chamber
cooling
jacket
to
turbine power, a control system senses thrust chamber bypass valve. The RL10 is one of few engines to employ a
system.
Closed-loop mixture-ratio control has been used in a number of operational rocket engines. This control is accomplished either with the main propellant valves or with a valve controlling propellant bypass around the pump. The MA-5 sustainer engine utilized control of the main
propellant
valve to regulate In engine
valves,
mixture
systems
that
'whereas
the
J-2 engine
has
an oxidizer-pump
bypass
use orifices
in the GG
feed
lines
for thrust
calibration
and a turbine
bypass for mixture ratio, there is a large change in thrust and subsequently discharge pressure as mixture ratio is varied (sec. 2.3.1.2). This type of control unsuitable for high-pressure exhibit wide variations in arrangement During
of the thrust
the
operating
design
staged-combustion pump discharge
chamber,
evaluation,
requirements
control
ratio.
of the
the
cycles, pressure
the turbine, influence
turbopump
because these systems as a consequence of
and the turbine of the
system
type
must
in pump system is
hot-gas
of engine
be evaluated.
basically the series
generator. control
Typical
system effects
on the of three
types of mixture-ratio control on pump discharge pressure requirements for a staged-combustion cycle are presented in figure 34. As shown, hot-gas valves in combination with a liquid-fuel valve decrease the pump discharge-pressure requirements relative to a system
with
a liquid-fuel
valve
alone.
However,
hot-gas
valves
have
not reached
the level
of
technology of liquid valves for control purposes and usually are avoided. Also, the systems with the added hot-gas valves have a slower response and are more complex. Therefore, the interaction between the type of control system and the turbopump design should be assessed before final system selections are made. Reference 30 gives the results of a detailed evaluation
of a control
system
for an engine
with a staged-combustion
89
cycle.
Mixture
Symbol
© ® ®
ratio
control
valves
--1
main
fuel
....
1 main
fuel
+ 1 hot
gas
main
fuel
+ 2 hot
gas
-----i
46O0
58OO
_- _ 420o
I:
380o
=
--',--_
5400
i
I
$
I
I
I
mixture (a)
34.
The dynamic of the engine
pump
-- Typica
effect
upon the constant,
Dynamic
of
type
pressure
of
mixture-ratio
(staged-combustion
Turbopump
transient slope of
[
response the pump
I
l
I
I
I
i
6 mixture ratio, off
Co)
Oxidizer
control
on
pusp
pump
cycle).
Analysis
conditions imposed on the turbopump system system need to be evaluated fully to ensure
requirements.
I
Englne
o/f
discharge
System
I
2600 5
ratio,
Fuel
Figure
these
M o
I
6 Engine
2.3.2
3000
[
system
characteristics
that
during the transient that the turbopump may
have
a strong
Operation will meet influence
of the engine system include pump type, turbopump time head/flow characteristic curve, pump NPSH requirements
during start, and turbine and pump arrangement. In addition to these factors, engine system factors such as engine starting method and propellant combination are highly important to the requirements imposed on the turbopump system.
2.3.2.1
START
Engine thrust buildup follows turbopump speed buildup. The turbopump be used as a measure of the turbopump speed buildup and the turbopump
time
constant
is a function
of pump
9O
rotational
inertia
time constant response rate.
and shaft
torque:
can The
r -
I
(15)
TqN where r = turbopump
time
I = turbopump
rotating
Tq
=
turbopump
constant,
shaft
N = turbopump
min.
mass moment torque
rotational
of inertia,
(design speed
value),
(design
in.-lbf
value),
The importance Of turbopump system characteristics with the exact engine system being considered. characteristics do not have a strong influence on because
this engine
utilizes
to the turbopump. minimize the effect pump
head-flow
a solid-propellant
lbm-in.
rpm
for engine system start varies For example, the turbopump the starting of the H-1 booster
gas generator
to supply
initial
The high-pressure, high-energy gases rapidly accelerate of turbopump inertia. The rapid start also minimizes
characteristic
and
of the pump
NPSH
requirements.
turbopump system characteristics are much more influential booster engine, which employs a tank-head type of start. The
On the
starting
greatly system engine, energy
the pumps and the effect of the hand,
the
on systems such as the tank-head start employs
other
F-1 the
pressure available from the vehicle tanks in order to initially supply propellants to the gas generator for the combustion gases to power the turbine. On this type of engine, reductions in start time can be achieved by reducing the turbopump time constant (eq. (15)). The time constant reduce
varies directly time constant.
with the rotating inertia and, therefore, Increases in turbine torque during start
decreases in inertia also will reduce the
will time
constant. The
ground-level
atmospheric
pressure
adversely
operating levels for the GG-type engine by reducing the turbine pressure ratio will always be high, developed for the same turbine can be increased by enlarging valve (which turbine nozzle On the
opens
is a positive area or turbine
an engine mixture
with ratio
affects
flowrate. For ground-level the nozzle area, thereby
at low power
levels but
restricts
developed
torque
at low
starts, turbine low-speed torque increasing the gas flow. A hot-gas
the flow
device for reducing start time. For flow may reduce start time as much two turbopumps, during the start
turbine
the operating pressure ratio. At altitude, and significantly more power will be
at mainstage)
in series
example, a 10-percent as 50 percent.
the relative turbopump transient. A relatively
with
the
increase
in
time constants can influence high time constant on one
turbopump will cause a lag in pump speed buildup and shift the mixture ratio during start. On a LOX/LH2 system, a bias to reduce the oxidizer speed can be very beneficial, since the bias will avoid
temperature
spikes
in the gas generator.
91
The shapeof the curvesfor pump head/flow characteristicscan be significant in engine transientbehavior.Steepslopeswill tend to stabilizeflowrate andreducedischargepressure variations.For a bootstrappingsystemwith a gasgenerator,variationsin the pump discharge pressuremay produce excessiveexcursionsin GG temperatureunless active controls are used. Whenliquid hydrogenis utilized asa thrust chambercoolant,the head/flow characteristics of the fuel pump mustbe carefully analyzed.During the start-transientbuildup, the relation betweenthe pump and the thrust chamberwill force the fuel pump to operateat reduced valuesfor the ratio of flowrate to rotational speed(Q/N). This is equivalentto operationat reducedflow coefficient _band can result in stall. The J-2 enginewasoriginally designedfor a tank-headstart. Digital computersimulationof the enginestart indicateda long slow start during which the fuel pump encountered the discontinuity due to stall in the head/flow curve.Stall in turn produceda headlossanda further reduction in Q/N. WhenQ/N dropped below approximately 1/3 of the designvalue,hydrogenvaporizedin the pump, a condition that resultedin a complete lossof dischargepressureand a total stoppageof fuel flow. To solve the problem, the tank-head start for the J-2 engine was abandoned and a pressurized-gas start systemwasusedsuccessfully. Pump cavitation during enginestart canaffect pump speedbuildup. Whena pump cavitates, lessenergyis required for fluid pumping and more is availablefor increasingthe rotational speed.An enginethat routinely experiencespump cavitationduring start canbe sensitiveto cavitation parameters.Typically, start-transient pump cavitation is observedduring high accelerations of suction-line flowrate. Such accelerations can be produced by a high-poweredturbine start or a rapid-openingvalvein the main propellant line. Under these conditions, start time and thrust buildup rates may vary with pump cavitation characteristicsandpropellant conditioning. Enginesystemstudiesconductedduring the past severalyearshavecomparedthe transient dynamics of turbine drive cycles (sec. 2.1.1.4) that differ from those actually usedin production rocket engines.Among thesecyclesare the thrust chambertapoff and the heat exchanger.For the thrust-chamber-tapoff cycle, with O2/H2 propellants and a separate turbine for eachpump, a parallel-flow turbine arrangementwill produce a fasterstart and also minimize the starting differences due to sea-leveland altitude environments. A series-flow turbine arrangement(becauseatmosphericpressurewould affect the exhaust pressureof the second turbine) would show a large difference in starting characteristics when sea-leveland altitude starts arecompared.The overallstart transientfor thi ssystemis primarily associatedwith the speed buildup of the oxidizer pump because,for a series turbine arrangement,the oxidizer turbine is usuallythe secondor downstreamturbine. The start dynamics of a 'heat-exchanger furnishes
heated
hydrogen
gas to power
cycle, in which a thrust-chamber the turbines, favors a series turbine
92
heat exchanger arrangement. In
this case,the seriesconfiguration achievesa fasterground-levelstart, sincethe overallstart transientis primarily associatedwith the speedbuildup of the fuel turbopump.
2.3.2.2 In
THROTTLING
future
applications,
Although concerning development dynamic
the
engine
thrust
no operational turbopump rocket engine throttling and
response
Throttling
from
analytical
aspects
response
be
varied
been both
important
to meet space vehicle requirements. determined by the turbopump time
(throttled)
during
the
mission.
throttled, considerable information from experimental rocket system
computer-simulation
of rocket-engine
becomes
may
engines have is available
studies.
This
section
will discuss
the
throttling. when
rapid
rates
of thrust
change
must
be achieved
Turbopump system response to throttling commands constant (sec. 2.3.2.1). As shown in equation (15),
is the
turbopump moment of inertia and the turbine driving torque are the factors that determine the time response of the system. Engine development and checkout for space launches make it desirable for the engine to exhibit similar throttling behavior for both altitude and ground-level simulation
of
environments. high altitude
turbine-drive
system,
a
arrangement,
because
the
pressure. therefore, cycle,
a series-turbine
desirable The
For low-thrust either turbine than
choice
capability response However, requires
This is especially true for large engine systems where environment would be costly. For a thrust-chamber-tapoff
parallel-turbine series
of the
type
is more
engines, altitude simulation arrangement can be utilized.
arrangement
a parallel
arrangement
arrangement
has been
shown
is
more
sensitive
desirable
can be performed For a heated-hydrogen
through
than
to variations
computer
a
the series
in atmospheric more easily and, heat-exchanger
simulation
to be more
arrangement. of throttling
control
system
largely
fixes
the
maximum
response
of the system. Generally, throttling of the hot turbine working fluid gives slower than direct throttling of the liquid-line valves to the main thrust chamber. the additional pressure drop necessary for throttling of the liquid-line valves the
turbopump
to operate
at a speed
higher
than
that
for
throttling
with
hot-gas
valves. Pump head/flow characteristics coefficients must be maintained Liquid-line
throttling
is attractive
increases pump discharge on the hydrogen pump. The ramifications feasibility
has been
are
particularly important high to prevent vapor from
the
standpoint
on the hydrogen formation within of
rapid
thrust
response,
pressure during throttling and can create flow coefficient To augment flow, hydrogen can be recirculated around
of hydrogen
recirculation
have
demonstrated.
93
not
been
completely
side. Flow the pump.
determined,
but
it
problems the pump. but
the
2.3.2.3
SHUTDOWN
Engine shutdown normally is accomplished by first cutting turbine power and then closing off the main propellant flows. With the main propellant valves downstream of the pumps, inlet pressure surges will be caused by both pump speed decay and the closure of the rfiain propellant valves. flow deceleration
Pump and
speed decay will tend to be exponential and produce an immediate suction pressure surge. Valve closure and the resultant flow
deceleration will depend on design details for the valve. A 500 msec cutoff engine system and necessitates very rapid closures of main-line valves. pressures
of several
hundred
psi
structure and must be designed is the sum of the main tank pressure resulting from the end The
fluid
surge
pressure
can
result.
The
• Inlet-line • Closure • Rate •
at engine
shutdown
geometry rate
(line diameter
of the propellant
needs
is the
is entirely
analogous
to the
most
critical
inlet pressure fluid column
water
hammer
outlet valve of a simple pressure conduit The following parameters influence the
of pump
of elasticity
and length
propellant of inlet-line
to be performed.
surge pressures than liquid hydrogen; has a low inertance that reduces surge.
upstream
of shutdown
valves)
valve
For long inlet lines, as in the forward rapid valve closure. In order to define conditions
typically
compressibility
of decay
Modulus
inlet
to withstand the overpressure. The maximum pressure, the cutoff surge pressure, and the of vehicle acceleration.
phenomenon that results from rapid closure of the carrying fluid flowing with steady uniform velocity. magnitude of the surge pressure: • Propellant
pump
is fast for a large Propellant surge
flow material
tank of a vehicle, high surge pressures can result from the maximum surge, a detailed analysis of transient
Liquid
oxygen
because
and
RP-1
generally
of the low hydrogen
will produce
density,
higher
the feed system
An engine system can also be shut off with main-line valves upstream of the pumPs,. In this configuration, the pumps will be forced into deep cavitation and will not experience high surge pressures. However, need to be considered.
the
surge
pressures
94
in the
inlet
duct
upstream
of the
valve
will
2.3.2.4 Pogo
SYSTEM
INSTABILITY
is a vehicle-system
instability
system, vehicle structure, frequency (5 to 25 Hz) pumps
are important
system
resonant
(POGO)
and along
involving
components
frequencies
of the
inlet conditions of small vapor
small which
vapor pockets markedly change a relatively high fluid inertance
system
overall
of inertance
system
and must
and inertance line, create
inertance
vary
pump
and
be considered
when
feed
total
speed,
and mode
The compliance is a inducer vanes; these
compressibility of the fluid
and would
manner in but pump inertance require a
terms.
at the pump suction, coupled with the relatively a resonant system with low damping. Compliance flowrate,
during a flight, the suction frequencies that are functions
propellants are consumed When a vehicle structural
displays both description
and compliance
compliance of a suction with
tanks
the effective fluid compressibility. The is formed is not completely understood,
that the fluid that a complete
The fluid inertance Therefore, resonant
propellant
are influenced by fluid compliance and inertance. pockets forming along the leading edges of the
testing has revealed Some data suggest
distributed
engine,
are calculated.
Pump result
dynamic effects.
the rocket
forward payload. The instability oscillation typically is low the vehicle longitudinal axis. The rocket engine propellanl
fluid
temperature,
system resonance of the structure
and
will vary. and mass
suction
high and
pressure.
The vehicle distribution.
also has As the
the vehicle mass decreases, resonant frequencies increase. matches a pump suction mode, a condition that can display
Pogo oscillations is created. Pump suction-pressure oscillations and the accompanying flow oscillations feed through to the thrust chamber and produce small thrust oscillations. These oscillations in turn feed into the structure and cause relative motion in the vehicle. The loop is then
reinforced
by a feedback
from
vehicle
motion
to propellant
acceleration
at the inlet
duct, thereby creating pump inlet-pressure oscillations. Since the vehicle mass is continuously changing, the tuning of the structure and feed system (if it occurs) is normally transitory. Because the vehicle geometry and operating role in a Pogo instability, potential changes cannot limited resulting
guarantee elimination to steep head/flow from
high pump
Analytical techniques through 34). Empirical and vehicle structural and design
changes
parameters at the pump in pump inlet conditions
inlet play such through design
a strong changes
of Pogo instabilities. Pump design considerations at present are characteristic curves, which reduce oscillation amplitudes
gain.
to predict the occurrence data are used to determine dynamics are obtained from
are then
investigated
analytically
95
of Pogo have been developed (refs. 31 pump compliance and inertance values, a structural analysis. System operational to determine
the effects
on Pogo.
2.3.3
.System Development
Testing
The turbopump system is tested during development to ensure that the system meets design requirements. A certain amount of the testing can be conducted on a turbopump test s_and, while the final testing as part of the engine system verifies that the turbopump system will operate
satisfactorily
in the operating
2.3.3.1
TURBOPUMP
environment
of the rocket
engine.
SYSTEM
Testing of the turbopump system usually subsystem testing. Volutes, pump back generally pressure tested individually modify the design ;,the tongue region
begins with a certain amount of component plates, manifolds, and similar components
or as subassemblies as early as possible in volutes often is critical. Brittle-lacquer
and are
to confirm techniques
locate the regions of maximum stress and strain gages to evaluate the stress generally employed. This testing for structural integrity is routine in nature, but proper attention can avoid serious problems. Pump
rotors
or turbine
disks
spinpit tested. Prototype determined by precise operated margins.
to
For critical is conducted
burst
are designed
to operate
components are first diameter measurements
speed.
This
procedure
standard
balancing
the
major
the liquid hydrogen traces, rapid weight
with capacitance-type were reduced from
testing
of the
means) is conducted performance before checked,
the
point after
design
and
and
limits
are to it
are generally
of general yielding testing) and then establishes
production but with
actual
(as are
safety
part. The spin test all holes and other
(if in high-stress regions). Spinning to produce yielding in such residual stresses and can increase the speed capability of the part. procedures,
lubricant (in place of photos of oscilloscope
Initial
tested to the before and
verifies
balanced at 1000 rpm on an adequate balancing to full design speed of 34 000 rpm was performed balancing was done in five planes in a high-speed
measured excursions
near the speed
applications, it is also customary to spin test each with critical surfaces in the semifinished state
discontinuities machined regions produces reverse In
that
or to
components
used). were
gap gages. With these 0.015 in. to 0.002 in.
turbopump
assembly
(or
cavitation-performance
runs
the
the
assembly
are
By use of quick-developing made. Shaft excursions were
procedures,
pump
only
integrity and The head/flow
are made.
96
then
machine. However, full-speed balancing up on a Mark 25 liquid-hydrogen pump. The vacuum chamber with a Freon-21 bearing
normally corrections
to establish the structural installation in the engine.
and
Testing
the
peak-to-peak
if it is driven
by
shaft
separate
to verify pump and turbine (H-Q) map of the pump is
must
proceed
very
carefully
at
first to preclude any rubbing problems or Sealleakagesthat could lead to catastrophic failure. Various test fluids can be used,and the designmust accommodateoperation with thesefluids. Air often is usedas the test fluid for all types of pumps,becauseof the costof propellanls and the size and complexity of drive equipment. Air testing is especially suitable for liquid-hydrogen pumps, since low-speedoperation with air can produce a compressibility effect comparablewith that of hydrogen. Low-temperatureeffects, of course,are absent with either air or water tests. Liquid nitrogen hasoften beenusedfor testingliquid-oxygen pumps to avoid the explosion hazardwhile providing low temperature.However, the high bearingand sealwear ratesthat result from liquid-nitrogen testingmust be consideredwhen the test programis formulated. The turbopump is properly instrumented for the turbopump-systemtests. Measurements normally include inlet and outlet pressures,temperatures,flowrates, and speeds.Also, accelerometersaremounted on the pump to detect high vibration levels. Tests on seals,bearings,and similar componentsand material compatibility testsare run individually prior to turbopump systemtesting. Then, in the turbopump-systemtests, the componentsareverified asa part of the entire assembly. One aspect of turbopump-systemtesting that bears additional emphasisis the "limits testing" concept.Testing accordingto this conceptexposesthe turbopump to the extremes of the operating environment to be experiencedin final engine-systemand flight testing._ The purposeof limits testing is to exposedevelopmentproblems early in the development cycle andto avoid costly catastrophicfailuresduring the engine-system test program.
2.3.3.2
ENGINE
The
turbopump
rated
thrust
turbopump complete
SYSTEM must
and
perform
operate
designer, means engine assembly.
to design
at
the
specifications
design
mixture
any rubbing
The operating hot-gas testing
engine
or high rolling
resistance
system
Engine-system
testing of the turbopump assembly The turbopump is usually well
engine-system tests. Measurements include flowrates, and speeds. Provisions for applying necessary to make sure that excessive turning detect
if the ratio.
is to meet testing,
as an integral instrumented
to
its the
part of the during the
inlet and outlet pressures, temperatures, turning torque by hand between test runs are resistance is not present. Torque checking can due to seal or bearing
problems.
environment within the .engine system can be very different from that in of the turb0pump system, especially as far as vibration levels are concerned.
The compression effect of the fluid entering be simulated more closely by engine-system
the pump that can lead to Pogo instabilities can testing, since the inlet-line geometry to a large
9?
extent resemblesthat of the flight system with the propellant tankagerelatively close coupledasit would be on the flight system.In turbopump-systemtests,the propellant tanks are kept more separatedwith the result that the inlet-line geometryusually doesnot closely simulate the flight system. Additional propellant shutoff valves and fire-extinguishing systemsareusedon engine-systemteststo help preventcatastrophicaccidents. The thermal conditioning requirementsof the turbopump systemare best defined through tests performed during enginetests.A fuel turbopump stall that occurredduring the start of the J-2 enginewas related to the thermal conditioning requirementsof the main thrust chamber. A solution to this problem was to incorporate thrust-chamber Conditioning techniquesto reduce fuel-systemresistanceduring enginestart. Thermalconditioning of the J-2 turbopump wasfound to benecessary,anda recirculation systemusingsmallelectrically driven pumpswas addedto the vehicle system.With this arrangement,propellant from the main tank is recirculated through the main pump andthe thrust chamberandthen returned to the tank. An intermediateposition on the main oxidizer valvewas provided to prevent excessivespeedbuildup of the oxidizer turbopump. Explosionsin liquid-oxygen pumpshaveoccurred during engine-systemtesting.Explosions on the H-1 turbopump werepreventedby first installing a shaftdeflection deviceto measure shaft radial movement.The results showedthat at start there were large deflections that were related to propellant main valve opening time and sequence.Inlet liners of Kel-F material wereincorporated. Rubbing betweenmetal rotating hardwareand the Kel-F liner wasnot detrimental,and the solution wassatisfactory. Several liquid-oxygen pump explosions occurred during engine-systemtesting of the F-I turbopump. Due to the high horsepowerand dynamic environmentof the pump, normally acceptedspline and pilot fits were insufficient to preventfretting and rubbing that ignited the pump materials in the liquid-oxygen environment. Fits on all rotating parts were tightened, and thermal techniques(e.g., heating the impeller so that it could be slippedon the unheatedshaft) wereusedto assemblethe pump so that at liquid-oxygen temperature all partswereoperatingunder an interferencefit. In the F-1 turbopump, leakageof the oxidizer past the primary sealin combination with leakagepast the fuel seal would result in propellant contact and subsequentlyin an explosion.The designsolution wasto use an intermediatesealpurgedby inert gasbetween the oxidizer and fuel seals.The purge gasis expelledfrom eachsideof the sealat a slot and drainedoverboardby a drain line. On the H-1 engine, momentaryleakageof
LOX through the LOX seal occurred at engine start. This leakage was caused by temporary pressure imbalance of the seal by a pressure surge. Holes were drilled through the seal housing so that pressure could act on the back side of the carbon seal coincident with the pressure surge on the carbon nose. This action prevented
the
separation
of the seal and the mating
98
ring and eliminated
the problem.
3. DESIGN
CRITERIA
Recommended 3.1
Practices
PRELIMINARY
DESIGN
3.1.1
System
3.1.1.1
PUMP HEADRISE
The
headrise
to produce To
and
determine
drops
its design the
Estimate
delivered
by the pump
discharge-pressure
downstream the pressure and
the preburner
the pump
FLOWRATE shall be adequate
for
the engine
thrust.
pump
(if applicable), across
AND
and flowrate
drops that occur chamber pressure jacket
Requirements
add
the
engine-system
pressure
of the pump discharge. For gas-generator cycles, add to the drops due to line losses, valve losses (if any), the regenerative
the
injector.
injector
headrise
requirement,
from
For
and the
staged-combustion turbine
cycles,
and the line losses
include between
the
pressure
them.
the expression
H=
144 [(Po)z-
(Po)]
(16)
Pl where
For
(Po)2
= pump
discharge
(Po)l
= pump
inlet
total
Pl = pump
inlet
propellant
a high-pressure
isentropic
enthalpy
hydrogen rise from
total
pressure,
pressure,
psia
density, pump
psia
lbm/ft
(above
the propellant
3
2000
psi pressure
properties
and
calculate
rise),
obtain
the
the corresponding
required head
rise from Hisen = J _hid)2-hl]
99
(17)
where Hisen= headrisefor an isentropiccompressionfrom (Po)l (hid)2 =
ideal
specific
enthalpy
at
hi = inlet
specific
enthalpy,
Btu/lbm
to (P0)2,
ft
Btu/lbm
(Po)2,
J = 778 ft-lbf/Btu
If the propellant is sufficiently compressible heating effect by applying equation (17) determine enthalpy
the actual enthalpy increment (isentropic
previous actual enthalpy, illustrated in figure 35. caused by the recirculation To determine the requirement from
in the application, account for the propellant at increments between (Po)l and (Po)a; i.e.,
at the beginning of a pressure enthalpy increment divided
increment by adding the actual by the pump efficiency) to the
and then sum the isentropic headrise increments. This procedure is When making this incremental calculation, include the heating of the thrust-balance-system flow.
volume
flowrate
requirements,
obtain
the
total
weight
flowrate
F
wE -
(18)
(Is)e where wE = engine
Then use the fuel pumps:
total
following
weight
flowrate,
equations
lbm/sec
to obtain
the
WE(MR) (Qo)e
=
volume
flowrates
for
the oxidizer
and
the
(448.8)
po (1 + MR)
(19)
WE(448.8) (Qf)p
-
of ( 1 + MR)
100
(20)
I I
zip
n
t
z_P 2
o o
0 0
AP 1
O3
sl
Specific
H.1sen
= J (Ahsl
Ahact
.
Figure
35.
-
Ah_
Illustration isentropic
entropy,
+
+
Ahs2
s,
Ahs2
+
of incremental headrise.
101
Btu/ibm-°R
+
Ahs3
Ahs5
method
+ ....
+
for
....
)
)
determining
where (Qo)e
= oxidizer
(Qf)P
=
fuel
pump
pump
Po = oxidizer
volume density,
0e = fuel density, MR = engine 448.8
= factor
volume
flowrate, lbm/ft
lbm/ft
mixture
flowrate,
ratio,
for converting
For preliminary design estimates, sizing, use the average or the local
gpm
gpm
a
3 (Wo/VVf)E ft 3/sec
to gpm
use the density.
inlet
If off-design operation is required, use the headrises and volume flowrates over the entire
3.1.1.2
NET POSITIVE
SUCTION
density.
For
more
detailed
flow-passage
procedure described above to determine engine operating range (sec. 2.3.1.2).
the
HEAD
The pump net positive suction head shall be suitable for the particular application, shall be adequate for stable and predictable pump performance, and shall minimize vehicle overall weight. Determine the methods for obtaining stable and predictable pump performance by consulting sections 2.2.1.1, 2.2.2.1, and reference 1. Then, if the engine contractor participates in the NPSH selection, conduct an optimization that considers the effects of pump inlet pressure and NPSH on vehicle tank and pressurization system weight, pump efficiency, turbopump weight, and system cost. The effects of NPSH on some of these factors are illustrated graphically in figure 36. Sections 2.1.1.6 and 2.2.1.1 discuss the effects of pump geometry on efficiency and suction performance. Convert the pump efficiency to turbopump equivalent weight (sec. 2.1.2.2). Then add the weights and select the optimum NPSH. If NPSH is specified to the engine manufacturers, then optimize the weight and performance at that NPSH.
3.1.1.3
PROPELLANT
PROPERTIES
The turbopump system design shall reflect the impact individual propellants and of the propellant combination. Because propellant design, use the considered:
properties have a major influence following checklist to make sure
102
of
the properties
on all aspects that all fluid
of the
of turbopump system property effects are
0
¢-J
_ _._ .rd
E--
I
I I
Y: ©
%d (t)
e_
o
_q
Figure
36. -
NPSH--------_
NPSH-------_-
NPSH.--.-.----_-
Effects
of variations
in pump
NPSH
on various
design
factors.
(1)
Density on arrangement
all aspects of turbopump design, (secs. 2.1.2.3 through 2.1.2.6 and
(2)
Material through
(3)
Corrosive, seal speed
(4)
Specific heat, specific heat ratio, on turbine geometry (see. 2.1.2.6).
(5)
Cavitation
characteristics
(6)
Two-phase
acoustic
(7)
Propellant saturation temperature rotor alignment (sec. 2.2.4.3), 2.2.4.6).
compatibility 10).
on
cooling, lubricating, limits (sees. 2.2.1.2
material
selection
on inducer
velocities
for
and viscosity and 2.2.1.3). and
on zero-NPSH
all
weight
overall turbopump 2.2.1.4).
wetted
characteristics
molecular
design
surfaces
on
bearing
of turbine
(refs.
DN
working
1
and
fluid
(see. 2.2.1.1). pump
on material and thermal
103
including the 2.2.1.1 through
capability
(sec.
2.2.1.1).
selection, turbopump conditioning (sees.
housing 2.1.1.8
and and
3.1.1.4
TURBINE
DRIVE
The turbopump
CYCLE
system
shall be compatible
with
the turbine
drive
cycle.
For cycles in which the turbine is in parallel with the thrust chamber (GG and tapoff cycles), minimize the required turbine flowrate by using high turbine pressure ratios (_ 15 to 25). Note that this practice requires some compromise with rotational speed, weight, efficiency, and turbine inlet temperature (sec. 2.2.3.1). Allow sufficient turbine discharge pressure for exhaust disposal (sec. 2.2.3.2). For cycles in which the turbine is in series with the thrust chamber (expander and topping cycles), maximize turbine flowrate, efficiency, and inlet temperature so as to minimize pressure ratio and engine weight. Using maximum values for if, r/, and (T0)l, calculate the pressure ratio fro,-, the expression (all parameters referred to turbine)
(21)
PR= 1-
r/J 550 Cp(To) (HP)
(ff
1
where PR = turbine
pressure
ratio,
(Po)l
= turbine
inlet total
(Po)2
= turbine
discharge horsepower
r/ = turbine
efficiency
Cp = specific (To)l
weight
pressure, pressure,
HP = turbine
= turbine
(Po) 1/(Po)2 psia psia
flowrate
heat
of turbine
= turbine
inlet
total
temperature,
")' = specific
heat
ratio
of turbine
550 = factor
for converting
working
fluid,
Btu/lbm-°R
°R working
hp to ft-lbf/sec
104
fluid
Add sufficient margin to this value to allow for efficiencies may not meet the initial predicted values.
3.1.1.5
THROTTLING
The turbopump Generate
the
fact
that
the
pump
and
turbine
RANGE operation
pump
the
shall be stable
head-versus-flowrate
over
the entire
characteristic
operating
required
range.
by
the
engine
during
its
maximum throttling excursion. Plot this engine requirement on the performance map for each pump candidate, as. shown in figure 7. If the pump stability limit crosses the requirement line during the engine excursion, do not use that pump design, because it will not meet the throttling requirement. In general, throttle to less than 50 percent of design.
3.1.1.6
use centrifugal
pumps
if the
engine
to
engine
is to
EFFICIENCY
The
turbopump
efficiency
shall
be
adequate
for
the
meet
its
requirements. For
engine
cycles
and
tapoff
cycles),
in which
the
determine
turbine the
is in parallel
maximum
F
allowable
with the thrust turbine
r(i )x/c-(ioE
1
Then
obtain
= thrust
chamber
the minimum
specific
allowable
impulse, turbopump
105
lbf-sec/lbm efficiencies
from
(gas generator the expression
(22)
where (Is)r/c
chamber
flowrate
from
w
_Vp
= (Wo)T
_V T
d- (_V_/f ) T
Hp
(23)
=
#
_p
T_ T
J
Cp(To)T
1 -- l\pR]
1
-r-_ i, .[
m
_/p
'T_p 7_T
J Cp(To)
Hp
T 1
--
p-R
where
fiT = total turbine (Wf)T = fuel-turbine
weight weight
(Wo)T = oxidizer-turbine weight
Hp = pump
headrise,
r/e = pump
efficiency
(To)T
1
flowrate,
efficiency
= turbine
inlet
in which cycles),
lbm/sec
flowrate,
lbm/sec
lbm/sec
ft
= turbine
For engine cycles staged-combustion
lbm/sec
flowrate,
weight
ffp = pump
7/T
flowrate,
total
temperature,
the turbine obtain the
°R
is in series with the effect of turbopump
thrust chamber efficiency on
requirements includes a
(fig. 14 is an example for a typical engine) mathematical steady-state representation
from an engine-system of all of the engine
Determine
the
chamber
chamber
constant-efficiency
pressure.
underestimates
Add
of system
sufficient pressure
curve
whose
margin
for
drop.
The
efficiency.
106
peak
overestimates result
pressure of turbopump
is the minimum
(expander and engine pressure analysis that components.
equals
the
efficiency
allowable
design and
turbopump
3.1.1.7
WEIGHT
The
weight
other
AND and
SIZE size
of
the
turbopump
the
reliability,
NPSH,
turbopump-system and
design
performance.
maximum
rotational
speeds
In so doing,
seal rubbing speed, turbine-blade pitchline velocity (sec. 2.2.1).
3.1.1.8
shall be minimal
consistent
with
requirements.
Maximize
Unless rotors.
system
centrifugal
turbopump-system
stress,
efficiency
within
consider and,
is critical,
the
inducer for
limitations
cavitation, geared
of life,
bearing
DN,
turbopumps,
do not use more
than
gear
two turbine
CONDITIONING
For restart applications with cryogenic require a minimum conditioning time
propellants, the and a minimum
turbopump system shall amount of conditioning
propellant. Minimize the pump temperature pump from the turbine: minimize minimize single-shaft
the turbine turbopumps
Evaluate
coating
the
material that will body of metal. Increase pumps,
3.1.1.9
take
mass relative (fig. 17). wetted a rapid
RELIABILITY,
to the
surfaces
the two-phase pumping if necessary (sec. 2.2.1.1).
LIFE,
rise during the contact
the
total
number
pump
of the
chill and
capability
AND
mass
pump
(sec.
with
also reduce
the
of the
2.2.4.6).
a thin rate
inducers
by thermally isolating at the contact points,
layer
of heat
and
research addition,
geared
and
of low-conductivity rejection
evaluate
of turbopump
units
the
life
and
reliability
the
to the
main
use of boost
to be built
and
the
requirements
intended
determine the effect of design variations on performance, weight, turbopump-system life and reliability requirements at a minimum low-speed of inducer
Evaluate
the and
COST
The turbopump system shall meet , mission at minimum overall cost: For
a shutdown period area, use insulation
designs, single-shaft arrangements fluid incidence angle to reduce and development costs to the assembly, manufacturing,
of
production
the
rate,
and cost. To meet the overall cost, consider
that eliminate gears, and the use of low values. cavitation erosion. In the cost analysis, evaluate
meet the handling
107
life and
and reliability requirements maintenance costs. Conduct
and, in a system
analysis to convert the performance and weight variations into payload variations. Then determine the cost per pound of payload variations by dividing the cost sums by the correspondingpayloads.Selectthe configuration that hasthe minimum cost per pound of payload.
3.1.2
Selection
3.1.2.1
NUMBER
OF UNITS
number
of turbopump
The
size and number Conduct
a cost
of System
of units
analysis
that
Type
units
per engine
produced
on total
considers
the
shall reflect
the impact
of turbopump
costs.
effects
of size
on
the
sum
of R&D
costs
and
production costs. Use a production learning curve to obtain average production costs for various numbers of production units. Then obtain total production costs for constant engine flowrate by multiplying the average unit costs by the number of units required to deliver the total
flowrate.
3.1.2.2
TURBOPUMP
The
turbopump
turbine For
engines
engine
EQUIVALENT-WEIGHT
(GG
system
flowrate in which and
design
and turbopump the
tapoff
specific
cycles,
shall weight
impulse
sec. 2.1.1.4),
FACTOR reflect to stage of the
evaluation
of
payload
weight.
turbine
use equation
exhaust (2)
the
equivalence
is less than
to determine
the
of
that
of the
turbopump
equivalent weight factor EWF. The factor OPL/O(Is)E in equation (2) is obtained from mission and vehicle analysis; note that PL refers to stage rather than vehicle payload. Then calculate the net effect on payload from equation (3). Use the parameters EWF and EW to evaluate For
turbopump
engines
in which
engine (expander essentially zero, weight
modifications. the
specific
impulse
of the
turbine
exhaust
and staged-combustion cycles), the turbopump and therefore the turbopump equivalent weight
(eq. (3)).
108
is equal
to that
of the
equivalent weight factor is equal to the turbopump
is
3.'1.2.3
ROTATIONAL
The
design
SPEED
rotational
hydrodynamic limitations. For a given coefficient,
speed
shall
performances,
pump turbine
the
and turbine type pitchline velocity,
reflect
evaluation
turbopump
weight,
(specified values etc.), determine
of
the
component
and
the
mechanical
of stage number, pump the effect of rotational
pump efficiency, turbopump weight, and turbine flowrate. Use equations determine the net effect of speed on payload. By means of a plot like that 16, determine weighing these
the effect of rotational speed on configuration. results against the other system requirements.
turbopump size (secs. 2.1.1.5 should never exceed the value If the
investigation
have potential,
3.1.2.4
of the
conduct
and 2.1.1.7) at optimum
given
a similar
TURBOPUMP
is unusually payload.
configuration analysis
shows
for each
(2) and (3) to shown in figure
Select design Unless throttling
critical,
that
inlet flow speed on
the
other
configuration
design
speed range
rotational
by or
speed
turbopump
configurations
and compare
the results.
ARRANGEMENT
The turbopump arrangement shall allow operation of the individual pumps and turbines at the speeds needed to produce the best overall system performance within the limitations of reliability and life. Use table size,
VII as a guide
propellant
type,
for screening
and
turbine
candidate
drive
cycle.
turbopump For
are dense, also consider the arrangement in figure in this arrangement, the turbine would reject preconditioning. geared) a prime with
multiple
In some candidate, candidates,
engines
based
in which
if the turbopump to both pumps,
both
predict
the
performances
and
weights,
the turbopump
and
use
on engine propellants
must be restarted; thereby reducing
cases, envelope restrictions make arrangement particularly for a small engine with restarts. For
(3) to determine the net effects on payload. Select these results against the other system requirements.
3,1.2.5
large 17(d) heat
arrangements
17(b) (pancake the applications equations
arrangement
(2) and
by weighing
PUMP CONFIGURATION
The
pump
maintaining
configuration the best
shall
compromise
For engines with thrust greater propellants other than hydrogen. single-entry
pump
would
deliver
the
among
required
the other
than 5000 lbf, use Consider double-entry
fall to the
right
of the
109
headrise system
and flowrate
requirements.
single-stage Centrifugal pumps for applications
peak-efficiency
while
specific
speed
pumps for in which a in figure
15.
Table
VII. - Guide for Screening Arrangements
Candidate
Turbopump
Recommended arrangement (fig. 17)
Engine Features Propellant type
Size
Turbine drive cycle Staged combustion & expander
Large
H2 fuel
(F > 25 000 lbf)
Dense oxidizer
Dense fuel
g
GG & tapoff
c,g
Staged combustion & expander
a,g
GG & tapoff
a,c,g
Dense oxidizer
Small
H2 fuel
(F < 25 000 lbf)
Dense oxidizer
Dense fuel
Staged combustion & expander
f,g
GG & tapoff
c,g,f
Staged combustion & expander
a,e,g
GG & tapoff
a,c,e,g
Dense oxidizer
Use shrouded impellers for liquid oxygen. shown in figure 19, and select the pump
For hydrogen, conduct an analysis similar to that type and number of stages by weighing the results
against the other system requirements. In general, use centrifugal pumps for hydrogen if the throttling ratio Qdes/Qmin is greater than 2. If the throttling ratio is very large, increase the number of centrifugal stages (figs. 13 and 19). If the ratio is less than 2, the axial pump also may be considered, particularly efficiency is very important. Do unshrouded titanium centrifugal
for missions with long operating durations where high not exceed approximate tip speed limits of 2800 ft/sec for impellers, 2000 (1700 to 2300 depending on design
specific speed, amount of sweepback, blade height, method of shroud attachment, etc.) ft/sec for shrouded titanium centrifugal impellers, and 1500 ft/sec for titanium axial rotors. For unshrouded centrifugal impellers, evalute the effects of tip clearance on efficiency (ref. 21) and the effects
of housing
weight
required
to maintain
thus
clearance.
Use the suction-performance relationships in section 2.2.1.1 to size the pump inlet to meet the NPSH requirements. Note that decreasing the inlet flow coefficient will not only decrease the NPSH requirement but will also decrease the pump efficiency (figs. 15 and 19). For
small
engines
are less than and (figs.
compare
400
(F < 5000 (figs.
other
lbf), initially
15 and
pump
types
analyze
18) or if impeller such
centrifugal tip
as partial-emission
18 and 20).
110
pumps.
diameters
If stage
are less than
centrifugal,
specific
speeds
1.5 in., analyze
Barske,
and
Rootes
3.1.2.6
TURBINE
The
CONFIGURATION
turbine
configuration
the best compromise Use axial turbines pressure-compounded
shall
among
the other
the
the required system
ratio
of the
do not turbine
pitchline
design'point
efficiency
curves
is unusually
critical,
velocity
Conduct
tradeoffs
between
similar
to the
(i.e.,
spouting
using
turbine
efficiency reference
more
than
two rotors.
cycles,
staged-combustion
consider
consider
temperature
and
weight
in which the the optimum
using
(secs.
Note,
2.2.1.4
partial-admission
reversing
the
the
ratio
however,
require additional and cooling. Figure
requirements
turbopumps to determine
21, use this velocity
to
turbine
and
turbines
so
that
that
supporting 23 may be
to determine
speed is combination
turbines in which the pitch diameter and, consequently, the pitchline must be reduced to keep a direct-driven, full-admission turbine
In
from
without causing axial thrust 4 for additional guidance.
that, in turn, for lubrication
flowrate
and inlet
in figure
For the
limit,
heights less 3 times the
(24)
speed,
minimum blade-height (figs. 17(e) and 17(0).
two-stage
velocity
diameter, small U/Co
ratio,
isentropic
to those
consider
diameter and, for centrifugal stress, pressure
rotors
1
large turbines require outboard bearings structure, an extra seal, and extra provisions used as a guide in selecting the type.
optimum rotor turbine-blade-root
achieving
requirements.
select the turbine type that will give the highest problems. Figure 23 may be used as a guide. Consult If efficiency
while
exceed 1800 ft/sec; do not use blade rotor diameters exceed approximately
/Co: From
horsepower
with, in general, no more than two and two-row (one-stage) velocity-compounded).
Use high pitchline velocities but than 0.15 in.; and do not let the pump impeller diameters. Calculate expression
deliver
the
limited by of rotor
2.2.3.1).
and
hot
velocity within
geared
gas
and the
turbines
from
the
precombustor can flow in through an annular axial inlet and, with the aid of a 180-degree annular elbow downstream of the turbine, the turbine discharge gas can flow back to the main combustion chamber through an annular axial discharge.
111
3:2
DETAIL
3.2.1
DESIGN
Limits
The and
turbopump low weight
curves
various
speed
2.1.2.2) funding
turbopump
limits
and
and
pump
configurational
is an example of this type with data relating turbopump
to compensate
turbine
efficiency,
3.2.1.1
not
exceed
required
efficiency
shall
remain
during the
free
from
the operational
inducer
suction
To
remove
preinducer optimization
information the
beyond
the
each limit.
example, assuming adequate 39 000 rpm is a reasonable
If the
curve
rotational
development selection for
speed
significantly
pump weight affects
on the plot.
cavitation speed
that
inducer
on inducer cavitation
speed
limits
upstream of and analysis
cavitation
overall
pump
profiles.
shown
turbopump a payload
in figure
24
at any
time
operation. optimization
If the pump inlet that considers
If long life (> 1 hr) is required, a be necessary. Reference 1 provides
limits.
on rotational
speed,
the main pump inlet. Use table to select the drive system for the
To pump propellants that are saturated inducer blade plus blade boundary-layer obtain the vapor-pumping capacity from
the
NPSH's near zero are desired, TSH limits expressed by equations (7) and
and tank weight. in figure 24 may
limit
impairs
and acceleration
specific
weight, pump efficiency, adjustment of the curves
detailed
to proceed
indicate
Use the information on the curves in and performance to vehicle payload (sec.
(8) (in combination with fig. 27) at any time during NPSH is not specified by the system, conduct
more
speed;
on figure 16; higher speeds would reduce bearings and seals, and would not reduce
two factors.
a turbine
vs rotational
the mission. For applications where operation should be considered. Do not exceed the NPSH
turbopump downward
performance limits of any
CAVITATION
inducer
performance
changes
speed. For requirements,
the other
include
INDUCER
The
for
efficiency
of plot. weight
turbopump optimization shown would require more complex
enough
during effects
weight the
shall provide for achieving good the mechanical or hydrodynamic
assembly.
as a basis for selecting pump and high vehicle-performance
the LOX efficiency,
Do
Speed
rotational speed without exceeding
in the turbopump
of
Figure 16 conjunction
INTEGRATION
to Rotational
component Plot
AND
place
either
a boost
pump
or a
VIII boost
as a guide in a system pump or the preinducer.
in the tank (zero tank NPSH), first determine the blockage (normally 20 to 40 percent) and then either figure 29 or from equation (9). For a narrow \
112
Table
VIII.
-
Drive system
Comparison of Typical and Preinducers
type
Drive
Advantages
Systems
for Boost
Pumps
Disadvantages
....
Gear (fig. 28(a))
Through-flow hydraulic turbine (fig. 28(b))
i
Positive speed control
Complex
Most efficient
Close coupled pump
Simple
Lags main pump start
Efficient
Close coupled pump
Relatively few rotating-shaft seals provide increased reliability
Limited
,
Recirculated-flow hydraulic
to main
to main
headrise
i
Allows remote location
Lags main pump start
In some appfications, more efficient than
Relatively
turbine
(fig. 28(c))
low efficiency
gas turbine
Gas turbine
Relatively few rotating-shaft seals provide increased reliability
Recirculated propellant may vaporize at turbine discharge
Allows remote
Complex
location
Allows pre-start
Requires source
separate
gas
In general, more efficient than recirculated-flow draulic turbine h
Electric
,,
motor
hy-
,
Allows
remote
location
Heavy electric
motor
(fig. 28(d)) Allows pre-start
Requires electrical energy source
Efficient
Limited
With a propellant.cooled motor, relatively few rotating.shaft seals provide increased reliability
113
headrise
range
of zero-NPSH
an (i//3)L
of 0.7;
operation,
use a value
for a wider
range,
vapor fraction and minimum allowable
the minimum inducer inlet
these
hydrogen
relations
pumping,
3.2.1.2
for
consult
references
BEARING
DN
bearings
shall
The
use
for (i//3)L of approximately
an (i//3)L
zero-NPSH annulus area.
and
oxygen.
operate
at DN
more
detailed
never
exceed
use the corresponding
tank-saturation pressure For zero inlet-line losses,
For
17, 23, 24, and
0.6 and
of 0.3 to 0.4. Then
to determine the figure 37 displays
information
on
two-phrase
25.
levels
that
will
permit
them
to meet
the life
requirements. Do
not
exceed
required
(10)
bearing
life including
a downward If the
the
preliminary
adjustment
turbine may
DN limits
be placed
(refs.
26 and
3.2.1.3
between
27) may
SEAL
the
calibration,
and
flight
the operating
operation
life. If
exceeds
1 hour,
size the shaft
the
speed
limit
on critical-speed for this
type
basis
(ref.
8). Equation
Of arrangement.
To permit
turbine bearing outboard (fig. 30(b)). Equation (11) speed limit. To permit even higher speeds, the pump
inducer
and
the
impeller
(fig.
30(c)),
maybe bearing
or a hydrostatic
bearing
be considered.
RUBBING
The seals shall
6 at any time during
may be necessary.
(fig. 30(a)),
to approximate
higher design speeds, place the used to estimate this rotational may
testing,
of the limits
is overhung
be used
in reference
operate
SPEED at rubbing
velocities
that will permit
them
to meet
time
during
the life
requirements. Do
not
exceed
the
seal speed
limits
given
in reference
7 at any
the
operating
life. If required life exceeds 1 hour, a downward adjustment of the limits may be necessary. Consider the use of static liftoff seals, which will eliminate dynamic rubbing seals. If the
turbine
8 provides are used,
is overhung,
detailed equation
equation
information (13) may
(12)
may
be used to estimate
the speed
limit.
Reference
on shaft sizing. If outboard bearings (figs. 30(b) and 30(c)) be used to estimate the limit. To permit operation at even
higher speed, advanced configurations are necessary (refs. 7, 26, and 28).
such
as liftoff,
114
hydrodynamic,
and
hydrostatic
seals
450
Ii00 11
45
10
5 40
400
i000
6
45
4
40
35 900
--
350
;
8oo -
o
300 700 20 <
-25
<
250
600,-2O
15
k
200 .,..q
500
--
400
--
Choked (max.
Ln
*/A)
Choked
0
150
(max.
0
O/A)
0
Flow
._
process
-
equilibrium
Tank conditions Solution method
lOO
:-.]
Official Subscript SO
isentropic
= saturated not critical
300
-Flow
liquid 200
minus
--
100
Line
vapor
20
30
volume
fraction, (a)
40
50
--
Po-PI
al'
percent
6O
0
refers
inlet
l i0 Line
l 20 vapor
40 fraction,
(b)
37.
-
Zero-NPSH for
oxygen.
pumping
capability
requirements
for
hydrogen
inlet
Oxygen
and
total
minus
pressure
l
30 volume
to
static
l
Hydrogen
Figure
liquid integration
Offical oxygen properties Subscript 0 , tank conditions 1 = inlet line conditions
0 t 10
= equilibriumisentropic
process
Tank conditions = saturated Solution method = numerical of momentum equation
Oo
hydrogen properties 0 = tank conditions 1 = inlet line conditions
Po-PI refers to inlet total inlet static pressure
©
l 50 al"
l 60 percent
l 70
3.'2.1.4
TURBINE-BLADE
The
turbine
which Do
not
blades
failure
exceed
CENTRIFUGAL shall
the turbine
N 2 Aa limits
given
a downward
adjustment
To optimize the turbopump that consider turbine inlet type,
pump
efficiency.
turbopump An iteration rotational
turbine
blade
Figure
speed for speed.
materials,
hot
that are below
the
4 at any time
of the limit may
turbopump flow charts
A pressure-ratio
in these
gas on
in reference
weight,
the levels at
which
is the
outside
can
cause
the
allowable
N2Aa
rotational
speed,
and
for determining
basic types of drive cycles. efficiency is a function of
result
of pressure ratios, is partially illustrated the centrifugal stress on the upstream
rotors
the operating
limit, conduct tradeoffs turbine pressure ratio,
procedures
limits for four because pump
optimization,
during
be necessary.
turbopump
of the analytical
and weight at the blade stress pump efficiency is indicated
compensate for the smaller limit somewhat. However, by
stresses
at the turbine-blade centrifugal-stress temperature, turbine inlet pressure,
38 presents
figure 38(a) over a range turbines are used, check temperature
at centrifugal
will occur.
life. If long life is required,
turbine
operate
STRESS
of applying
the
logic
in
in figure 31. If multi-rotor rotors, because the higher
to drop
enough
to more
than
annulus area. Turbine blade cooling can be utilized to relieve this before using turbine cooling, consider the thermal stresses caused and
cold
cryogenic
propellant
on the
inside;
for repeated
rapid
starts, these stresses can cause cracking due to thermal fatigue. If repeated starts cause thermal fatigue even without cooling, consider hollow blades with hot gas ducted to the hollow center. Hollow tapered blades can be used to increase the allowable N2Aa limit without
affecting
the
outer
contour
of the
blade
and, therefore,
without
penalizing
turbine
performance.
3.2.1.5
GEAR
The gears
PITCHLINE shall operate
levels at which In designing tooth
bending
a geared stress,
failure
VELOCITY at pitchline
velocities
and tooth
stresses
that are below
the
will occur.
turbopump, and tooth
observe compressive
the
recommendations
stress
116
given in reference
on gear 5.
pitchline
velocity,
TORBINE INPUT VELOCITY TURBINE
INPUT
PITCHLINE
VELOCITY,
INLET PRESSURE, INLET
RATIO,
O/C o
TURBINE EFFICIENCY, U
_T
INLET TEMPERATURE, (To)T1 TURBINE TYPE
(Po)TI
TENPERATURE, (To)T1
PRESSURE RATIO,
PR
TURBINE TYPE N2Aa LIMIT
"VELOCITY " TURBINE N2Aa
RATIO,
U/C o
EFFICIENCY,
_T PI'I_HLINEVELI_ITY,
LIMIT FOR TURBINE Pt_P
INLET PRESSURE,
INPUT
PRESSURE HEADRISE, TURBINE
FLO_ATE,
WT
EFFICIENCY, ANNULUS
ROTATIONAL
AREA,
SPEED,
TURBOPUMP TURBOPUMP WEIGHT, ROTATIONAL WTpSPEED,
TURBINE ANNULUS AREA, (Aa)T
I
ROTATIONAL
!
SPEED,
,1
NT
TORBOPOI_ TURBOPUMP ROTATIONAL WEIGHT, WTpSPEED,
NTp
(b) Staged-combustion (a) Gas-generator
and
tapoff
HI)
FLO_ATE, _p I Pt_P INPUT EFFICIENCY, _p
_p
I
NTp I
HEADRISE,
FR
Wp
1
(Aa) T
NT
RATIO,
U' (Po)T1
Hp
FLO_ATE,
"-4 TURBINE
FOR _BINE
cycles
Figure 38. - Flow charts for determining turbopump speed and weight at the turbine-blade centrifugal-stress limit.
i
and
expander
cycles
3.2.2
Pump Design
3.2.2.1
INDUCER
INLET
FLOW COEFFICIENT
The inducer inlet flow good pump efficiency. Do
not
design
at inlet
coefficient
tip flow
inlet flow coefficient, generate 19. Reference 1 provides more
3.2.2.2
shall provide
coefficients curves detailed
for good
much
below
suction
0.07.
To
performance
determine
and
the
impact
at the design point that are similar to those information on inducer suction performance.
of
in figure
STABILITY
The pump operating
shall
be stable
and shall
have
predictable
performance
over
the
entire
range.
For axial pump rotors and stators, do not exceed a blade diffusion factor DR of 0.70 mean blade diameter (or station) at any point within the design operating range:
W2
DB,ms
Wl
u
--
W2
at the
u
= 1 ----+----
(25)
W 1
2O
W 1
where D n,ms = blade wl
= inlet
diffusion
factor
fluid velocity
w2 = discharge
fluid
at blade
relative
velocity
to the
relative
station
blade
to the blade
wl u = tangential
component
of inlet relative
w 2 u = tangential
component
of discharge
= blade
solidity
(ratio
of chord
For centrifugal pumps, design such that zero-slope points on the performance-map pump
mean
always
For more references
operates
in the
detailed information 2 and 3.
negative
relative
length the
slope
velocity velocity
to spacing)
entire speed
operating lines or,
region
of the pump
on centrifugal
pump
118
stability
range falls to the right of tile in other words, such that the speed and
line (figs.
axial
pump
12 and stall,
13).
consult
3.212.3
TIP SPEED
The pump range
impellers
of rotational
shall
maintain
mechanical
over
the entire
operating
speed.
Design pump impellers with an adequate and the tip speed at which burst occurs; pump impellers. capability, the
integrity
For forged approximate
margin between the maximum operating tip speed reference 2 may be used as a guide for centrifugal
titanium, tip speed
which limits
is the material are 2800 ft/sec
with the highest tip-speed for unshrouded centrifugal
pump impellers, 2000 ft/sec for shrouded centrifugal pump impellers (1700 to 2300 ft/sec depending on design specific speed, amount of sweepback, blade height, method of shroud attachment, hole in the center, etc.) and 1500 ft/sec for inducers and axial pump rotors. Note
that
impellers references
3.2.3
3.2.3.1
the
tip-speed
limit
decreases
with
degree
of blade
sweepback
and that hydrogen pumps are the only pumps that approach 1, 2, and 3 for more detailed information on tip speed limits.
Turbine
for
these
centrifugal
limits.
Consult
Design
PERFORMANCE
OPTIMIZATION
The turbine shall provide for low weight and high efficiency necessary to maintain mechanical integrity. For a tip-speed-limited turbine design (i.e., effect of inlet temperature on allowable temperature on various design 39(a). If the turbine diameter
a design pitchline
within
the limitations
limited by rotor stress), velocity. Consider the
determine the effect of inlet
factors and conduct an optimization as indicated in figure becomes disproportionately large (greater than 3 times the
pump impeller diameter) or the moment of inertia becomes too large for rapid starts, it may be necessary to back off from the tip-speed limit. This condition often occurs with the oxidizer
turbopump
in a hydrogen-fueled
application.
For designs limited by turbine-blade centrifugal stress, determine the effect of inlet temperature and pressure ratio on the allowable rotational speed (sec. 2.2.1.4). Consider the effects of inlet temperature on various design factors and conduct an optimization as indicated
in figure
hydrogen
turbopumps.
To
determine
the
39(b).
number
Note
that
of stages,
the
use
blade
stress
equivalent
optimum tradeoff weight and performance. Unless engine, do not use so many stages that an outboard reference 4 for more detailed information on turbine
119
limit
generally
weight
(sec.
is encountered
2.1.2.2)
only
to determine
the payoff is highly significant turbine bearing is necessary. rotor stress limitations.
in
the
to the Consult
Rotational Pressure
speed ratio
N = constant" PR = constant
o .4 0 o
Pitthllne
o
l_wbine
q_
$i o
veZecity
U • coas_t
ratio
presmare
PR = (Po)T1/PT2
!
k
hO 0
I
-,4 O
| _0
O
o
Turbine
inlet
o
temperature------_
Turbine
(a) At the turblne-blade
tip-speed
Figure
39.
-
inlet
temperature
Turbine
-----e"
(b) At the turblne-blade
limit
Effect in
inlet temperature
of
turbine
turbine
inlet
optimization
temperature (gas-generator
on
various cycle).
design
factors
Turbine
inlet te_erature_
centrifugal-stre_s
limit
3.2.3.2
EXHAUST
The
turbine
performance To
avoid
cycles, choke
PRESSURE shall
exhaust
and sufficient
unpredictable
engine
design for a turbine between the turbine
engine turbine
level
for the engine
to meet
performance
3.2.4
into the
thrust
Turbopump
When
solving
particular
3.2.4.1 The
bearing
simplest
flareback
in GG and
tapoff
engine
for sufficient plumbing and
Integration
design
shall
components
reflect
problem,
evaluation
also
consider
in the turbopump
location
shall
without
provide
exceeding
turbopump
design
the
maximum
of" component
the
effects
of
that
system.
individual with
no
support
bearing
more
than
inboard of the turbine if the resulting bearing DN DN is not within limits or the turbine has more
the
thrust.
PLACEMENT
bearing outboard of the component development arranging
engine
pressure great enough to permit the gas flow to the end of the exhaust duct. To meet design
mechanical-design
on all other
manner
a single-shaft
bearing bearing
mechanical
BEARING
its design
predictable
chamber.
a particular
solution
for
and expander engine cycles, design flow to pass through the downstream
the
Mechanical
The turbopump in teraction.
sufficient
or exhaust
discharge static discharge and
thrust for staged-combustion discharge pressure to permit
the injector
For
at a pressure
turbine
turbine. (pump
relative
To simplify future alone and turbine
for
the
rotor
in the
limits. two
turbine
rotors,
place
the
aft
is within limits. If the resulting than two rotors, place the aft
development alone prior
effort, consider individual to combining them) when
to the pump.
For a single-stage pump, place the forward bearing inboard of the pump impeller: If the pump has two or more stages, place the forward bearing between stages. If the pump has a separate inducer stage, place the pump and use the inducer stators the pump have more own sets of bearings geared
turbopump
than three and Seals.
forward bearing to support the
between the inducer and the rest of the bearing housing. If both the turbine and
rotors, consider designing each as a separate unit with its For engines with thrust less than 10 000 lbf, consider the
arrangement.
121
_'.2.4.2
TURBINE
The turbine effects of concentricity, be simple;
ROTOR
ASSEMBLY
and shall be easy
Concentric
pilots
to assemble
may
various components, the pilots. (2)
Disk mounting coefficient that
(3)
Even
disk
from
not
and disassemble. thermal
growth,
be dependable
or if thermal
consider
if different
growth
and
conditions,
clamping
Disks will deflect the other.
axially
the
disks
bolts
must
if the
temperature
may
adapt
operate
to the
If the rotor tip speeds are approaching the stress limits, designing without a center hole. To permit the two sides
following
when
are used
for the
force
act together
a thermal
at different
resulting
on one
the
materials
centrifugal
(or clamp) bolts may loosen if they have is much different from that of the disks.
at steady-state
If so, the growth. (4)
ATTACHMENT
rotor assembly and its attachment to the main shaft shall adapt to the centrifugal stress and thermal growth on deflections, normality, alignment, and clearances, shall transmit the torque reliably; shall
To minimize problems resulting designing the turbine rotor assembly: (1)
AND
on
expansion
temperatures.
differences
side is different
in radial
from
that
on
minimize centrifugal stresses by of such disks to be symmetrical,
use a drive coupling such as a curvic coupling to attach the disks to the shaft and to each other. For curvic couplings, place the clamping bolts at the coupling mean diameter in order to avoid disk deflection. In general, use a disk diameter that is less than 4 times the coupling diameter; if a greater diameter disk stability and a reasonable
ratio is desired, conduct a rotor unit loading for the coupling (ref.
dynamics analysis 8). If the stresses
to ensure are within
the material limits, the torque may be transmitted by torque pins (a minimum of three), the clamping bolts, or a spline (with radial pilots) on the extended hub of the disk. To attach the disks to the main shaft, use clamping bolts if the disks do not have a hole in their centers. If they have a center hole, mount the disks on clamping bolts or a center clamping nut on the shaft.
a shaft
and secure
them
with
either
To attach the turbine rotor to the main drive shaft of a single-shaft turbopump, coupling in combination with an involute spline alone, a curvic coupling alone, torque-pin joint. The main considerations are normality and concentricity.
use a curvic or a bolt and For a geared
turbine,
requirements
consider
gear
separation
cause problems, consider using pump are separate units, use deviation.
and
allowances
for
misalignment.
If these
a quill shaft arrangement (fig. 4). If the turbine and the a coupling that is adaptable to misalignment and axial
122
If the turbine rotor consistsof individual disks,use an idiot pin or asymmetricalbolt holes to prevent backward mounting. To facilitate handling and positioning, install the disks vertically (vertical turbopump shaft). If the aft bearing is outboard, consider stackup tolerances,concentricity tolerances,runouts,and thermaldeflectionsduring assembly. If the pump has a thrust balancepiston, position the turbine rotor axially relative to the main driveshaft to avoid rubbing of the rotors andstators. Consultreference8 for detailedinformation on designof shaftsandcouplings.
3.2.4.3
TURBOPUMP
The
turbopump
pump
and
permit
housing
the
components To
HOUSING
turbine
and avoiding
a single-shaft
the cold pump and use an arrangement without
losing
thermal
than
growth
If the turbine avoid contact struts
of
the
to small
and have
to adapt
to the
large
to the
the pump housing.
the
the
various
pump.
If the
difference
for the turbopump casing can slide cylindrical
should remain This is the point
between
housing or unrestrained
housing
is used,
large
torque
turbopumps
are
for the radial
by
passing
them
more
subject
to unsymmetrical
pins to withstand. sides of the rotor, design manifold or by shielding through
hollow
vanes;
the casing to the support use
wherever possible. This design maintains a constant relative position between bearing supports and ensures alignment between the nozzles and the rotor blades. deflection is staill excessive, use a self-centering linkage arrangement to connect to the bearing To
avoid
regions.
high-temperature
insulation the two If thermal the casing
carrier.
cracking To
the
ambient and, therefore, the at which the cylinder should
casing. Use a structural cone to attach the hot turbine The radial pin method is less applicable to large
because
bearing
between
between
temperature
either use a long cylinder on which the hot turbine
is a separate unit with bearings on both with the hot gas by using a discharge discharge
differences
alignment
failure.
relation
a larger
temperature proper
along the cylinder will remain constant.
and bolted onto to a cylindrical
turbopumps
structural
turbopump
its concentric
to the
maintaining
the hot turbine, of radial pins
temperature somewhere diameter at that point be flanged manifold
shall adapt while
of
avoid
the sealing
housing,
minimize
problems
as
material well
regions.
123
as
cross cracking,
sections do
in high-temperature not
use
flanges
in
3.2.4.4
BEARINGS
The
bearings,
AND
SEALS
seals, and
requirements, shall minimize
bearing
Provide lubrication to all bearings; If the pumping fluid is hypergolic lubricant To
from
avoid
a separate
early
with a slight so that more
system
bearing
lubrication is more or if the turbine
that
difficult if the aft bearing is a separate unit, supply
failure,
prevent
sliding
by installing
the
bearing
the
bearing
seal
that
DN is within
will
limits,
permit
consider
the
seal
the effects
The thrust overloading
to absorb centrifugal
races
is outboard. the bearing
on the
to operate
shafts
bearing, Obtain
effectively,
of the various
rubbing velocity is within limits. These limits are a function and type of lubricant. See sections 2.2.1.2 and 2.2.1.3.
AXIAL
axial
and
operating
and radial and axial .loads. In general, use is a separate unit. Consult reference 6 for
In some configurations, particularly large ones, use a flexible bearing housing to the turbine manifold. If critical speed or shaft try changing the bearing preload so as to change the bearing spring
If the
life
of the rotor,
hot turbine gas from entering the bearing, install a seal between the bearing and source. In general, use a shaft-riding seal and a face-contact seal. Check to ensure
that the seal drop, material,
3.2.4.5
the turbopump
interference fit. If axial or radial loads are severe, consider a split-race balls may be inserted. Use a seal material compatible with the lubricant.
conditions such as temperature, kind of lubricant, ball bearings to absorb axial loads if the turbine more information on selection of bearing type. To prevent the hot-gas
shall meet
source.
the pressure balance across the particularly if the seal is dynamic. To ensure
lubrication
shall satisfy the axial and radial load requirements rotor dynamic problems.
thrust
THRUST
it. If it is not, pumps), which,
diaphragm to attach the whirl becomes a problem, rate.
BALANCE
balance system the bearings. is within
of seal pressure
shall
resist
the load-carrying
the
axial
ability
thrust
of
the
of ball bearings,
rotor
without
use ball bearings
consider balance ribs on the back faces of the impellers by causing the propellant to spin with the impeller, decrease
net axial thrust on the rotating assembly. If the axial again, if the pump is centrifugal, consider a labyrinth the impeller back face in combination and, therefore, the axial thrust.
with
a cavity
124
thrust seal vent
is too great arrangement
to reduce
alone (for the
for balance ribs and, to form a cavity on
the pressure
in the cavity
If either the variation in axial thrust between operating points or a reasonblemargin for error in predicted thrust are too great to be carried by the bearings,use bearingsthat are free floating axially in combination with a thrust balancepiston (refs. 2 and 3). To avoid balance-pistonrubbing, use the bearingsaslimit stopsor, if the bearingDN's are too high to carry the loads,userubbing stops.For a double-ringorifice-type balancepiston in hydrogen (fig. 33), the orifice rings can be usedas rubbing stops.However,in oxygen,a limit stop with a lower rubbing speedthat is located elsewherein the turbopump may be necessary. Locate the limit stops close to the balancepiston (axially) to minimize the influence of thermal and pressuregradients on the axial distancebetween the stop and the balance piston. To minimize axial length and thereforeturbopump weight and to minimize leakagelosses, use balancepistonsthat areintegralwith the pump impellersor rotors. Analyzehousingand impeller deflections during designbecauseintegral balancepistonsaresensitiveto clearance variations. Separatebalancepiston mounted elsewhereon the turbopump shaftmay beused for caseswhich havedeflection programsor where axial length is not a prime concern. To obtain more predictable forces and minimize shaft flexing, use shroudedimpellerson centrifugal pumps. If impellersare unshrouded,momentarydifferencesin the flow patterns in the flow passages will generateadditional unbalancedforcesthat mustbe accommodated in the design.
3.2.4.6
THERMAL
Thermal
barriers
temperature simpler
BARRIERS between
rise during
and more
the
pump
shutdown
and
the
turbine
so as to make
shall
minimize
subsequent
restart
the
pump
procedures
efficient.
For applications in which cryogenic turbopumps are restarted, isolate the hot turbine rotor from the pump rotor by connecting them with an involute spline together with a quill shaft, or with a ball spline coupling, or, if stress permits, with a coupling made out of a different material (e.g., plastic). To prevent heat flow from a hot turbine housing to a cold pump housing, material which
isolate between cryogenic
the housing by using the the two housings, or using fluid may
be recirculated
radial pin arrangement, a thermal barrier such
after
125
shutdown.
clamping insulating as a manifold through
3.2.4.7
ASSEMBLY
The turbopump design features assembly shall be compatible. If
the
turbopump
is designed
and
for
long
the procedures
service
life
and facilities
and
overhaul
access to all points. To avoid mistakes during reassembly, noninterchangeable fasteners. Make provisions for extracting, must
be separated
for
turbopump
capability,
provide
index components without damage,
easy
and use parts that
by force.
If extremely close clearances between the rotor and the housing must be maintained, consider using matched assemblies. Consider fabrication and maintenance when selecting the turbopump
materials.
use materials Assemble class-100 rocket
that
For
require
example, annealing
the turbopump 000, condition-C engine
for parts
in clean
turbopumps.
that
and reheat
are likely
treatment
after
a temperatureand room is recommended
Class
100 000
means
that
to require
weld
repair,
do not
welding.
dust-controlled environment. for the assembly of all types the
room
must
satisfy
one
A of
of two
criteria: (1) no more than 100 000 particles per cubic foot if they are 0.5 micron (#) and larger, or (2) no more than 700 particles per cubic foot if they are 5/_ and larger. Condition C means that the-temperature is 72 -+ 5 ° F. During assembly, check seal leakage rates, rotor axial and radial clearances, and bearing drag. Use locking devices to secure all bolts, nuts, and other fasteners. To avoid damage due to broken pieces, never twice. After assembly,
use lockwire within the turbopump and never use a locking device check the torque to ensure that there is no interference between the
rotor
and the housing
and that
3.2.5
System
Interfaces
3.2.5.1
PUMP INLET
The
pump
inlet
and
the bearing
upstream
and
seal drags
ducting
shall
are within
specifications.
enhance
the
pump
suction
performance. Minimize
losses
by minimizing
bends
and changes
in duct
cross section.
Use turning
vanes
in
elbows that are too close to the pump (within approximately 15 to 20 pump inlet diameters upstream of the pump). Consider tangential pump inlets if there is sufficient NPSH and if the overall result is a more compact unit. Use straightening vanes at the tank discharge to minimize
fluid rotation
reference
1 for more
at the pump detailed
inlet
information
and to prevent
fluid vortexing
on flow-distortion
126
effects
in the tank. and pump
Consult
inlet design.
3_2.5.2
PUMP DISCHARGE
The pumt_ shall meet
discharge connections and downstream ducting shall be leak tight and the requirements of light weight, minimum pressure drop, and ease of
servicing. Use a bolted O-rings for
flange to connect the pump discharge noncryogenic propellants and with
to the discharge pressure-actuated
line. Seal the joint with seals for cryogenic
propellants. If the flanges become disproportionately large and heavy (this may occur for small line sizes), consider welding the discharge line to the pump discharge. Use a diffuser at the pump discharge to minimize losses drop against line weight to obtain' the diffuser area more detailed
3.2.5.3
ratio to match pump information on pump
TURBOPUMP
The
turbopump
in the downstream optimum line size
discharge discharge
expansion large pad turbopumps
3.2.5.4
use close and with
mounts
shall support
coupled,
GAS-GENERATOR
The gas-generator
consider
and
weight
reference
10 for
and all loads applied
at one end of the turbopump other end. expansion
mounting
CONNECTION
connections
expansion
expansion
Consult
without causing unacceptable or the thrust chamber systems.
rigid pads
AND
shall be leak
deflections
and,
and
to accommodate
For cast structures, consider one and contraction. For very small
at the pump
discharge
flange.
MOUNTING tight
and shall adapt
to a high degree
and contraction:
Since the gas connection to the turbine turbine manifold and select its material thermal
the turbopump
contraction, ball joints at the integral keys to accommodate
with cast volutes,
of thermal
velocities.
MOUNTING
to the turbopump during operation distortions in either the turbopump In general,
and line lines.
plumbing. Trade line pressure and, therefore, the appropriate
contraction.
manifold also supports to withstand the added For
ease
of assembly
the gas generator, design the loads, including those due to and
disassembly,
use
bolted
flanges with pressure-actuated metallic seals at the connection point. To obtain greater reliability and lower cost, consider welded connections if the production rate is to be high. Use X-ray and penetrant detection methods to check the quality of the welds.
127
3.2.5.5
TURBOPUMP
The turbopump without removal Provide provide
for for
ON THE
ENGINE
system design shall provide for routine from the engine or major disassembly.
replacement of the seals without disturbing instrumentation replacement, leak checks,
possible,
provide
bearings, of rotors
rubbing, and and seals.
3.2.5.6
SERVICE
for
TURBOPUMP
The turbopump
hand
turning
worn
seals.
the
turbopump
Consider
the
servicing
the bearing inspections,
to permit
assembly. In addition, and torque checks. If
early
use of boroscope
and inspection
detection
ports
of damaged
to allow
inspection
OVERHAUL system
design
shall provide
for easy and inexpensive
overhaul.
Because overhaul
the turbopump replacement requires expensive engine recalibration, plan for rather than replacement. Plan for easy replacement of life-limited parts. When
deciding facilitate
whether repair,
to integrate parts to facilitate original manufacture or to separate parts to weigh the cost of original manufacture against the probability of failure. To
permit replacement of either unit without the turbine rotor assemblies to be balanced
3.2.6
The
the other,
try to design
the pump
and
Start Systems
start
system
shall
repeatability,-response, and turbine exhaust integrity. Consider leading
disturbing individually.
meet
the
vehicle
requirements
of start
time,
restart
time,
light weight, and simplicity," shall match the turbine type disposal method," and shall not endanger vehicle structural
main propellant tank-head and candidates to meet the combined
solid-propellant engine and
detailed trade study to make a final selection below in sections 3.2.6.1 through 3.2.6.4.
128
of a start
start-cartridge start vehicle requirements. system,
using
systems Conduct
the methods
as a
given
3.2.6.1
MAIN-PROPELLANT-TANK
The start start
system
transient
Consider
employing
without
this
start
propellant-expander
HEAD main-propellant-tank
undue system
drive
suitable
cycles.
Conduct
of a computerized mathematical control system requirements for simple ope.n-loop enough to cause transient occur,
tank pressure
model a main
head shall provide
(weight)
a satisfactory
penalty.
for
gas-generator,
a detailed
analysis
staged-combustion, of the engine
start
and
with
the aid
to determine the most satisfactory sequence and propellant tank-head start. For a GG cycle, use a
valve sequence system unless the propellant inlet pressure range is large wide variations in the engine start transient. If wide variations in the start provide closed-loop control of the GG oxidizer flow (controlling GG
temperature). For staged-combustion and expander turbine drive cycles, conduct a detailed computer analysis to determine if closed-loop control is required during the start transient. The start of a staged-combustion-cycle engine is more sensitive to transients than that of a GG-cycle
system,
3.2.6.2
and the need
PRESSURIZED-GAS
The pressurized provide
START
gas in the start
satisfactory
starts
Consider pressurized-gas start low-pressure-ratio (high-flow) Conduct determine
for closed-loop
under
control
should
tanks
shall be sufficient
all operating
tanks a suitable turbines used
mathematical
considered.
TANKS
model
in energy
and
quantity
starting method for GG cycles but on staged-combustion or expander
of the
engine
system mission.
system
is required
for repressurizing the start tank when long orbital Compensate for pressure variation with temperature
that, under normal environmental a relief valve or regulator that pressure-regulator Minimize
the
type number
possible; specify Class and visual inspection). transients.
of device of leak I welds Insulate
conditions, is set higher
temperature than the
or a narrow-band
paths
by combining
relief valve
not for cycles.
all operating requirements, to possible leakage of the stored for this analysis.
Provide the capability for increasing the gas storage pressure to a higher value than in the analysis in the event that testing indicates that additional energy is required. backup of the
to
conditions.
a detailed analysis of the start system, considering the required stored-gas energy. Account for any
gas. A computerized
be carefully
(and normal
coasts are a requirement by storing the gas such pressure) operating
valve for the functions.
that used Provide a
relief
will increase. Use level. Specify a system.
Use welded
joints
where
(i.e., welds that require radiographic, penetrant, dimensional, the tank and utilize reflective coatings to minimize thermal
129
3.2.6.3
LIQUID-PROPELLANT
The
liquid-propellant
unacceptable
start-tank
TANKS
system
shall
provide
satisfactory
for
systems
start
without
overpressures.
Liquid-propellant complexity,
START
start
tank
probably
are
systems
not
can
suitable
be used
for new
engine
GG
systems.
but,
because
However,
of system
if liquid-propellant
start tanks are suitable, control the start-tank pressurization rate to prevent overpressures in the GG feed system. A start-tank pressurization time of 3 seconds is satisfactory for most GG systems. rate. Conduct
Provide
a detailed
orificing
capability
heat-transfer
in the
system
and flow analysis
for
adjustments
of the start-tank
of pressurization
system
during
chilldown
to determine required flowrate through the system; size the fill and vent system accordingly. Under steady-state liquid conditions during fill, maintain some pressure (5 to 15 psig) in the start tank by restricting restriction experimentally. this purpose. Provide
a bleed
hole
the
in one
vent Design
system. If chilldown time is critical, the vent system oversize and include
of the
start-tank
isolation
check
valves.
determine the exact orifice provisions for
Determine
the hole
by experimentation, since the required size depends on volume of trapped propellant, of propellant, and environmental conditions. For ground-mounted start-tank systems, the bleed hole in the check valve in the GG supply line from the main feed system°
3.2.6.4
SOLID-PROPELLANT
The solid-propellant provide a satisfactory Consider
the
START
start-cartridge start transient
solid-propellant
start
thrust-chamber-tapoff cycles. Conduct considering all operating requirements,
size type place
CARTRIDGE combustion rate and propellant under all operating conditions. cartridge
a suitable
a detailed to determine
starting
analysis of the required
grain size shall
method
for
GG
and
the engine start system, solid-cartridge burn time
and flowrate. A computerized mathematical model of the engine system is required for this analysis. Control the grain temperature so that the start transient is within the safe operating range
of the
engine
and control installation.
of vehicle Conduct
temperature system less
variation sensitive
system.
Minimize
grain
temperature
variation
boattail temperature and by conditioning a detailed heat-transfer analysis to define
by
insulation,
coatings,
the start cartridge prior to the design for minimum
and to determine the range of grain temperatures. Make the turbine drive to grain temperature effects by utilizing a low-energy start cartridge
(providing acceleration to the 50 +- 10 percent level), thus allowing overlap of the power sources; then either provide closed-loop control to sense start-cartridge burnout and then turn on the GG, or use a check valve to isolate the start-cartridge turbine drive system from
130
the mainstage turbine drive system, thereby providing an automatic transition from start-cartridgepower to mainstagepower. If start-time repeatability is critical, utilize an electric heaterblanket to maintain the propellant grainat someconstanttemperaturehigher than the maximumexpectedenvironmentaltemperature. To avoid damagingdetonationsin the oxidizer feed system,avoid contamination by using (1) an inert-gaspurge through the oxidizer systemwhile the solid-propellantstart cartridge is burning, (2) a poppet-type injector designdiscussedin reference9,or (3) a check valve isolating the liquid-propellant GG from the solid-propellant start cartridge. To prevent thrust overshoot for cases in which the solid-propellant start cartridge ignites the liquid-propellant GG, provide pyrotechnic igniters in conjunction with the start-cartridge gasesto minimize the overlap and use the start-cartridgegasesonly for ignition of the igniters. Maintain the ignition sourcefor t to 2 seconds to ensure reliable ignition of the liquid
propellants.
3.3
DESIGN
3.3.1
3.3.1.1
EVALUATION
Engine-System
DESIGN-POINT
Characteristics
SYSTEM
The power requirements shall be consistent with engine system.
BALANCE
and output flow the requirements
characteristics of the turbopump system imposed during operation of the rocket
Determine the pump headrise, pump flowrate, and turbine performing a design-point balance of the turbopump system rocket engine system. Include in the system balance detailed engine
components.
pumps fulfilled
and by
for the
Write the power turbopump
steady-state available system
equations from during
for
the turbine. steady-state
available from the turbine is equal to the horsepower to be used when performing the power calculations engine thrust at the desired mixture ratio balance can be found in reference 35. The
design-point
system
engine-system requirements. turbopump-system design.
balance If
power The basic operation
whether
engine-system
131
and
flow
power is that
required
the
A detailed
example
turbopump
requirements
are
system not
by the
condition to be the horsepower
required by the pump. The is that required to produce
of the propellants.
determines the
the
flowrate requirements by when it is operating in the characteristics of all of the
met,
pump flow the design of a system
meets modify
the the
3.3.1.2
OFF-DESIGN
SYSTEM
The
turbopump
system
over
the entire
operating
Obtain
the
range
BALANCE
shall operate
reliably
range.
of planned
Determine
turbopump-system
the
operating
In order to assess properly the requirements, analytically determine other turbopump parameters. examples of which are shown and has
the 95 or 99.5
requirements
operation
from
the
engine-system
balance
tolerances from existing engines to determine components at the extremes of the planned
envelope
within
which
95 percent
effects of off-design operation the simultaneous changes in pump
of the engines
on system design head and flow and
Present this information in the form of operating ellipses, two in figure 40. Each figure shows three ellipses, marked 50, 95,
99.5 percent. These percentages of falling within these ellipses.
either
performance
range.
(sec. 2.3.1.1) and then use known component the effects of off-design engine and turbopump operating will fall.
and meet
percent
are the probabilities A turbopump system
that any normally
normal operating point should be designed to
limits.
Percent
58
x103
/
72
99.5
Percent
xlO 2
99.5 37
,_
36
_
5s
._
70 68
_
66
Q
62-
I
!
I
I
"85
87
89
91
Pump
(a)
Pump
flovrate,
head
vs
Figure
point
operating
Nominal point
53
5 operating
-Nominal
64
xlO 2
I
I
I
I
I
26.5
27
27.5
28
28.5
gpm
pump
40.
Turbine
flowrate
-
Probability
(b)
ellipses
132
for
turbopump
Turbine
operation.
speed,
power
xlO 3
rpm
vs
turbine
speed
Compute the ellipses by normal probability techniques. Treat : each of the performance-determining parameters of the components as an independent, randomly'distributed variable with a mean equal to its nominal value, and a standard deviationdeterminedfrom its tolerance.The standarddeviationsof eachengine-dependent variable should be calculated from a table of enginelinear influence coefficients and'the toleranceassignedto the engineindependentvariable.The bivariateprobability distribution shouldbe calculatedby firgt computing the correlationcoefficient betweenthe turbopump variablesbeing considered(ref. 36). Finally, the effectsof the predictablyvarying operating conditions (e.g.,the variation of pumpinlet pressureduring flight)should be determinedfor each dependent variable and added algebraically to the 95-percent range to yield the 95-percentlimiting valuesshown. If the turbopump designdoesnot meet the reliability andperformancerequirementsat the extremesof this envelope,modify the designandrecheckit. 3.3.1.3 The
CONTROL turbopump
engine,
CONSTRAINTS system
shall be simple,
Use system balances (sec. effects of control system
controls
shall
and shall cause
meet
the
minimal
2.3.1. l) and system types and locations
response
performance
requirements and weight
dynamic models (sec. 2.3.2) on pressure drops, control
of the penalties.
to determine the ranges, transient
responses, and stability. In addition, determine the corresponding effects of any variations in pump pressure requirements on turbopump weight and performance. Conduct tradeoffs of simplicity against resultant performance and weight penalties and, finally, select the optimum control requirements are
system. illustrated
staged-combustion
engine
3.3.2
3.3.2.1 The
System
Typical effects in figure 34
of complexity for a range of
on pump discharge-pressure engine mixture ratios for
a
cycle.
Dynamic
Analysis
START turbopump
support
system
the attainment
moment
of inertia
of the engine-system
To improve turbopump system reducing the turbopump-system
transient moment
and
low-speed
start-transient
torque
133
shall
objectives.
response, reduce the time of inertia and increasing
torque. To reduce the turbopump moment of inertia, with the limitations discussed in section 2.2.1. Also
capability
constant (eq. the low-speed
design at the highest select materials that
(15)) by turbine
speed consistent, have the highest
Strength-to-densityratios andareadequatefor the load andlife requirements.Increasingthe low-speed turbine torque will also reduce the turbopump time constant. To provide increasedturbine torque at low speedsfor GG rocket engine systemsthat start with availablepropellant tank head pressure,use a hot-gasvalvein seriesbeforethe turbine. This valvewill be open during enginestart to provideaddedturbine inlet pressure.At mainstage conditions, the hot-gasvalvewill be closedto achievethe steady-stateoperatingconditions of the engine. For systemshaving separateturbopumpsfor the oxidizer andfuel, the fuel turbopump time constant should never be greaterthan the oxidizer turbopump time constant.This helps preventtemperaturespikesduring enginestart. Checkthe steepnessof the head/flow curveof the pump.A steepcurvewill tend to stabilize the system.Sweepbackof the bladesin a centrifugalpumpwill producea steeperhead/flow curve. Use analytical digital or analog engine start models to verify stable turbopump operation during transient conditions. These models are generally for a specific engine configuration and caution should be used in trying to generalizeresults. Evaluate the structural adequacyof the turbopump designto withstand transient overloadsthat may occur during the start.
3.3.2.2
THROTTLING
Turbopump
throttling
control
systems
shall
meet
the
required
engine-system
response. Consider (thrust before
control
of the
turbine
working
fluid
for systems
requiring
moderate
response
rates
change up to 20 percent per second). Assess hot-gas valve technology carefully using hot-gas valves. Use throttling of the main liquid propellant valves when faster
ramp rates are needed (thrust change greater than 20 percent per second). Also reduce the turbopump system time constant (sec. 3.3.2.1) for better response to all commands. If the pump inlet flow coefficient becomes low enough during throttling to cause suction-performance
3.3.2.3
problems,
prewhirl
(ref.
37) or simple
flow recirculation:
SHUTDOWN
The pump
structural
design
Compute surge pressures at engine-system configuration, geometry.
consider
Design
the inlet
shall accommodate
the pump inlet, using valve characteristics,
to withstand
these surge
134
shutdown
surge
pressures.
mathematical models fluid characteristics, pressures.
that reflect the and suction-line
3.3.2.4
scope
INSTABILITY
LoW-frequency Investigate
To
verify
the
volume
system
change
per unit
shall be defined. pressure
change)
compliance
and
the
dynamic
of the turbopump system. Test the turbopump the operating fluid system with a measured
and
pressure) 33.
gain,
conduct
with a flight-type pulse of varying
/
System
3.3.3.1
TURBOPUMP
The
Development
turbopump
system
and
anticipated
extremes
Conduct
structural
testing. confirm
the
testing
of ac_ual
testing
the
that
are required
ensure
that
demonstrate
assembly engine
of certain
turbopump-system system,
shall
turbopump
Conduct hydrostatic the adequacy of the disks
Testing
SYSTEM
components
engine
(the
pump-cavitation
3.3.3
For
compliance
of the turbopump
partial derivative of discharge pressure with respect to inlet design following the methods recommended in reference
tests exciting
frequency.
turbine
characteristics
pump-cavitation
frequency-resp°nse inlet duct by
(POGO)
response
the dynamic gain (the for the turbopump-system
_e pl
dowl
SYSTEM
under
operating
testing design.
to operate
components of pump Conduct
near material
testing
conducted
the
instrumentation
adequacy that
of
critical
simulate
the
conditions.
turbopump
pressure structural
the conditions
before
volutes spinpit
prior
to turbopump-system
and turbine manifolds testing of pump rotors
to and
limits. integration
is adequate
of the turbopump to verify
pump
into and
the
turbine
performance. To do this, hold a formal design review of the entire turbopump system to ensure that the instrumentation requirements are met before making a hardware release. Critical measurements include fluid flowrates, pressures, and temperatures, and rotor speeds, accelerations, To
avoid
positions, engine-system
turbopump. range potential
and
To
failures
do this,
of engine-system
torques.
calculate operation.
due the
to turbopump extremes
Then
test
operation,
of turbopump the
turbopump
apply
limits
operation at these
testing
from
the
extremes
to the planned
to expose
problems.
_lianc_
If pump tests are conducted prior to the turbopump-system tests, consider variable-speed electric motor drive for low horsepower units. For high-horsepower may be necessary to use the same turbine as will be used on the engine. tests, use the suitable pump fluids from table VII so as to reduce propellant pump
tests,
use the design
fluid.
135
using pumps,
a it
For early pump costs. For later
Term
Definition
or Symbol
cryogenic
fluids
or conditions
at low temperatures,
usually
at or below
-150 ° C
•(222 ° R) D
diameter,
DB
blade
ft or in.
diffusion
factor DHIA
D
S
DN
specific
diameter,
fts/4/gpm
bearing
speed-capability
mm and rotational
drag pump
index,
the product
of bearing
bore
size (D) in
speed (N) in rpm
pump whose rotor consists of a disk with many short radial blades. The flow enters radially and is carried within the blade passages around the disk and is discharged
radially
EW
equivalent
weight,
Ibm
EWF
equivalent
weight
factor,
F
engine
GG
gas generator
thrust,
acceleration
g
1/2 , Ds = -Q,Iz
a port.
lbm/(lbm/sec)
lbf
due to gravity,
H
head
or headrise,
HP
shaft horsepower,
hp
h
specific
Btu/lbm
ZXhv
latent
I
rotating
mass moment
Is
specific
impulse,
i
fluid incidence
i/_
ratio
enthalpy, heat
through
ft/sec 2
ft or ft-blf/lbm
of vaporization,
Btu/lbm
of inertia
lbf-sec/lbm angle, deg
of fluid incidence
140
angle to impeller
blade
angle
Definition
Term or Symbol
inertance
the impeding effect fluid-filled conduit
J
converstion
k
thermal
KD N
empirically
derived
coefficient
for bearings
Kss
empirically
derived
coefficient
for seals
L/D
length-to-diameter
MR
mixture
N
rotational
NPSH
net positive
factor,
of fluid inertia
of oscillations
in a
J = 778 ft-lbf/Btu
conductivity,
ratio:
on transmission
Btu/(hr-ft-°R)
ratio ratio
speed,
of oxidizer
to fuel, MR = _Vo/gZf
rpm
Po - Pv suction
head,
ft or ft-lbf/lbm,
NPSH -
at pump
inlet
P
M S
specific
o/f
ratio
P
pressure,
psi
Pc
chamber
pressure,
Pv
vapor
Po
total
Pitot
pump
speed,
rpm-gpmV2/ft
of oxidizer
to fuel
pressure, pressure,
3A, N s -
psi
psi psi
pump in which a rotating liquid ring is created inside pressurized fluid is scooped from this ring by stationary ducted to the outside.
PL
stage payload,
PR
pressure
Q
volume
lbm
ratio flowrate,
gpm
141
a rotating drum; Pitot heads and
TermorSymbol R Rootes
Definition
reaction pump
rotary
pump
produce S
consisting
of
two
a positive-displacement
clearance
between
rotor
intermeshing
pumping
and casing,
cam-like
rotors
that
action
in.
NQ V2 a S
s;
suction
specific
speed, rpm-gpml/2/ft
3_, S s -
(NPSH) 3_
as
corrected
suction
specific
speed,
S's -
(1 - u2) _
Seq
torsional subject
stress shaft,
in a solid
shaft
of the
same
outside
diameter
as the
psi
SS
seal speed,
SSME
space shuttle
s
specific
stall
loss of pumping capability surface of the blades
T
temperature,
To
total
temperature,
T/C
thrust
chamber
Tq
torque,
in.-lbf
Tesla pump
pump similar to a centrifugal configuration, which consists energy to the flow by friction
TSH
thermodynamic
time constant
time for a variable
ft/sec main
entropy,
engine Btu/(lbnr-°R) as a result
of flow separation
on the suction
°R °R
suppression to reach
142
pump with the exception of the rotor of many closely spaced disks that add
head,
ft
63% of its final value
Term
or symbol
Definition
U
turbine
U/Co
isentropic
velocity
u
tangential
velocity,
ut
blade
VC
velocity
pitchline
velocity,
ft/sec J
vane pump
tangential
pump
velocity
consisting
(tip speed),
ft/sec
of a rotor
with
sliding
vanes
that
is mounted
housing
W
weight,
;v
weight
w
fluid velocity
Z
number
lbm flowrate,
lbm/sec relative
to blade
of blades
vapor volume factor
percent
at the pumping
for thermodynamic
gas specific incremental ratio
ft/sec
compounded
eccentric
A
ratio
of inlet
I?
efficiency
%)T.s
turbine
P
ratio
P
density,
o
blade
limit
suppression
head; blade
to discharge
tip diameter,
heat ratio change tip diameter
total-to-static
to inlet tip diameter,
lbm/ft 3 (ratio
143
Dtl 6 =-Dt2
efficiency
of inlet hub diameter
solidity
angle
of chord
length
to spacing)
Dh 1 u =---Dtl
in an
Term
or symbol
Definition
turbopump
time constant,
I _- = --TqN
Gin
_o
flow coefficient,
ref. to blade
tip speed,
_0= -tl t
stage head coefficient,
gH ref. to blade tip, ff =--
ut
SUBSCRIPTS act
actual
des
design
E
engine
f
fuel
h
hub
I
inducer
id
ideal
isen
isentropic
L
liquid
m
meridional
min
minimum
ms
mean station
o
oxidizer
P
pump
si (i = 1,2,3,etc)
specific
component
value for entropy
144
(fig. 35)
stmsc urrs stall
stall
T
turbine
T/C
thrust
TP
turbopump
t
blade
u
tangential
v
vapor
0
total
1
inlet
2
discharge;
chamber
tip component
exhaust
Identification
MateriaP
A-50
50/50
Armalon
trade
Bearium
B-10
blend
of N2I-I _ and UDMH
name of E. I. duPont
trade name bronze
of
CRES
corrosion-resistant
Freon
trade
GN:_
gaseous
Hastelloy
name
Bearium
Metals
Corp.
Teflon
(Rochester,
N.Y.)
for
leaded
steel
of E.I. duPont
nitrogen
Haynes-Stellite
C
Co. for glass-filled
Co. for a family
of fluorinated
hydrocarbons
per MIL-P-27401 Corp.
designation
for
a nickel-base
high-temperature
alloy hydrazine
N2H4,
1Additional
information
on
Plaza,
York,
in
Defense, Society
New
N.Y.;
Washington,D.C., for
Metals
(Metals
metallic
materials
MIL-HDBK-5B, Sept.
Park,
1971 OH),
; and
propellant herein
Metallic in
Metals
can
grade per MIL-P-26536 be
found
Materials Handbook
1961.
145
in
and (8th
the
1972
Elements ed.),
SAE for
Vol.
Handbook,
Aerospace
1 : Properties
SAE, Vehicle and
Two
Pennsylvania
Structures,
Selection
Dept.
of Metals,
of Am.
Material
Inconel
Identification
trade
625,718
names
IRFNA
inhibited
Kel-F
trade
of International
red fuming name
of
Nickel
nitric
Co. for austenitic
acid, propellant
3M Corp.
for
grade
nickel-base
alloys
per MIL-P-7254
'
a high-molecular-weight
polymer
of
chlor otrifluorethylene "K"
monel
LH2 LOX,
LO 2
trade
name
alloy
containing
liquid
hydrogen,
liquid
oxygen,
N2H4
hydrazine,
N204
nitrogen
RJ-1
ram grade
of International
propellant
tetroxide,
jet
copper,
propellant
propellant
fuel,
grade
per MIL-P-27201
per MIL-P-25508
grade
a high-density,
per MIL-P-26539
kerosene-like
high-energy
name
of General
Teflon
trade
name
of E.I. duPont
UDMH
unsymmetrical
321,347
austenitic
431
martensitic
2219
wrought
6061
wrought
41
age-hardenable
per MIL-P-26536
propellant
trade
Rene
for a wrought
and aluminum
grade
grade
Co.
hydrocarbon,
propellant
per MIL-F-25558
kerosene-base MIL-P-25576
RP-1
nickel,
Nickel
hydrocarbon
Electric
Co. for a polymer
nickel-chromium-iron
aluminum
grade
nickel-base
per
alloy
of tetrafluorethylene
propellant
grade
per MIL-P-25604
steels
steel alloy with copper
aluminum
propellant
Co. for an austenitic
dimethylhydrazine,
stainless
fuel,
alloy
with
as principal
magnesium
and
alloying silicon
element as principal
alloyil_g elements
Vehicles
Identification
and Engines
J
Agena
YLR81-BA-
11
engine for Agena upper stage; manufactured by Bell Aerospace
146
15 000 lbf thrust; uses IRFNA/UDMH; Company, Division of Textron
Vehicles
and Engines
Identification
Atlas
launch
A-7
engine system for Redstone; manufactured by Rocketdyne
Centaur
high-energy
vehicle
using MA-5 engine
upper
stage
for
system
78 000 Division, Atlas
lbf thrust; used North American
and
Titan;
uses
LOX/alcohol; Aviation, Inc.
two
RL10A-3-3
engines F-1
engine
for S-IC; 1 500 000 lbf thrust;
Rocketdyne H-1
engine
Division,
Rockwell
for S-IB; 205
uses LOX/RP-1
; manufactured
by
uses LOX/RP-1;
manufactured
by
uses LOX/LH2;
manufactured
by
designed
and
International
000 lbf thrust;
Corp.
Rocketdyne J-2
engine for S-II; Rocketdyne
J-2S
uprated
J-2;
developed
230 000
265000
lbf
lbf
thrust;
thrust;
uses
LOX/LH2;
by Rocketdyne
Aerojet engines for first stage of the Titan • the -3 uses LOX/RP-1, and develops • the -5, -7, -9 uses N204/A-50
vehicles 150 000 lbf thrust
, and develop
215 000 lbf thrust
LR-91-A J-3, -5, -7, -9
Aerojet engines for the second stage of the Titan vehicles • the -3 uses LOX/RP-1, and develops 80 000 lbf thrust • the -5, -7, -9 use N204/A-50 , and develop 100 000 lbf thrust
MA-5
five-engine 1 sustainer
system for Atlas vehicle containing 2 booster, 2 vernier, and engines; boosters provide 330 000 to 370 000 lbf thrust;
sustainer, 57 000 by Rocketdyne MB-3
engine
system
manufactured
to 60 000
for Thor
lbf thrust;
vehicle;
uses LOX/RP-1
170 000
lbf
thrust;
; manufactured
uses
LOX/RP-1;
by Rocketdyne
Redstone
early
RL I 0A-3-3
engine for Centaur; 15 000 lbf thrust; uses LOX/LH 2 ; manufactured Pratt & Whitney Aircraft Division of United Aircraft Corp.
Saturn
launch
V
launch
vehicle
vehicle
using A-7 engine
for Apollo
147
manned
system
mission
to the moon
by
Vehicles andEngines
Identification
S-IB
booster
S:IC
first
using a cluster stage
(booster)
of eight of
the
H-1 engines Apollo
Saturn
V vehicle;
uses
five
F-1
engines S-II
second
stage
of the Apollo
Saturn
V vehicle;
uses a cluster
of five J-2
engines S-IVB
third
Thor
launch
Titan
X-8
I, II, III
stage of the Apollo vehicle
Saturn
using MB-3 engine
family
of launch
rocket
engines
vehicles
developed
using
V vehicle; system the
by Aerojet
experimental throttleable rocket LOX/LH2 ; developed by Rocketdyne
148
uses a single J-2 engine
LR-87-AJ Liquid engine;
and LR-91-AJ
Rocket 90000
series of
Co. lbf
thrust;
uses
REFERENCES .
Anon.:
Liquid
SP-8052, 2.
Anon.:
Anon.:
Engine
Turbopump
Inducers.
Space Vehicle
Design
Criteria
Monograph,
NASA
Design
Criteria
1971.
Liquid
Monograph, 3.
Rocket
May
Rocket
NASA
Liquid
Engine
SP-8109,
Rocket
Centrifugal
December
Engine
Axial
Flow
Turbopumps.
Space
Vehicle
1973. Flow Turbopumps.
Space Vehicle
Design
Criteria
Monograph
(to
be published). 4.
Anon.: January
Liquid 1974.
Rocket
Engine
Turbines.
5.
Anon.:
Liquid
Rocket
Engine
Turbopump
SP-8100, 6.
Anon.:
March
Rocket
March
Gears.
Engine
Turbopump
Turbopump
8.
Anon.:
Turbopump
10.
* 11.
Anon.: March
Liquid
Liquid 1972.
13.
Propellant
September
Balje,
O.E.:
Performance
A Study
Flexible
Discharge
American
A. J.: Centrifugal
Pump
Klassen,
Vehicle
Monograph,
Design
NASA
SP-8110,
Criteria
Monograph,
NASA
Space Vehicle
Design
Criteria
Monograph,
NASA
on Design and Matching
Criteria
Shafts
and
Space Vehicle
Tubing,
Flow
Aviation,
and Axial
Rotating-Shaft
Seals.
Couplings.
Space
Vehicle
Design
Criteria
Space
Vehicle
Design
Criteria
1972.
Gas Generagors.
Pump
Div., North
Stepanoff,
84, 1962, 14.
Engine
SP-8101,
A. E.: Induced
Rocketdyne 12.
Rocket
NASA
Anon.: Liquid Rocket Lines, Monograph (to be published). Marks,
Criteria
1971.
Anon.: Liquid Rocket Engine Monograph (to be published).
9.
Design
Space
Bearings.
7.
Monograph,
Vehicle
1974.
Liquid
SP-8048,
Space
Design
Bellows,
Oscillations
and
Filters.
and Matching
NASA
Vehicle
SP-8081,
Design
H-Q Operation.
Criteria
LAP 67-100,
1967.
John Wiley & Sons,
of Turbocomponents.
MOnograph,
Space
Due to Position
Inc., unpublished,
Flow Pumps.
Criteria
Inc.,
of Turbomachines: J. Eng. Power,
1957.
Part Trans.
B-
Compressor
ASME,
Series
and A, vol.
pp. 103-114. H. A.:
Mean-Diameter
Cold-Air
Investigation
Single-Stage
Axial
of Effects
Flow Turbine.
of Partial NASA
Admission
TN D-4700,
on Performance
August
*Dossier for design criteria monograph "Turbopump Systems for Liquid Rocket Engines." Unpublished. material available for inspection at NASA Lewis Research Center, Cleveland, Ohio.
149
of 3.75-Inch
1968. Collected source
v
15.
Stenning,
A. H.: Design
79, Dynamic
of Turbines
Anal. and Control
for High-Energy-Fuel,
Lab.,
Mass. Inst.
Tech.,
Low-Power-Output September
Applications.
1953.
16.
Balje, O.E.: A Study on Design Criteria and Matching of Turbomachines: Relations and Design Criteria of Turbines. J. Eng. Power, Trans. ASME, Series 83-102.
17.
Wong, G. Turbopumps-
S.;
and Second
Wagner, Interim
W. R.: Thermodynamic Report. R-7585, Rocketdyne
Report
Improvements Div., North
Part A- Similarity A, vol. 84, 1962, pp.
in Liquid Hydrogen American Aviation, Inc.,
July 1967. 18.
"19.
Huzel,
D. K.;
SP-125,
1971.
Dinges,
R. A.:
and
Huang,
Mark
Inc., unpublished,
D. W.: Design
19 Final
August
Report.
MacGregor, C. A.; and Csomor, Rocket Engines. NASA CR-72965,
21.
Hoshide,
R. K.; and Nielson,
Report. 1972.
NASA
CR-120815,
Gelder,
T.
Ruggeri,
Measured D-3509, 23.
Bissell,
F.;
Pressure July
TDR-3117-3011,
A.: Rotating Rocketdyne
C. E.: Study
R.
and
Propellant
Rocket
Rocketdyne
Engines.
ed., NASA
American
Aviation,
Div., North
Second
1963.
20.
22.
of Liquid
and Positive Displacement Div., Rockwell International
of Blade
R-8806,
Rocketdyne
S.;
Moore,
and
Temperature
Clearance Div.,
R. D.:
Depressions
Effects
North
Cavitation
for Low-Thrust February 1974.
on Centrifugal
American
in Cavitated
Pumps Corp.,
Rockwell
Similarity Regions
Pumps Corp.,
Considerations of
- Final
November
Based
Freon-114.
on
NASA
TN
Pumps
for
1966.
W. R.; Wong,
O2/H 2 Rocket
G. S.; and Winstead,
Engines.
J. Spacecraft
T. W.: Analysis
Rockets,
24.
King, North
J. A.: Design of Inducers for Two-Phase American Rockwell Corp., July 1970.
25.
King,
J. A.:
North
American
Design
of Inducers
Rockwell
Corp.,
Operation
for Two-Phase January
of Two-Phase
vol. 7, no. 6, June
Oxygen
- Final
- Final
Flow
in
LH2
pp. 707-713.
Report.
R-8283,
Rocketdyne
Div.,
Report.
R-8832,
Rocketdyne
Div.,
1972.
26.
Wong, G. S.: High Pressure Pumping Technology Aviation, Inc., October 1964. (Confidential)
(U).
27.
Young,
of Hydrostatic
W. E.; and Reddecliff,
1970,
J. M.: Investigation
Cryogenic TurbopumpsFinal Report (U). United Aircraft Corp., May 1967. (Confidential)
AFRPL
R-5833,
TR-67-130,
Rocketdyne
Bearings Pratt
Div., North
for Use in High Pressure & Whitney
*Dossier for design criteria monograph "Turbopump Systems for Liquid Rocket Engines." Unpublished. material available for inspection at NASA Lewis Research Center, Cleveland, Ohio.
150
American
Aircraft
Div.,
Collected source
q
28. Anon.:
Final Program R-5652-4P, Rocketdyne
29. Lieblein, Limiting
30. Anon.:
S.; Schwenk, Blade
31. Anon.: (Pogo).
32. Anon.: Analysis
33. Anon.:
H-1 Engine R-6813,
34. Anon.: Criteria
Diffusion
Design
Corp., System
System
Investigation
Broderick,
R.
in Axial-Flow-Compressor
Rocketdyne
Program.
Rocketdyne
J. C.; and
Engine
Aircraft
Engine
Engineering, Task I, Turbomachinery Inc., May 1965. (Confidential)
Loadings
Advanced
Div., United
Progress Report for Product Div., North American Aviation,
July
Functions
Div., North
Prevention
of
NASA
Coupled
Loop
Rockwell
for Estimating
NACA
RM-E53D01,
FR-2488
Structure-
SP-8055,
October
Inc., December
of S-V Vehicle
American
Instability Corp.,
August
Propulsion
Losses
and
1953.
,
Pratt & Whitney
Aircraft
Longitudinal
Stability
1966.
Longitudinal
Aviation,
Stability
Inc., March
on S-II Stage
(Pogoa
1967.
of Saturn
V. R-7970,
1969. Instability
(Pogo).
Space
Vehicle
Design
1970.
35. Wieseneck, Model
Factor
of S-IVB Vehicle
Aviation,
for Support
Div., North
17 Hz Closed American
Monograph,
PWA Report
for Support
American
Functions
Rocketdyne
of
Div., North
Bell (AEB).
Elements.
1967.
Transfer
Transfer
R-6929,
Study,
L.: Blade
(U).
H. C.: RSS-8314-1, Space Shuttle Main Program. R-8892P (Enclosure 1), Rocketdyne
Engine, Phase B, Engine System Div., North American Rockwell
A.F2
McGraw-Hill
Performance Corp., April
1971. Mood,
Introduction
Wong, G. S.; Macgregor, Throttled Turbopumps.
to the Theory
of Statistics.
C. A.; and Hoshide, J. Spacecraft Rockets,
Book
Co. (New
York),
1950.
R.K.: Suppression of Cavitation and Nonstable vol. 2, no. 1, Jam-Feb. 1965, pp. 73-80.
151
Flow
in
C-I
NASA SPACE VEHICLE DESIGN CRITERIA MONOGRAPHS ISSUED TO DATE
ENVIRONMENT SP-8005
Solar Electromagnetic
SP-8010
Models
of Mars Atmosphere
SP-8011
Models
of Venus
SP-8013
Meteoroid March
Radiation,
(1967),
Atmosphere
Environment
Revised
May 1971
May 1968
(1972),
Revised
Model-1969
September
(Near
Earth
1972
to Lunar
Surface),
1969
SP-8017
Magnetic
Fields-Earth
and Extraterrestrial,
SP-8020
Mars Surface
SP-8021
Models
SP-8023
Lunar
SP-8037
Assessment
and Control
of Spacecraft
SP-8038
Meteoroid Environment October 1970
Model-1970
SP-8049
The Earth's
SP-8067
Earth
SP-8069
The Planet
SP-8084
Surface Revised
SP-8085
The Planet
Mercury
SP-8091
The Planet
Saturn
SP-8092
Assessment June 1972
and
SP-8103
The Planets
Uranus,
SP-8105
Spacecraft
Models
of Earth's Surface
(1968),
May 1969
Atmosphere
(90
Models,
March
and Emitted
Jupiter
(1970),
Atmospheric June 1974
km),
Revised
March
1973
July
December
1971
Control,
153
(Launch
March June
September
1970
and Planetary),
1971
and
Transportation
Areas),
1972
1972
of Spacecraft
Neptune,
Fields,
(Interplanetary
Radiation,
(1971),
Control
Magnetic
1971
Extremes
(1970),
Thermal
to 2500
1969
May 1969
Ionosphere,
Albedo
March
and Pluto May 1973
Electromagnetic
(1971),
November
Interference,
1972
STRUCTURES SP-8001
BuffetingDuringAtmospheric Ascent,Revised November 1970
SP-8002
Flight-Loads Measurements DuringLaunchandExit, December 19_4
SP-8003
Flutter,Buzz,andDivergence, July1964
SP-8004
PanelFlutter,Revised June1972
SP-8006
LocalSteadyAerodynamic LoadsDuringLaunchandExit, May1965
SP-8007
BucklingofThin-Walled CircularCylinders, Revised August1968
SP-8008
Prelaunch GroundWindLoads,November 1965
SP-8009
Propellant SloshLoads,August1968
SP-8012
NaturalVibrationModalAnalysis, September 1968
SP-8014
EntryThermal Protection, August1968
SP-8019
BucklingofThin-Walled Truncated Cones, September 1968
SP-8022
Staging Loads,February1969
SP-8029
Aerodynamic andRocket-Exhaust HeatingDuringLaunchandAscent May1969
SP-8030
Transient LoadsFromThrustExcitation,February1969
SP-8031
SloshSuppression, May1969
SP-8032
Buckli:,gof Thin-Walled DoublyCurvedShells, August1969
SP-8035
WindLoadsDuringAscent,June1970
SP-8040
FractureControlofMetallicPressure Vessels, May1970
SP-8042
Meteoroid Damage Assessment, May1970
SP-8043
Design-Development Testing, May1970
SP-8044
Qualification Testing,May1970
SP-8045
Acceptance Testing,April 1970
SP-8046
LandingImpactAttenuationfor Non-Surface-Planing Landers,April 1970 154
SP-8050
StructuralVibrationPrediction, June1970
SP-8053
NuclearandSpace Radiation EffectsonMaterials, June1970
SP-8054
Space Radiation Protection, June1970
SP-8055
Prevention of CoupledStructure-Propulsion Instability(Pogo),October 1970
SP-8056
FlightSeparation Mechanisms, October1970
SP-8057
StructuralDesignCriteriaApplicable to aSpace Shuttle,Revised March 1972
SP-8060
Compartment Venting,November 1970
SP-8061
Interaction withUmbilicals andLaunchStand,August1970
SP-8062
EntryGasdynamic Heating, January1971
SP-8063
Lubrication, Friction,andWear.June 1971
SP-8066
Deployable Aerodynamic Deceleration Systems, June1971
SP-8068
Buckling Strengthof Structural Plates,June1971
SP-8072
AcousticLoadsGenerated by thePropulsion System, June1971
SP-8077
Transportation andHandling Loads,September 1971
SP-8079
Structural Interaction withControlSystems, November 1971
SP-8082
Stress-Corrosion Cracking in Metals, August1971
SP-8083
Discontinuity Stresses inMetallicPressure Vessels, November 1971
SP-8095
PreliminaryCriteria for the FractureControl of SpaceShuttle Structures, June1971
SP-8099
Combining AscentLoads,May1972
SP-8104
StructuralInteractionWith Transportation andHandlingSystems, January1973
GUIDANCE ANDCONTROL SP-8015
Guidance andNavigation forEntryVehicles, November 1968
155
SP-8016
Effectsof StructuralFlexibilityonSpacecraft ControlSystems, April 1969
SP-8018
Spacecraft Magnetic Torques, March1969
SP-8024
Spacecraft Gravitational Torques, May1969
SP-8026
Spacecraft StarTrackers, July1970
SP-8027
Spacecraft RadiationTorques, October1969
SP-8028
EntryVehicleControl,November 1969
SP-8033
Spacecraft EarthHorizonSensors, December 1969
SP-8034
Spacecraft MassExpulsion Torques, December 1969
SP-8036
Effectsof StructuralFlexibilityon LaunchVehicleControlSystems, February1970
SP-8047
Spacecraft SunSensors, June1970
SP-8058
Spacecraft Aerodynamic Torques, January1971
SP-8059
Spacecraft Attitude ControlDuringThrustingManeuvers, February 1971
SP-8065
TubularSpacecraft Booms(Extendible, ReelStored),February 1971
SP-8070
Spaceborne DigitalComputer Systems, March1971
SP-8071
Passive Gravity-Gradient LibrationDampers, February 1971
SP-8074
Spacecraft SolarCellArrays,May1971
SP-8078
Spaceborne Electronic ImagingSystems, June1971
SP-8086
Space VehicleDisplays Design Criteria,March1972
SP-8096
Space VehicleGyroscope Sensor Applications, October1972
SP-8098
Effectsof StructuralFlexibilityon Entry VehicleControlSystems, June1972
SP-8102
Space VehicleAccelerometer Applications, December 1972
156
CHEMICAL PROPULSION SP-8087
LiquidRocketEngineFluid-Cooled Combustion Chambers, April 1972
SP-8109
LiquidRocketEngineCentrifugal FlowTurbopumps, December1973
SP-8052
LiquidRocketEngineTurbopump Inducers, May1971
SP-8110
LiquidRocketEngineTurbines, January1974
SP-8081
LiquidPropellant GasGenerators, March1972
SP-8048
LiquidRocketEngineTurbopump Bearings, March1971
SP-8101
LiquidRocketEfigineTurbopumpShaftsandCouplings, September 1972
SP-8100
LiquidRocketEngineTurbopump Gears, March1974
SP-8088
LiquidRocketMetalTanksandTankComponents, May1974
SP-8094
LiquidRocketValveComponents, August1973
SP-8097
LiquidRocketValveAssemblies, November 1973
SP-8090
LiquidRocketActuatorsandOperators, May1973
SP-8080
LiquidRocketPressure' Regulators, ReliefValves,CheckValves,Burst Disks,andExplosive Valves, March1973
SP-8064
SolidPropellant Selection andCharacterization, June1971
SP-8075
SolidPropellantProcessing Factorsin RocketMotorDesign,October 1971
SP-8076
SolidPropellant GrainDesign andInternalBallistics, March1972
SP-8073
SolidPropellant GrainStructuralIntegrityAnalysis, June1973
SP-8039
SolidRocketMotorPerformance Analysis andPrediction, May1971
SP-8051
SolidRocketMotorIgniters,March1971
SP-8025 SP-8041
•SolidRocketMotorMetalCases, April 1970 Captive-Fired Testingof SolidRocketMotors,March1971 •It,U.S.
GOVERNMENT
157
PRINTING
OFFICE:
1975--635-049/73
I
Z 0 m
i..._m ,m[
I< n,,
i[
,mc
m
I-(n m
Z m
5
L,J L,J
C3 < UJ
LLJ
(,1 < a. u)
Q Z < (n 0 I
A
Z 0
I... immmB
< Z 0
m
....I
I..I.I ---' C.3 •-_ m
•_
a..
"" U.J _.=
0 I=_ a.. ...j
m
m...--,,,w
m
r.n
r._
.._
Z
r._
_
r,.-3 v
m
._.J
n.J
< ,.I
L,_.
< Z 0 m
I-< Z
FOREWORD
NASA
experience
Accordingly,
has indicated
criteria
are being
a need
for uniform
developed
criteria
in the following
for the design
of space
vehicles.
areas of technology:
Environment Structures Guidance and Control Chemical Individual
components
of this
work
Propulsion
will be issued
as separate
monographs
as soon
as they
are completed. This document, part of the series on Chemical Propulsion, is one such monograph. A list of all monographs issued prior to this one can be found on the final pages of this document. These
monographs
are
to be regarded
as guides
to design
and
not
as NASA
requirements,
except as may be specified in formal project specifications. It is expected, however, that these documents, revised as experience may indicate to be desirable, eventually will provide uniform design practices for NASA space vehicles. This monograph, "Turbopump Systems for the direction of Howard W. Douglass, Chief, project written
management by A. J.
Corporation
and
Liquid Design
Rocket Criteria
Engines," was prepared under Office, Lewis Research Center;
was by Harold W. Schmidt and M. Murray Bailey. The monograph was Sobin and W. R_ Bissell, Rocketdyne Division, Rockwell International was
edited
by Russell
B. Keller,
Jr. of Lewis.
To assure
technical
accuracy
of this document, scientists and engineers throughout the technical community participated in interviews, consultations, and critical review of the text. In particular, W. W. Heath of Aerojet Liquid Rocket Company; H. M. Gibson of Pratt & Whitney Aircraft Division, United Aircraft Corporation; and D. D. Scheer of the Lewis Research Center reviewed the text in detail. Comments
concerning
the
technical
National Office),
Aeronautics and Space Cleveland, Ohio 44135.
August
1974
content Administration,
of this
monograph
Lewis
Research
will be welcomed Center
(Design
by the Criteria
For saie by the National Springfield,
Virginia
Technical
22 i 51
$6.25
Information
Service
GUIDE
TO
THE
USE OF THIS
MONOGRAPH
The purpose of this monograph is to organize and present, significant experience and knowledge accumulated in programs to date. It reviews firm guidance for achieving
and assesses current greater consistency
product, and greater efficiency in the design major sections that are preceded by a brief references. The State of the Art, section identifies which design elements
use in design, the and operational
design practices, and from them in design, increased reliability effort. The introduction
2, reviews are involved
for effective development
establishes in the end
monograph is organized into two and complemented by a set of
and discusses the total design in successful design. It describes
problem, succinctly
and the
current tecnnoiogy pertaining to these elements. When detailed best available references are cited. This section serves as a survey
information is required, the of the subject that provides
background Recommended
for the Design
The
material and Practices.
Design
limitation, successful project The
Criteria,
shown
a proper
in italics
or standard must design. The Design
to use in guiding
Recommended
Practices,
a design
also
clearly
and briefly
Criteria
what
rule,
sections
have
been
positive
or in assessing
guidance
organized
can be followed
through
3, state
criteria monograph or a design manual.
to
the
into decimally
both
correspond of subject sections
and
guide,
on each essential design element to assure serve effectively as a checklist of rules for the
in section
within similarly numbered subsections the Contents displays this continuity
The design specifications,
3, state
base
its adequacy. how
to satisfy
each
of the
possible, the best procedure is described; when this cannot be done references are provided. The Recommended Practices, in conjunction
Design Criteria, provide successful design.
design
technological
in section
be imposed Criteria can
manager
Whenever appropriate
Both
prepares
practicing
numbered
designer
subsections
on
how
so that
to achieve
the subjects
from section to section. The in such a way that a particular
as a discrete
criteria.
concisely, with the
format for aspect of
subject.
is not intended to be a design handbook, a set of It is a summary and a systematic ordering of the large and
loosely organized body of existing successful its merit should be judged on how effectively to the designer.
design techniques and it makes that material
iii
practices. available
Its value and to and useful
CONTENTS
Page 1
INTRODUCTION
2.
STATE
3.
DESIGN
.............................
OF THE ART CRITERIA A - Conversion
APPENDIX
B - Glossary
REFERENCES NASA
.........
2
..................
and Recommended
APPENDIX
1
Practices
.................
Units
to SI Units
of U.S. Customary
99 137
............
139
............................
................................
Space Vehicle
Design Criteria
149 Monographs
SUBJECT
Issued
STATE
to Date
153
.............
OF THE ART
DESIGN
CRITERIA
2.1
12
3.1
99
2.1.1
12
3.1.1
99
Pump Headrise and Flowrate Net Positive Suction Head
2.1.1.1 2.1.1.2
13 15
3.1.1.1 3.1.1.2
Propellant Properties Turbine Drive Cycle
2.1.1.3 2.1.1.4
18 20
3.1.1.3 3.1.1.4
2.1.1.5 2.1.1.6
26 28
2.1.1.7
30
3.1.1.5 3.1.1.6 3.1.1.7
and Cost
2.1.1.8 2.1.1.9
30 31
3.1.1.8 3.1.1.9
Type
2.1.2
32
3.1.2
108
2.1.2.1 2.1.2.2 2.1.2.3
33 33 34
3.1.2.1 3.1.2.2
108 108
2.1.2.4
35
3.1.2.3 3.1.Z4
2.1.2.5 2.1.2.6
39 44
3.1.2.5 3.1.2.6
109 109 109
PRELIMINARY System
DESIGN
Requirements
Throttling Efficiency Weight
Range
and Size
Conditioning Life, Reliability, Selection
of System
Number Turbopump Rotational
of Units Equivalent-Weight Speed
Turbopump Arrangement Pump Configuration Turbine Configuration
Factor
99 102 102 104 105 105 i07 107 107
III
SUBJECT
STATE
OF THE ART
DESIGN
CRITERIA
2.2
48
3.2
112
2.2.1
48
3.2.1
112'
2.2.1.1 2.2.1.2
48 57
3.2.1.1 3.2.1.2
112 114
2.2.1.3 2.2.1.4 2.2.1.5
59 61 63
3.2.1.3 3.2.1.4 3.2.1.5
114 116 116
2.2.2
63
3.2.2
118
2.2.2.1
63
Stability
2.2.2.2
64
3.2.2.1 3.2.2.2
118 118
Tip Speed
2.2.2.3
64
3_2.2.3
119
2.2.3
65
3.2.3
119
2.2.3.1
65
2.2.3.2
66
3.2.3.1 3.2.3.2
119 121
2.2.4
67
3.2.4
121
2.2.4.1 2.2.4.2
67 68
3.2.4.1
121
2.2.4.3 2.2.4.4
72 74
3.2.4.2 3.2.4.3
122 123
2.2.4.5 2.2.4.6 2.2L4. 7
75 77 78
3.2.4.4 3.2.4.5 3.2.4.6
124 124 125
2.2.5
79
3.2.5
126
2.2.5.1 2.2.5.2
79 80
3.2.5.1
126
3.2.5.2
127
2.2.5.3 2.2.5.4
80 81
3.2.5.3 3.2.5.4
127 127
2.2.5.5 2.2.5.6
81 82
3.2.5.5 3.2.5.6
128 128
2.2.6
82
3.2.6
128
2.2.6.1
83
2.2.6.2 2.2.6.3 2.2.6.4
84 85 86
3.2.6.1 3.2.6.2 3.2.6.3 3.2.6.4
129 129 130 130
DETAIL
DESIGN
Limits
AND INTEGRATION
to Rotational
Inducer
Cavitation
Bearing DN Seal Rubbing Turbine-Blade Gear Pitchline Pump
Speed Centrifugal Velocity
Stress
Design
Inducer
Turbine
Speed
Inlet
Flow
Coefficient
Design
Performance Optimization Exhaust Pressure Turbopump Bearing Turbine
Mechanical
Integration
Placement Rotor Assembly
and Attachment
Turbopump Housing Bearings and Seals Axial Thrust Balance Thermal
Barriers
Assembly System
Interfaces
Pump Inlet Pump Discharge Turbopump Mounting Gas-Generator Connection Turbopump
Service
Turbopump
Overhaul
and Mounting
on the Engine
Start Systems Main-Propellant-Tank Head Pressurized-Gas Start Tanks Liquid-Propellant
Start Tanks
Solid-Propellant
Start Cartridge
vi
3.2.4.
7
126
SUBJECT
DESIGN
STATE
EVALUATION
Engine-System
Characteristics
Design-Point System Balance Off-Design System Balance Control Constraints System
Dynamic
Analysis
Start Throttling Shutdown
DESIGN
CRITERIA
2.3
87
3.3
131
2.3.1
87
3.3.1
131
2.3.1.1 2.3.1.2 2.3.1.3
87 88 89
3.3.1.1 3.3.1.2 3.3.1.3
131 132 133
2.3.2
90
3.3.2
133
2..3.2.1 2.3.2.2
90 93
3.3.2.1 3.3.2.2
133 134
Instability
(Pogo)
2.3.2.3 2.3.2.4
94 95
3.3.2.3 3.3.2,4
134 135
Development
Testing
2.3.3
96
3.3.3
135
2.3.3.1 2.3.3.2
96 97
3.3.3.1 3.3.3.2
135 136
System System
OF THE ART
Turbopump System Engine System
vii
LIST OF FIGURES
Title
Figure 1
J-2 centrifugal-flow
2
J-2 axial-flow
3
RL10A-3,3
4
MA-5 booster
5
YLR87-AJ-7
6
F-1 turbopump
7
Typical
8
Speed
limits
for 5500
psi LOX turbopump
9
Speed
limits
for 4300
gpm LOX turbopump
oxidizer
turbopump
fuel turbopump turbopump turbopump
pump
13
Effect
of stage design
centrifugal 14
Effect
of turbopump
pressure 15
Effect
pumps
of centrifugal
tip diameter
to pump
specific
speed
with sweptback efficiency
16
Effects
17
Schematics
18
Efficiency
pump
of rotational
21
on pump discharge = 6) .................
throttling
on pump
stage specific
pressure 25 26
.................
on throttling impeller
characteristics
of basic turbopump speed
of 27
blades ................
discharge cycle)
pressure
vs chamber 28
...............
speed on pump
efficiency
..................
speed on required
vs stage specific
17
..................
(staged-combustion
> 10 in.)
16
drive cycles ....................
limits
requirements
14
...................
Effect of drive cycle and chamber pressure (oxygen/hydrogen engine, mixture ratio
turbine
and flowrate
il
..................
map .......................
11
Headrise
10
,
Schematics
12
9
......................
10
of basic
8
..........
.......
performance
7
....................
......................
assembly
assembly
6
...............
.............
assembly
turbopump
assembly
assembly
assembly
Page
(impeller 29
........ configuration
arrangements
for a LOX turbopump
36
......
37
..................
for various
types
..°
V111
of pumps ........
....
39
Title
Figure 19
Estimated
effect
for a pump
of pump
specific
Ns-D s diagram
for various
21
Typical
point
design
22
on efficiency
speed of 1000 ......
20
relative
configuration
Page
kinds
for 1-,2-,
as a function
of pressure
Influence
25
Influence of inducer design inlet flow coefficient efficiency ..............................
27
Thermodynamic vapor
of application design
of empirical
data
28
Schematics
29
Vapor
on suction
30
Schematics
31
Typical effects turbine-blade
capability
of bearing
for driving
support
of turbine centrifugal
Typical effect of type of mixture-ratio (staged-combustion cycle) .....
37
Zero-NPSH
engine
turbines
performance
LH 2 pump.
assembly
of incremental
of variations pumping
method
capability
50
......
hydraulic
of various
pumps
propellants
and inducers
as a function
.
51
of
inducers
.
.
55
................
56
................
57
.................. at the limit of
62 69
....................
76
..........................
in pump
47
.....
5O
pressure ratio on design stress .......................
34
Effects
on inducer
for various
arrangements
Balance
36
of rocket
on cavitation
an inducer
for two-phase
33
Illustration
temperature
54
Mark 29 experimental
35
kinds
performance
head factor
32
piston
ratio (inlet
.............................
of four methods
pumping
for various
inlet flow coefficient
suppression
pressure
(supersonic
46
24
Summary
turbines
45
Approximate
26
43
3-, and 4-stage
23
of inducer
41
..........................
Typical turbine velocity ratios of 1500 ° F) ............................. regions
throttling
of pumps ...................
efficiencies
Mach number)
and engine
................
control on pump .................... for determining
NPSH on various requirements
isentropic
design
factors
for hydrogen
ix
discharge
pressure 90
headrise
.......
............ and for oxygen ......
101 103 115
Figure 38 39 40
Title
Page
Flowchartsfordetermining turbopump speed andweightattheturbine-blade centrifugal-stress limit..........................
117
Effectof turbineinlettemperature onvariousdesign factorsin turbine optimization (gas-generator cycle)......................
120
Probability ellipses forturbopumpoperation ...................
132
LIST OF TABLES Title
Table Features
of Operational
Turbopump
II
Chief Features
of Operational
Turbopumps
III
Chief Features
of Operational
Turbines
IV
Effect
!
Chief
of Propellant
Density
Page Assemblies
on Turbopump
Design
.......................
V
Advantages
and Disadvantages
of Major Turbine
VI
Comparison
VII VIII IX
Guide
for Screening
Comparison Fluids
Candidate
of Typical
Suitable
for Pump
Testing
5
(Comparison
of Oxidizer
and 19
Limits
Turbopump
Drive Systems
,4
...................
for J-2 Engine)
and Design
3
...............
Fuel Pumps
of Operational
.............
Drive Cycles on Turbine
Arrangements
for Boost .....................
xi
Pumps
..........
Drive Cycles
22 ......
........... and Preinducers
23 110
......
113 136
_mnlmm
TURBOPUMP FOR
LIQUID
SYSTEMS
ROCKET
ENGINES
1. INTRODUCTION
The turbopump assembly for a modern liquid propellant rocket engine is a complete system in itself. It consists of many components, some of which are themselves subsystems (e.g., the
pump
and the turbine).
selection of the proper into a working system. pumps, turbines, monographs (refs. Rocket
engine
This monograph
system Details
the turbopump
as a system,
type for each application and integration on the design of the various components
gears, and bearings 1 through 10).
turbopumps
deals with
have
may
be
demonstrated
found
in
excellent
the
covering
of the components including inducers,
appropriate
reliability
component
in service.
However,
because of the strong emphasis on light weight and high performance, many of the. turbopump components are designed near the limits of the state of the art. Therefore, many problem areas must be avoided if the turbopump is to be reliable and compatible with the vehicle. For must operate
example, at high
to meet the performance speed and, consequently,
and weight requirements, the turbopump must have bearings and seals that will satisfy
life requirements at high speed; in addition, the pump must have a high-suction-performance inducer so that tank pressure and weight are minimized. At the same time, the turbopump housing must accommodate the wide variations in temperature between the pumps and the turbine without affecting alignment or imposing excessive radial or axial loads on the bearings. The design problem is made more complex cycles and by a constantly advancing state of the art. The
monograph
a turbopump
is organized system
from
to follow
the logical
preliminary
design
by
succession through
the
wide
of events
testing
on
range
in the the
of possible
duty
development
rocket
engine.
of This
process normally begins with a preliminary design phase in which the turbopump size, the component types, and the component arrangement are selected to meet the system requirements. The next phase is detail design and integration, in which the final rotational speeds limits,
are selected within the the pump and the turbine
constraints of the various mechanical and hydrodynamic are optimized within the constraints of the mechanical and
fluid dynamic limits, and the components assembly. The final phase is design evaluation, by both computer simulation and experimental
are integrated in which the testing.
into an turbopump
overall design
turbopump is evaluated
2. STATE OF THE ART
The chief features of operational in tables I, II, and Ill, respectively. To minimize
inert
for a liquid machinery. even that
weight
turbopump
and thereby
increase
pumps,
the delivered
and
payload,
turbines
are displayed :,
the turbopump
system
rocket engine has the highest power-to-weight ratio in the entire field of rotating Specific horsepower (hp/lbm)*, which was high in the mid-1950's, has increased
further in more recent engines. Table have been used on flight vehicles
illustrates
assemblies,
this progress.
In approximately
I, in which rocket engine are listed in approximate 10 years,
the
specific
turbopump assemblies chronological order,
horsepower
increased
from
2.22 for the Redstone turbopump assembly to values greater than 10 for the Saturn V (F-1 and J-2 engines) turbopump assemblies. The turbopumps for Space Shuttle Main Engine (SSME), which will undergo initial development testing in 1974, will increase the specific horsepower
by another
order
of magnitude
to a value
greater
than
100.
r
This
increase
systems
in specific
that
can
horsepower
operate
reliably
has been
achieved
at
speed.
high
developed and utilized in systems that inducers then were developed to enable pump inlet pressurization requirements,
primarily
by developing
High-speed
bearings
turbopump
and
seals
were
exert low axial thrusts on the bearings. High-speed high-speed pump operation without any increase in a very important condition because of the increase
in inert weight involved in raising the pressure level in large propellant tanks. As shown for the F-1 and J-2 turbopump assemblies, the high-speed technology has made it no longer necessary to use gearing to couple As a result, the latest designs have In general, centrifugal design
are
a high-volume pump the pump connected
the pumps for all propellants other than hydrogen designs (table II). Simplicity, light weight, and good the
reasons
for
this
selection;
the J-2
figure 1 is a good example of the current state requirement has made it necessary to employ centrifugal-stress limits stage) axial design was was
used
with an efficient, high-speed directly to the turbine.
for
the
centrifugal-flow
of the art. For multiple stages
have been performance oxidizer
hydrogen, in order
pump
turbopump
with the fuel and oxidizer has both pumps and the
pumps turbine
LR81-BA-11
had
for
converting
U.S.
shown
in
the high headrise to stay within the
(fig. 3). The
Atlas
MA-5
booster
turbopump
(fig. 4) has the fuel and oxidizer pumps on the same shaft and uses a double from the drive turbine. The Titan II YLR87-AJ-7 (fig. 5) has a geared-turbine
*Factors
single-stage, on and off
for pump impellers. As a result, an eight-stage (including the inducer used for the J-2 fuel pump (fig. 2) and a two-stage centrifugal design
RL10
engine
turbine.
have
customary
two
units
on separate shafts. The F-1 turbopump on the same shaft. The turbines for rotors,
to the
the
International
2
design
System
being
of
Units
either
(SI units)
assembly
gear reduction arrangement
assembly (fig. 6) all but the Agena
velocity
are given
compounded
in Appendix
or
A.
Table
I. -
Chief
Features
of Operational
Turbopump
Assemblies
Turbopump
Engine Chamber Application
Designation A-7
Redstone
MB-3
Thor
LR87-AJ-3
Titan
Thrust,
I,
lbf
Assembly Specific
Efficiency,
pressure, psia
Arrangement
78 000
318
170 000
594
i Geared
turbine
150 000
585
Geared
turbine
I st stage
t
Weight, Ibm
percent
Single shaft, turbine in middle
and
horsepower, hp/lbm
Start
system
26.4
332
2.22
Liquid tank
46.0
562
5.40
Solid propellant
45.8
720
5.11
Liquid propellant
start tanks
start
monopropellant
start
start
cartridge
pumps 80 000
682
Geared
O 2 pump
34.0
204
7.25
Liquid propellant
205 000
702
Geared
turbine
47.0
520
7.98
Solid propellant
start cartridge
Atlas
57 000
706
Geared
turbine
35.0
229
7.27
Solid propellant
start cartridge
MA-5 booster
Atlas
330 0002
577
Geared
turbine
48.0
875
3.59
Solid propellant
start cartridge
F-1
Saturn
Single shaft, turbine on end
44.6
3150
YLR8 I-BA-11
Agena
Geared
LR91-A J-3
Titan l, 2nd stage
H-1
Saturn
MA-5 sustainer
IB
L_ IC
1 522 000
1122
16 000
506
turbine
and
20.0
turbine
and
38.1
tanks
16.6
Tank head
5.81
Solid propellant
start cartridge
484
10.70
Solid propellant
start
256
8.30
Solid
start cartridge
9.03
Tank head Pressurized-gas
60.5
pumps YLR87-AJ-7
Gemini-Titan
215 000
784
1st stage
cartridge
pumps
Gemini-Titan
YLR9 I-A J-7
Geared
1O0 000
804
Geared
N204
15 000
400
Geared
02 pump
42.0
787
Dual turbopump, series turbines
37.4
305
7.73
44.9
369
21.60
56.5
555
50D
58.5
701
108.9
pump
propellant
2 nd stage RLIOA-3-3
Centaur
J-2
Saturn S-II and S-1VB
SSME (EPL*) 3
230
Space Shuttle
000
512 300
3237
high pressure
IBased 2Two
on
the
engines,
best each
available developing
data
as of 165
000
mid-t973.
Numbers
presented
Dual turbopump, parallel turbines
are
those
for
a turbopump
hp.
3Not operational, but presented for comparative purposes.
*Emergency po_,er level.
operational
76.1
system.
Tank
head
start tank
Table II. - Chief Features of Operational Turbopumps* Rated Number
Propellant Engine designation
Propellant
A-7
MB-3
LR87-AJ-3
density, lbm/fta( i )
Oxygen Alcohol (41
71.4 56.6
Oxygen RJ-I
71.4 53.2
Oxygen RP-I
71.4 50.5
Pump type
of stages
Centrifugal
I
Discharge
inlet
Head
Weight
Volume
Rotational
pressure, psia
pre_ure, psia
rise, ft
flowrate, lbm/sec
flowrate, gpm
speed, rpm
SPSHmin, ft 12)
NPSHc rit, riO)
Efficiency,
Power,
percent
hp
356
49.8
616
205
1 290
4 718
18
11
72.0
32O
464
42.5
1 139
150
1 190
4 718
40
35
70.0
418
867
53.0
1 651
456
2 870
6 303
55
79.0
1830
913
48.0
2 337
202
1 700
6 303
34
72.0
1 210
798
53.0
t 510
412.7
2 600
7 949
40
1034
22.0
2 881
183.3
1 630
8 780
30
i LR91 -A J-3
Oxygen RP-I
71.4 50.5
Oxygen RP-I
70.8 50.5
Oxygen RP-I
i H-1
MA-5 sustainer
MA-5 booster
F-I
YLR81-BA-I
YLR87-AJ-7
YLRgI-AJ-7
RL I0A-3-3
1
35.0
1613
175.6
1 100
8 945
31
42.0
3 024
74.1
659
25 207
100
980
65.0
1 851
537
3 410
6 680
35
25
77.8
2 340
1020
57.0
2 719
240
2 130
6 680
35
28
71.8
1 670
71.4 50.5
982
53.0
1 879
193.2
1 200
10 160
30
14
64.2
1 018
996
77.0
2 616
91.6
745
10 160
85
60
64.5
Oxygen RP-I
71.4 50.5
877
50.0
1 679
839
73.0
2 184
Oxygen
71.4
t600
65.0
RP-I
50.5
1856
45.0
IRFNA (s)
98.2
949
24.0
1 360
UDMH (6)
49.4
749
24.0
2 110
N204 A-50 (7) r
90.3 56.1
1182
84.0
1 740
1363
33.5
N204 A-50
90.3 56.1
Mixed Flow Mixed Flow
1112 1201
Oxygen
68.8
Cent rifugal
_r
597
60.5
t120
Centrifugal
2
990
30.0
31 800
1
1114
39.0
2 185
460.4
2 920
8 753
25
1238
30.0
38 000
83.6
8 530
27 130
130
31 000
(9)
37 400
(9)
Hydrogen J-2
I I
819 1097
Oxygen Hydrogen
4.35 70.8 4.4
Centrifugal Axial
7+ inducer
458
62O
2 862
6 314
40
211
1 867
6314
33
74.3
1800
73.6
I 151
3097
4070
25 200
5 488
65
5 168
1715
15 250
5 488
70
,60
74.6
30 200
55
72.6
22 I00
39.3
180
25 389
12
15.3
139
14 410
34
550
2 700
8 382
44
68.0
2 560
3 381
274
2 180
9 209
43
68.0
2 480
41.0
1 713
207
1 010
8 405
30
67.4
960
44.5
2 981
115
904
23 685
100
57.1
1090
I
352 352
28.2
184
12 100
17
62.9
5.6
581
30 250
132
55.0
592
94
18
80.0
2 358
75
73.0
7 977
(9)
78.1/ 69.6
27 400
(9)
74.1
76 400
stage SSME (EPL)**
Oxygen
70.4
Centrifugal
1/2 Is)
high pressure Hydrogen
4.38
Centrifugal
3
5174/ 8491
379/ 4940
9640/ 7100
6981
188
193 900
1137/ 120
7250/ 633
160.5
16 450
*Bared on the best data as of mid-1973. Numbers presented are those for a turbopump operational system. (l)At temperature specified by the application.
(6)Unsymmetrical dimetbylhydmzine,
(2)Contraentatly specified pump NPSI{, maximum acceptable.
(7)50 percent hydrazine and 50 percent UDMH,
(3)NPSII at a given drop in punlp discharge pressure, generally 2 percent.
(8) I0 percent of the flow goes through a second stage; numbers below slash are _r _eond stage alone_
(4)75 percent alcohol, 25 percent water.
(9)Boost pump upstream.
(5)Inhibited red fuming nitric acid.
(CH3) 2 NNH 2.
**Not operational, bat presented for comperative purpo-_.
Table
Number Engine
Working
designation A-7
fluid H202
I*
-
Chief
temperature, Type 2-row velocity
Features
of Operational
Turbines
_
Inlet
Inlet
of stages
Ill.
pressure, psia
°F
Pitchline Pressure
Flowrate,
Rotational
ratio
Ibm/see
speed,
740
37I
21.2
1204
523
17.6
1334
413
17.8
1240
560
28.2
622
17.7
6.68
rpm
Efficiency, percent
Power, hp
veloc!ty, ft/sec
4 718
37.2
758
412
15.4
30 540
-
3 162
1200
11.7
25 172
63.6
3 678
1035
25 207
63.1
1480
1020
32 700
70.2
4 141
1290
compounded MB-3
O2/RJ-I
2
pressure compounded
LR87-AJ-3
O2/RP-I
pressure
2
conrpounded LR91-AJ-3
O2/RP-I
2
pressure
6.31
compounded 18.0
H-1
O2/RP-1
2
pressure compounded
1200
MA-5 sustainer
O2/RP-I
2
pressure
1075
692
25.0
9.73
38 000
46.3
1 663
995
1240
497
15.7
14.07
30 986
66.9
3 140
1205
1450
920
16.4
5 488
60.5
52 900
840
compounded MA-5 booster
O2/RP-I
2
pressure compounded
F-1
02/RP-I
1*
2-row velocity
171.8
compounded
_q YLR81-BA-I
1
YLR87-AJ-7
IRFNA/UDMH
I
impulse, partial admission
14130
480
37.7
1.58
24 800
41.0
352
855
N204/A-50
2
pressure
1667
443
17.8
12.85
23 992
56.0
5 180
980
1650
420
29.0
5.46
23 685
53.0
2 122
964
106
698
1.42
5.35
30 250
74.0
687
782
2.5
5.0
8 753
48.4
2 358
590
7.3
7.0
27 130
60.1
7 977
1480
31000
72.9**
27 400
1363
37 400
79.0**
76 400
1661
compounded YLR9 I-A J-7
N204/A-50
2
pressure compounded
RL10A-3-3
Hydrogen
2
pressure compounded
J-2 - Oxidizer
O2/H 2
1"
2-row
velocity
760
89.5
compounded Fuel
O2/H 2
1"
2-row
velocity
1200
657
compounded SSME (EPL) (21 High Pressure Oxidizer Fuel
O2/H 2
2
reaction
1101
5880
159"*
63.3
O2/H 2
2
react ion
1391
5820
1.56"*
159.0
IBased on the best available data as of mid-1973.
Numbers presented are those for a turbopump
,xhal from the information given above. 2Not operational, but presented for comparative purposes.
system. Similar tables in monographs on system components (e.g., ref. 4) present design values that may differ some*l-stage because it has 1 nozzle, 2-row because it has 2 rotor blade rows. **Total-to-total fluid condition basis.
Turbine nozzles
Pump volu'
inlet -- Ind ivid ua I
turbine
rotors
Thrust balance cavit' i Impeller ! 1
package
/ bearing
Inducer inlet
/ ;i
am---
coupling
Vent bearing
for lube Forward
bearing Radial pins for thermal
Turbine
Figure
1. --J-2 centrifugal-flow
oxidizer turbopump
6
assembly.
inlet
volute Pump stator
Individual turbine rotors
Aft bearing
carrying clampingbolt
_or aft bearing lube
Forward r for
bearing
growth Inducer stator and forward bearing support
Pump rotoI
turbine inlet manifold
Thrust balance piston
Figure 2. --J-2 axial-flow
fuel turbopump
i
assembly.
pins thermal
Oxidizer
inducer
inducer
Hydrogen
Fuel pump impellers
pu_p bearing
Oxidizer gear
inlet manifold
drive
Turbine bearing
Turbine
Figure 3. - RL10A-3-3
turbopump
rotoz
assembly.
Imp • 11 er backface :ribs
Oxidizer impeller
pump Fuel
pump impeller
inducer Oxidizer
Turbine inlet manifold •
Gear reduction
Turbine quill shagt
Figure 4. - MA-5 booster turbopump
assembly.
rotors
1
Oxidizer
pump impeller
pump impeller
o
Oxidizer pump drive
gear cooler
Turbine inlet manifold
Oil re
Turbine rotors Fuel gear
pump drive Oil :pump Figure 5. - YLR87-AJ.
7 turbopump
assembly.
Turbine inlet mmifoid pump
Turbine rotors
.... '
impeller • i
Forward
Fuel inducer
bearings Oxidizer impeller
Oxidizer inducer
coupling
Purge¸ seal package
Radial thermal
Figure
6. -
pins for growth
F-1 turbopump
assembly.
pressure variety To
compounded;
the
of approaches
meet
future
LR81-BA-11
for turbopump requirements
had
design for
high
a single-stage
impulse
and arrangement chamber
turbine.
has been
pressure
and
Thus,
a wide
used. wide
throttling
range,
turbopump systems will continue to use single-stage centrifugal pumps for propellants tither than hydrogen and multistage centrifugal pumps for most hydrogen applications. Two-phase pumping
capability
will
become
increasingly
pressurization system by permitting the Future attitude control and auxiliary accumulators
charged
by
small
turbopump will have to start seconds. As a result, rapid
important
cryogenic* propulsion
turbopumps.
as a means
to simplify
the
tank
propellants in the tank to be saturated. engines may be fed through gaseous
To
minimize
the
accumulator
size,
each
thousands of times and, after each start, operate for only a few and efficient pump preconditioning will become increasingly
important.
2.1
PRELIMINARY
2.1.1
The
System
turbopump
will be used.
Requirements
system The
DESIGN
basic
requirements requirement
stem
from
is that
the
the engine turbopump
system
in which
operate
engine system while delivering propellants at the conditions chamber; more specifically, the various sytem requirements
the vehicle
Because
pump
to deliver speed
is
a proportionately a
performance, pump rotational important design parameters
strong
higher
influence
in
the turbopump the limits
of the
required by the engine thrust are evaluated individually and
collectively for their impact on the turbopump design. In turbopump a strong emphasis on attaining the lowest weight, because a relatively enable
within
system design, there is low system weight will
payload. all
phases
of turbopump
design
and
speed N and pump specific speed Ns evidence themselves as in the evaluation process. These parameters are related by the
expression
Ns -
*Terms
and
symbols,
materials,
and
vehicles
and engines
NQ y2
(1)
H3/_
are defined
12
or identified
in Appendix
B.
where Ns= specificspeed,rpm-(gpm)v2/ft N = rotational Q -- volume
speed,
An
increase
generally increase
reduces in speed and,
gpm
ft
in the
efficiently
rpm
flowrate,
H = headrise,
3/_
rotational
the head can reduce if
the
speed
increases
the
pump
specific
coefficient and, therefore, increases the number of pump stages required
specific
speed
is below
2000
to
3000,
speed,
which,
in turn,
the throttling range. An toJ provide a given head can
increase
the
pump
efficiency. In addition, increasing the speed will often increase the turbine efficiency and will almost always reduce the turbopump weight. On the negative side, increasing the speed increases the pump inlet pressurization requirements, decreases pump life, and increases the cost required to attain a given reliability.
2.1.1.1 Pump
PUMP HEADRISE headrise
and
AND
flowrate
FLOWRATE
are the
basic
requirements
imposed
on the turbopump
system,
because the specified pressure and quantity of propellant must be delivered to the engine thrust chamber if it is to develop its design thrust. Pump designers generally prefer to work with h6adrise and volume flowrate because for a given pump design these terms are •independent consequence
of propellant density, of this independence,
whereas a pump
flowrate (fig. 7) applies to all propellant as in high-pressure hydrogen pumps.
pressure rise and performance map
densities
unless
Pump,developed head at the design value for propellant the difference between the discharge head and the (discharge pressure) must be sufficient to overcome propellant example,
system
and
deliver
in a regeneratively
pressure must cooling jacket,
propellant cooled
engine
with
required density
in a pump or, more
chamber
frequently accurately,
compressibility
pressure
chamber
are present,
at the required cycle,
the
pressure. pump
will be less than
by dividing propellant
the
design
For
discharge
line losses, the pressure drop and the engine chamber pressure.
is approximated by the average
13
effects
flowrate (engine design thrust) is suction head. The discharge head the hydraulic resistances in the
a gas-generator
equal the sum of the propellant the pressure drop in the injector,
pressure is insufficient, the engine design thrust will not be reached. The headrise the propellant
to the thrust
weight flowrate are notl As a based on headrise and volume
in the If the
value,
and
the pressure rise by density. The most
l =-
,-r,
I
I Volume
I
flowrate
Q,
I
I
gpm
Figure 7. - Typical pump performance map.
accurate
method
that will headrise.
propellant
produce the required For situations in
high-pressure in increments adding
is to use the
hydrogen with the
the
actual
to determine
the isentropic
pressure rise. This enthalpy rise which propellant compressibility
pumps), enthalpy
increment
properties
increment
divided
by the
rise
can then be converted is significant (e.g.,
the procedure involving isentropic enthalpy at the beginning of a given increment being
(isentropic
enthalpy
pump
to in
rise is applied determined by
efficiency)
to the
previous actual enthalpy. This procedure is necessary because the propellant heating caused by the inefficiency in one increment decreases the flow density at the beginning of the next increment. As a result, the headrise based on isentropic enthalpy rise as the sum of the isentropic increments is roughly half-way between the headrise determined from th e inlet density
and
the
headrise
heating effect had discharge pressures propellant
heating
based
on
a single-step
little influence on were approximately is a very
significant
the
factor
RL10 1000
isentropic
for 6000-psi
14
enthalpy
rise.
This
and J-2 hydrogen pumps and 1250 psia, respectively. hydrogen
pumps.
propellant because the However,
The volume flowrates required from the pumps are establishedfrom the engine thrust, engine specific impulse, engine mixture ratio, and the density of the propellant being considered.Thereis often somequestionasto whether to usethe inlet, the discharge,or the averagedensity. The inlet density is easierto determine,but the dischargeor the average density is more accurateif the headriseis to be calculatedfrom the flow-passage geomet_ and the volume flowrate. A generalpolicy is to use inlet density for preliminary design where accuracyis not critical, and local densityfor the detail designof bladeflow passages. During off-design operation, the requirementsgenerally are determined over the entire operatingrange, becausea turbopump system often encountersits most severeoperating conditionsduring off-designoperation. Figures8 and 9, which presentvariouskinds of pumpspeedlimits for specifiedconditions, illustrate the effect of pump flowrate and headrise on the turbopump design speed. Regardlessof the mechanicalor hydrodynamic limit (sec.2.2.1) at which the turbopump is designed,increasingthe flowrate decreasesthe allowabledesignspeed.If the turbopump is designedat the bearing DN or the seal speed limits, increasingthe headrise(discharge pressure)also decreases the allowabledesignspeed.However,if the pump is designedat a givenvalueof specificSpeedto obtain the characteristicsof a giventype of pump, increasing the headrisewill increasethe speed.As far as the other limits are concerned,discharge pressurehaslittle effect on speed.
2.1.1.2
NET POSITIVE
By definition, the head due
SUCTION
HEAD
net positive suction head NPSH is the difference, at the pump inlet, between to total fluid pressure and the head due to propellant vapor pressure; it is
expressed in feet of the propellant being pumped. In vehicle, the NPSH is determined from an optimization
the preliminary that considers
design phase the weights
of the of the
vehicle tank, the tank pressurization system, the pressurization gas, and the feed line and, in addition, the system cost, the pump efficiency, and the turbopump weight. The trade often is made without the last two items because, in most cases, vehicle considerations far outweigh engine considerations. If the
NPSH
is less than
a certain
critical
value,
cavitation
will occur
in the
pump
inlet,
and
the pump headrise will be less than the design value. This critical NPSH is usually the value where the headrise is 2 percent less than the noncavitating value. Three methods have been used to correct the problem of an NPSH insufficient for the pump to meet design requirements:
(1)
increasing
the
tank
pressure,
which
increases
the
NPSH
supplied
but also
increases the required tank wall thickness and the tank weight; (2) decreasing the pump design speed, which decreases the NPSH required but also decreases the pump efficiency and increases the turbopump weight; and (3) redesigning the pump inlet by increasing the diameter and lowering the flow coefficient, a step that decreases the NPSH required but can decrease
the
pump
efficiency.
(Method
(3) was used
15
to meet
the
engine
requirements
with
Propellant = liquid oxygen Pump discharge pressure = 5500 psi Turbine centrifugal stress = 35 000 Seal DN
= 1.5
Shaft 100
rubbing
speed
= 325
psi
ft/sec
X 106
stress
= 40 000
psi
N s = 2000
OOC
= 63
000
_/o)
10 000
ill I
1000 !000
I
I
I
¸
I
I
I
I
I I Itl
I I 10 000 Flowrate,
Figure
8. -
Speed
limits
for
5500
i00 gpm
psi LOX
16
turbopump
(ref.
2).
000
100
000
Turbine
SO
000
20
000
(low
Turbine
NPSH
NPSH
O O
10
stress
=
SO
pressure
ratio_
stres_e
ratio)
ft
= 30
ft
000
NPSH
=
I0
I
ft
SO00 Propellant = liquid Pump flowrate - 4500 Turbine centrifugal 55 000 psi Seal rubbing DN= 1.5 xl0 Shaft stress 2000
Ns =
200
400
s_eed=525 =
I000
9.
-- Speed
limits
40
000
=
ft/sec psi
2000
2000
4000
Headrise,
Figure
oxygen gpm stress
for
4300
17
ft
gpm
LOX
turbopump.
10
000
20
0@0
the
J-2 hydrogen
pump
speed,
2.1.1.3
pump.)
and pump
A more design
PROPELLANT
detailed
discussion
of critical
NPSH,
great influence on turbopump system design-point requirements (i.e., headrise, the bearings and seals, the corrosive,
design. NPSH, cooling,
is contained
of the
in section
interrelation
2.2.1.1.
PROPERTIES
Propellant physical properties have a Propellant density affects the turbopump volume flowrate, and horsepower). In
lubricating, and viscosity characteristics of the propellant affect the speed limits (secs. 2.221.2 and 2.2.1.3), the method of lubrication, bearing-and-seal package. The size of the turbopump arrangement (sec. 2.1.2.4). material selection and thermal-conditioning
bearing-and-seal The propellant requirements
specific heat, s_ecific heat ratio, and molecular the blade heights and pitchline velocities, the (secs.
2.1.1.6
Density Densities
material selection, the and the size of the
package can also influence the saturation temperature affects for the pump. In the turbine, the
weight of the turbine working fluid affect type of staging, and the number of stages
and 2.1.2.6).
is the propellant of rocket engine
acid (IRFNA) to 4.4 lbm/ft illustrated by a comparison
property propellants
that has the greatest range from 98 lbm/ft
3 for liquid hydrogen of the J-2 liquid-oxygen
effect on turbopump design. 3 for inhibited red fuming nitric
(LH2). The effects and liquid-hydrogen
of this wide range are turbopumps (table
IV). As shown, the headrise whole LO2 pump, and
in the LH2 inducer alone is more than twice the headrise of the overall headrise for the LH 2 pump is almost 20 times the LO2 value,
even though the pressure difference in headrise, the
rises for tip speed
the two pumps are very similar. As a result of the of the LH 2 pump is more than twice as great as the tip
speed of the LO2 pump ; if the LH2 pump were a single-stage the LO2 pump), the tip speed would have to be four times addition, the LH2 much horsepower, times
pump weighs 64 pounds even though the weight
centrifugal type (i.e., similar the value for the LO2 pump.
to In
more and requires more than three times as flowrate of the LO2 pump is more than five
greater.
The
differences
the
fact
that
in size between hydrogen
LH2
turbopumps
turbopumps
and
can be rotated
other
faster
turbopumps
because
the
are minimized bearing
DN and
by seal
speed limits are higher. Because of the relatively high density of LO2, the LO2 pump usually can be designed to deliver the required pressure rise in a single stage at the optimum rotational speed without encountering any limitations on component speeds that cannot be alleviated. However, the cavitation characteristics and the density of LO2 are such that a preinducer would be required in order to avoid high the low density of LH2, a LH2 turbopump usually would
produce
the highest
payload,
because
tank pressures. Conversely, cannot be designed at the
the centrifugal
18
stress
on the turbine
because of speed that blade
limits
Table
IV.
-
Effect
of Propellant
(Comparison J-2 Engine)
Density
of Oxidizer
Pump characteristic
centrifugal
axial
Number
1
7 + inducer
lbm/sec
460.4
83.6
gpm
2920
8530
rpm
8753
27 130
1075
1208
of stages flowrate, flowrate, speed,
Pressure
rise, psi
Inducer
headrise,
Pump
headrise,
NPSH,
Power,
density,
RP-1
are
assumed discussion
-For
rocket
rotational same
shaft,
LO2/LH2
2.2.1.4).
Since
not
widely
for
such
on
engines
Ibm
However,
cavitation
known,
as
in the pumps
F-I
not
until
the and engine
generally
(no inducer)
5050
2185
38 000
18
75
390
865 - -
305
369
2358
7977
has
excellent
require
test
of
data
sufficient consult
two the
fuel
system. are
are
lacking, data
When
mounted
other
applications, NPSH
than
at
LH2,
like
obtained.
and
that
For
a
the the
tank
L02,
and
of water
a more
is
detailed
2.2.1.1.
oxidizer
are
similar
pumps
propellants on separate
19
zero
performance are
densities
and
some
at
propellants
test
characteristics
For
feasible
section
propellant
cavitation
a preinducer.
operation
characteristics
because
similar,
LH2
make
performance,
in which
), the
may
hydrogen
propellants
are
since
pump of
the
cavitation
speeds
ft/sec
hp
a hydrogen
2.2.1.1).
tip speed,
weight,
characteristics
(sec.
ft
discharge
Turbopump
cavitation
ft
ft
Impeller
low
for
Pump type
Rotational
(sec.
Design
Pumps
Liquid hydrogen, 4.4 lbm/ft 3
Volume
speed
Turbopump
Fuel
Liquid oxygen, 70.8 lbm/ft 3
Weight
the
on
and
(e.g.,
generally differ
shafts
widely so that
LO z/RP-1), are
pump
mounted
on
in densities each
can
operate
the
(e.g., at
its optimum rotational speed, as in the J-2 engine. The RL10 engine used only one turbine by.gearing the LO2 pump to the LH2 turbopump shaft (fig. 3). This arrangement avoided the low efficiency of a small, low-speed oxidizer turbine. Specific heat on isentropic
Cp and specific heat ratio "y of the turbine spouting velocity Co and, therefore, on
two-stage (LO2/RP-1
pressure-compounded and N204/A-50),
and
high-energy exceptions
hydrogen-fuel to this rule are
combinations the Redstone
working turbine
fluid have a large influence design (ref. 4). As a result,
turbines are used for low-energy-fuel two-row velocity-compounded turbines (table III). Low-energy-fuel A-7 and the F-1 turbines,
combinations are used for
turbines which were
that are two-row
velocity-compounded turbines because they had low pitchline velocities. The RL10 and the planned SSME turbines are high-energy-fuel exceptions because the turbine drive cycles for these engines call for pressure ratios of only 1.4 to 1.6 and, therefore, lower isentropic spouting velocities. Section 2.1.2.6 provides additional information on turbine relationships.
2.1.1.4 The
TURBINE
method
used
DRIVE
CYCLE
to drive
the
turbine
has
a direct
effect
on the
pump
headrise
and
power
requirements and on the pressure ratio and flowrate available to the turbine for supplying the necessary power. Thus, the turbine drive cycle used on a liquid rocket engine affects the design requirements of the turbine as well as those of the propellant pumps. Figure 10 (ref. 4) shows
typical
flow
schematics
for the
basic
types
of turbine
drive
cycles.
If the turbine
flow is in parallel with the combustion chamber (gas-generator and thrust-chamber-tapoff cycles), the pump head and power requirements are relatively low. However, if the turbine is in series with the combustion chamber (expander and staged-combustion cycles), the pressure
drop
across
the
therefore, discharge
the head and pressure is more
turbine
must
be
added
power requirements sensitive to changes
to
below
the turbine
pump
discharge
are high; in addition, in the pump and turbine
To date, most rocket engines have used the gas-generator which the turbine working fluid is derived by combustion at a temperature
the
temperature
limits.
pressure
the required efficiencies.
(GG) cycle (figs. 10(a) of the main propellants If the turbine
exhaust
the turbine a relatively
pump
and (b)), in in the GG
is afterburned
by the introduction of additional oxidizer, higher performance can be obtained GG cycle. The J-2S development engine utilizes a variation of the GG cycle called cycle (fig. 10(c)), in which injector at a location where
and,
from the the tapoff
working fluid is tapped off near the face of the cool gas is available. In the expander (or hot-fuel
tapoff) cycle (fig. 10(d)), which is used for the RL10 engine, the hydrogen that is evaporated and heated in the thrust-chamber regenerative jacket is used to drive the turbines. The turbine exhaust gas is then fed to the main chamber for combustion. In the staged-combustion then discharges
into
cycle the
(fig. 10(e)), a preburner generates main combustion chamber, where
Occurs.
20
the turbine the second
working fluid, which stage of combustion
% (a)
(b)
Bipropellant gas generator
Honopropel 1ant gas generator
(c)
Thrust tapoff
chamber
0 o Oxidizer pump J F Fuel pump T Turbine GG Gas generator P Preburner
I
(d)
Figure
The V;
general this
advantages
table
compares
the
turbine
For
any
cycle,
and
the
turbine by
pressure
turbine. pump the
must
drive
the
discharge staged-combustion
figure cycle
pressure,
the
also,
because cycle,
In
1500
the
chamber
chamber the
GG
is in series pressure
21
can
of
tapoff
cycles,
fluid
with
the
as high
the
pump
penalties
is relatively expander
be
potential
imposed
working the
growth
on
low
on
cycle
chamber
combustion as 3000
turbine
account
to entering requires
of the
a higher
chamber. psi
the
discharge
the on
on
engine,
of high
prior
VI
pressure,
limitations
the
in table Table
requirements
and
pressure
listed
cycle.
performance,
pressure
to avoid
are
each
and
chamber
turbine
pressure,
turbine
flowrate
with
in order the
level,
pressureCratio and
cycles of
control,
and
the
engines
for
drive
thrust
of cycle
the
chamber
turbine
characteristics
pressure
psi
available Same
basic the
drive cycles (ref. 4).
mixture-ratio
within
effects 11.
below
of heating for
the
pump The
kept
expander
amount However,
in are
and
meet
the
system.
illustrated
In
limited
pump
of combination;
throttling cycles.
engine
of the
propellant
drive
must
generally
flowrates. the
the
the
are
pressures
of
combustion
of basic turbine
comparison
temperature,
of the
turbine
disadvantages
concise
effect
working-fluid each
a
Staged
10. - Schematics
and
gives
(e)
Expander (fue 1)
because
For the
Table
Cycle General
Monopropellant (a)
advantages
Extensive
GG
experience
Turbine
-
Advantages
Bipropellant (a)
Extensive
and
power
ent of main
independ-
engine
Disadvantages
experience;
(a)
response
multiple (d)
(e) bo bo
for
Cycles
Thrust
expander
(a)
performance
(e.g., Thor, Jupiter, Atlas, Titan II, S-IC,
(b)
S-IB, S-IVB)
(c)
Component insensitive
Minimum
(b)
performance to other com-
Simple,
reliable
(d)
Clean,
throttling
Good uprating
(d)
Control and operation over a wide thrust and
components/
Staged
combustion
(a)
Maximum
(b)
Allows
performance
working
Component performance insensitive to
pressure without
other some
penalty
components; interaction
high chamber and throttling performance
be-
tween T/C and turbine drive
noncorrosive
turbine
(c)
Minimum
tapoff
components
fluid
(c)
Good uprating capability
chamber
high reliability
ponents, i.e., little interaction
restarts
Simple control
Drive
system (b)
Quick
High
Turbine
used
operation (c)
of Major
Propellant
GG
most frequently
with H202 (b)
V.
Good
uprating
capability
capability
(d)
Simple control
(a)
More complex thrust chamber design
(a)
Interaction
(b)
Limited uprating capability
(c)
Engine performance sensitive to component
throttling
mixture ratio range relatively easy General
(a)
Low
(a)
performance
Low performance when
disadvantages
(b)
Additional system
propellant
required
one of main
(except
afterburning
(a)
is
(b)
1000
Carbon
buildup
on tur-
bine nozzles may occur with some fuel-rich
(c)
Heavy
propellant
combinations
(d)
Limited
Additional
combustor
(b)
pressures
because
performance
loss
of
(c)
feed system over tapoff cycles
psia because available fuel
and
High systempressures Poor uprating capability
required and expander
(d)
Engine sensitive design
Requires
higher
discharge
pressures
pump
of from
(b)
tapoff thrust
performance to component
between
system and chamber assem-
bly increases opment
(c) to low chamber
to approximately
power heated
is a monopropellant and complex
in chamber
pressure
used)
unless
propellants
Limited
risk
devel-
design
Table
__
Turbine
_
VI.
-
Comparison
of Operational
and
Design
Limits
on
Turbine
Drive
Cycles
Drive
Cycle
Monopropellant
GG
Bipropellant
Propellant
GG
expander
Thrust
chamber
tapoff
Staged
combustion
Constraint Propellant combination
(a)
H 2 02,
(a)
N 2H4, or similar
monopropellant used with pellant
can be
any main
Can be used propellant
ance
(h)
Fuel-rich
cycle usually is not considered unless one of
blades
main
major
level
(a)
Becomes
passage
(b)
problem
(a)
le_s attractive
increasing
level
of bipropellant
Equally
corrosive, blades
chamber is a
pumps
(a)
for all turbo-
are practical
as performance
and weight
penalties
but carbon
and thrust
walls
some
where
System
not feasible
(a)
at
high thrust
level,
transferred
per pound
propellant
pumped
used with
is a major
Equally thrust
as heat
(h)
cham-
problem
propellants
suitable levels
turbopumps
of
engine
there
is little
be.
perform-
ance gain
passage
some
an all-
and non-
on turbine
ber coolant with
levels
Not
monopropellant
gases are high-
deposits
with
suitable
thrust
thrust
Fuel.rich
est performing
walls
(a)
cause
carbon
and thrust
any
combination
fuel
propellants
with system tO
engine
Can be used with propellant
on turbine
coolant
is a
(a)
with
and nonbut
deposits
monopropellant Thrust
systems
I-I2 or light hydrocarbons
GG is highest
corrosive,
of low perform-
and high weight,
propellants
Limited'to as the main
combination
Because
(a)
any
properforming
(b)
with
combination
for all
(a)
Fuel-rich
is highest
and noncorrosive,
but carbon
deposits
on turbine
blades
and thrust
coolant
passage
problem
with
some
propellants
At low thrust
levels
and high
where
pressure,
are practical
ciencies
decreases
preburner
performing
chamber walls is a major
decreasing drive
pump
up pump
effi-
discharge
pressure_
in-
crease Performance
(a)
Lowest-performance when
used with
main
engine.
family
system
(a)
order
Hydrazine
gives much
performance
than
Low performance
at lO00-psia
peroxide
sure.
Loss
chamber Pressure
(a)
Pump
discharge
approximately chamber
pressure
is
(a)
I s on the
of 1/3 to
higher
Pump
pressure
rage over
to
creases
other
with
chamber
pressure
discharge
chamber
pres.
proportional
No loss in engine
pressure
perform-
(a)
is
(a)
1.5 times
Pump
same
as bi-
(a)
GG cycle
No engine advantage creases
in-
System
applicable
chamber
(c)
at all
pressures
(b)
Cycle
is equally
at all chamber
performance with
Performance
loss very
increases
large
chamber
pressure
at high
pressures
although
increasing
(b)
System
pressure
is
(a)
All comments propellant
2.5 times
for bi-
(a)
GG cycle
limited
to
(b)
chamber
sure by power ments
other
systems
increasing
in-
chamber
Pump
discharge
pressure
linear
function
of chamber
sure and may
pressure
1000-psia loss
over
loss due Performance
pressure
increasing
discharge
chamber
applicable
bleed.
with
chamber
(b)
performance
to propellant
pressure
approxilnately
pressure
systems
Performance propellant
ance due to propellant bleed. Performance advan.
1 percent
chamber
approximately
1.5 times
(a)
system.
Loss in engine
bipropellant
pres.
require-
and power
Turbine
pressure
available lower
may
than
pres-
2.0 times
pressure
ratio
be slightly
with
exceed
is a non-
GG
(h)
Cycle
has an upper
ber pressure
limit
of cham-
operation
available
pressure (continued)
Table
_
VI. -
Comparison
of Operational
and Design
Limits
on Turbine
Drive Cycles
(concluded)
cycTUrbine Drive Bipropellant
Monopropeltant GG
le
GG
Staged
Thrust chamber tapoff
Propellant expander
combustion
Constraint Working-fluid temperature
(a)
Temperature should be as high as practical to max. imize performance. Complex, cooled turbines increase performance and reduce weight
(a)
increase performance and complexity and reduce weight
(b)
t,_
Throttling and mixture.ratio
(a)
(b)
(c)
Growth potential
(a)
Controlof quired
four valves re-
Turbine bypass used for deep throttling to minimize GG injector throttling and GG control complexity Performance loss with throttling highest of cycles
Easily uprated ifmonopropellant storage is sufficient
Temperature should be as practical (_ 1500 ° F with uncooled turbine and GG) to maximize performance, Cooled turbines and GG
(a)
(a)
(b) Power balance and system pressures strongly affected by temperature
(c)
Temperature may be dictated by coking charac, teristics of certain propel. lants Simultaneous
control of
(a)
(0
(a)
Excess turbine flow in deep.throttled condition and additional injector pressure drop at nominal condition increase turbine flow and performance loss Easily uprated, flexible system
Allcomments for bi. propellant GG system are applicable.
(b)
Temperature may be dictated by coking characteristics of certain propellants
Temperature may be dictated by coking characteristics of certain propellants
Simultaneous control of turbine i(a) flow and oxidizer valves required
(b) Turbine bypass used for deep throttling to minimize gas generator injector throttling and GG control complexity
(a)
(a)
Performance not affected by working-fluid temperature
(b)
Power balance, system pressures, and maximum chamber pressure are strongly affected by temperature. Temperature should be as high as practical to minimize system pressures, maximize chamber pressure
(c) • Temperature may be dictated by coking characteristics of certain propellants
main propellant gas generator valves may be required (b)
Performance not affected by working-fluid temperature
(0
Low PR unchoked turbines provide coupling for possible feed system and combustion instability
Cb)
No performance throttling
(c)
loss with
(d) Very high fuel system AP and fuel pump discharge pressure required for deep throttling
(a)
Limited uprating capability. Limited by heat-transfer rate
(a)
Simultaneous control of main propellant valves required; no bypass Thrust chamber must provide acceptable tapoff gas over wide throttle range for deep throttling ! Performance loss(for throttling capability) at nominal conditions (on the order of 0.5 to 1.0 sec) same as GG cycle, lower in throttled condition (no excess flow)
Easily uprated, flexible system
(a)
Control of five valves required
(b)
Low.PR unchoked turbines provide coupling for possible feed system and combustion instability
(c)
Very high preburner injector AP or complex injector design required for deep throttling
(d)
No performance
(a)
Limited uprating capability. Cycle generally operated near maximum chamber pressure for maximum per. formance
loss with throttling
s000
40oo
_'
2oo0 _
Gas-genez'ator
cycle
i
i
i
t
I
500
1000
1500
2000
2500
Chamber pressure,
Figure
1 1. -- Effect
of drive
discharge ratio
preburner
provides
for this efficiency, Allowing injector,
the resultant
and
psia
chamber
pressure
(oxygen/hydrogen
engine,
on pump mixture
= 6).
high-energy
cycle because, the required for pressure
cycle
pressure
fluid
for the turbine;
3000
psi is an approximate
required
discharge
pressures
can be as high as 7000
and the
tapoff cycles, turbine pressure ratios of approximately flowrate of turbine working fluid and thereby maximize
impulse.
For
the
engine
illustrated cycle. For pressure Relative
expander
weight
by comparing these engines,
pump
discharge
ratios
and
staged-combustion
cycles,
the
to 8000
psia.
20 are required to the engine specific
optimum
turbine
pressure
cycles,
(sec.
2.1.1.7).
Some
of the
differences
between
these
cycles
are
the J-2 engine, a GG cycle, and the RL10 engine, an expander the chamber pressures are 787 and 400 psia, respectively; the ratios pressure
are 19 and
to GG
for expander result from efficiency
limit
generally less than 1.5, because of the large quantities of turbine working fluid In .design, expander and staged-combustion cycle turbine pressure ratios are in order to maximize chamber pressure and minimize turbopump weight, thereby
minimizing
of fuel
upper
for given values of turbine inlet temperature and turbopump pump discharge pressure rises steeply at higher values (fig. 11). drops in the lines, regenerative jacket, preburner, turbine, and
For GG minimize ratios are available. minimized
3000
to chamber
pressure
are
1.6 and
2.5; and the overall
turbine
1.4. turbopump
weight
and
horsepower
generally
are
somewhat
greater
cycles and are, much greater for staged-combustion cycles; these differences differences in the pump discharge pressure requirements. To meet the high requirements,
pumps
in expander
and
25
staged-combustion
cycles
must
either
operate at higher speedor havemore stages-thanthosein GG cycles.To meet the turbine pressure-ratio requirements, GG and tapoff cycles generally incorporate two-row velocity-compounded turbines, and expander and staged-combustioncycles generally incorporateeither two-stagepressure-compounded or reaction turbines.
2.1.1.5 In some
THROTTLING future
RANGE
applications
such
auxiliary propulsion systems, As a result, the turbopumps (fig.
as main
engines
for reusable
vehicles
and accumulator-fed
the engine thrust may be varied (throttled) during the mission. will be required to operate over a range of head and flow values
12).
Throttling
range
==2 120
/ 100 G
/
80 o
60
o
40
"r.
20
0
20
I
I
I
I
40
60
80
100
Flovtate
Figure
12.--Headrise
fztction,
and
flowrate
26
q/qde$,pet,
limits
to
cent
pump
throttling.
120
Pump throttleability unstable if the slope Therefore, zero-slope
is limited, because of the constant speed
for centrifugal pumps, which point is generally considered
a rocket-engine/pump combination can become lines on the pump H-Q map are positive (ref. 11). generally have discontinuity-free the stability limit (fig. 12). For
speed lines, axial pumps,
the the
stability limit occurs when the head drops abruptly or, in other words, when the pump stalls. Relative to the engine, the throttling limit occurs when the engine operating line, which is primarily linear, crosses the pump stability line. Consequently, pump throttleability improves
as the
stability-limit
flowrate
line moves
fraction
at the
stability
to the left relative
For centrifugal pumps, the decreases and, for sweptback head coefficient decreases
throttling impellers as stage
limit
to the design
decreases
or, in other
words,
as the
point.
cap_/bility increases as stage design head coefficient with a constant discharge blade angle, stage design specific speed increases (ref. 12). Consequently,
centrifugal pump throttling capability generally increases with and, therefore, with pump design rotational speed. As shown, down to zero flow if the specific speed is above approximately
stage specific speed (fig. 13) stable throttling is possible 2500. However, heating of
the trapped propellant makes operation at complete shutoff inadvisable except for very short durations (< 10 sec.). For axial pumps, rotational speed has little effect on throttleability. As shown in figure 12, centrifugal pumps generally have approximately twice the
throttling
range
of axial
pumps
and
therefore
are primary
candidates
engines.
In el
3.0_
IN
lqs
'
"_-
3400
=0
2.0
--
0
4.1 o_1 U .,'4 ID O U "CI gt
Figure
13. -
I
I
I
0.5
1.0
1.5
Flow
coefficient
of
stage design
Effect
characteristics impeller
fraction,
specific
of centrifugal
blades.
27
_/_
des
speed on throttling pumps
with
sweptback
for throttleable
2.1.1.6 High
EFFICIENCY turbopump
efficiency
(the
product
of
the
pump
and
the
turbine
efficiencies)
is
important to all types of pump-fed liquid rocket engines. It is important for engines with GG or tapoff cycles because engine specific impulse increases with decreasing flowrate of turbine working fluid. For engines with expander or staged-combustion cycles, increasing the turbopump efficiency increases the chamber pressure attainable discharge pressure (fig. 14). This increase in chamber pressure makes the size and weight of the thrust chamber and, therefore, of the engine.
(_mmhel"
Figure
14. -
Effect
In current
engines,
overall
(table I). As shown, the approaching 60 percent. meet this goal. Figure efficiency -tip flow velocity), maximum specific
15,
a plot and
of
stage
coefficient centrifugal value, then speed
efficiency
pressure
turbopump
efficiency
efficiencies
for a centrifugal
efficiency
in general
vs specific pump.
(ratio of fluid axial pump efficiency increases decreases as stage specific ranges
28
discharge
pressure
(staged-combustion
development pumps and
_Itl
at maximum
on pump
requirements
SSME now under High-specific-speed
pump
design
pressure
of turbopump
vs chamber
for a given pump it possible to reduce
speed,
cycle).
range
from
35
to 48 percent
will have turbopump reaction turbines are
illustrates
For a given
value
the
relation
efficiencies required to
of inducer
between design
inlet
velocity at the inlet to blade tangential with increasing speed until it reaches a speed increases (refs. 2, 12, and 13). Stage,
from
1300
at q_itl
= 0.05
to 2500
at
_Itl
=
9O 0.2
!.°
.IS
|
_o
_-
60
/
40
I
30
cn
,
300
I ,,,,I 5_
I lO00
Stage
Figure
15.
-
Effect
0.20. Consequently, maximum efficiency increase
(impeller
-
2000
specific
of centrifugal
efficiency
,
10
I,
000
speed
pump tip
SO00
'
stage
specific
diameter>
10
speed
on
pump
in.).
if a pump has a design stage specific speed that is less than that at the and has a constant number of stages, an increase in design speed will
efficiency.
With axial pumps, the efficiency will decrease 2000. At these low specific speeds, the blade rotor diameter, and consequently the blade height. Above a stage
as the stage specific speed is decreased heights become very small in relation
the blade tip clearance becomes specific speed of 2000, axial-pump
below to the
a significant fraction of stage efficiencies remain
at high levels up to specific speeds of 10 000 (refs. 12 and 13) (fig. 15 for _itl > 0.20). Consequently, if an axial pump has a stage specific speed less than 2000 and has a constant number of stages, the efficiency increases with increasing design rotational speed. Above 2000,
changes
in design
For high-flowrate efficiency, because
speed
do not have
high-horsepower turbines, the turbine can be full
given values is primarily
of working a. function
are designed
with
For
requiring
engines
rotational
rotational admission
fluid inlet temperature of pitchline velocity.
pitchline
velocities
turbines
effect
on axial-pump
speed has little with a reasonable
effect blade
efficiency. on turbine height. For
and turbine pressure ratio, turbine efficiency For this reason, many rocket engine turbines
close to the rotor
of low
much
horsepower,
because they have such small blade heights that of the blade height. In this case, partial-admission
29
stress
limit.
full-admission
turbines
are inefficient,
the tip clearances are a significant fraction turbines are used (refs. 14, 15, and 16).
For
example,
a partial-admission'turbine
The efficiency increases with
is used
of this kind of turbine decreasing diameter, and for More
low-thrust-level engines, detailed information on
2.1.1.7
AND
SIZE
Light
weight
payload.
is an important
Size is minimized
turbopump so that
and
YLR81-BA-11
turbine turbine
design
handling
the
engine
with arc of admission, in turn decreases with
speed. Therefore, rotational speed. 2.1.2.6.
WEIGHT
on
increases diameter
efficiency efficiency
requirement mounting
(table
III).
arc of admission increasing rotation
increases with design is contained in section
because
of the direct
are simplified.
Weight
effect
on
and size are
directly related to the turbopump design speed. An increase in the design speed decreases the diameters of both the pump and the turbine, because the allowable tip speeds, which are set by stress-limiting
factors
and
the discharge-pressure
requirements
of each
component,
essentially constant for a given application. To maintain proper geometrical the turbopump length also will decrease. Consequently, both the turbopump the turbopump weight decrease with increased rotational speed. Turbine turbines added added
are
relationships, envelope and
efficiency can be increased by using a large number of stages. However, multistage are heavier and longer because of the added rotors and nozzles and because the rotating mass must supporting structure.
2.1.1.8
be supported
by
an outboard
bearing
that,
in turn,
must
have
CONDITIONING
If a pump
for cryogenic
propellants
is not properly
flash into vapor on entering the pump, and consequence is that the turbopump bearings be in danger some way requirements
of failure. of
Therefore,
chilled
prior
to start,
the pump will become will not be lubricated
applications
with
cryogenic
the propellants
vapor locked. or cooled and
propellants
will
Another thus will
require
either
prechilling the pump or a pump surface that chills very rapidly. also exist for restart applications with cryogenic propellants because,
These during
the engine shutdown period, heat soaks back from the hot turbine to the cold pump and, if the shutdown time is long enough, the pump can reach temperatures as high as 0 ° to 100 ° F. The development of upper stages employing cryogenic propellants (e.g., the Saturn S-II stage, the Saturn S-IVB stage, and the Centaur stage) has provided solutions to the problem of turbopump conditioning. The solutions were particularly important for the Saturn S-IVB stage (J-2 restart. In the loops
engine)
Saturn for
and
S-IVB
turbopump
for
stage,
some
both
chilldown.
Centaur
the
oxygen
Electrically
stages
(RL10
engine)
and the hydrogen driven
3O
secondary
because
feed systems pumps
force
the
engines
include propellants
must
coolant from
the tanks through the inlet lines and the pumps. Return lines connectedfrom the pump dischargesthen permit the heated propellants to return to the tanks. This system has performed satisfactorily. However,propellant can be consumedduring chilldown (because the returned heated propellant can causeadditional tank venting), the restartsare by no meansinstantaneous,and the addedcomplexity of the secondarypumps,lines,andvalvesis undesirable. The loss of chilldown propellant and the chilldown time can be reduced by coating the surfaceswetted by pumped fluid with an insulating material and, with hydrogen, by increasingthe two-phasepumping capability (ref. 17). Analytical and experimentalstudies indicate that a thin layer of low conductivity material applied to a metal surfacewill make possiblea rapid surfacechill and will reduce the heat rejection rate from the mainbody of the metal (ref. 17).Two-phasepumpingcapability canbe improvedby designingthe inducer with a high ratio of fluid incidenceangleto blade angle,minimum blade blockage,and a large inlet annulusarea.Boost pumps,having low tip speedsand large inlet diameters,can further increasethe vapor pumpingcapacity(sec.2.2.1.1). The chilldown problem can be made less severeby minimizing the rate of heat soakback (ref. 17). To this end, the number and size of the contact points betweenthe turbine and the pump are minimized, and the remainingcontact points areinsulated(sec.2.2.4.6); in addition, the turbine massrelative to the pump massis minimized by using single-stage turbines rather than multistage turbines and by using smallerdiametersand lower turbine inlet temperature.However,reducingthe diameterandthe temperaturewill adverselyaffect performance. Gearedand single-shaftturbopumpsaresuitedto restartapplicationsbecause,in both cases, one turbine rejects heat to two pumps, thereby producing lower pump temperatures.In addition, for single-shaftturbopumps, one pumpinsulatesthe turbine from the other pump and, for gearedturbopumps, the gearbox insulates the turbine from both pumps and providesan additionalheat sink.
2.1.1.9 The
LIFE,
prime
RELIABILITY,
objective
in the
AND
COST
optimization
and
design
of a new
rocket
engine
system
is to
minimize the cost per pound of payload while meeting the mission reliability requirements. The turbopump plays an important part in this optimization process. The turbopump cost can be reduced by designing for low rotational speed, state-of-the-art components, and the minimum number of pump and turbine stages. The manufacturing, and maintenance costs. In addition,
reduced complexity minimizes assembly, the amount of development required to
attain a given reliability is reduced. Low-speed designs also may have a built-in uprating capability so that redesign is not necessary each time the thrust requirement increases. In
31
general,however,modifications aimed at low cost will increaseturbopump systemweight, decreaseperformance, and can produce a net increasein cost per pound of payload. Theretbre,an optimum turbopump systemexistsfor eachapplication.The designrotational speedat this optimum is a function of the mission,the number of enginesproduced, and other parameterssuchasthe reliability goal andenginereuse: A compromiseon rotational speedwas made for the J-2 hydrogenturbopump (table II). The number of pump stagesand the weight could have been reduced by designingat a higher speed.However,the resultinghigher valuesof bearingDN and sealrubbingvelocity would haverequiredgreaterdevelopmentcoststo meetthe life and reliability requirements. Therefore, the speedwas limited to 27 000 rpm. Another exampleof a designcompromise to obtain greaterreliability is the F-1 turbopump, in which both pumpsaremounted on the sameshaft as the turbine. This arrangementincreasedreliability by eliminating the gearbox. However,relativeto the gearedturbine arrangement(e.g.,the MB-3 andMA-5 turbopumps), the fuel pump inlet pressurerequirementsarehigherbecause,with the turbine mounted on oneend, the fuel pump inlet could no longerbe axial. In
general,
designing
for
increased
life
will
increase
the
weight.
Design
/ procedures
for
increasing life include designing at relatively low speeds, avoiding resonant frequencies, using low stress levels to avoid fatigue limits, using low turbine temperature, and using low values of inducer inlet incidence angle and inducer tip speed to reduce cavitation erosion. Information on designing to obtain the presented in references 1 through 10.
2.1.2
In the
Selection
selection
turbopump
of System
of system
components
type, and
the various
system.
Headrise
NUMBER and
life
in turbopump
system
components
is
Type
system
requirements
Optimization
requirements are conflicting. Even after and design judgment must be exercised.
2.1.2.1
desired
are used
is often
necessary
many
decisions
optimization,
to select
because
the best
some
of the
are still not obvious,
OF UNITS
flowrate
requirements
can be met
more turbopumps, each delivering a fraction to be considered, then both R&D (research considered in the selection of turbopump
either
by a single
turbopump
or by two
or
of the total flowrate. If more than one pump is and development) and production costs must be size. Research and development costs increase
with size because of the increased hardware costs, increased propellant costs, and higher test facility costs. Production costs for an individual unit will also increase with size; however, for a constant than one pump
flow, two pumps, each pumping that is pumping all the flow.
32
one-half
the
flow,
usually
will cost
more
Researchand developmentcostsandunit production costsare determinedasa function of size. A production learning curve is used to obtain averageproduction costs for various numbers of production units. Typically, an equivalent93-percentlearning curveis used(a 93-percentlearning curve reducesthe averagecost per unit by 7 percent for eachdoubling of the number of units). Total production costsarethen determinedfor constanttotal flow for various size turbopumps by multiplying the averageunit cost by the number of units required to give the total flow. Total costsareobtainedby addingthe R&D andproduction costs.Sizecan then be selectedon the basisof minimum costs. For low production rates,total costswill be minimal at smallpump sizes,becausethe R&D costswill be predominantfor this type of requirement.As the production rate increasesand production costs become the predominant cost factor, the minimum costswill occur at largerand largerpump sizes.
2.1.2.2 The
TURBOPUMP
EQUIVALENT-WEIGHT
best-performing
with
given
turbopump
values
of
is the
thrust
level,
one
FACTOR that
mission
results
velocity
in the heaviest
increment,
gross
payload
for a vehicle
weight,
and
specific
impulse. Because the turbopump weight is part of the stage burnout weight, a reduction in turbopump weight allows an increase in the vehicle payload. In addition, a decrease in turbine flowrate for a GG cycle increases the vehicle payload by increasing the engine overall
specific
specific The
impulse
equivalent
expressed
impulse. lower weight
in pounds
This than
effect
that
factor
of stage
occurs
because
of the thrust EWF
is used
payload
the turbine
chamber to convert
per pound
exhaust
for a GG cycle
has a
exhaust. turbine
per second
flowrate
of turbine
to payload
and
is
flowrate:
(2) (EWF)T
-
v - O(-_s)E
(IS)E .]
F
where (EWF)T
e = turbopump PL = stage
(Is)F. = engine
equivalent
payload, specific
weight
factor,
Ibm impulse,
lbf-sec/lbm
33
lbm/(lbm/sec)
(Is)x2 = turbine exhaustspecificimpulse,lbf-sec/lbm F = enginethrust, lbf 0PL O(Is)E
= f (missionvelocity increment,systemgrossweight,systemspecific impulse,andstagepropellant fraction)
For GG cycles,EWF rangesfrom5 lbm/(lbm/sec)fOr boosterenginesto 200 lbm/(lbm/sec) for upper-stageengines(ref. 18). For staged-combustioncycles,(Is)w2 = (Is)E, andtherefore the correspondingEWF is essentiallyzero. The total effect of the turbopump on stagepayloadis determinedby addingthe turbopump weight to the product of EWF and the turbine weight flowrate.The sumis the turbopump equivalentweight: (EW)TI' = WTv +
(EWF)T
p(WT )
(3)
where (EW) T v = turbopump
This
W T p
_-
turbopump
WT
=
turbine
equivalent
weight
equivalent weight,
weight
can
weight,
Ibm
Ibm
flowrate,
be reduced
lbm/sec
by
decreasing
either
the
turbopump
weight
or the
turbine turbine
flowrate. Turbine inlet temperature,
flowrate, in turn, is decreased by increasing turbopump efficiency, and turbine pressure ratio. For the staged-combustion cycle,
turbine
inlet
and turbopump
temperature
efficiency
have
a direct
effect
upon
engine
weight
and delivered payload. High turbine and pump efficiencies will reduce the required pump discharge pressure and reduce the engine system weight. High turbine inlet temperature (within the capabilities of the turbine materials) also will decrease the required pump discharge pressure and reduce engine weight.
2.1.2.3
ROTATIONAL
SPEED
Design speed has more influence on turbopump parameter. It influences all the turbopump design selection of the turbine, the pump, for speed that is the best compromise complex decision facing a turbopump
system design than any other single requirements and in addition affects the
and the turbopump configuration. among all the important factors system designer.
34
Selecting is perhaps
the value the most
To select the speed,its effect on eachof the designrequirementsmust be known. These various effects can then be weighed, and the decisioncan be made.The effects must be consideredcarefully because,if the speedis too low, the weight will be too high, and the turbopump efficiency and consequentlythe enginespecificimpulse will be too low. As a result, the vehiclemight not meet the payloadrequirements.On the other hand,if the speed is too high, the resultingturbopump may not meet the reliability and life requirements.This failure would result in a costly and time-consumingdevelopmenteffort to modify the hardwareand then provethat the modification waseffective. In the process of increasing rotational speed to improve vehicle performance, several mechanicaland hydrodynamic speedlimitations may be encountered(sec.2.2.1). Figures8 and 9 illustrate these limits for a given propellant as a function of pump flowrate and headrise.As shown in figure 16, the turbopump configuration must be modified in order to exceedeach of thesespeedlimits and achievesuccessfuloperation at a higher speed.For example,if the speedis limited to 13000 rpm, the turbopump will not require any special components,will weigh 800 lbm, and will have a pump efficiency of 67 percent.If the speed is increasedto 27 000 rpm, the weight decreasesto 500 lbm and the efficiency increasesto the maximum value of 77 percent. However,this designrequiresa preinducer and an outboard turbine bearingthat, in turn, requiresanadditional sealandprovisionsfor cooling; it might be better to designat 18000 rpm and settle for 600 lbm and74 percent. On the other hand, if weight is critical, it may be desirableto designat speedsabovethe maximum-efficiencyvalue of 27 000 rpm. The upper limit is the speedat which the benefit of the decreasingweight is balancedby the penalty Of the decreasingefficiency or, in other words,it is the speedat the optimum payload(sec.2.1.2.2). In some cases,a turbopump with a different type of component is also a potential candidate.For example, a two-stagepump for the application on figure 16 would have a different stage specific speed and, therefore, from figure 15, would have an efficiency characteristicdifferent from that in figure 16.Consequently,a thorough speedoptimization often requiresanalysisof severalturbopump configurations,eachwith its own pump type and turbine type.
2.1.2.4 The
TURBOPUIVlP
term
turbines
"turbopump for both
the
ARRANGEMIENT arrangement" fuel
refers
and the oxidizer
to the (fig.
physical
17). The
relation
arrangement
of the pumps has a strong
and
the
effect
on
vehicle payload because it influences the weight and the speeds (and, consequently, the efficiencies) at which the various components can operate. However, it is rather difficult to generalize on the best arrangement because, as shown in figure 17; there are many options. In addition, and current
the selection is influenced technology. For example,
by a number of factors, including prior experience a gear drive might have been necessary 10 years ago
35
Speed
limiting
_/'/_//////_/'_/_/
r////////Pre_'n'ducer Pump
= centrifugal
Stages
= I
Flowrate
= 4200
Headrlse
= II
////i
factors
Outboard
t.urbin_
_e:a_i_n_
i////_,
components I
required Special
}
stress Turbine limits
///High- PR////
gpm )0 ft
100
O
80-
U .,4 U
60-
/
40 O
20-
1000
M ,-4
\
80O
d ,z= O_ .e4 ¢)
600
400 0
200
0
1
I
1
10
20
30 Design
Figure
16. -
Effects
of rotational
a LOX
turbopump.
40 rotational
I
I
50
60
speed,
speed on required
36
70
80xl 03
rpm
configuration
for
_
Single
(a) Pumps
back
Dual
Geared
shaft
to back
(b)
Pancake
(c)
shaft
Turbines
in
series
_3
(e (d) Turbine
between
.ne_
pumps
SYMBOLS
--VTurbine
Pump (f)
Single
geared
(g) Turbines
pump
Gears
Figure 17. - Schematics
of basic turbopump
arrangements.
in parallel
because of a lack technology would
of technology for dealing permit an entirely different
If the propellant 17(a) and 17(d)) only one turbine.
densities are similar offers the advantages Within this category,
as in the F-1 engine In some
cases,
the
a disproportionately
large
between
arrangement
diameter
the same shaft and the turbine (fig. 17(e)). This arrangement
permits
optimum
(fig.
in a turbine
In this event,
17(d))
more
advanced
that
as in the A-7 engine.
has very
the two pumps
short
blades
or
may be placed
on
height and diameter. The MA-5 booster (fig. 4) engine utilize turbopumps of this type. The
(fig. 5) engine and the YLR81-BA-11 (fig. 17(b)) in which each pump is geared
configuration components.
today
on a separate shaft that is geared to the pump shaft the turbine to operate at a higher rotational speed,
thereby alleviating the problems of blade and sustainer engines and the YLR-91-AJ-7 YLR87-AJ-7 arrangement
the pumps
results
or both.
placed permits
whereas
(e.g., LOX/RP-1),the single-shaft arrangement (figs. of less complexity and less weight because it requires the turbine may be mounted on one end (fig. 17(a))
(fig. 6) or mounted single-shaft
with cavitation, selection.
rotational
speeds
engine separately for
each
utilize the "pancake" to a single turbine. This of
the
three
turbopump
For small hydrogen-fueled upper-stage engines, the single geared turbopump (fig. 17(f)) can be used to avoid the efficiency penalty of having too small an arc of admission in the oxidizer In
the
turbine.
The
dual-shaft
optimum particularly
speed true
RL10
engine
arrangement and, for
utilizes
(figs.
this arrangement.
17(c)
therefore, the hydrogen/oxygen
and
17(g)),
each
pump
can
overall pump efficiency will engines where the propellant
be designed be higher. densities
at its This is are very
different. Within this dual-shaft category are two turbine drive arrangements: series turbines (fig. 17(c)), and parallel turbines (fig. 17(g)). Series turbines permit the initial turbopump to rotate faster and reduce the turbine flowrate requirements by permitting a larger overall turbine
pressure
The dual-shaft such as starting, the problem development.
ratio.
The
J-2 and J-2S engines
utilize
this arrangement.
parallel-turbine arrangement has great flexibility throttling, and making mixture-ratio excursions; of separating the liquid This arrangement is also
propellants and efficient because
for off-design operation in addition, it eliminates
simplifies advances
turbopump in bearing,
system seal, and
cavitation technology in combination with increases in headrise requirements have made it possible to run the pump at the same speed as the turbine. The experimental X-8 engine (ref. 19) utilized parallel turbines, and this arrangement is used in the SSME now under development.
38
2.1.2.5
PUMP CONFIGURATION
A rocket lightweight
engine and
pump receives propellant delivers the propellant
at low pressure at high pressure
so that propellant so that it can
tanks can be flow into the
high-pressure combustion chamber. Pumps are divided into two basic categories: [1) positive-displacement pumps, in which fluid is forced into a high-pressure region by reducing the volume of a chamber that is momentarily sealed off from a low-pressure region, and (2) nonpositive-displacement
pumps,
and diffusing the kinetic nonpositive-displacement generally the axial
in which
the
fluid
pumps rocket
is raised
by alternately
energy. As shown in figure 18, the design specific pumps (axial, centrifugal, Barske, Tesla, Pitot, and
higher than those for positive-displacement and centrifugal pumps having the highest
rocket engine (eq. (1))for
pressure
have been either engines are high.
adding speeds drag)*
for are
pumps (Rootes, vane, and piston)*, values. Up to the present time, almost all
axial or centrifugal The high speeds
because the pump specific speeds are necessary because of the high
I00 _Inducer
(_-
0.3)
8O
g
O
/
,7-,
/
/
,,.o.,,
4C
2O
Tesla
l
I ] til[ll
2
S
I 10
i i llZJtl
20
SO
I I00
Stage
---o
I i Itlilt
200
SO0
specific
speed,
i 1000
I I litlli
2000
5000
I 10
000
20
I I
000
Ns
Figure 18. - Efficiency vs stage specific speed for various types of pumps.
*Pumps these
named pumps
are is given
described in
reference
briefly
in
Appendix
B;
sketches
of
20.
39
the
pumps
appear
on
figure
20.
Detailed
information
on
flowrate requirementsand the high rotational speedsrequired to minimize turbopump weight. Pumpsother than axial or centrifugalbecomemore competitivewhen flowratesare of the order required for enginethrust levelslessthan 5000 lbf. Sincealmostall pump-fed rocket engineshave been larger than this (table I), the pump selection for enginethrust greaterthan 5000 lbf will be discussedfirst. For large engines(F > 5000 lbf), centrifugalpumpshavebeenusedfor all propellantsother than
hydrogen
able
to meet
because the head
they
are simple
requirements
and
flexible
efficiently
and, in relatively
in a single
stage.
dense
However,
propellants,
are
the low density
of
hydrogen requires overall pump headrise so high that multistaging is necessary to obtain a stage specific speed sufficient for high efficiency (fig. 15). For this reason, the RL10 engine uses a two-stage centrifugal hydrogen pump and the J-2 engine uses a seven-stage (plus one inducer
stage)
that, for single-stage three-stage order In
axial
both
the efficiency
preliminary
hydrogen, the best coefficient
Another
reason
for multistaging
hydrogen
pumps
design
of
requirement a single-stage
and
the tip-speed
centrifugal
pump
is for A in
limitation. for
propellants
other
than
the inducer inlet tip flow coefficient _bit 1 must be selected such that it provides compromise between efficiency and NPSH requirements. Increasing the flow increases the efficiency (fig. 15) but decreases the suction specific speed (sec.
2.2.1.1) and therefore the added complexity operation (above NPSH penalties,
increases the NPSH requirement. This problem of a preinducer (boost pump) can be tolerated.
can be eliminated if To permit high-speed
the peak-efficiency specific speed, fig. 15) without suffering efficiency or a double-entry impeller can be used. Impellers in liquid-oxygen pumps
usually are shrouded avoid rubbing against For and
pump.
pump discharge pressures greater than 2000 to 2500 psi, the tip speeds pumps would exceed the allowable stress limits for existing materials. pump has been designed for the SSME high-pressure-hydrogen application
to meet
the
hydrogen
in order to obtain the housing.
hydrogen, the preliminary the number of stages and
high efficiency
design involves selecting resolving the axial versus
with the large
the inducer centrifugal
clearances
required
to
inlet flow coefficient question. For a pump
specific speed (based on pump head rather than stage head) of 1000, the effects of these variables on engine throttling ratio and pump efficiency are illustrated in figure 19. If throttling to low thrust levels (high throttling ratio) is required, a centrifugal pump is the primary candidate because axials cannot be throttled much beyond 2 to 1. As discussed above, the more efficient high-flow-coefficient designs require more NPSH. The use of inducers with low inlet flow coefficient has less effect on the efficiency of axial pumps than on that axial
of centrifugal
pumps
Another
have more
factor
approximate 2000 ft/sec
that
pumps
because
than
one stage.
influences
the inducer
hydrogen-pump
tip-speed limits are 2800 ft/sec for shrouded titanium centrifugal
flow coefficient
design
is the
affects
impeller
only one stage
tip-speed
limit.
and
The
for unshrouded titanium centrifugal impellers, impellers (1700 to 2300 ft/sec, depending on
40
1oo o_ .
J _
(engine line between design point and origin
_ I __/
_Itl O.1S
0 IS O*
_
IOI_
assumed
to
be
!.
linear)
"
0.05
Axial
--_
C_triNgal i
4_
1
3
5
7
9
3
tl Number of stages inducer stage for
(includes
S
7
9
U
axial)
Figure 19. - Estimated effect of pump configuration on efficiency and engine throttling for a pump specific speed of 1000.
design
specific
attachment,
speed,
hole
in the
amount
of
center,
etc.),
blade and
sweepback, 1500
These tip-speed limits and parametric analyses the basis for hydrogen-pump selection. For (approximately
7000
psi in liquid
ft/sec
blade for titanium
height, axial
method rotors
of (sec.
shroud 2.2.2.3).
similar to those illustrated in figure 19 form example, assume a headrise of 200 000 ft
hydrogen),
an overall
pump
specific
speed
of 1000
to
correspond to figure 19 (which means a constant rotational speed set by one of the rotational speed limits), a pump inlet flow coefficient of 0.15, and a centrifugal pump to meet a wide throttling range (> 2); then 1-, 2-, and 3-stage pumps would have tip speeds of 3300, 2300, and 1900 ft/sec and efficiencies of 78, 84, and 86 percent, respectively. A comparison with the impeller because the tip speed would impeller. three-stage
limits indicates that a single-stage pump would be impossible far exceed the tip-speed limit for an unshrouded titanium
The tradeoff then becomes one shrouded pump, The unshrouded
simple crossover inlet. However,
tubes to carry the it would be heavier
between a two-stage unshrouded pump and a two-stage machine would have fewer parts and
flow from the first-stage discharge to the second-stage due to the larger diameter and to the heavier housing
required to minimize housing deflections that would cause excessive leakage losses at the impeller blade tip; reference 21 shows the effect of blade tip clearance on the efficiency of an unshrouded centrifugal pump. The shrouded three-stage machine would be more efficient and easier to balance for axial thrust. However, the three stages would make a nose-to-back arrangement between the probably capabilities.
necessary discharge
would
be
Analytical
to avoid sealing problems and to avoid extremely of one stage and the inlet of the next. In this based
on
studies
engine-system for
the
SSME
41
performance high-pressure
analyses
complex case, the and
liquid-hydrogen
ducting selection
fabrication pump
demonstratedthat the three-stageconfiguration had an efficiency advantageof 6 to 7 percentagepoints over the two-stage,becausethe pump specific speedswereabout half of thoseshown in figure 19.This madethe selectionof a three-stagemoreobviousthan that in the aboveillustrative examplewherethe advantagewasonly 2 percentagepoints. For low-flow applications(F < 5000 lbf), the pump clearances,tolerances,andflow-passage heights become smaller, and manufacturing becomesmore difficult. To stay within manufacturinglimits, the rotational speedmust be held essentiallyconstant as flowrate is decreasedbelow a given minimum level. As a result, the pump design specific speeds decreaseand low-specific-speeddesignssuch as Barske, Pitot, vane, and piston pumps become more attractive. The regionsof application for some of these low-specific-speed candidatesare shownin figure 18. Also, to obtain reasonablebladeheightsandefficiencies for the turbines, partial-admissionor geared turbines or both becomenecessaryat low flowrates. These turbopump configurations are lessefficient than higher flowrate designs, andconsequentlythe turbine flowrate penalty is.greaterfor smallengines. The subjectof low-flowrate limits for pump-fedsystemsis well exploredfor rotating pumps (ref. 13). The pumping action of the rotor and the lossesin the pump flow path are interrelated with the similarity parametersNs (defined in equation (1)) and Ds, specific diameter,definedby DH¼
(4)
D s -
where Ds. = sPecific D = impeller
diameter,
fts/4/gpm
diameter,
1/2
ft
Thus, the pump efficiency becomes a unique function of Ns and D_ ; its maximum value for given N_ and D_ can be determined by differentiating the equations for the interrelations between losses and geometry and solving for the geometry that yields the minimum losses for
given
operating
conditions
(Ns
D_ values).
This
procedure
is bes_t performed
by
high-speed digital computers (ref. 13). The resulting data can then be presented in the form of Ns - D_ diagrams (fig. 20) in which lines of constant efficiency and lines of constant geometrical parameters (representing the pump geometry) are plotted as functions of specific speed and specific diameter. The validity of the calculated data depends only upon the accuracy of the functions used to interrelate losses with pump geometry. The data presented for high Ns values at the right-hand side of figure 20 show that axial pumps attain highest Dhl/Dtl
efficiencies at high specific speeds and that decreasing values are desired for increasing specific speeds. Centrifugal
medium-specific-speed
regime
efficiently;
increasing
42
values
of the
rotor
of the hub ratio pumps cover diameter
ratio
u = the 6-
(_[
"_aJ) sduand
_o spu!_
sno!JeA Jo_ uJeJ6e!p
s(] -- sN -- "OZ aJn6!-4
sN 000
OOZ
000
001
I
000
OZ
I
000
0I
I
000Z
I
O001
OOZ
I
I
OOI
I
OZ
I
OI
I
tO'O I_n_=_uoD
Dtl/Dt2 are partial-emission
desired with pumps and
pump type, the Tesla pump, speed regime. In single-stage and is directly interrelated application The
of rotating
wear-ring
are scaled are likely
pumps
centrifugal
is indicated
by the dashed
can become
TURBINE
The
turbine
the
fluid
(open
the latter
regime, Another
offers a fair efficiency potential in the medium-to-low specific rotating pumps, the head coefficient has certain limiting v_tlues with specific speed and specific diameter. The limit for the
and tip clearances
Tesla (although
pressure. drive the
speeds. In the low-specific-speed the highest efficiency potential.
or shrouded
two
types
line in figure
disproportionately
down in size. Thus, losses resulting from to be large for small pumps. Pump
conventional
2.1.2.6
decreasing specific Pitot pumps offer
high as conventional
wear-ring and blade-tip types that share this
impeller
do not have
20.
or partial
emission),
the wear-ring-clearance
pumps
clearance problem Barske,
leakage are the drag,
and
problem).
CONFIGURATION
receives
working
to mechanical
fluid
at high
energy,
The mechanical energy pump. The turbine must
and
temperature
exhausts
the
and pressure, spent
fluid
is delivered to the turbopump remove the maximum amount
converts
at lower
the energy
in
temperature
and
shaft where it is used to Of energy from each pound
of working fluid; thus for GG and tapoff drive cycles the turbine flowrate is a minimum, and for topping and expander drive cycles the turbine pressure ratio is a minimum. At the same
time,
the
turbine
must
not
impose
an unacceptable
weight
penalty
on the
turbopump
system. As shown being
in table
two-stage
III, nearly
all rocket
engine
pressure-compounded
turbines
for
have had
low-energy
two axial
propellants
rotors, (fuels
the designs other
hydrogen), two-row velocity-compounded for high-energy propellants (hydrogen GG and tapoff cycles, and two-stage reaction for expander and staged-combustion Exceptions to these rules are the A-7 and the F-1 turbines, which have two-row compounded low. Another an expander Turbine
turbines
propellants
because
the turbine
pitchline
is a function
of the
turbine
pitchline velocity U to the isentropic shown in figure 21. Note that these
because include
they are for subsonic relative clearance losses. Since radial
turbine
U/Co's
are
generally
21(a)) for rocket engine same U/Co, it is heavier.
less than
applications. Therefore,
type,
the
number
(fig.
22),
axial
Another problem with rocket engines have used
44
are
turbine turbine.
and the ratio
in
of
spouting velocity Co. These design-point curves are for illustrative purposes only,
Mach numbers (Mach turbines are difficult 0.4
of stages,
fuel) in cycles. velocity
velocities
exception is the RL10, which has a two-stage pressure,compounded cycle because the turbine pitchline velocity is too low for a reaction
efficiency
the turbine trends are
for low-energy
than
number losses = 0) and do not to multistage and since rocket turbines
are more
efficient
the radial turbine is that, axial turbines exclusively.
(fig. at the
100 50%
reaction
50%
reaction
80 y
Ires sure compounded
60
(zero
reaction)
40
20
/_
t m'bine
//_
0
I
,
I
'
I
,
I_
I
,
I
,
I
,
I
[-,(a)
-j o
Single
stage
(b)
.°.F
100
50%
Two
stage
(two
row
for
VC)
reaction
/_ction Pressure
60
_
/y_
Velocity
.
compounded
/ v.,ooit f
/!
0
0.2
(C)
Three
0.4
stage
0.6
(three
row
compounded
,
l 0.2
,
(d)
Four
stage
0.8
for
Turbine
VC)
velocity
ratio,
I 0.4
J
(four
,
I 0.6
row
for
I 0.8
VC)
U/C °
Figure 21. --Typical design point efficiencies for 1-, 2-, 3-, and 4-stage turbines (supersonic relative Mach number).
To maximize the energy available per pound GG and tapoff cycles, pressure ratio generally
of flow, the turbine inlet temperature are maximized. Since this procedure
and, for produces
a high turbine isentropic spouting velocity Co and since a high U/Co is required to obtain a high efficiency (fig. 21), turbine pitchline velocities are generally maximized. However, because of centrifugal stress, these velocities are limited to a range of 1500 to 1800 ft/sec. Additional limitations result from the fact that, to obtain proper proportions for the turbopump, the pitch diameter is limited to approximately diameter and, to avoid excessive leakage losses, the approximately 0.15 in. Two-stage (two-row for velocity compounded) turbines offer higher efficiency than single-stage turbines (cf.
45
three times the pump impeller lower limit on blade height is
generally figs. 21(a)
are selected and 21(b))
because without
they the
1.6
1.4
1.2
O
.5 r-,
0.8
°_
0.6
U,
ft/sec
0.4 1500
1000
I O2/RP-I
1000 02/H 1500
2I_______
o
lo I
Turb£ne
4;
I
pressure
ratio,
(Po)T1/PT2
Figure 22. - Typical turbine velocity ratios as a function of pressure ratio (inlet temperature of 1500 ° F).
added
complexity
and
weight
of turbines
with
three
or more
stages
(three
compounded). As shown in figure 21(b), two-row velocity-compounded as efficient or nearly as efficient as two-stage pressure-compounded ratio is less than 0.2. In addition, velocity-compounded turbines
rows
because they have a turning vane, rather than a nozzle, between the rotors. nearly zero) pressure drop across the turning vane also results in a lower axial velocity,compounded design. Therefore, are used for rocket engines with turbine between 0.2 and 0.34, two-stage pressure
is little
or no pressure
drop
across
The zero (or thrust for the
two-row velocity-compounded turbines generally velocity ratios less than 0.2. For velocity ratios compounded (impulse staging) turbines generally
are used; they are more efficient up to 0.3 (fig. 21(b)) efficient between 0.3 and 0,34, they have less axial thrust there
for velocity
turbines are either turbines if the velocity are shorter and higher
the
rotor.
46
Above
and, than velocity
even though slightly less reaction turbines because ratios
of 0.34,
reaction
TURBINE
HARDWARE
SYMBOL
o
2
row,
compounded
velocity
Pressure Reaction
compounded
Velocity
I
3
2 Number
compounded
of
rotor
4
blade
rows
Figure 23. --Approximate regions of application for various kinds of rocket engine turbines.
turbines generally are used because, these ranges agree reasonably well
as shown, they are more with existing hardware.
efficient. It should
Figure23 be noted
shows that that these
ranges are not efficient than
absolute. For example, for two stages, a 25-percent-reaction turbine is more either a pressure-compounded or a 50-percent-reaction turbine at a velocity
ratio
therefore,
thrust
of
0.31; relatively
less
it would important.
be selected On
the
other
47
if efficiency hand,
were
a two-stage
very
important
and
pressure-compounded
axial
turbine could be selectedat a velocity ratio of 0.36 if axial thrust were relatively more important than efficiency.This overlapis shownby the designdatapoints on figure23. Turbine efficiency can be increasedby using more than two stages.However,as shown in, figure 21, the rate of increasedecreasesas stagesareadded.For example,at a U/Co of 0.3, typical 1-, 2-, 3-, and 4-stageturbines (50-percentreaction) haveefficiencies(for relative Mach numberslessthan 1.0) of 68, 80, 85, and88 percent;respectively.In addition, these multistage (more than two stages)turbines are more complex, are heavier, and require outboard bearingsthat, in turn, require an extra sealand specialprovisionsfor cooling and lubricating. Figure 23 alsoshowsthe approximateregionsof application for theseturbines. An additional factor consideredin turbine designis that, for a given rotational speed,a diameter tradeoff between weight and efficiency exists. In addition, a tradeoff among turbopump weight, turbine pressureratio, and turbine inlet temperatureexistsif the speed is limited by turbine-bladecentrifugalstress.Theseandother tradeoffsarediscussedin more detail in sections2.2.1.4and 2.2.3.1. For low-horsepowerapplications, partial admissionor gearedturbines often are usedto obtain bladesthat are high enough to avoid excessiveclearancelosses.Someearly RL10 turbines were partial admission and the YLR81-BA-11 turbopump uses both partial admission and gearing. Geared turbines also eliminatesome of the controls that are necessaryfor dual-shaftarrangements.This reduction is desirablefor smallengines,because control componentweightsgenerally do not scaledown to smallsizeaswell asturbopumps do. For staged-combustioncycles, reductions in system pressuredrops and interconnection complexity may be achievedby reversingthe turbine so that the high-pressurestageis outside of the low-pressurestage.The hot gas from the precombustion can then enter through an annularaxial inlet and,with the aid of an annular180-degreeelbow downstream of the low-pressurestage,the turbine dischargegascan flow back to the maincombustion chamberthrough anannularaxial discharge. 2.2
DETAIL
2.2.1
2.2.1.1 The pump
DESIGN
AND
Limits to Rotational
INDUCER
inducer impeller
is the
INTEGRATION Speed
CAVITATION low-pressure
on (usually)
pumping
the main
element
turbopump
48
shaft.
located
immediately
Its purpose
upstream
is to permit
of the
the pump
to
operate
at a lower
inlet
NPSH
(sec.
2.1.1.2)
or, for a given
pump
inlet NPSH,
to operate
at
higher speed. If the inducer is operated above its suction specific speed limit, excessive cavitation will occur and the pump will not deliver the design headrise. This limit to inducer suction specific speed encountered in design.
generally
Suction
Ss is a useful
speed,
specific
speed
flowrate,
and
is the
net positive
first
and
suction
(lowest)
limit
significant
design
to turbopump
parameter
rotational
that
relates
speed
rotational
head: NQ v2
ss =
(NPSH)
(5) _
where Ss = suction NPSH Corrected
specific
= net positive suction
specific
speed,
suction speed
with zero inlet hub diameter speed, rotational speed, and made by numerically at the inlet:
rpm-gpm
head,
'_/ft _
ft
S's is the
suction
specific
speed
of a hypothetical
inducer
that operates with the same inlet fluid axial velocity, inlet minimum required NPSH as the test inducer. The correction
increasing
the flowrate
to compensate
for the area blocked
tip is
by the hub
Ss
S's -
(6) (1 - v2) _
where inlet hub v = inlet
diam.
tip diam.
Dh Dt 1
Selection of inducer design is based on the suction performance and efficiency at the design point (figs. 24 and 25, resp.). The inducer design-point suction-performance correlations in figure inducer
24 were obtained by drawing curves designs that were tested in hydrogen,
through oxygen,
the peaks of the test curves for 20 and water; these basic test curves are
shown in figure 26 (adapted from ref. 2). The efficiency correlation in figure 25 was obtained by drawing curves through the experimental best-efficiency points of 20 inducer designs, some of which were the same as those listed in figure 26. The
allowable
suction
at a lower inlet pumps presented this
action
specific
speed
of an inducer
flow coefficient, as shown in figure 24. (Also shown
reduces
the inducer
efficiency
can be increased
by designing
by the correlation of test is that the data peak at q_It (fig.
49
25)
and,
particularly
data
the inducer on high-speed However,
1 "_ 0.05.)
at high
pump
specific
-----------200
correlation deviation from
correlation
X 103 !_,_
Range
Test
-Ji NPSH
[
Ss '
Propellsnt
Cm2/2g "--'___
I
hydr°gen
8160/_Ttl
I
OXY' en
4850/',t I
2(cm212g)
0 .,-I
-
_
-
100 --
0 o1-1 4J
3(Cm2/2g)(.O721/_Itl)
I
u_ _:1
2,
II
50-
U ¢) I-I 0 CJ
I
2O 0.03
I 0.05
Design
I
I
inlet
I
tip
I
I 0.I0
flow
I 0.20
coefficient,
_Itl
Figure 24. -- Influence of inducer design inlet flow coefficient cavitation performance.
on
100 Q) .,.4 fJ .PI q4 q_ U
-_Itl
h£dS-_'_S
(high
)
_
_O@itl
-
0.1
G)
50
Cflat plate> m
0
30 0.03
I
Design
Figure 25. -
I 0.05
I
inlet
I
1
I
I 0.10
tip flow
coefficient,
I 0.20 _Itl
Influence of inducer design inlet flow coefficient inducer hydraulic efficiency.
50
on
4t9
0 co _>
;'.7
E 0
O_
0
¢L
J
°
iI
il iIII
--
--_¢
._-
-o
=.>o
¢_
-
=.
x
•
•
•
.....
r
.............. - _o. _ ._'_"_= .......
N
o_ °_,.,,, _
/
,,
0
//N
. u
I1_ 4.J
_o -0
O_ v
D-
O_ CJ
.c_ O
_-_-_
_
._-
J_':,'TZ/" w._'_
U
E E E bO
I
--"
t'N
P 03 LL
I
1
I _/l_(_a-l)sS
I
I
I = s)S
I _paads
I
I
o!._!oads
I
I
uo!_,ons
1
I
pal:)aJJO::)
"0 cco
CL
speeds, relation.
the overall pump efficiency The original pump inducer,
(fig. which
15). The J-2 hydrogen pump had an inlet diameter of 7.25
illustrates this in. and an inlet
fl0w coefficient of 0.095, did not meet the NPSH requirements. The inducer was then redesigned for an inlet diameter of 7.8 in., the result being a reduction of the inlet flow coefficient to 0.073 and an increase in the suction specific speed capability (fig. 24). This change
decreased
the pump optimized
the
NPSH
requirement
efficiency. Therefore, in terms of turbopump
suction
specific
become
an important
speeds
are often design
(TSH). suction
derated
to reduce
cavitation
limit of suction performance, thereby reducing the vapor
This effect permits the The difference between
effects and is approximately above-mentioned reduced
level,
but
damage,
and
pump to operate the basic NPSH,
equal to the value of NPSH
vaporization pressure and
decreased
therefore
life can
within the inducer increasing the local
satisfactorily at a reduced value of inlet which is independent of thermodynamic
NPSH in water at room temperature, and the is called the thermodynamic suppression head
This parameter varies with propellant type and is the reason performance shown in figure 24; at low inducer tip speeds,
higher suction were correlated
also slightly
factor.
During operation close to the chills the surrounding liquid, NPSH. NPSH.
to an acceptable
for a given application and mission, the configuration, is weight, pump efficiency, and tank weight. For long life,
for the variation TSH would result
in in
specific speeds than those shown. Theoretical relations from reference 22 with data on the J-2 oxidizer pump, J-2 fuel pump, Atlas-sustainer oxidizer
pump and Thor oxidizer pump that include these thermodynamic
to produce the following empirical expressions for NPSH effects (all parameters referred to inducer inlet):
114.7 tc NP-SH m 2/2g
/
_ 0.931
/
_)t 4/9
(Ot/Z)
0"16
/_hydrogen
(7) _t 2
Ut 115
hydrogen
6.35
(Dt/Z)
0"16
/3 other
(8) _NPSH) m2/2g
_ 0.931 other
_t 2 Ut
_t4/9
where cm
=
fluid axial velocity
g = acceleration
at inducer
due to gravity,
q_t = inducer
inlet
tip flow
Dt = inducer
inlet
tip diameter,
inlet, ft/sec 2
coefficient in.
52
ft/sec
1"15
Z = numberof inducerbladesat inlet /3 = thermodynamic ut = inducer
The
first
term
suppression
inlet
on
the
tip speed,
right
side
head
factor
(fig. 27)
ft/sec
in each
of these
equations
(room-temperature water) to the inlet fluid axial velocity entering fluid expressed in feet of head) and is derived from correlation (fig. 24) and the suction specific speed equatiQn each head
equation is the ratio of the thermodynamic of the pumped fluid. The thermodynamic
propellant vapor pressure and is shown NPSH to inlet velocity head is greater ratio and the inlet axial velocity head. two-phase-flow must inlet
Mach
number
possible, principles of NPSH
ratio
of the
basic
suppression head to the inlet axial velocity factor _3is a function of propellant type and
than
(refs. Mach
1.0,
the
inlet
line is choked,
and
23, 24, and 25). If the ratio is less than number is less than 1.0, two-phase
the NPSH can be less than the inlet velocity apply. (These principles are discussed at the end to inlet velocity head is greater than 1, equations
diameter
and
blade
is contained
number
are slight.
in references
1 and
Additional
discussion
the
NPSH
1,0 and the pumping is
head, and the two-phase-flow of this section.) When the ratio (7) and (8) show that the pump
NPSH requirements may be adjusted by changing tip speed and vapor pressure inlet flow coefficient. Because of the small exponent on the (D/Z) term, the head
NPSI_
head (kinetic energy of the the water suction-performance (eq. (5)). The second term in
for various propellants on figure 27. If the ratio of than 1.0, the NPSH is equal to the product of the If the ratio is less than 1.0 but the inlet equilibrium
is greater
equal the inlet velocity head equilibrium two-phase-flow
is the
of thermodynamic
as well as effects of suppression
22.
The inducer speed limit due to cavitation preinducer upstream of the main inducer
can be essentially removed by placing to provide the required inlet pressure
a low-speed to the main
inducer. A preinducer can be driven by several methods, four of which are illustrated schematically in figure 28. The gear drive (fig. 28(a)) is the most positive and efficient but, because of the gear train, may be the most complex and least reliable. The through-flow hydraulic
turbine
drive
(fig.
28(b))
is both
simple
hydraulic turbine drive (fig. 28(c)) can be installed the pump inlet, whereas the gear and the full-flow efficient than the other two drive systems. The installed at a location remote from the pump
and
efficient.
The
recirculated-flow
in the engine at a location remote from drives cannot; however, this drive is less electric drive (fig. 28(d)) can also be inlet. The electric motors often are
propellant-cooled to save weight and reduce the number of rotating seals. This drive is attractive for low-thrust applications. Another configuration that can be installed at a remote location is the gas-turbine drive, the type used for the RL10 engine in the Centaur i!/.
stage.
!!;i!
each is gear driven by its own hydrogen-peroxide gas turbine. The gasrturbine boost pump system is complex and usually requires either a separate gas generator or a catalyst bed, but has the advantage (as used on the Centaur) that it can be started before the main pumps.
The
Centaur
boost
pumps
(fuel
and oxidizer)
t
53
are located
in the propellant
tanks,
and
i00
I
2 gJ ( A hv) = _hydrogen
PL
Pv _other
o
=PL
k_
| (adapted from ref. 22)
_h v (Cp)L
(Thor
data
correlation)
I0
o
,= o .,-_ ul u_
k
HYDROGEN (LH,
u
N204 o
1.0
METHANE (CH
WATER
(H20) (LF 2)
(CH6N 2 OXYGEN
(h02)
I i0
0.i I
Propellant
vapor
Figure 27. --Thermodynamic propellants
i00
pressure,
suppression
as a function
54
Pv' psia
head factor for various
of vapor pressure.
___Preinducer
/-
Gear
drive Hydraulic
Inlet-_
i
__L-_"_ turbine__ _t
____-_ {
///
¢-
!_
_=__j/
___L_.A. 4_I_= (a)
Gear
"°w
shaft
(b)
drive
..... .ydrau_ic
ta
_Discharge
//# l
Inlet flow
-"
_
Throughflow
Electric
motor
hydraulic
turbine
stator-
7
drive
Elec.%ric r°tor r
--
/
Inle&
.,o. =>r v
/__i
II Irr (c)
// _Discharge
//___A_---_
Recirculated-flow
flow
__.____=--=_-_
_ _hydraulic
Figure
L_ turbine
28.
"_I-L
Inducer
(d)
drive
-- Schematics
of
four
methods
for
driving
an
____LS__ v_[J _
Electric
inducer.
motor
drive
motor
Recent developments (refs. propellants can be pumped
23, and,
24, and therefore,
25) have indicated that two-phase (vapor-liquid) that cryogenic propellants that are saturated in
the tank (zero tank NPSH) can be pumped. from figure 29 or from the expression
Vapor-pumping
o_= 1-
1-B
capacity
can
be determined
J
(9)
where o_ = vapor
volume
fraction
at the
(i//3)L = 1 -- arc tan 4_L = ratio (for a pure liquid) _L = flow coefficient B = blockage boundary The
values
for pumping
expansions from for the increased
of incidence
at inducer
fraction layer
inlet,
vapor-volume/mixture-volume
angle
to blade
inlet (for a pure
(normal
capacity
pump
to flow
with
a saturated liquid are used to size size, two-phase inducers are similar
at the inducer
of the inducer
the thermodynamic
tJ
60 4a
c%j
so 40
_'_
o_
20 -
lo -
_o
0 Blade
Figure
":
,0×NNN
4a
29. -
10 plus
Vapor
20
30
40
bound&ry-layer
pumping
capability
56
blade
plus blade
relationships
for
the pump inlets for zero NPSH. Except to other inducers. In oxygen, low-speed
80
f-t
inlet
liquid)
direction)
in combination
angle
50
60
blockage,
70
80
B, percent
for two-phase
inducers.
,
boost pumps accommodate
generally are the extremely
to avoid
choking
2.2.1.2
BEARING
The
second
required in order to obtain the large inlet areas necessary to low acoustic velocity of two-phase oxygen in equilibrium (i.e.,
in the inlet line).
DN
turbopump
speed
bearing. This parameter, (rev/min), is proportional the inner race. If the DN
limit
reached
the product of to the tangential limit is exceeded,
is usually
the
the bearing may fail because of overheating and contact contains the information on bearing capabilities, including engine Lower
applications in which DN values are necessary
in figure
of the
wear and fatigue. Reference 6 the DN limits for typical rocket
to shaft diameter, bearing location.
it becomes a function of everything The most common bearing locations
30.
(a)
Inboard
(c)
Outboard
bearings
(b)
(d)
bearings
Outboard
Outboard inducer
Inducer (e)
Outboard
rolling-contact
the life requirement and the radial and axial loads are low. if long life and high load-carrying capacity are required.
Since bearing DN is proportional affects shaft diameter, including shown
DN limit
bearing bore D (mm) and rotational speed N velocity of the bearing at the inside diameter of lubrication and cooling will be insufficient, and
ptmp
turbine
bearing
bearings
without
_Turbine
bearing l_lmp
B
Bearing
Figure 30. - Schematics of bearingsupport arrangements.
57
that are
The
overhung-turbine
rocket
engine
bearing
turbopumps
arrangement
in which
(fig.
the
pump
30(a))
has been
is mounted
used
almost
on the same
shaft
exclusively
for
as the turbine;
the turbopumps for the J-2, RL10, and F-1 engines (figs. 1, 2, 3, and 6) are examples of this arrangement. This configuration avoids the additional supporting structure and the separate lubrication and sealing systems required for the outboard turbine bearing (figs. 30(b), 3_(c), and 30(d)). For the overhung arrangement, the shaft is often sized so that the turbopump design
speed
is below
the
lowest
simple
shaft
bending.
This
value
sizing
of shaft
critical
is particularly
speed,
important
which if the
in general turbopump
over a wide speed range or, in other words, if the turbopump application The following equation is used to estimate this limiting speed:
is that must
requires
due
to
operate
throttling.
(DN)4 / 3 N -
(10) KDN (HP) 1/3
where DN = bearing HP = shaft
bore
× bearing
horsepower,
KDN = empirically
speed,
mm-
hp
derived
coefficient
for bearings,
Number of pumps on shaft
The final design speed equation (10), because for existing spring rate,
the speed limit pump impeller "educes length, diameter
and, therefore,
Propellant
and and
KDN
t
LH2
325
1
Dense
374
2
Dense
478
because weight
not agree with the value from correlation of the shaft sizes
critical speed is influenced of the components. If this
by factors such as bearing deviation causes difficulty,
such as bearing spring rate, may be adjusted without having to redesign the whole turbopump.
can be obtained by placing (fig. 30(e)) and using the
the overhang increases the
as follows
based on a critical-speed analysis may the equation is based on an empirical
turbopumps shape, size,
some of these factors, critical-speed problem
rpm
the pump bearing between inducer stator to support
(relative to that shown in fig. 30(a)) and, bearing span. Both of these factors would would
increase
the speed
58
limit.
so as to avoid the Some increase in the inducer and the it. This arrangement
for a given turbopump permit a smaller shaft
Figure 8 showsthat the bearing limit for this arrangement(labeled critical speed)differs from the other limits in that turbopump geometricsimilarity (constantspecificspeed)is not maintainedwhen scalingwith flowrate. The reasonfor this is that bearingsandsealsdo not scale as the other turbopump components do. Consequently, turbopumps for low-thrust-level engineshave proportionally longer shafts and more overhangthan db turbopumps for high-thrust-levelenginesand therefore have a more severecritical-speed problem. By placing the turbine bearingoutboard (fig. 30(b)), the shaft diameter can be sizedby torsional stress; since this practice generally produces a smaller shaft diameter, the turbopump therefore can be designedfor a higher speed.The following equationexpresses this speedas a function of the torsional stressthat would occur in a solid shaft equalin diameter to the shaft diameter at the bearing(an equivalent-shaftstressof 25 000 psi has producedgood correlationwith final design): N=
1.378×10-
5
Seq 1/2 (DN) 3/2
(11)
(Hp) 1/2
where Seq = torsional stress.in a solid shaft at the bearing, psi
of the same
outside
diameter
as the pump
shaft
Even higher speeds may be obtained by moving the bearing on the pump end to a location between the inducer and the impeller (fig. 30(c)). This placement minimizes the required bearing
diameter
by
minimizing,
the
torque
transmitted
through
the
boost pump, a main pump inducer may not be needed, because provide sufficient NPSH for the main pump impeller. In this case, placed The
outboard arrangement
options
pads that act leakage losses,
expensive and contamination
bore.
With
a
would can be
(fig. 30(d)). in figure
30
have
thus
far
rolling-contact bearings for nearly all applications. become restrictive, hydrostatic bearings (refs. 26 pressurized minimize
bearing
the boost pump the pump bearing
against these
permitted
sufficient
If rolling-contact and 27), which
latitude
bearings function
to use
eventually by having
the shaft to keep it centered, can be used. ttowever, bearings have very close clearances and therefore
are subject to the various and differences in thermal
problems growth
to are
associated among the
with close clearances (e.g., shaft, the bearing, and the
speed limit that generally occurs at approximately DN limit is the seal rubbing speed. Seal rubbing
the same rotational speed speed is the speed at which
housing).
2.2.1.3
SEAL
A turbopump as the bearing
RUBBING
SPEED
59
the rotating mating ring on the shaft rubs againstthe stationary sealnosepiece.If the seal speedlimit is exceeded,the nosepiecewill weardown too rapidly or the sealmay fail due to overheating.The sealspeedlimits in commonpropellantsmay be found in reference7. The turbopump rotational speedat the sealspeedlimit dependson the shaftdiameter.For a turbopump with an overhung turbine (fig. 30(a)) and a shaft sizedon critical speed,the following equationis often usedto estimatethe limiting rotational speed: Kss(SS)4/3
N=
(12)
(HP)1/ 3
where SS = seal rubbing Kss = empirically
speed, derived
ft/sec coefficient
for seals,
Number of pumps on shaft
as follows
Propellant
SS
LH2
240
Dense
208
Dense
164
As with equation (10), equation (12) is based on an empirical correlation of shaft diameters for various turbopump configurations that have overhung turbines and it does not scale along a constant specific speed line (fig. 8). The seal speed limit, like the bearing DN limit, can be alleviated by utilizing a turbopump configuration with outboard turbine bearings (fig. 30(b)) instead of an overhung-turbine configuration. The limit then is expressed by
N = 4.37
Seq 1/2 (as) 3/2
(13)
(HP) _!2
As before, Seq is the (equivalent) usually 25 000 psi.
torsional
If the turbopump is to be designed configuration in figure 30(b), special
stress
in a solid
shaft
of the same
diameter
and is
at a rotational speed in excess of that allowed by the seals are required. The seal function can be broken into
6O
two parts: static, anddynamic.The liftoff seal,which sealsduring static conditionsandlifts off when pressurizedduring operation, can perform the static function. The dynamic or operatingfunction canthen be performedby a secondsealthat is in serieswith the first and that doesnot haveasseverea speedlimitation asthe rubbing type of seal.This arrangement is being used in the SSME high-pressurehydrogen turbopump in which the operating function is performedby a stepped-labyrinthseal. A type of sealthat performs both functions is the hydrodynamic face seal,which is being developedfor the SSMEhigh-pressureoxygen turbopump. During static conditions,sealing is accomplishedconventionally by having the spring-loadednosepressagainstthe mating ring on the shaft. However,during operation, groovesin another part of the noseassembly trap pump fluid andforce the noseawayfrom the matingring. As a result, the noserideson a film of liquid duringoperation andthereforeis not subjectto the rubbing-sealspeedlimit. More information on the various kinds of seals,including the hydrostatic seal,canbe found in references7, 26, and 28.
2.2.1.4
TURBINE-BLADE
Turbine-blade centrifugal
CENTRIFUGAL
centrifugal force
on the
stress
refers
blades
during
to the
STRESS stress
turbine
at the
operation.
turbine-blade This
stress
roots
caused
is proportional
by the to the
square of the turbine rotational speed times the turbine annulus area (N 2 Aa). When this stress exceeds the stress limit of the turbine blade material, the turbine blades will pull off the turbine rotor. The N 2 Aa limits for candidate materials as a function of temperature may
be found
in reference
4.
The previously mentioned methods for alleviating allow the turbopumps for most propellants to maximize payload. However, for liquid-hydrogen
component speed limits are sufficient operate at rotational speeds needed turbopumps, the rotational speed
to to for
maximum payload generally exceeds the turbine-blade centrifugal-stress limit. Because there is no simple way of alleviating this limit, hydrogen turbopumps usually are designed close to the turbine stress limit; this practice generally results in a speed less than the maximum payload value. The J-2 hydrogen turbopump, cavitation limits, is an example of such a design.
which
also
operates
To optimize the turbopump design at the centrifugal-stress limit, tradeoffs among turbine inlet temperature, turbine pressure ratio, blade materials, turbopump blade materials, turbopump weight, and speed are required. For example, raising the turbine inlet temperature turbine required flowrate but will also decrease the allowable rotational the allowable N 2 Aa. This speed decrease, in turn, will increase the weight.
In
addition,
it will
tend
to decrease
the
61
efficiency
close
to the
inducer
some rather involved turbine type, turbine turbopump rotational will decrease the speed by decreasing turbopump size and
of centrifugal
pumps
and
to
increasethe number of stagesfor axial pumps.The sametype of optimization is necessary to determinethe optimum turbine pressureratio (fig. 31) because,asshown,the maximum speedand consequentlythe minimum turbopump weightboth occur at a low pressureratio, where the dischargeannulus areaAa is minimum, whereasthe minimum turbine flowrate occursat a much higher pressureratio and consequentlyat a larger dischargeannulusarea (assuminga constantinlet pressure).
(T°)TI
o2/n2l
-
Propellants '
2000°R
l
hp - 1o ooo
I
°OO00[
i_
g z5
40 000
30
000
;
2°°°I
i
4a
°
10
S
.2
i0000[
.o
O" 0
• i0
I 30
i 20
I 40
_
0
|
0
(Po)T1/PT2
Figure
31.
-
Typical limit
•
30
40
(Po)T1/PT2
effects of
|
20
I0
of
turbine
turbine-blade
pressure
centrifugal
ratio
on
design
at the
stress.
q
Cooling
the
turbine
blades
speeds above the stress determine the optimum cooling Turbine
with
the propellant
limit for uncooled amount of cooling
being
pumped
blades. A payload and to determine
to allow an increase in turbine cooling is common in jet-engine
inlet temperature turbines and may
turbopumps. One problem with turbine outside of the blade and cold cryogenic
cooling is thermal propellant on the
stress resulting from the severe thermal and therefore can adversely affect life.
gradient
can cause
In some cases when multiple restart is required, even if the blade is not cooled. The temperature blade blades
and with
the hot
cold gas
center ducted
of the to the
blade hollow
makes
may
62
optimization the proper
to operate
at
is necessary to tradeoff between
and cooling to raise the speed. be used in future rocket engine stress inside.
caused by hot gas on the For repeated rapid starts,
cracking
due
cracking due to thermal gradient between the
be sufficient
center
it possible
to cause
can alleviate
the
to thermal
fatigue
fatigue can occur hot surface of the problem.
this problem.
An
Hollow additional
benefit of hollow bladesis that the blade thicknesscanbe taperedto reducethe masswith radial distance from the base without affecting the outer contour of the blade. This techniquecanpermit a higherlimit on centrifugalstresswithout penalizingperformance.
2.2.1.5
GEAR
PITCHLINE
VELOCITY
For geared turbopumps, the limit limits results in an upper limit turbopump.
The
pitch diameter. rubbing velocity, at the
contact
For a given be increased
gear
on gear pitchline velocity in combination on rotational speed for the high-speed
pitchline
velocity
is the
If the upper limit on po0r lubrication (high
points),
and heat
tangential
can cause
rapid
2.2.2
There
on limits
limit, the rotational the limit on tooth
speed at that point is the of the turbopump. Reference
to gear pitchline
teeth
at the
velocity
and gear-tooth
speed bending
can or
upper limit on 5 contains the
stress.
Pump Design
are
several
performance, inducer inlet the maximum
2.2.2.1
pump
design
decreasing stability, flow coefficient, the pump
INDUCER
impeller
INLET
generally
To meet the requirements have helical blades with
mounted
Qit i (ratio
of
flow coefficient pump efficiency
on the
fluid
turbopump
axial velocity
decreases inducer (fig. 19). However,
flow
until
performance of 0.05 are
decreases. The extremely difficult to manufacture
inducer
design
oxidizer
and fuel pumps
cannot
of liquid very gradual
performance
inlet
that
be
exceeded
without
penalizing
are the minimum pump stages, and
FLOW COEFFICIENT
impeller. inducers coefficient
are
limits
or risking failure. Among these limits stability limits for axial and centrifugal
tip speeds.
Inducers
design overall
gear
a combination of high the amount of lubricant
wear or failure.
horsepower transmitted at this pitchline-velocity by decreasing the gear pitch diameter until
information
of the
this velocity is exceeded, centrifugal forces decrease
compressive stress is reached. The rotational rotational speed for the high-speed component design
velocity
with tooth stress component of the
coefficient
flow
value
coefficients
have
inducer
shaft
upstream
of the
to blade
tangential
velocity).
efficiency (fig. 25), which, a decreasing flow coefficient
of 0.05
first-stage
pump
rocket engines for low pump inlet pressure, curvature and low values for the inlet flow
is reached
(fig. 24); below
Decreasing
the
in turn, decreases increases suction that,
the suction
thin blades required to operate at a flow coefficient and are often inadequate structurally. Consequently, are inlet
seldom
less
flow coefficients
63
than
0.07.
of 0.11
For
example,
and 0.073,
the
respectively.
J-2
2.2.2.2
STABI LITY
If an axial-pump stage is operated at a flow coefficient, the pump will stall, thereby causing performance to become poor and unpredictable. the
original
problem
existed
with
problem, addition
it was allowed
necessary to add a seventh stage identical to the original the pump to deliver the same flow at a lower rotational
resulting margin.
higher
operating
Stall may be predicted 29). In the J-2 engine
flow
J-2
coefficient much less than the design flow the developed head to drop (fig. 12) and the In addition, damage can occur. This .stall
hydrogen
coefficient
pump,
unloaded
which
had
the blades,
six stages.
thereby
To solve stages. speed,
increasing
centrifugal
pumps,
operation
in the positive-slope
region
of the head/flow
(figs. 12 and 13) has been known to cause instability during operation Although the effect is not nearly as drastic as it is in axial pumps, that region avoided. Therefore, centrifugal pumps generally are designed such that the coefficient
2.2.2.3
during
This The
the stall
by using a blade loading parameter called the diffusion factor (ref. program, it was found that axial-pump stall occurred when the blade
diffusion factor at the mean flow-passage diameter exceeded 0.70. Therefore, are designed such that during operation the maximum blade diffusion factor diameter is not larger than this value for either the rotors or the stators. For
the
operation
falls to the right
of the zero-slope
point
axial pumps at the mean
characteristic in the engine. of operation is minimum flow
on the head/flow
curve.
TIP SPEED
For a given stage in a centrifugal pump, and the square of the impeller discharge
headrise is proportional tip speed:
to the stage head
coefficient
J(ut:) 2 H =
(14)
g
where = stage head ut2
----
coefficient
stage impeller
g = acceleration
discharge due
tip speed,
to gravity,
ft/sec
ft/sec 2
Because stage head coefficient for a centrifugal pump generally falls between 0.4 and 0.6, the primary component of stage headrise is tip speed, and consequently high-headrise stages require high tip speeds. However, if the tip speed is too high, the centrifugal forces will exceed
the
material
strength
of the
impeller,
and
64
it will burst.
Therefore,
although
high tip
speedsare often desired to minimize the number of stages,pump impellers are always designedwith an adequatetip-speedmargin betweenthe maximumoperatingvalue andthe burst value. For titanium, which has the highest tip-speed capability of the candidate impeller materials, the approximate tip-speed limits are 2800 ft/sec for unshrouded centrifugalpump impellers,2000 ft/sec for shroudedcentrifugalpump impellers,and 1500 ft/sec for inducers and axial pump rotors. (Thesevaluesare difficult to generalizebecause the blade,sweepback,the designspecificspeed,the blade height, and the method of power transmission- all affect the allowabletip speed.)Multistagingis sometimesrequiredto stay within theselimits. It must be noted that hydrogen pumps generally are the only pumps that approachthe tip-speedlimits becausehydrogen with its low density is the only propellant that has headriserequirementsgreater than the capability of a singlecentrifugal impeller. It must alsobe noted that multistagingis usedto obtain a higherstagespecificspeedandthereforea higher efficiency (fig. 15). As a result, somemultistage hydrogen pumpshavetip speeds much less than the limits. For example, the J-2 and the RL10 (the only flight-proven hydrogen pumps) havetip speedsof approximately 900 and 800 ft/sec, respectively;the impellersweremadeout of "K" monel andaluminum,respectively.
2.2.3
Turbine Design
There
are
several
performance, and turbine tradeoffs
design
limits
that
decreasing stability, or risking blade centrifugal stress limits, to
requirements
2.2.3.1
turbine
obtain
optimum
imposed
PERFORMANCE
Among form
are
conducted,
the interrelated
moment
start
without
penalizing
these limits are the rotor disk the boundaries within which and
the
exhaust
pressure
OPTIMIZATION
effects
of inertia,
exceeded
configuration.
Tip speed is a function of rotor disk diameter and the centrifugal stress on the rotor will exceed the fail. Therefore, turbines are optimized within the best combination of turbine flowrate, turbopump treat
be
failure. which
performance
by the engine
cannot
of tip speed,
time,
inlet
rotational speed. If tip speed is too high, material strength of the rotor, and it will rotor disk stress limitations to obtain the weight, and start time. The optimizations
tip clearance,
temperature,
and
efficiency, material
flowrate,
weight,
strength.These
rotating
interrelations
may be illustrated by considering the effects of an increase in tip speed for a stress-limited candidate with constant rotational speed. In this case, an increase in the tip speed (1) increases
the
and
therefore
the
diagram
turbine the
diameter
turbopump
efficiency
(ref.
and therefore start 4);
time; (4)
the weight;
(3)increases
decreases
65
the
(2) increases the velocity
blade
height,
the moment ratio
U/Co
which,
for
of inertia
and therefore a given
tip
clearance,increasesthe tip leakagelosses;(5) increasesthe materialstresslevel,which makes it necessaryto reduce the inlet temperature to obtain a higher allowable stresS;and (6) causesa net changein turbine flowrate as a result of the changesin diagramefficiency, tip leakagelosses,and inlet temperature.Turbopump equivalentweight (sec.2.1.2.2) is often usedin suchan optimization to convertthe turbine flowrate to equivalentvehicleweight. The turbine blade centrifugal-stresslimit, discussedin section2.2.1.4,generallyappliesonly to hydrogen turbopumps andinvolvesan evenmore complex interrelation becauseit affects rotational speedand therefore alsoaffects the pump design.As with the tip-speedanalysis, optimizations involving turbopump equivalent weight are usedto determine the optimum combination of factors. Another tradeoff is that of weight and efficiency to determine the optimum number of stages.Multiple stagesincreasethe efficiency (fig. 21), but also increasethe weight and require an outboard turbine bearing with its added supporting structure, seals, and lubrication system.Again, turbopump equivalentweight is usedin determiningthe optimum numberof stages.All existing flight turbopumps(table I) haveusedno more than two stages becausethe relatively short flight times havemadeweight relatively more important than turbine efficiency.
2.2.3.2 The
EXHAUST
turbine
PRESSURE
exhaust
pressure
is the
static
pressure
of the
turbine
gas as it leaves
the
last
turbine blade row. For GG cycles, this pressure is low so that a high turbine pressure ratio and therefore a low turbine flowrate are achieved. However, the exhaust pressure must be high enough to permit an efficient and stable exhaust. For example, the flareback of turbine exhaust gases into the boattail of early ballistic missiles occurred because of the subsonic discharge problem. On
engines
turbine turbine. have
of the
turbine
gas. A higher
with
conical
or bell nozzles
gas disposal can affect If the turbine exhaust this
nozzle
"choked"
turbine
that
discharge
employ
the available is discharged
(operated
could
or tapoff
turbine pressure ratio into a separate nozzle,
above
discharge pressure will not be affected flight. This practice will increase the
a GG
pressure
the
critical
pressure
by changes in atmospheric turbine discharge pressure
have
cycle,
avoided
the
this
method
of
and the design of the it may be desirable to so that
turbine
pressure during and affect the
ratio),
vehicle turbine
pressure ratio. If the turbine exhaust is discharged into the main nozzle, as on the J-2 engine, the point at which the turbine exhaust is put into the nozzle will determine the turbine back pressure.
When
the
turbine
exhaust
is used
to cool
the expansion
nozzle
(flowing the turbine exhaust down the coolant passages and discharging expansion at the exit being used to increase performance), the pressure must be considered in establishing the turbine discharge pressure.
66
by "dump
cooling"
the gas at the. exit, loss in this circuit
For staged-combustionand expandercycles (sec. 2.1.1.4), the turbine dischargepressure must be sufficient to permit the flow to passthrough the downstreamducting and the injector into the thrust chamber; therefore, the turbine dischargepressureexceedsthe enginechamberpressureinsteadof being a smallfraction of the chamberpressure,asin the GG cyclel As a result, the turbine pressureratios for expanderandstaged-combustion cycles are minimized to minimize the pump dischargepressurerequirements;however, for _G cycles, which show little effect of turbine pressureratio on pump dischargepressure,the pressureratios aremaximized within the previouslydiscussedlimits. The RL10 engineis an example of the expander cycle arrangement. No operational engines employ the staged-combustioncycle,but this cycle is usedin the SSMEnow under development. 2.2.4
Turbopump
Mechanical
In turbopump mechanical attachment of the various design
objectives.
problems are pointed out. interdependent considered
2.2.4.1 From 30)
In the
integration, the detailed construction, arrangement, and turbopump components are planned to best meet the turbopump following
section,
the
major
options
in solving
the
basic
design
discussed, and the options that in general provide the best solutions are However, because there are many possible requirements and many components, solutions other than the best general one often must be
and used.
BEARING the design
are
PLACEMENT standpoint,
primarily
the turbopump
a function
of the
single-shaft turbopump design, shaft that usually is supported and one aft toward the turbine. is within usually
Integration
limits
and
is overhung
overhung simplify housing. overhung this case
arrangement
if the
of the
the
has been
bearing lubrication and However, if the turbine
has,
in general,
aft bearing
used
almost
minimize has more
turbine generally is too large and also for very-high-speed
housing.
In
and the placement and
turbine
of bearings
components.
For
(fig. the
selecting
the
no
inboard exclusively
the than
more (fig.
than 30(a)).
two
rotors,
At the present
for nongeared
turbines
the
turbine time,
the
in order
to
structural requirements of the turbopump two rotors, the shaft required to support an
to permit the bearing to stay within DN limits. For designs, the aft bearing usually is placed outboard.
This arrangement is more complex because seal between it and the turbine, provisions turbopump
pump
the pump impellers and the turbine disks are mounted on a by two bearing sets, one located forward within the pump If the rotational speeds are low enough that the bearing DN
turbine
by placing
arrangement
sizes
an outboard aft bearing requires an additional for lubrication, and structural support by the
bearing
arrangement,
the
complexity
required
operate at a higher speed often must be weighed against the potential benefits of increased pump efficiency or decreased weight. In short, the added complexity must be justified.
67
to
On the forward end, the bearing for a single-stagepump usually is placedinboard of the pump impeller, as in the J-2 oxidizer turbopump (fig. 1). If the pump has two or more stages,the forward bearinggenerallyis placedbetweenthe first and the secondstagesasin the RL10 fuel pump (fig. 3); this location makespossiblea reduction in the shaftdiameter and consequentlythe bearing DN by reducingthe amount of torque transmitted thrmagh the bearing.If the pump has a separateinducer stage,the bearingoften is placedbetween the inducer and the remainder of the pump, with the inducer statorsactingas a bearing support, asin the J-2 fuel (fig. 2) andthe Mark-29fuel pumps(fig. 32). /
For
cases in which
and
the
turbine
both can
the
be
turbine
designed
as separate
seals. Under these circumstances, and later attached to one another The
turbopump
configuration
and the pump
units,
the pump and by a coupling. may
have
each
the
incorporate
more
than
with
three
its own
impellers,
the pump
sets of bearings
turbine
are developed
as separate
a reduction
gear between
the
and units
turbine
and
the pump, thereby permitting each to be designed at its optimum rotational speed. In general, geared turbopumps are restricted to small sizes in which the turbine speed must be high enough to obtain reasonable blade heights and in which multiple turbines would require
excessive
important
2.2.4.2
control
because
TURBINE
In general,
the
elements
controls
ROTOR
ASSEMBLY
considerations
involved
turbopump depend on whether separate unit. For a single-shaft the turbine considerable combination
in terms
do not scale
of size and
down
weight.
in size as rapidly
AND
The
latter
in the
relation
the turbopump configuration,
of the
turbine
to the
whole or a of
rotor to the main drive shaft, in which case the drive coupling receives attention. This drive coupling (ref. 8) may consist of a curvic coupling in with an involute spline as in the Mark-29 fuel turbopump (fig. 32); a curvic as in the
involute (fig. 2).
spline Here,
alone, or just a bolt and torque-pin joint as in the later the main considerations are normality and concentricity
proper
operation
F-1 (fig.
of the
6), J-2 oxidizer
turbine
rotor
(fig.
relative
MA-5 sustainer turbopump or through a quill bearings as in the turbopumps for H-l, MB-3, some
gear is mounted separation, into
misalignment
between
directly on the turbine the turbine drive shaft.
the
1), and early
to the pump
If the turbine is geared to the pump, the turbine rotor this case, power may be transmitted by a gear attached
allows
rotor
configuration is single shaft, geared, the main consideration is the attachment
alone
shaft
is
ATTACHMENT
coupling
ensure
consideration
as turbopumps.
gear
68
and
torque
turbopumps;
the
turbine
deflections
an
J-2 fuel turbopumps limits in order to
rotor.
is considered to be a separate directly to the turbine shaft
shaft into a power and MA-5 booster
shaft,
J-2 fuel
unit. In as in the
gear mounted on separate (fig. 4). The use of a quill drive
shaft.
If the power
may be introduced
by gear
Cylindrical for thermal
growth
bearing
I_ducer star and forward
Turbine inlet manifold
support
Inducer Curvic couplings
Intermediate drive coupling
Involute spline Ch
Forward bearing
.Centrifugal impeller
Turbine rotors Thrust balance piston
Figure 32. - Mark 29 experimental
LH 2 pump.
tf the speed,
turbine is a separate the drive coupling
unit and is mounted in line with is of major importance. Such
transmitting the torque from the turbine to deflection in regard to alignment and to duration thrust in the turbine and the pump.
the pump. It must also be adaptable to in axial distance that may result from axial
tip speed low, the
Redstone A-7, the turbopump
Centaur RL10 (fig. 3), and MA-5 booster (fig. 4) turbopumps, and therefore shaft can pass through them. However, if the tip speeds are high, the rotor
may
have
to be solid
influences rotor disks
operating at the same must be capable of
Turbine rotor tip speeds are
disks
strongly turbine
the pumps a coupling
at the center
turbine rotor assembly can have holes through
in order
to withstand
and attachment. If the their centers, as in the
the high centrifugal
in the J-2 (figs. 1 and 2) and F-1 (fig. 6) turbopumps. Under these partway out on the disks are required for attachment to the turbopump. Extreme with a
thermal cryogenic
difference rotors range
conditions may exist within a turbopump assembly fluid and the turbine with an extremely hot
between
the
are mounted occurs in the
the centrifugal demand utmost
turbine
and
the
pump
may
be as high
on the same drive shaft as the pump rotor assembly. These severe temperature
forces,
circumstances,
as
bolts
if the pumps operate gas. The temperature
as 2000 ° F. If the
turbine
impeller, the total temperature gradients in combination with
stresses introduced by the high tip speeds of the turbine attention and care during the design of the rotor assembly.
and
pump
rotors
Thermal growth is an important factor in the rotor assembly. Pilots designed
that must be considered in maintaining concentricity to maintain concentricity among the various rotor
components components
if different materials are used for those centrifugal force act together on the pilots.
designed neglected. the not
may not be dependable or if thermal growth and to maintain concentricity may The material of the mounting
properly if thermal considerations are bolts, which fasten the turbine disks to
main drive shaft, is carefully selected so that the difference in thermal expansion result in loss of the clamping torque. Such loss of torque occurred during
development adequately
of the J-2 hydrogen pump. tightened. As a result, the
coupling
teeth
improvements Differential are
not operate (or clamp)
used,
cracked
in growth
because
and
The original bolts were first-row turbine disks
fretted.
Stronger,
tighter
thermal
distribution
within
the
turbine
may
temperature downstream
on the upstream side of a turbine disk may side, deflection of the disk is also a consideration.
to high
a uniform
in
tip speed,
the
disk profile,
rotor which
disks
do not
means
that
70
the clamping between the
ha_ze a hole both
sides
be
cause
will the to be curvic
combination
curvic design and disk material processing eliminated the between turbine disks is also considered even though similar
hotter than the other. In this event, to this differential radial growth
If, due
not strong enough cracked, and the
bolts
considerably themselves
have
various A pilot
one
with problem. materials
disk to operate
bolts are designed to adapt turbine disks. Since the higher
in the center,
are symmetrical
than
that
on
the
it is desirable
to
in coutour.
For
thesecases,the disks are attachedto the shaft by an intermediatedrive coupling,asin the Mark 29 experimental hydrogen turbopump (fig. 32). Since the torque that is to be transmitted from the turbine to the pump usuallyis large,the diameterof the intermediate coupling at which the turbine disks are attachedusually is large. A large diameter also contributes to disk stability and reduces the unit loading. Such an intermediate drive coupling is attachedto the main driveshaft with an involute splineand concentricpilots. Curvic couplings(ref. 8) frequently areused to attach the turbine disks to eachother (fig. 32). This type of coupling providesmaximum torque-carryingcapability andconcentric and normality relationshipsamong the individual components.The curvic coupling is usedin connection with clamping bolts, usually through-bolts usedin combinationwith clamping nuts. These clamping bolts are generally positioned at the mean diameter of the curvic couplingin order to avoiddeflection of the turbine disk dueto clampingforces. If torque-Carryingrequirements are relatively small, a key or pin drive often is used. Frequently, the clampingbolts are modified in sucha way that they are able to carry the torque as well as clamp the disk to the main drive shaft, asin the later J-2 fuel turbopump (fig. 2). Sometimesadditional radial pilots areusedto maintain concentricity betweenthe maindrive shaftand the turbine disk. However,this function canalsobe performedby drive pins; if so, at least three pins are used.Another method for maintainingconcentricity is to cut the involute splineandradial pilots directly on the extendedhub of the turbine disk and attach the disk directly to the main drive shaft. Clamping bolts are usedto attach the turbine disk to the main shaft at a suitablebolt-circle diameter.If the radial stressesof the turbine disk are moderate,a hole in the centerof the ttirbine disk is often used.In this case, the disksare mounted on a shaftthat passesthrough their centersandareeither attachedto each other with clamping bolts or shoulderedagainstthe main shaft and clampedwith a centerclampingnut. Torque pinsoften areusedfor power transmissionin suchanassembly. Two distinctly different turbine rotor designshavebeenemployed: one in which the rotor disks are individual, as in the J-2 (figs. 1 and 2), MA-5 booster(fig. 4), YLR87-AJ-7 (fig. 5), and F-1 (fig. 6) turbopumps; and the other in which they are made from one piece or welded together into one unit, as in the RL10 (fig. 3) and A-7 turbopumps.If the rotor consistsof individual disks, the mounting of the disks to the drive shaft is a step-by-step procedure,Such a configuration is designedso that the diskscannot be mounted onto the drive shaft backwards.Backwardmounting is preventedby asymmetricalpositioningof the bolt holes or by the addition of a so-called"idiot pin" (a locating pin that matchestwo parts in correct orientation). Since turbine disks are usually heavy and very difficult to handlebecauseof their delicatenature, they areusuallyinstalledvertically. This procedure ensuresaferhandling and better control for positioning the disk centerrelativeto the shaft center. In an installation where the disks are mounted directly to the drive shaft of the 'pump and the bearing is inboard, the shaft is already mounted in the bearingsand is therefore positioned.However,if an outboardbearingis used,additional stackuptolerances, concentricity tolerances, runouts, and thermal deflections have to be considered.The
71
possibility the
of backwards or counter-rotating drive shaft usually does not exist
main
has
a
thrust
positioned the rotor
balance
interference
that
TURBOPUMP
For
single-shaft
allows
axial
travel
of the turbine rotor in relation to is made from one unit. If the pump of
the
to the main drive shaft because, may be reduced. This reduction
and consequent
2.2.4.3 a
piston
axially relative and the stator
installation if the rotor
loss of power
shaft,
the
turbine
rotor
is
if neglected, the spacing between would introduce the danger,of
in the turbine.
HOUSING turbopump,
joining
the
pump
housing,
which
is superchilled
for
cryogenics, to the superheated turbine casing constitutes a major design problem. The interface between those two elements must be designed to adapt readily to the temperature differential.
The
differential
turbopumps. Structural thermal growth of the most
existing
large
In the structural configuration on the heated along
shrinkage
turbopumps method,
have used
a relatively
results in a cone smaller side. Theoretically, the
this cone
and,
therefore,
The
lbf thrust)
mechanical
turbopumps
(figs.
in diameter temperature at that
pronounced
in
large
method.
is used
for the turbopump
housing.
on the chilled side and larger should be equal to ambient location
should
remain
This
in diameter somewhere
constant.
It is at
housing is flanged and bolted onto the pump casing. The is the key to solving the problem of differential thermal was used for the Mark 29 fuel (fig. 32) and the uprated F-1
for accommodating
1, 2, and
differential
thermal
to
is used, the turbine hot gas is hotter
that it may freely expand is attached to the manifold
the turbine discharge manifold within the
to this system,
be accommodated strains and deformation
growth
however,
by
the in the
anchored. Many mechanical problems have the pins were attached directly to the hot-gas
When a structural cylinder containing the high-pressure such cone
6). A drawback
turbine has may introduce
the radial pins are brackets containing
interstage
long cylinder
more
is an arrangement
or keys on which the hot turbine casing can slide unrestrained without losing relation to the pump; this method was used in the J-2 and standard F-1
torque of the accommodation
mounted structural
are
are available for ensuring unrestrained method presently is favored, although
turbopumps.
method
of radial pins its concentric
growth
the mechanical
the diameter
this point that the cylindrical ambient-temperature interface growth. This structural method ( 1.8 million
and
and mechanical methods turbine casing. The structural
side of the structural outside turbine casing, this structural
72
the
full reactive
arisen when the mounting manifold or torus.
manifold is attached to it; since the manifold than the surrounding shell, the manifold is
relative on one
stators.
is that
radial pins or keys; this mounting brackets in which
to the cylinder. In this mounting, end and, on its larger-diameter end, cylinder. In addition to suspending cone may also be used to support
a to the the
The structural cylinder has been usedalso to alleviatethermal-distortion problems. For example, the rear bearing support on liquid-hydrogen pumps is generally subjected to liquid-hydrogen temperaturesnear the bearingand to high temperaturesnearthe turbine mount. On one of the early versionsof the J-2 hydrogenpump, gradualdistortion of the bearingsupport occurred.The distortion wasdue to large,repeatedthermalstressesddring eachfiring that resulted from the radial temperaturegradientthrough the bearingsupport. The solution was to changeto a strongermaterial. However,as previously discussed,the solution for a similar designproblem on a later pump, the Mark-29 hydrogen,is a more fundamentalone. Here, the bearingsupport and turbine mount function areseparatedinto two assemblies.The turbine mount is arrangedin the form of a thin cylindrical part. The temperaturegradient:is then essentiallyaxial, andthermalstressesarevery low sincegrowth is unrestrained. If the turbine is a separateunit with bearingson both sidesof the rotor, the turbine rotor is fixed relative to the turbine casingby the shaftbearingmounts, while the manifold with the gasdischargenozzle is flexible relative to the turbine casing.The problem is not so much misalignmentbetween the manifold and the rotor, but the relative positions of the two bearingsupports,which are more or lessindependentmembersbecausethey arepositioned in housingsthat are connected to each other by the casing.Hot gasshould not contact either the casingor the bearingsupport structure because,if it does,the resulting thermal deflectionswill produce bearingmisalignment.The passageof the hot-gasdischargethrough the bearingsupport often is avoidedby using a dischargemanifold that exhauststhrough a hole in the casing.If the hot-gasdischargepassesthrough the bearingsupport, the bearing support struts are shieldedby havingthem passthrough vanesthat contact the hot gas.The spacebetweenthe vanesand the struts may be filled with insulation.Insulation is alsoused betweenthe hot-gaslnanifold andthe casing. For large multistage turbines in which some thermal deflection is unavoidable,a linkage arrangementfrequently connectsthe casingto the bearingcarrier at the turbine discharge. This linkage keepsthe bearingcenteredrelative to the casing,regardlessof the amount of thermalgrowth in the casing. Material cross sectionsare heId to a minimum wherever high temperature is expected. Flangeswith bolt holes or threadedholes are avoidedin hot-gas regions,becausethose flangesare usually of heavycrosssectionandaresubjectto thermal deflection andcracking. The reason for this is that the core of the heavy section will not heat as rapidly as the surfacematerial and therefore high thermal stresseswill be introduced. In addition, hot-gas flanges have to be sealed, a requirement that is often difficult to satisfy becauseof differential growth anddistortion of the sealsand the flanges.
73
2.2.4.4
BEARINGS
If the turbine
AND SEALS
and pump
are in a single-shaft
bearing considerations for arrangement shown in figure propellant
is tapped
off
at
a
and
the pump the pump
the
turbine
and the turbine
high-pressure
propellant passes through both venting to either the pump inlet cooling method is used for both bearing
unit
the turbine are similar 30(a) with propellant-cooled region
to
bearing
is inboard,
seal and
those for the pump. For the bearings, a portion of the puml_
such
as
the
pump
discharge.
This
bearings into a low-pressure region that is obtained by or a low-pressure region in the impeller flow passage. This J-2 turbopumps (figs. 1 and 2). A seal between the turbine
minimizes
leakage
from
the
pump
to the
turbine.
The
pressure
on
side exceeds the pressure on the turbine side so that turbine gas cannot leak into and affect the bearing cooling. For an oxidizer turbopump, such as the J-2 oxygen
pump (fig. 1), fuel-rich turbine hydrogen
pump
propellant
back
purge Seals generally are used to prevent any oxidizer leakage into the gas. If the support for the pump bearing is the inducer stator, as on the J-2 (fig. 2), the
If an outboard
static
pressure
through
the bearing.
turbine
bearing
(fig.
rise across
30(b))
is used,
the
stator
can be used
lubrication
may
to recirculate
be a problem.
pumping fluid is hypergolic, then a separate lubricant flow has to be source. A separate turbine unit does not have an internal source for Therefore, lubricant and cooling flow have to be introduced into the removed. Bearing races on the turbine shaft or on the main drive shaft
supplied bearing bearing usually
If the
by another lubrication. cavities and are installed
with a slight interference fit in order to prevent looseness and possible sliding of the bearing race on the beari_ag journal Of the drive shaft. Such sliding generates heat and contributes to an early selected;
failure of the bearing. If axial and radial loads are severe, a split-race bearing often in this design, additional balls can be inserted into the bearing, thereby making
is it
capable of absorbing higher loads. Selecting the proper material for sealing the lubricant flow from the hot gas in the turbine is critical, because a sealing material that is compatible with one lubricant may not be compatible with another. The cavity pressure of the lubricant flow
and
operate
the
and
DN radial
in the
this step
balance,
beating
axial
pressure
effectively;
on pressure The
cavity
turbine
are
is particularly
important
bearings
limit
depends
loads.
have
have no axial-load To
prevent
hot
bearing and shaft-tiding assembly
If
on operating
the
turbine
approximately
capacity turbine
gas from
rings
three
gnd are often
the hot-gas chambers. seal and the other
of carbon
established
when
so that
a dynamic
the
seal, which
seal
will
operates
is used. conditions
is part
made
entering
such
as temperature,
of a single-shaft
bearing may be a roller bearing. If the turbine likely to be ball bearings because they have Roller
carefully
lubricant,
configuration,
the
and
turbine
times
is a separate assembly, the bearings are most to absorb axial loads as well as radial loads.
subject
the
radial-load
the
to end wear bearing
cavity,
capacity and skewing
of ball bearings (ref.
a seal is installed
6). between
This seal arrangement usually consists of two seals, a face-contact type of seal. The shaft-tiding seal in segments
around
74
the
periphery
and
bound
but
together
the one a is an by
small coil springs.It acts like a labyrinth sealin that it causesa pressuredrop in the turbine gas.That portion of the gasthat may leak through the sealis collectedin the cavity between the two sealsand drained. The second seal,which is located close to the bearing, is a dynamic seal that necessitatesa mating ring. This sealis activated statically by built-in springs that press the seal ring against the mating ring. During operation, a carefully calculatedpressurebalanceensuresconstant contact of the seal-ringface with the mating ring. Sincethe sealring is alwaysin contact with the matingring, the peripheralvelocity is limited. This rubbing-velocitylimit is a function of the pressuredrop acrossthe sealandthe propertiesof the bearinglubricant. In some turbine configurations, the bearingand sealhousingsare alsopart of the turbine casing.In others, particularly in largeconfigurations,a flexible diaphragmis attachedto the turbine manifold and bolted to the bearinghousing.In that case,the bearinghousingis a separatecomponent.If shaft whirl or critical speedbecomesa problem, it canbe alleviated by changingthe bearingpreloadso asto changethe bearingspringrate.
2.2.4.5 For
AXIAL
sm_ill
However,
THRUST
turbopumps, for
the
high
BALANCE axial
thrust
bearing
DN's
may
be
in liquid
controlled rocket
by
antifriction
turbopumps,
the
bearings
alone.
axial-load-carrying
capacity of the bearings generally is too low to carry all the load without some other design provisions. For example, the H-l, the MA-5 booster (fig. 4), and the MA-5 sustainer turbopumps have ribs on the backface of the centrifugal-pump impellers. These ribs cause the propellant to spin with the impeller, thereby decreasing the pressure with radial distance from the impeller tips. The effect is to reduce the net axial thrust on the rotating assembly. For centrifugal pumps in which the axial load is too great to be controlled by balance ribs, it can be controlled by putting a shoulder on the impeller backface that mates with a labyrinth seal (fig.
1). The
cavity
surrounded
the turbopump, thereby reducing reducing the axial thrust. This sensitive solution.
than
balance
by that
the pressure acting wear-ring arrangement
ribs to impeller
Open-faced and shrouded acting against the housing,
seal is vented
axial position
centrifugal pump impellers and thus centrifugal pump
on
to a lower
pressure
region
the impeller backface has the advantage of
and therefore
produce shrouding
is generally
within
and being
thus less
the preferred
very different pressure forces strongly affects axial thrust.
In general, shrouded impellers have more predictable pressure distributions and are less likely to produce forces that will cause the turbopump shaft to flex: However, pumps with unshrouded impellers have been designed to accommodate the different forces and have been operated successfully. In high-speed, high-pressure pumps, the high bearing DN's reduce the bearing capacity at the same time that the high pressures increase the axial loads. If either
75
axial-load the thrust
variation betweenoperating points or a reasonablemargin for error in thrust prediction exceedsthe bearingaxial-loadcapacity,the bearingsaredesignedto movefreely in the axial direction, and a balancepiston (fig. 33) is usedto centerthe rotating assembly.The balance piston accomplishesthis by varying its axial force with axial position. With a series-type balancepiston as shown in figure 33, if movementis to the right, a high-pressureorifice (in
Highpressure seal Transducer Stator
assy
Support
assy
Low-pressure seal _ssure
orifice Spring assy
High-pressure orifice
Balance
piston
Rotor
assy
Labyrinth
Figure
Turbine seal
33.
-
end
Balance
piston
bearing
assembly.
hydrogen, rub rings) opens, and a low-pressure orifice closes such that the pressure acting against the piston increases and provides a restoring force in the opposite direction. If movement is in the other direction (to the left in fig. 33), the high-pressure orifice closes. and the direction position. balance
low-pressure decrease,
The J-2 hydrogen pistons, and current
-for balance The rub exhibited occasional
orifice opens such that the pressure and thereby permitting the net force to return (fig. 2) and Mark 29 hydrogen designs for the SSME turbopumps
the force the rotor
acting to the
in that neutral
(fig. 32) turbopumps utilize now under development call
pistons. rings low
of the friction
impact
J-2 hydrogen balance system originally were carbon, a material that and associated low wear. However, rotordynamic transients caused
against
the
carbon
rings
and, on a few occasions,
76
the carbon
cracked.
This
failure led to a materials testing program in liquid nitrogen. From it, (1) leaded bronze (BeariumB-10)ringswith an Inconel 718 piston and(2) glass-filledTeflon (Armalon) with a titanium piston wereselected.Thesetwo combinationshave.workedwell. Balancepistons are often made integral with the pump impellersso as to minimize axlal length and therefore weight. Also, by acting on fluid that is in an impeller clearancespace, the integral arrangementdoes not require an additional supply of fluid and, therefore, reducesleakagelosses.The integral balancepiston is sensitive to housing and impeller deflectionsand therefore requiresanalysisof thesedeflectionsduring design.To avoid the complication of deflection analysis,a separatebalancepiston may be mounted elsewhereon the turbopump shaft. However, as previously discussed,this arrangementincreasesboth axial length andleakagelosses. Ball-bearingsystemsfor turbopumps haveundergonemuch development.If the bearing DN's are not too high, the bearingaxial springsareshimmedin sucha way that the bearing loads are controlled within acceptablelimits as long as the rotor axial motion doesnot exceedthe "free motion" betweenthe rub rings. This practiceallowsthe rub ringsto carry large transient loads during turbopump starts and stops without bearing overload (for periods of 1 or 2 sec).On the other hand, if significant rub-ring wear occurs,the bearings pick up a larger shareof the load andprovide someoverloadcapacity.Duplex bearingpairs, which are free to float axially, arid rubbingstops,which carry the excessaxial loads,canbe usedto eliminateall axial,loadson the bearings. In pumps with a single spring-loadedangular-contactbearing at eachend in combination with a balance-piston system, a serious problem arises in designing and setting the axial-spring loads. Differential thermal shrinkage of rotor and stator assemblies,axial shrinkageof the rotor at high speedfrom Poisson'seffect, anddifferential axial growth from pressureeffects - all enter into the axial bearingloads under variousoperatingconditions. This combination of effects has often led to operation of unloaded ball bearings,an undesirablecondition for high-speedoperation. One solution to this problem that has worked well on two models of LH 2 pumps is to use duplex bearings at each end, the bearings
being
loaded
against
each
other
(1) the inlet differential-growth
bearing set problems
can be mentioned
The effect increased.
spring
can
2.2,4;6
on
THERMAL
rates
be
through
allowed above,
turbopump
springs.
a great
to float axially, and (2) the radial advantage
if shaft
Two
advantages
thus spring critical
bypassing rates are speeds
result: all the doubled.
need
to be
BARRIERS
When a hot-gas turbine is used in combination from the hot turbine to the cryogenic pump After
Belleville
shutoff,
the pump
temperature
77
with a cryogenic pump, thermal soakback becomes a problem if restarts are required. rises rapidly
if the turbine
is not thermally
isolated from the pump. This temperaturerise causesthe cryogenicfluid in the pump to evaporate,thereby making the restart of the pump very difficult. Consequently,specialcare is taken to prevent thermal paths betweenthe turbine and the pump. Various mechanical meansare availableto accomplishthis. Couplingshavebeenusedbetweenthe turbine and the pump rotor; e.g., an involute spline together with a quill shaft or, if centrifugal stress permits, a coupling from a dissimlarmaterial suchasplastic.Another useful couplingis the ball joint; this joint has the advantagethat, assoon as torque transmissionceases,the balls sink back into their sockets,thereby disengagingthemselvesfrom the pump drive coupling and essentiallydisengagingthe two shafts.Cooling the turbine bearingwith the cryogenic fluid is often helpful in preventingtemperaturecreepbackinto the pump through the drive shaft. The housingsmay be attachedto eachother with an arrangementof radial pins that act as thermal insulators; an insulating material may be clampedbetween the turbine and the pump housing;or a thermal barrier may be added.The thermalbarrier may consistof a manifolding devicethrough which cryogenicfluid is circulatedafter turbine shutoff so that the fluid may absorbsomeof the heat. This fluid may then be carried in the form of gas back to the cryogenictank andusedfor tank pressurization.
2.2.4.7
ASSEMBLY
To avoid
contamination
engine turbopumps manufacturers have the
clean
meet
room
critical
Several
that
is temperature
assembly
other
the turbopump the assembly checked and
wear rates
controlled
to permit
and possible
environment. In addition
verification
explosions,
rocket
For this purpose, most to being dust controlled,
of critical
dimensions
example,
seals
and
to
dimensions.
requirements
are also met
during
assembly.
For
positioned
in
are checked to verify that leakage will not exceed the amount specified in specification. Rotor axial and radial clearances and the bearing drag are verified. Since every component is subjected to considerable vibration, all
bolts, nuts, and other however, is never used could
can lead to excessive
are assembled in a dust-controlled an area designated a "clean room."
do considerable
fasteners are secured within the turbopump damage;
lockwire
with locking devices or lockwire. because pieces of wire that failed
usually
is used
externally
or in areas
Lockwire, in fatigue
where
there
is
no possibility that broken pieces can escape. As a safety measure,.locking devices are never used twice. After assembly, a torque check is performed on the rotor to ensure that there is no interference between the rotor and the housing and that the drag on the seals and bearings Some
is within rotor
Matched maintained.
the specification.
assemblies
assemblies For
are often
example,
particularly in a multiple-stage of the rotor after assembly components is the best
is difficult answer.
matched, are
and
necessary
close
individual if extremely
clearances
between
components
cannot
accurate the
be interchanged.
relationships
rotor
tip
rotor assembly, frequently require grinding to avoid tip clearance losses. Interchangeability
to achieve
in such
a case, and therefore
78
have
and
the matched
the
to
be
housing,
the tip diameter of individual assembly
often
The selection Of
materials
for
the
turbopump
as well as operational requirements. some period of operation, a process but
not
with
others.
Some
For that
materials
components
must
consider
fabrication
needs
example, a turbine may require weld repair after may be easily accomplished with some materials
(e.g.,
Hastelloy)
do not
require
annealing
after
weld
repair, whereas others (e.g., Ren6 41) must be annealed if weld repair has taken place bn a finished machined component. The annealing process, however, may produce deflection in such a component and therefore may make its reusability questionable. The selection of tolerances may either ease the higher rate of rejection. Often,
manufacturing process or make it more difficult by causing a the use of magnetic material eases the manufacturing process
by enabling for machining
to use magnetic
the manufacturer those parts.
If the turbopump
is designed
for long service
of considerable importance. Periodic procedure should be possible without components reassembly. selected assembly
are Easy
chucks
and
other
life and overhaul
inspection difficult
capability,
holding
devices
ease of assembly
of some components often disassembly and reassembly.
indexed against each other so that mistakes access without complicated tooling is provided
so that they are are made different
magnetic
cannot be to all points.
not interchangeable. For example, fasteners in thread size in order to prevent installation
is
is necessary; this Therefore, rotor made during Fasteners are
within a given of a given bolt in
an area that may require a longer or shorter bolt. Components often have to be separated by force as a result of thermal deflections that have taken place during high-temperature operation. Provisions are made for extracting those components without damanging them, usually by adding threaded the disassembly sequence
holes to the clamping and used as extricators.
flanges so that bolts can be inserted during Interference tolerances often are selected
for components that must maintain close relation to each other. In such a case, the assembly process calls for chilling down one component or heating up the other before assembly, so that they are easily joined. Provisions for disassembly of such joints are also necessary.
2.2.5
The
System
interfaces
between
vehicle and engine the fluid entering contraction
Interfaces
Interfaces desired
turbopump
performance the inducer
can cause
(3) maintenance costs serviced after use.
2.2.5.1
the
system
and
the
engine
and vehicle
systems
affect
and maintenance costs. For example, (1) the flow pattern of affects pump suction performance, (2) thermal expansion and
mechanical can become
connections
to fail if the connections
high if the
turbopump
system
is not
are not flexible, designed
and
to be easily
PUMP INLET
between fluid
flow
the
propellant
pattern
at the
feed
line and
pump
inlet.
79
the The
pump flow
inlet
are designed
pattern
entering
to produce a pump
the
strongly
affects the pump suction performance, and a poor flow pattern can causeexcessive cavitation. Therefore, bends and changesin inlet ducting crosssection are minimized. If elbowstoo closeto the inlet (approximately 15 to 20 pump-inlet diametersupstream)are required, turning vanesare usually used.If there is sufficient NPSH,tangentialpump inlets often are usedto minimize ducting weight and complexity. Straighteningvanesat the talak exits are employed to minimize swirl of the pump inlet flow (ref. 1) and to prevent fluid vortexing in the tank.
2.2.5.2
PUMP DISCHARGE
For ease of assembly discharge
line.
To
pressure-actuated discharge lines
and disassembly, seal
this joint,
pump
O-rings
seals generally are are often welded
disproportionately
large
and heavy
discharges
are often
chamber The
vehicle thermal
ball-joint
propellants,
and
small pumps, the flanges become
discharge line and the downstream plumbing, of the pump volute tongue. The amount of because too little diffusion produces a large
and too much diffusion produces be large and therefore heavy.
a discharge
line
the connections between the turbopump assembly and the loads due to the weight of the turbopump assembly and react inertia, propellant inertia, engine gimbaling, fluid pressure
flange forces, expansion and
common end
types
and gyroscopic forces. The contraction of the turbopump
connection
of turbopump
arrangement ball-ended
is Close-coupled, struts at the other for the
J-2S hydrogen
This method
struts, tension
rigid pads at one end to accommodate pump
systems
dimensional
ball-ended in pure
shrinkage. The thick sections required for cast structure. If the pump is very flange.
mounting
to accommodate
growth or shrinkage. Theoretically, structure by loading the members
discharge
to the
mounts also adapt to the assembly and the thrust
assembly.
most
was used
for noncryogenic
bolted
MOUNTING
The turbopump mounts are engine. These mounts support to loads due to machinery differentials, differential
and
for small line sizes.
pressure drop due to high fluid velocity that, because of low fluid velocity, must
TURBOPUMP
used
are flanged
used for cryogenic propellants. For to the pump discharge, because
To minimize the pressure losses in the diffusers generally are used downstream diffusion is determined by optimization,
2.2.5.3
usually
with
tolerances
and
with
at least
differential
one
thermal
end of the turbopump and one or two the dimensional variations. One large pad
rectangular,
for designs
80
struts
arranged triangularly, yield the lightest or compression. However, a simpler
for this type small, it often
is best suited
utilize
integral
keys
to accommodate
of design make it particularly is mounted directly through with
cast volutes.
the
suitable the pump
2.2.5.4
GAS-GENERATOR
CONNECTION
Gas generators usually are influence on the turbopump
are
connections
often
interrelated,
of the
The major temperatures are used
and
discharge
better to relatively small support the GG weight.
to connect
the
gas manifold, and and contraction in some
duct
cases
also serve
turbines,
problems with and pressures
MOUNTING
close coupled to the turbopump design. The GG connections join
GG to the hot-gas inlet of the turbine loads and adapt to thermal expansion mounts
AND
because
(e.g.,
as the GG the
stiffer
and, therefore, have some the hot-gas discharge of the
the GG mounts (ref. 9). These the
J-2
mount.
turbopump) Such
manifold
support the 'GG connections and the
a system
structures
welded
lends
can more
itself easily
the connections are that they must seal gas at extremely high (e.g., 1550 ° F and 920 psia on the F-1 engine). If bolted flanges GG
to the
turbine,
pressure-actuating
metallic
seals
generally
are
used. Dual sealing lands increase reliability if a vent to low pressure is provided between the lands (thus providing a high pressure drop only across the pressure-actuating portion of the seal) in combination with a secondary low-pressure seal that precludes external leakage even if the primary seal develops slight leakage. Welded connections between the GG and the turbine
manifold
provide
a
more
reliable
joint
if
thorough
X-ray
crack-detection methods are employed. Welded connections are difficult the GG or turbine is damaged, and therefore may not be advanced-development systems. However, these connections are more expensive
for high production
rates
once
the system
and
penetrant
to repair in case best suited for reliable and less
is developed.
Gas-generator mounts must adapt to a high degree of thermal expansion and contraction. Solid-propellant GG's for start systems can produce a severe tendency to low-cycle fatigue. Maximum transient gas temperatures can be as high as 2400 ° F (above 2000 ° F for about 1 sec). On the J-2S liquid-hydrogen caused a Hastelloy-C manifold greater elongation configuration was gradients.
2.2.5.5 When
resulted _ in warped made of Inconel 625
TURBOPUMP a
reassembly,
turbopump may
turbopump, to crack after
have
SERVICE
to undergo
flanges with with a zirconia
ON THE
is disassembled,
the steep temperature ten starts. A change
it
gradients during starts to 347 CRES to obtain
consequent leakage. lining to reduce the
ENGINE must
a full-speed
be
removed
operational
from
test.
the
costly removal from the engine may tasks is performed. In the extreme, testing may be necessary.
81
engine
It is obviously
from a time standpoint to be able to replace seals, instrumentation, and without major disassembly. If provisions for performing these procedures the turbopump design, these relatively simple followed by operational
The final temperature
and,
perhaps are not
be necessary each major turbopump
after
advantageous bearings made in
time one of disassembly
it is highly desirableto designthe turbopump system such that main shaft sealsmay be replaced without disturbing the bearing assembliesand that leak checks following the installation of sealscan be madeto verify the correctnessof the installation. In addition, provision for hand turning the turbopump permits a simple torque checkfor early detection of damagedbearings,excessiverubbing, and worn seals.Boroscopeports are someti_nes providedfor inspectionof turbopump condition prior to reuse.
2.2.5.6 For
TURBOPUMP
major
overhaul
turbopumps (e.g., they are reinstalled
OVERHAUL or
disassembly,
the
turbopump
is removed
from
the
engine.
Most
those of the J-2 engine) then require a full-speed operational test before on the engine. If the turbopump overhaul involves other than a simple
change of bearings and seals, the engine should then be hot fired for recalibration. Therefore, simple turbopump overhaul is less costly than replacement, because a full-speed operational test of a turbopump is generally less costly than an engine recalibration. The cost of replacement parts is an important consideration in the design decisions as to whether to integrate certain parts for ease of original manufacture. For example, fir-tree-mounted turbine buckets can be replaced individually; if the turbine buckets are integral
with
bucket
the
disk,
is damaged.
weighed
against
The turbine individually.
however,
In deciding
entire
which
the probability
rotor This
the
turbine
design
wheel
to use,
the
may
have
cost
of original
assembly and the pump permits the replacement
2.2.6
Energy
the shaft
if one
manufacture
is
of failure. rotor assembly of either unit
are often designed without rebalancing
the case of the J-2 fuel turbopump (fig. 2), this approach replacement of the entire turbine assembly (manifold, stators, disturbing
to be replaced
seals or bearings
of the
to be balanced the other. In
is extended to permit the and rotor assembly) without
turbopump.
Start Systems
storage
systems
that
supply
propellant main tanks, pressurized-gas and solid-propellant start cartridges.
initial
drive
start tanks, All of these
power
for
the
turbine
include
pressurized-liquid-propellant systems have been used
on
the
liquid
start tanks, production
flight engines and are within the current state of the art. The type of start system on specific engines is shown in table I. The selection of an energy storage system, for starting an engine is dependent on several factors including maximum thrust), repeatability of starts, number of environmental conditions, weight, commonality valves,
and type
of turbine
drive
cycle.
82
allowable start time (time to 80-percent starts, tanked propellant conditions, of start and normal-operation control
The primary areasof impact of the start systemon turbopump designare provision of adequatestall marginduring the pump accelerationtransient,and possibleprovision of small boost pumpsupstreamof the main turbopumps.Adequatepump stall margin normally can be provided by adjustment of the energy level of the start system, although it may,be necessaryalsoto provide greatermargin through pump design.The J-2 enginesystem,which uses a pressurized-gas start system, was initially designedfor a storagepressurelevel of 800-+50 ° psia and a temperature of 200-+50 ° R. Thesevalueswere selectedto permit the start tank to be refilled, for restart capability, from the thrust-chamber fuel injector manifold. During the initial phase of the J-2 engine developmentprogram, there was a tendency for the fuel pump to stall when the oxidizer systemwasprimed under the above start-tank conditions. This problem was corrected by increasingthe fuel pump speedto move the operatingpoint during the start transientfarther from the pump stall curve.The additional stored energy required to increase the fuel pump speed was obtained by increasingthe start-tankpressurelevel to 1250 + 50 psia. Small
boost
pumps
are
considered
for
liquid-propellant
main-tank
start
systems
when
marginal start power is available. These small turbopumps have small rotating inertia low torque requirements and can be designed to operate with a wide range of propellant conditions to minimize thermal preconditioning requirements.
2,2.6.1 This
MAIN-PROPELLANT-TANK
start
system
uses
the head
for initial turbopump concept, is the simplest One of the principal results from start
HEAD available
in the main
rotation. This method of the start systems.
advantages repeatability
and inlet
propellant
is used
of this sytem requirements
in the
is minimization that make
separate start tanks, operating under blowdown components for refilling in order to accomplish
conditions, in-flight
tanks F-1 and
to provide RL10
the energy
engines
and,
in
of weight. The reduced weight it necessary for systems with to have restarts.
either These
multiple tanks weight-reduction
or
benefits are" amplified for engines that operate on a closed cycle with turbines that normally operate with a high flowrate and low pressure ratio; providing these conditions with propellants from separate start tanks would require large, heavy systems. Propellants a bipropellant
from
the
main
preburner
tanks
can be used
or gas generator
for engines
or the
thrust
that utilize chamber
either
cooling
the exhaust fluid
from
as a source
of
energy to drive the turbines. Although the exhaust from a bipropellant GG operating under tank head was used with the F- 1 engine, it should be noted that this kind of circuit normally is a low-flow circuit, and the GG should not be operated greatly off its design mixture-ratio conditions.
However,
for
engines
staged-combustion cycle (turbine main-propellant-tank start system is designed for liquid propellants,
with
low-pressure-ratio
turbines
operating
on
working fluid exhausting to thrust chamber injector), may develop combustion instability if the GG (preburner) since in effect it would be operating under deep throttled
83
a a
conditions. Solution of this problemrequiresa sophisticatedinjector or control system.Use of the thrust-chambercooling fluid as a sourceof energyto start the turbopumps,asin the RL10 engine, posesdifferent problems. The enginemust be conditioned within certain limits from what may be widely varying initial conditions to provide start repeatability. Under certain initial conditions, only marginal power may be availablefor turbopump breakawaytorque. Other problems associatedwith main-tank-headstarts for certain types of enginesand mission requirementsare exemplified by early developmentof the J-2 engine.The start method was eventually changedfrom tank-headstart to a pressurized-gas start systemfor the following reasons: (1) The start transient wastoo slowandunrepeatablefor the clusteredenginesof the SaturnS-II stage(5 to 6 secfrom start signalto mainstage). (2)
Because the engine was to be used in several different applications and had to meet orbital start requirements, engine inlet-pressure requirements were low and varied over a wide range (fuel pressure = 27 to 46 psia, oxidizer = 33 to 48 psia). As a result not
(3)
of the
wide
range
of start
conditions,
a common
start
sequence
was
feasible.
A common large
Development
start
difference and
main-tank-head
sequence in turbine
qualification
start
at sea level and altitude
systems
was not feasible
because
exit pressure. of
the
to other
SSME
types
will
of engine
advance cycles
the and
state wider
of
the
ranges
start requirements. This advance will be accomplished by placing a portion transient under closed-loop control and using a digital-computer controller monitor and command engine operation. The SSME also will incorporate
of fast-start
2.2.6.2
PRESSURIZED-GAS
art
for
of mission of the start package to low-pressure
turbopumps upstream of the main turbopumps. These low-power turbopumps with a wide range of inlet propellant conditions and, being low in rotating capable
of the
can operate inertia, are
transients.
START
TANKS
This system, employed on the J-2 engine system, utilizes high-pressure gas stored in a small engine-mounted tank to initially spin the turbines during engine start. This system typically includes a tank; fill, vent, relief, discharge, and check valves; and a duct connecting the start tank to the turbine. Both inert gases (e.g., nitrogen and helium) and fuels (e.g., hydrogen) have is that
been the
used start
successfully tank
as the turbine
can be replenished
developed on the J-2 engine; however, to refill the start tank. The alternatives
drive during
fluid. engine
The advantage operation.
of using This
the engine
procedure
has
fuel been
up to 60 seconds of mainstage operation is required are to use multiple tanks or a single large tank for
84
restart capability. The latter of thesealternativesresults in successivelylower start-tank pressuresfor eachrestart; this lower tank pressuremust be compensatedfor to achieve repeatablestarts. Pressurized-gas start tanks are alsoat a disadvantagewhen consideredfor closed-cycle engines because the high-flow, low-pressure-ratioturbines result in large fluid-storagerequirements. Tank volume is establishedon the basisof the energyrequiredto start the engine.Pressureis the most significant independentparameter for this type of energystoragesystem.It has been empirically determinedthat, for minimum weight, the gasstoragesystemshould be designedfor pressuresbetween 1500 and3000 psia.Becauseof thesehigh pressuresandthe largenumber of components(leakagepaths),the systemis susceptibleto leakage. A critical component for pressure/temperaturecontrol of cold storedgasis the relief valve. Relief valvesarenormally consideredas safety devices.Therefore,toleranceson relief and reseat pressuresare large,usually +5 percent on relief pressures, with the reseat pressure being 10 to 15 percent below relief pressure. control of start tanks where consistent start device
or a narrow-band
2.2.6.3
relief
These energy
tolerances is required.
are not adequate for pressure A pressure-regulator type of
valve is required.
LIQUID-PROPELLANT
START
TANKS
Liquid-propellant start tanks have been used on the Atlas MA-5 sustainer early versions of the Thor MB-3 engine. The start tanks can be ground based devices for single-start engines or they can be mounted within the vehicle.
engine and on with disconnect Very repeatable
starts
tanks.
this
(from
engine
type
of
pressurization, State-of-the-art
to
system
engine) is very
fill, vent
relief,
can
be provided
complex and tank
liquid-propellant
and
by liquid-propellant requires
discharge
start-tank
start
high-pressure
However,
propellant
tanks
and
valve systems.
systems
have
been
used
with
the
LOX/RP-1
propellant combination; however, many of the design elements are applicable to other propellant combinations. The tanks are filled with propellants and pressurized prior to engine start. Tank capacities and pressure levels are established to accelerate the turbine to a speed sufficient into mainstage This
method
to produce pump discharge pressures operation. The ratio of tank pressures of engine
start
has
not
been
used
where
that will allow the engine governs the mixture ratio engine
restart
to bootstrap during start.
is required.
Repeatable
refilling of start tanks with cryogenic propellants such as liquid oxygen and liquid hydrogen is difficult because of rapid vaporization due to warm tanks and unpredictable heat transfer rates. Since of the
the two liquid turbopumps,
propellants a separate
must small
be burned combustor
85
to provide
the energy
for initiating
is required
for engine
cycles
that
rotation do not use
a gas generator to provide a turbine drive fluid. As with pressurized-gas start systems, liquid-propellant start tanks areat a disadvantagewhen consideredfor closed-cycleengines becausethe high-flow, low-pressure-ratioturbines result in largefluid-storagerequirements. 2.2.6.4
SOLID-PROPELLANT
START
CARTRIDGE
Solid-propellant start cartridges have been developed into very reliable start systems for the YLR87-AJ-7, YLR81-BA-11, J-2S, and current versions of the MB-3 and MA-5 engine systems. Restart capability is available by using a separate cartridge for each start. The system
for the
of residue closed-cycle Design
J-2S
engine
provides
contamination, engines in which
considerations
the the
involved
three
starts
with
solid-propellant turbine exhausts in the integration
separate
cartridges.
However,
start technique cannot be to the thrust-chamber injector. of a solid-propellant
because used
with
GG into a turbopump
start system include determination of burn duration and flowrate, control of grain thermal environment, integration with the ignition system for the liquid-propellant GG, and selection of the solid-propellant start-cartridge method for multiple-start systems. A detailed analysis must be made of the engine start system solid-propellant burn duration and flowrate. Both parameters repeatability,
prevention
of thrust
cartridge on the J-2S engine 50 percent of their mainstage propellants in the pumps, propellants from the
overshoot,
and start
reliability.
starts the engine by accelerating speed. During this acceleration because the vehicle propellant
at burnout.
The performance because the grain
The
total
start-cartridge
propellant
effect engine
start
J-2S
weight
engine,
an
orbital
heat-transfer
after
analysis
was
temperature range of the start-cartridge engine with a three-start mission use
between
engine
The
is approximately
installation
maximum expected was a multiple-start starts.
recirculation to achieve The nominal grain burn rate from 2.3 lbm/sec
is strongly dependent on propellant is conditioned to 50 ° + 10 ° F for
installation on the engine and then is maintained, 75 ° F by the vehicle boattail environment. the
solid-propellant
the pumps to approximately period, there are mixed-phase ducts contain mixed-phase
engine time is to 9.9
13 lbm.
of the solid,propellant start cartridge is sensitive to grain temperature, burns faster and provides a higher flowrate as the grain temperature
increased. Grain-temperature start cartridge for the H-1
For
The
the required to start time,
that must pass through the pumps and be replaced with subcooled propellants propellant tanks prior to burnout of the solid-propellant start cartridge. The J-2S
engine does not require propellant bleeding or propellant start, since the pumps will start on mixed-phase propellants. 2.3 seconds and the flowrate increases at an exponential lbm/sec
to determine are critical
starts
satisfactorily
with
86
composition. The 24 hours prior to
on the engine,
conducted
is
at 40 ° to
to determine
the
grain. This experimental engine and up to 10 hours coast time
a grain-temperature
range
of-50
° F to
+140° F. However,start time varied from 3.2 to 4.2 sec.The J-2Sturbine drive systemwas made lesssensitiveto grain temperatureby (1) utilizing a checkvalvein the combustion chamber tapoff line, a step that allowed automatic changeof the power source,and (2) using a low-energysolid-propellantstart cartridgethat allowedpower overlapwithout thrust overshoot. , During engines,
the initial development contamination of the
Combustible start-cartridge
products forced into gas caused detonations
purge for the gas-generator H-1 engine gas generator. pressure
level and
The ignition solid-propellant RP-1
of the solid-propellant start cartridge for the MA-5 and H-1 injector for the liquid-propellant GG was a major problem.
flowrate
of GG start
enter
the
the LOX injector when LOX entered
LOX manifold A series of tests to prevent
combustor.
To
was developed was conducted
contamination
liquid propellants cartridge, which
manifold by the manifold.
the solid-propellant A gaseous nitrogen
to prevent contamination to determine the required
and maintain
reliable
reliable
ignition
and
minimize
overlap, two pyrotechnic igniters are used to ignite the gas generator. autoignited by the heat from the start cartridge and burn for approximately a redundant heat source to ignite the liquid propellants.
2.3 DESIGN 2.3.1 If
the
engine
start
cartridge
These igniters are 2 sec to provide
EVALUATION
system system
turbine power The turbopump meet a nominal system turbopump
and
is to
must
meet
be within
specifications,
the
on-
certain
The
turbopump
engine system
and
off-design system
operation must
of the
provide
the
component
tolerances
and
efficiency
are
considered
during
the
design.
DESIGN-POINT system mixture
SYSTEM
BALANCE
balance, known ratio, and thrust
as pump efficiencies, turbine efficiencies, turbine inlet temperature, and various pump
limits.
necessary to match the required pump power at all test and flight conditions. system must also be designed such that all delivered engine systems will performance specification. Therefore, the effects of variations in turbopump
In a design-point nozzle area ratio,
headrise,
the
by the LOX and
Engine-System Characteristics
turbopump
2.3.1.1
ignition.
in the H-1 engine system is accomplished burns for approximately 100 msec after
ensure
in the purge
flowrate,
and
turbine
factors such are combined turbine engine
flowrate
87
as engine type, chamber pressure, with initially assumed factors such
pressure ratio (for a GG or a tapoff cycle), performance factors to predict the pump requirements.
In making
this
balance,
the
turbine power is equatedto the sumof the oxidizer pump power,the fuel pump power, and any auxiliary power that might be in this system. A design-pointsystembalanceof the final design-pointcharacteristicsof the turbopump is also necessarybecauserefinements and changesthat occur during the detail designo,f the rocket engine systemin turn can affect the turbopump system operation. For example, during the developmentof the MA-5 sustainerengine, the turbine and pump efficiencies when tested were below the initial values usedin the engine-systemdesignbalance.This condition necessitatedchangesto increasethe turbine drive pressureavailablefrom thegas generatorso that the enginewould meet its ratedthrust value, i 2.3.1.2
OFF-DESIGN
SYSTEM
BALANCE
Frequently, an off-design operating for a particular component than operating range is not carefully at the design point may either
condition of the turbopump system may be more the nominal design condition. Therefore, if the
severe entire
investigated, a turbopump system that operates satisfactorily fail or perform unsatisfactorily during off-design operation.
An example of the effect of operating range on turbopump be found in the J-2 upper-stage engine. To meet flight 230 000 lbf thrust while operating at an oxidizer-to-fuel
system design and operation can requirements, this engine delivers mixture ratio of 5.5; the thrust
then
ratio
decreases
requirement requirement
to a value
of 170 000 pounds
at a mixture
of 4.5. The
for the fuel pump occurs at a mixture ratio of 4.5, while for the oxygen pump occurs at a mixture ratio of 5.5.
The first step in defining the operating range is to determine A system balance (sec. 2.3.1.1) that includes the predicted turbopump is used to determine engine throttling and mixture-ratio
the
(which are due to engine throttling, and tolerances and the
turbopump
range
must
power
the maximum
power
of planned operation. characteristics of the
entire range of turbopump operation over the planned excursions. In addition to the predictable excursions of
turbopump operation atmospheric pressure, due to manufacturing operating
the range operating
maximum
be
known variations of pump inlet pressure, mixture ratio), engine component variations effect of these variations on the required
allowed
for
in the
initial
design.
These
effects
are
treated by first obtaining known component tolerances from existing liquid rocket engines as a basis for estimating, for example, the hydraulic-resistance tolerances of lines, valves, and nozzle coolant passages; the tolerances in the pump head/flow and efficiency/flow relations; and
the
2.3.1.3), normal
tolerances
of the
etc. The (Gaussian)
tolerances of all of the components are assumed to have a statistically distribution. The effects of all of the component tolerances are not
algebraically probability taking the
added,
since
main-engine
this
would
nozzle,
the
be
of occurrence. The component square root of the sum of the
a
turbine-exhaust
condition
the controls
with
(sec.
a very
low
tolerance effects are summed statistically squares of the effect of each component.
by The
88
"worst-case"
nozzle,
individual component effects are determined either by performing an engine balance assumingthat the given component is at the extremeof its tolerancerangeor by using a linearizedversion of the engine-system equationsto determinethe influencecoefficientsfor eachof the enginecomponentsunderconsideration.
2.3.1.3 The
CONTROL
control
affect
points
the
majority
CONSTRAINTS used
required
for regulation
pump
of engines
discharge
to date
have
of engine pressure
used
thrust
and
a simple
and
thus
the
open-loop
mixture
ratio
turbopump system
can
system
markedly
design.
to control
The
turbopump
power. Either calibration orifices were used to control the flowrate to the turbine GG or a pressure regulator was used to control the oxidizer flow to the GG. The MA-5 and MB-3 rocket engines are examples of the latter. The J-2, F-l, and H-1 rocket engines all use calibration orifices in the turbine gas-generator lines to set power level. The RL10 engine used
fuel
(hydrogen)
heated
power the turbine; pressure and adjusts closed-loop
by being
to control a turbine
thrust-control
passed
through
the
thrust-chamber
cooling
jacket
to
turbine power, a control system senses thrust chamber bypass valve. The RL10 is one of few engines to employ a
system.
Closed-loop mixture-ratio control has been used in a number of operational rocket engines. This control is accomplished either with the main propellant valves or with a valve controlling propellant bypass around the pump. The MA-5 sustainer engine utilized control of the main
propellant
valve to regulate In engine
valves,
mixture
systems
that
'whereas
the
J-2 engine
has
an oxidizer-pump
bypass
use orifices
in the GG
feed
lines
for thrust
calibration
and a turbine
bypass for mixture ratio, there is a large change in thrust and subsequently discharge pressure as mixture ratio is varied (sec. 2.3.1.2). This type of control unsuitable for high-pressure exhibit wide variations in arrangement During
of the thrust
the
operating
design
staged-combustion pump discharge
chamber,
evaluation,
requirements
control
ratio.
of the
the
cycles, pressure
the turbine, influence
turbopump
because these systems as a consequence of
and the turbine of the
system
type
must
in pump system is
hot-gas
of engine
be evaluated.
basically the series
generator. control
Typical
system effects
on the of three
types of mixture-ratio control on pump discharge pressure requirements for a staged-combustion cycle are presented in figure 34. As shown, hot-gas valves in combination with a liquid-fuel valve decrease the pump discharge-pressure requirements relative to a system
with
a liquid-fuel
valve
alone.
However,
hot-gas
valves
have
not reached
the level
of
technology of liquid valves for control purposes and usually are avoided. Also, the systems with the added hot-gas valves have a slower response and are more complex. Therefore, the interaction between the type of control system and the turbopump design should be assessed before final system selections are made. Reference 30 gives the results of a detailed evaluation
of a control
system
for an engine
with a staged-combustion
89
cycle.
Mixture
Symbol
© ® ®
ratio
control
valves
--1
main
fuel
....
1 main
fuel
+ 1 hot
gas
main
fuel
+ 2 hot
gas
-----i
46O0
58OO
_- _ 420o
I:
380o
=
--',--_
5400
i
I
$
I
I
I
mixture (a)
34.
The dynamic of the engine
pump
-- Typica
effect
upon the constant,
Dynamic
of
type
pressure
of
mixture-ratio
(staged-combustion
Turbopump
transient slope of
[
response the pump
I
l
I
I
I
i
6 mixture ratio, off
Co)
Oxidizer
control
on
pusp
pump
cycle).
Analysis
conditions imposed on the turbopump system system need to be evaluated fully to ensure
requirements.
I
Englne
o/f
discharge
System
I
2600 5
ratio,
Fuel
Figure
these
M o
I
6 Engine
2.3.2
3000
[
system
characteristics
that
during the transient that the turbopump may
have
a strong
Operation will meet influence
of the engine system include pump type, turbopump time head/flow characteristic curve, pump NPSH requirements
during start, and turbine and pump arrangement. In addition to these factors, engine system factors such as engine starting method and propellant combination are highly important to the requirements imposed on the turbopump system.
2.3.2.1
START
Engine thrust buildup follows turbopump speed buildup. The turbopump be used as a measure of the turbopump speed buildup and the turbopump
time
constant
is a function
of pump
9O
rotational
inertia
time constant response rate.
and shaft
torque:
can The
r -
I
(15)
TqN where r = turbopump
time
I = turbopump
rotating
Tq
=
turbopump
constant,
shaft
N = turbopump
min.
mass moment torque
rotational
of inertia,
(design speed
value),
(design
in.-lbf
value),
The importance Of turbopump system characteristics with the exact engine system being considered. characteristics do not have a strong influence on because
this engine
utilizes
to the turbopump. minimize the effect pump
head-flow
a solid-propellant
lbm-in.
rpm
for engine system start varies For example, the turbopump the starting of the H-1 booster
gas generator
to supply
initial
The high-pressure, high-energy gases rapidly accelerate of turbopump inertia. The rapid start also minimizes
characteristic
and
of the pump
NPSH
requirements.
turbopump system characteristics are much more influential booster engine, which employs a tank-head type of start. The
On the
starting
greatly system engine, energy
the pumps and the effect of the hand,
the
on systems such as the tank-head start employs
other
F-1 the
pressure available from the vehicle tanks in order to initially supply propellants to the gas generator for the combustion gases to power the turbine. On this type of engine, reductions in start time can be achieved by reducing the turbopump time constant (eq. (15)). The time constant reduce
varies directly time constant.
with the rotating inertia and, therefore, Increases in turbine torque during start
decreases in inertia also will reduce the
will time
constant. The
ground-level
atmospheric
pressure
adversely
operating levels for the GG-type engine by reducing the turbine pressure ratio will always be high, developed for the same turbine can be increased by enlarging valve (which turbine nozzle On the
opens
is a positive area or turbine
an engine mixture
with ratio
affects
flowrate. For ground-level the nozzle area, thereby
at low power
levels but
restricts
developed
torque
at low
starts, turbine low-speed torque increasing the gas flow. A hot-gas
the flow
device for reducing start time. For flow may reduce start time as much two turbopumps, during the start
turbine
the operating pressure ratio. At altitude, and significantly more power will be
at mainstage)
in series
example, a 10-percent as 50 percent.
the relative turbopump transient. A relatively
with
the
increase
in
time constants can influence high time constant on one
turbopump will cause a lag in pump speed buildup and shift the mixture ratio during start. On a LOX/LH2 system, a bias to reduce the oxidizer speed can be very beneficial, since the bias will avoid
temperature
spikes
in the gas generator.
91
The shapeof the curvesfor pump head/flow characteristicscan be significant in engine transientbehavior.Steepslopeswill tend to stabilizeflowrate andreducedischargepressure variations.For a bootstrappingsystemwith a gasgenerator,variationsin the pump discharge pressuremay produce excessiveexcursionsin GG temperatureunless active controls are used. Whenliquid hydrogenis utilized asa thrust chambercoolant,the head/flow characteristics of the fuel pump mustbe carefully analyzed.During the start-transientbuildup, the relation betweenthe pump and the thrust chamberwill force the fuel pump to operateat reduced valuesfor the ratio of flowrate to rotational speed(Q/N). This is equivalentto operationat reducedflow coefficient _band can result in stall. The J-2 enginewasoriginally designedfor a tank-headstart. Digital computersimulationof the enginestart indicateda long slow start during which the fuel pump encountered the discontinuity due to stall in the head/flow curve.Stall in turn produceda headlossanda further reduction in Q/N. WhenQ/N dropped below approximately 1/3 of the designvalue,hydrogenvaporizedin the pump, a condition that resultedin a complete lossof dischargepressureand a total stoppageof fuel flow. To solve the problem, the tank-head start for the J-2 engine was abandoned and a pressurized-gas start systemwasusedsuccessfully. Pump cavitation during enginestart canaffect pump speedbuildup. Whena pump cavitates, lessenergyis required for fluid pumping and more is availablefor increasingthe rotational speed.An enginethat routinely experiencespump cavitationduring start canbe sensitiveto cavitation parameters.Typically, start-transient pump cavitation is observedduring high accelerations of suction-line flowrate. Such accelerations can be produced by a high-poweredturbine start or a rapid-openingvalvein the main propellant line. Under these conditions, start time and thrust buildup rates may vary with pump cavitation characteristicsandpropellant conditioning. Enginesystemstudiesconductedduring the past severalyearshavecomparedthe transient dynamics of turbine drive cycles (sec. 2.1.1.4) that differ from those actually usedin production rocket engines.Among thesecyclesare the thrust chambertapoff and the heat exchanger.For the thrust-chamber-tapoff cycle, with O2/H2 propellants and a separate turbine for eachpump, a parallel-flow turbine arrangementwill produce a fasterstart and also minimize the starting differences due to sea-leveland altitude environments. A series-flow turbine arrangement(becauseatmosphericpressurewould affect the exhaust pressureof the second turbine) would show a large difference in starting characteristics when sea-leveland altitude starts arecompared.The overallstart transientfor thi ssystemis primarily associatedwith the speed buildup of the oxidizer pump because,for a series turbine arrangement,the oxidizer turbine is usuallythe secondor downstreamturbine. The start dynamics of a 'heat-exchanger furnishes
heated
hydrogen
gas to power
cycle, in which a thrust-chamber the turbines, favors a series turbine
92
heat exchanger arrangement. In
this case,the seriesconfiguration achievesa fasterground-levelstart, sincethe overallstart transientis primarily associatedwith the speedbuildup of the fuel turbopump.
2.3.2.2 In
THROTTLING
future
applications,
Although concerning development dynamic
the
engine
thrust
no operational turbopump rocket engine throttling and
response
Throttling
from
analytical
aspects
response
be
varied
been both
important
to meet space vehicle requirements. determined by the turbopump time
(throttled)
during
the
mission.
throttled, considerable information from experimental rocket system
computer-simulation
of rocket-engine
becomes
may
engines have is available
studies.
This
section
will discuss
the
throttling. when
rapid
rates
of thrust
change
must
be achieved
Turbopump system response to throttling commands constant (sec. 2.3.2.1). As shown in equation (15),
is the
turbopump moment of inertia and the turbine driving torque are the factors that determine the time response of the system. Engine development and checkout for space launches make it desirable for the engine to exhibit similar throttling behavior for both altitude and ground-level simulation
of
environments. high altitude
turbine-drive
system,
a
arrangement,
because
the
pressure. therefore, cycle,
a series-turbine
desirable The
For low-thrust either turbine than
choice
capability response However, requires
This is especially true for large engine systems where environment would be costly. For a thrust-chamber-tapoff
parallel-turbine series
of the
type
is more
engines, altitude simulation arrangement can be utilized.
arrangement
a parallel
arrangement
arrangement
has been
shown
is
more
sensitive
desirable
can be performed For a heated-hydrogen
through
than
to variations
computer
a
the series
in atmospheric more easily and, heat-exchanger
simulation
to be more
arrangement. of throttling
control
system
largely
fixes
the
maximum
response
of the system. Generally, throttling of the hot turbine working fluid gives slower than direct throttling of the liquid-line valves to the main thrust chamber. the additional pressure drop necessary for throttling of the liquid-line valves the
turbopump
to operate
at a speed
higher
than
that
for
throttling
with
hot-gas
valves. Pump head/flow characteristics coefficients must be maintained Liquid-line
throttling
is attractive
increases pump discharge on the hydrogen pump. The ramifications feasibility
has been
are
particularly important high to prevent vapor from
the
standpoint
on the hydrogen formation within of
rapid
thrust
response,
pressure during throttling and can create flow coefficient To augment flow, hydrogen can be recirculated around
of hydrogen
recirculation
have
demonstrated.
93
not
been
completely
side. Flow the pump.
determined,
but
it
problems the pump. but
the
2.3.2.3
SHUTDOWN
Engine shutdown normally is accomplished by first cutting turbine power and then closing off the main propellant flows. With the main propellant valves downstream of the pumps, inlet pressure surges will be caused by both pump speed decay and the closure of the rfiain propellant valves. flow deceleration
Pump and
speed decay will tend to be exponential and produce an immediate suction pressure surge. Valve closure and the resultant flow
deceleration will depend on design details for the valve. A 500 msec cutoff engine system and necessitates very rapid closures of main-line valves. pressures
of several
hundred
psi
structure and must be designed is the sum of the main tank pressure resulting from the end The
fluid
surge
pressure
can
result.
The
• Inlet-line • Closure • Rate •
at engine
shutdown
geometry rate
(line diameter
of the propellant
needs
is the
is entirely
analogous
to the
most
critical
inlet pressure fluid column
water
hammer
outlet valve of a simple pressure conduit The following parameters influence the
of pump
of elasticity
and length
propellant of inlet-line
to be performed.
surge pressures than liquid hydrogen; has a low inertance that reduces surge.
upstream
of shutdown
valves)
valve
For long inlet lines, as in the forward rapid valve closure. In order to define conditions
typically
compressibility
of decay
Modulus
inlet
to withstand the overpressure. The maximum pressure, the cutoff surge pressure, and the of vehicle acceleration.
phenomenon that results from rapid closure of the carrying fluid flowing with steady uniform velocity. magnitude of the surge pressure: • Propellant
pump
is fast for a large Propellant surge
flow material
tank of a vehicle, high surge pressures can result from the maximum surge, a detailed analysis of transient
Liquid
oxygen
because
and
RP-1
generally
of the low hydrogen
will produce
density,
higher
the feed system
An engine system can also be shut off with main-line valves upstream of the pumPs,. In this configuration, the pumps will be forced into deep cavitation and will not experience high surge pressures. However, need to be considered.
the
surge
pressures
94
in the
inlet
duct
upstream
of the
valve
will
2.3.2.4 Pogo
SYSTEM
INSTABILITY
is a vehicle-system
instability
system, vehicle structure, frequency (5 to 25 Hz) pumps
are important
system
resonant
(POGO)
and along
involving
components
frequencies
of the
inlet conditions of small vapor
small which
vapor pockets markedly change a relatively high fluid inertance
system
overall
of inertance
system
and must
and inertance line, create
inertance
vary
pump
and
be considered
when
feed
total
speed,
and mode
The compliance is a inducer vanes; these
compressibility of the fluid
and would
manner in but pump inertance require a
terms.
at the pump suction, coupled with the relatively a resonant system with low damping. Compliance flowrate,
during a flight, the suction frequencies that are functions
propellants are consumed When a vehicle structural
displays both description
and compliance
compliance of a suction with
tanks
the effective fluid compressibility. The is formed is not completely understood,
that the fluid that a complete
The fluid inertance Therefore, resonant
propellant
are influenced by fluid compliance and inertance. pockets forming along the leading edges of the
testing has revealed Some data suggest
distributed
engine,
are calculated.
Pump result
dynamic effects.
the rocket
forward payload. The instability oscillation typically is low the vehicle longitudinal axis. The rocket engine propellanl
fluid
temperature,
system resonance of the structure
and
will vary. and mass
suction
high and
pressure.
The vehicle distribution.
also has As the
the vehicle mass decreases, resonant frequencies increase. matches a pump suction mode, a condition that can display
Pogo oscillations is created. Pump suction-pressure oscillations and the accompanying flow oscillations feed through to the thrust chamber and produce small thrust oscillations. These oscillations in turn feed into the structure and cause relative motion in the vehicle. The loop is then
reinforced
by a feedback
from
vehicle
motion
to propellant
acceleration
at the inlet
duct, thereby creating pump inlet-pressure oscillations. Since the vehicle mass is continuously changing, the tuning of the structure and feed system (if it occurs) is normally transitory. Because the vehicle geometry and operating role in a Pogo instability, potential changes cannot limited resulting
guarantee elimination to steep head/flow from
high pump
Analytical techniques through 34). Empirical and vehicle structural and design
changes
parameters at the pump in pump inlet conditions
inlet play such through design
a strong changes
of Pogo instabilities. Pump design considerations at present are characteristic curves, which reduce oscillation amplitudes
gain.
to predict the occurrence data are used to determine dynamics are obtained from
are then
investigated
analytically
95
of Pogo have been developed (refs. 31 pump compliance and inertance values, a structural analysis. System operational to determine
the effects
on Pogo.
2.3.3
.System Development
Testing
The turbopump system is tested during development to ensure that the system meets design requirements. A certain amount of the testing can be conducted on a turbopump test s_and, while the final testing as part of the engine system verifies that the turbopump system will operate
satisfactorily
in the operating
2.3.3.1
TURBOPUMP
environment
of the rocket
engine.
SYSTEM
Testing of the turbopump system usually subsystem testing. Volutes, pump back generally pressure tested individually modify the design ;,the tongue region
begins with a certain amount of component plates, manifolds, and similar components
or as subassemblies as early as possible in volutes often is critical. Brittle-lacquer
and are
to confirm techniques
locate the regions of maximum stress and strain gages to evaluate the stress generally employed. This testing for structural integrity is routine in nature, but proper attention can avoid serious problems. Pump
rotors
or turbine
disks
spinpit tested. Prototype determined by precise operated margins.
to
For critical is conducted
burst
are designed
to operate
components are first diameter measurements
speed.
This
procedure
standard
balancing
the
major
the liquid hydrogen traces, rapid weight
with capacitance-type were reduced from
testing
of the
means) is conducted performance before checked,
the
point after
design
and
and
limits
are to it
are generally
of general yielding testing) and then establishes
production but with
actual
(as are
safety
part. The spin test all holes and other
(if in high-stress regions). Spinning to produce yielding in such residual stresses and can increase the speed capability of the part. procedures,
lubricant (in place of photos of oscilloscope
Initial
tested to the before and
verifies
balanced at 1000 rpm on an adequate balancing to full design speed of 34 000 rpm was performed balancing was done in five planes in a high-speed
measured excursions
near the speed
applications, it is also customary to spin test each with critical surfaces in the semifinished state
discontinuities machined regions produces reverse In
that
or to
components
used). were
gap gages. With these 0.015 in. to 0.002 in.
turbopump
assembly
(or
cavitation-performance
runs
the
the
assembly
are
By use of quick-developing made. Shaft excursions were
procedures,
pump
only
integrity and The head/flow
are made.
96
then
machine. However, full-speed balancing up on a Mark 25 liquid-hydrogen pump. The vacuum chamber with a Freon-21 bearing
normally corrections
to establish the structural installation in the engine.
and
Testing
the
peak-to-peak
if it is driven
by
shaft
separate
to verify pump and turbine (H-Q) map of the pump is
must
proceed
very
carefully
at
first to preclude any rubbing problems or Sealleakagesthat could lead to catastrophic failure. Various test fluids can be used,and the designmust accommodateoperation with thesefluids. Air often is usedas the test fluid for all types of pumps,becauseof the costof propellanls and the size and complexity of drive equipment. Air testing is especially suitable for liquid-hydrogen pumps, since low-speedoperation with air can produce a compressibility effect comparablewith that of hydrogen. Low-temperatureeffects, of course,are absent with either air or water tests. Liquid nitrogen hasoften beenusedfor testingliquid-oxygen pumps to avoid the explosion hazardwhile providing low temperature.However, the high bearingand sealwear ratesthat result from liquid-nitrogen testingmust be consideredwhen the test programis formulated. The turbopump is properly instrumented for the turbopump-systemtests. Measurements normally include inlet and outlet pressures,temperatures,flowrates, and speeds.Also, accelerometersaremounted on the pump to detect high vibration levels. Tests on seals,bearings,and similar componentsand material compatibility testsare run individually prior to turbopump systemtesting. Then, in the turbopump-systemtests, the componentsareverified asa part of the entire assembly. One aspect of turbopump-systemtesting that bears additional emphasisis the "limits testing" concept.Testing accordingto this conceptexposesthe turbopump to the extremes of the operating environment to be experiencedin final engine-systemand flight testing._ The purposeof limits testing is to exposedevelopmentproblems early in the development cycle andto avoid costly catastrophicfailuresduring the engine-system test program.
2.3.3.2
ENGINE
The
turbopump
rated
thrust
turbopump complete
SYSTEM must
and
perform
operate
designer, means engine assembly.
to design
at
the
specifications
design
mixture
any rubbing
The operating hot-gas testing
engine
or high rolling
resistance
system
Engine-system
testing of the turbopump assembly The turbopump is usually well
engine-system tests. Measurements include flowrates, and speeds. Provisions for applying necessary to make sure that excessive turning detect
if the ratio.
is to meet testing,
as an integral instrumented
to
its the
part of the during the
inlet and outlet pressures, temperatures, turning torque by hand between test runs are resistance is not present. Torque checking can due to seal or bearing
problems.
environment within the .engine system can be very different from that in of the turb0pump system, especially as far as vibration levels are concerned.
The compression effect of the fluid entering be simulated more closely by engine-system
the pump that can lead to Pogo instabilities can testing, since the inlet-line geometry to a large
9?
extent resemblesthat of the flight system with the propellant tankagerelatively close coupledasit would be on the flight system.In turbopump-systemtests,the propellant tanks are kept more separatedwith the result that the inlet-line geometryusually doesnot closely simulate the flight system. Additional propellant shutoff valves and fire-extinguishing systemsareusedon engine-systemteststo help preventcatastrophicaccidents. The thermal conditioning requirementsof the turbopump systemare best defined through tests performed during enginetests.A fuel turbopump stall that occurredduring the start of the J-2 enginewas related to the thermal conditioning requirementsof the main thrust chamber. A solution to this problem was to incorporate thrust-chamber Conditioning techniquesto reduce fuel-systemresistanceduring enginestart. Thermalconditioning of the J-2 turbopump wasfound to benecessary,anda recirculation systemusingsmallelectrically driven pumpswas addedto the vehicle system.With this arrangement,propellant from the main tank is recirculated through the main pump andthe thrust chamberandthen returned to the tank. An intermediateposition on the main oxidizer valvewas provided to prevent excessivespeedbuildup of the oxidizer turbopump. Explosionsin liquid-oxygen pumpshaveoccurred during engine-systemtesting.Explosions on the H-1 turbopump werepreventedby first installing a shaftdeflection deviceto measure shaft radial movement.The results showedthat at start there were large deflections that were related to propellant main valve opening time and sequence.Inlet liners of Kel-F material wereincorporated. Rubbing betweenmetal rotating hardwareand the Kel-F liner wasnot detrimental,and the solution wassatisfactory. Several liquid-oxygen pump explosions occurred during engine-systemtesting of the F-I turbopump. Due to the high horsepowerand dynamic environmentof the pump, normally acceptedspline and pilot fits were insufficient to preventfretting and rubbing that ignited the pump materials in the liquid-oxygen environment. Fits on all rotating parts were tightened, and thermal techniques(e.g., heating the impeller so that it could be slippedon the unheatedshaft) wereusedto assemblethe pump so that at liquid-oxygen temperature all partswereoperatingunder an interferencefit. In the F-1 turbopump, leakageof the oxidizer past the primary sealin combination with leakagepast the fuel seal would result in propellant contact and subsequentlyin an explosion.The designsolution wasto use an intermediatesealpurgedby inert gasbetween the oxidizer and fuel seals.The purge gasis expelledfrom eachsideof the sealat a slot and drainedoverboardby a drain line. On the H-1 engine, momentaryleakageof
LOX through the LOX seal occurred at engine start. This leakage was caused by temporary pressure imbalance of the seal by a pressure surge. Holes were drilled through the seal housing so that pressure could act on the back side of the carbon seal coincident with the pressure surge on the carbon nose. This action prevented
the
separation
of the seal and the mating
98
ring and eliminated
the problem.
3. DESIGN
CRITERIA
Recommended 3.1
Practices
PRELIMINARY
DESIGN
3.1.1
System
3.1.1.1
PUMP HEADRISE
The
headrise
to produce To
and
determine
drops
its design the
Estimate
delivered
by the pump
discharge-pressure
downstream the pressure and
the preburner
the pump
FLOWRATE shall be adequate
for
the engine
thrust.
pump
(if applicable), across
AND
and flowrate
drops that occur chamber pressure jacket
Requirements
add
the
engine-system
pressure
of the pump discharge. For gas-generator cycles, add to the drops due to line losses, valve losses (if any), the regenerative
the
injector.
injector
headrise
requirement,
from
For
and the
staged-combustion turbine
cycles,
and the line losses
include between
the
pressure
them.
the expression
H=
144 [(Po)z-
(Po)]
(16)
Pl where
For
(Po)2
= pump
discharge
(Po)l
= pump
inlet
total
Pl = pump
inlet
propellant
a high-pressure
isentropic
enthalpy
hydrogen rise from
total
pressure,
pressure,
psia
density, pump
psia
lbm/ft
(above
the propellant
3
2000
psi pressure
properties
and
calculate
rise),
obtain
the
the corresponding
required head
rise from Hisen = J _hid)2-hl]
99
(17)
where Hisen= headrisefor an isentropiccompressionfrom (Po)l (hid)2 =
ideal
specific
enthalpy
at
hi = inlet
specific
enthalpy,
Btu/lbm
to (P0)2,
ft
Btu/lbm
(Po)2,
J = 778 ft-lbf/Btu
If the propellant is sufficiently compressible heating effect by applying equation (17) determine enthalpy
the actual enthalpy increment (isentropic
previous actual enthalpy, illustrated in figure 35. caused by the recirculation To determine the requirement from
in the application, account for the propellant at increments between (Po)l and (Po)a; i.e.,
at the beginning of a pressure enthalpy increment divided
increment by adding the actual by the pump efficiency) to the
and then sum the isentropic headrise increments. This procedure is When making this incremental calculation, include the heating of the thrust-balance-system flow.
volume
flowrate
requirements,
obtain
the
total
weight
flowrate
F
wE -
(18)
(Is)e where wE = engine
Then use the fuel pumps:
total
following
weight
flowrate,
equations
lbm/sec
to obtain
the
WE(MR) (Qo)e
=
volume
flowrates
for
the oxidizer
and
the
(448.8)
po (1 + MR)
(19)
WE(448.8) (Qf)p
-
of ( 1 + MR)
100
(20)
I I
zip
n
t
z_P 2
o o
0 0
AP 1
O3
sl
Specific
H.1sen
= J (Ahsl
Ahact
.
Figure
35.
-
Ah_
Illustration isentropic
entropy,
+
+
Ahs2
s,
Ahs2
+
of incremental headrise.
101
Btu/ibm-°R
+
Ahs3
Ahs5
method
+ ....
+
for
....
)
)
determining
where (Qo)e
= oxidizer
(Qf)P
=
fuel
pump
pump
Po = oxidizer
volume density,
0e = fuel density, MR = engine 448.8
= factor
volume
flowrate, lbm/ft
lbm/ft
mixture
flowrate,
ratio,
for converting
For preliminary design estimates, sizing, use the average or the local
gpm
gpm
a
3 (Wo/VVf)E ft 3/sec
to gpm
use the density.
inlet
If off-design operation is required, use the headrises and volume flowrates over the entire
3.1.1.2
NET POSITIVE
SUCTION
density.
For
more
detailed
flow-passage
procedure described above to determine engine operating range (sec. 2.3.1.2).
the
HEAD
The pump net positive suction head shall be suitable for the particular application, shall be adequate for stable and predictable pump performance, and shall minimize vehicle overall weight. Determine the methods for obtaining stable and predictable pump performance by consulting sections 2.2.1.1, 2.2.2.1, and reference 1. Then, if the engine contractor participates in the NPSH selection, conduct an optimization that considers the effects of pump inlet pressure and NPSH on vehicle tank and pressurization system weight, pump efficiency, turbopump weight, and system cost. The effects of NPSH on some of these factors are illustrated graphically in figure 36. Sections 2.1.1.6 and 2.2.1.1 discuss the effects of pump geometry on efficiency and suction performance. Convert the pump efficiency to turbopump equivalent weight (sec. 2.1.2.2). Then add the weights and select the optimum NPSH. If NPSH is specified to the engine manufacturers, then optimize the weight and performance at that NPSH.
3.1.1.3
PROPELLANT
PROPERTIES
The turbopump system design shall reflect the impact individual propellants and of the propellant combination. Because propellant design, use the considered:
properties have a major influence following checklist to make sure
102
of
the properties
on all aspects that all fluid
of the
of turbopump system property effects are
0
¢-J
_ _._ .rd
E--
I
I I
Y: ©
%d (t)
e_
o
_q
Figure
36. -
NPSH--------_
NPSH-------_-
NPSH.--.-.----_-
Effects
of variations
in pump
NPSH
on various
design
factors.
(1)
Density on arrangement
all aspects of turbopump design, (secs. 2.1.2.3 through 2.1.2.6 and
(2)
Material through
(3)
Corrosive, seal speed
(4)
Specific heat, specific heat ratio, on turbine geometry (see. 2.1.2.6).
(5)
Cavitation
characteristics
(6)
Two-phase
acoustic
(7)
Propellant saturation temperature rotor alignment (sec. 2.2.4.3), 2.2.4.6).
compatibility 10).
on
cooling, lubricating, limits (sees. 2.2.1.2
material
selection
on inducer
velocities
for
and viscosity and 2.2.1.3). and
on zero-NPSH
all
weight
overall turbopump 2.2.1.4).
wetted
characteristics
molecular
design
surfaces
on
bearing
of turbine
(refs.
DN
working
1
and
fluid
(see. 2.2.1.1). pump
on material and thermal
103
including the 2.2.1.1 through
capability
(sec.
2.2.1.1).
selection, turbopump conditioning (sees.
housing 2.1.1.8
and and
3.1.1.4
TURBINE
DRIVE
The turbopump
CYCLE
system
shall be compatible
with
the turbine
drive
cycle.
For cycles in which the turbine is in parallel with the thrust chamber (GG and tapoff cycles), minimize the required turbine flowrate by using high turbine pressure ratios (_ 15 to 25). Note that this practice requires some compromise with rotational speed, weight, efficiency, and turbine inlet temperature (sec. 2.2.3.1). Allow sufficient turbine discharge pressure for exhaust disposal (sec. 2.2.3.2). For cycles in which the turbine is in series with the thrust chamber (expander and topping cycles), maximize turbine flowrate, efficiency, and inlet temperature so as to minimize pressure ratio and engine weight. Using maximum values for if, r/, and (T0)l, calculate the pressure ratio fro,-, the expression (all parameters referred to turbine)
(21)
PR= 1-
r/J 550 Cp(To) (HP)
(ff
1
where PR = turbine
pressure
ratio,
(Po)l
= turbine
inlet total
(Po)2
= turbine
discharge horsepower
r/ = turbine
efficiency
Cp = specific (To)l
weight
pressure, pressure,
HP = turbine
= turbine
(Po) 1/(Po)2 psia psia
flowrate
heat
of turbine
= turbine
inlet
total
temperature,
")' = specific
heat
ratio
of turbine
550 = factor
for converting
working
fluid,
Btu/lbm-°R
°R working
hp to ft-lbf/sec
104
fluid
Add sufficient margin to this value to allow for efficiencies may not meet the initial predicted values.
3.1.1.5
THROTTLING
The turbopump Generate
the
fact
that
the
pump
and
turbine
RANGE operation
pump
the
shall be stable
head-versus-flowrate
over
the entire
characteristic
operating
required
range.
by
the
engine
during
its
maximum throttling excursion. Plot this engine requirement on the performance map for each pump candidate, as. shown in figure 7. If the pump stability limit crosses the requirement line during the engine excursion, do not use that pump design, because it will not meet the throttling requirement. In general, throttle to less than 50 percent of design.
3.1.1.6
use centrifugal
pumps
if the
engine
to
engine
is to
EFFICIENCY
The
turbopump
efficiency
shall
be
adequate
for
the
meet
its
requirements. For
engine
cycles
and
tapoff
cycles),
in which
the
determine
turbine the
is in parallel
maximum
F
allowable
with the thrust turbine
r(i )x/c-(ioE
1
Then
obtain
= thrust
chamber
the minimum
specific
allowable
impulse, turbopump
105
lbf-sec/lbm efficiencies
from
(gas generator the expression
(22)
where (Is)r/c
chamber
flowrate
from
w
_Vp
= (Wo)T
_V T
d- (_V_/f ) T
Hp
(23)
=
#
_p
T_ T
J
Cp(To)T
1 -- l\pR]
1
-r-_ i, .[
m
_/p
'T_p 7_T
J Cp(To)
Hp
T 1
--
p-R
where
fiT = total turbine (Wf)T = fuel-turbine
weight weight
(Wo)T = oxidizer-turbine weight
Hp = pump
headrise,
r/e = pump
efficiency
(To)T
1
flowrate,
efficiency
= turbine
inlet
in which cycles),
lbm/sec
flowrate,
lbm/sec
lbm/sec
ft
= turbine
For engine cycles staged-combustion
lbm/sec
flowrate,
weight
ffp = pump
7/T
flowrate,
total
temperature,
the turbine obtain the
°R
is in series with the effect of turbopump
thrust chamber efficiency on
requirements includes a
(fig. 14 is an example for a typical engine) mathematical steady-state representation
from an engine-system of all of the engine
Determine
the
chamber
chamber
constant-efficiency
pressure.
underestimates
Add
of system
sufficient pressure
curve
whose
margin
for
drop.
The
efficiency.
106
peak
overestimates result
pressure of turbopump
is the minimum
(expander and engine pressure analysis that components.
equals
the
efficiency
allowable
design and
turbopump
3.1.1.7
WEIGHT
The
weight
other
AND and
SIZE size
of
the
turbopump
the
reliability,
NPSH,
turbopump-system and
design
performance.
maximum
rotational
speeds
In so doing,
seal rubbing speed, turbine-blade pitchline velocity (sec. 2.2.1).
3.1.1.8
shall be minimal
consistent
with
requirements.
Maximize
Unless rotors.
system
centrifugal
turbopump-system
stress,
efficiency
within
consider and,
is critical,
the
inducer for
limitations
cavitation, geared
of life,
bearing
DN,
turbopumps,
do not use more
than
gear
two turbine
CONDITIONING
For restart applications with cryogenic require a minimum conditioning time
propellants, the and a minimum
turbopump system shall amount of conditioning
propellant. Minimize the pump temperature pump from the turbine: minimize minimize single-shaft
the turbine turbopumps
Evaluate
coating
the
material that will body of metal. Increase pumps,
3.1.1.9
take
mass relative (fig. 17). wetted a rapid
RELIABILITY,
to the
surfaces
the two-phase pumping if necessary (sec. 2.2.1.1).
LIFE,
rise during the contact
the
total
number
pump
of the
chill and
capability
AND
mass
pump
(sec.
with
also reduce
the
of the
2.2.4.6).
a thin rate
inducers
by thermally isolating at the contact points,
layer
of heat
and
research addition,
geared
and
of low-conductivity rejection
evaluate
of turbopump
units
the
life
and
reliability
the
to the
main
use of boost
to be built
and
the
requirements
intended
determine the effect of design variations on performance, weight, turbopump-system life and reliability requirements at a minimum low-speed of inducer
Evaluate
the and
COST
The turbopump system shall meet , mission at minimum overall cost: For
a shutdown period area, use insulation
designs, single-shaft arrangements fluid incidence angle to reduce and development costs to the assembly, manufacturing,
of
production
the
rate,
and cost. To meet the overall cost, consider
that eliminate gears, and the use of low values. cavitation erosion. In the cost analysis, evaluate
meet the handling
107
life and
and reliability requirements maintenance costs. Conduct
and, in a system
analysis to convert the performance and weight variations into payload variations. Then determine the cost per pound of payload variations by dividing the cost sums by the correspondingpayloads.Selectthe configuration that hasthe minimum cost per pound of payload.
3.1.2
Selection
3.1.2.1
NUMBER
OF UNITS
number
of turbopump
The
size and number Conduct
a cost
of System
of units
analysis
that
Type
units
per engine
produced
on total
considers
the
shall reflect
the impact
of turbopump
costs.
effects
of size
on
the
sum
of R&D
costs
and
production costs. Use a production learning curve to obtain average production costs for various numbers of production units. Then obtain total production costs for constant engine flowrate by multiplying the average unit costs by the number of units required to deliver the total
flowrate.
3.1.2.2
TURBOPUMP
The
turbopump
turbine For
engines
engine
EQUIVALENT-WEIGHT
(GG
system
flowrate in which and
design
and turbopump the
tapoff
specific
cycles,
shall weight
impulse
sec. 2.1.1.4),
FACTOR reflect to stage of the
evaluation
of
payload
weight.
turbine
use equation
exhaust (2)
the
equivalence
is less than
to determine
the
of
that
of the
turbopump
equivalent weight factor EWF. The factor OPL/O(Is)E in equation (2) is obtained from mission and vehicle analysis; note that PL refers to stage rather than vehicle payload. Then calculate the net effect on payload from equation (3). Use the parameters EWF and EW to evaluate For
turbopump
engines
in which
engine (expander essentially zero, weight
modifications. the
specific
impulse
of the
turbine
exhaust
and staged-combustion cycles), the turbopump and therefore the turbopump equivalent weight
(eq. (3)).
108
is equal
to that
of the
equivalent weight factor is equal to the turbopump
is
3.'1.2.3
ROTATIONAL
The
design
SPEED
rotational
hydrodynamic limitations. For a given coefficient,
speed
shall
performances,
pump turbine
the
and turbine type pitchline velocity,
reflect
evaluation
turbopump
weight,
(specified values etc.), determine
of
the
component
and
the
mechanical
of stage number, pump the effect of rotational
pump efficiency, turbopump weight, and turbine flowrate. Use equations determine the net effect of speed on payload. By means of a plot like that 16, determine weighing these
the effect of rotational speed on configuration. results against the other system requirements.
turbopump size (secs. 2.1.1.5 should never exceed the value If the
investigation
have potential,
3.1.2.4
of the
conduct
and 2.1.1.7) at optimum
given
a similar
TURBOPUMP
is unusually payload.
configuration analysis
shows
for each
(2) and (3) to shown in figure
Select design Unless throttling
critical,
that
inlet flow speed on
the
other
configuration
design
speed range
rotational
by or
speed
turbopump
configurations
and compare
the results.
ARRANGEMENT
The turbopump arrangement shall allow operation of the individual pumps and turbines at the speeds needed to produce the best overall system performance within the limitations of reliability and life. Use table size,
VII as a guide
propellant
type,
for screening
and
turbine
candidate
drive
cycle.
turbopump For
are dense, also consider the arrangement in figure in this arrangement, the turbine would reject preconditioning. geared) a prime with
multiple
In some candidate, candidates,
engines
based
in which
if the turbopump to both pumps,
both
predict
the
performances
and
weights,
the turbopump
and
use
on engine propellants
must be restarted; thereby reducing
cases, envelope restrictions make arrangement particularly for a small engine with restarts. For
(3) to determine the net effects on payload. Select these results against the other system requirements.
3,1.2.5
large 17(d) heat
arrangements
17(b) (pancake the applications equations
arrangement
(2) and
by weighing
PUMP CONFIGURATION
The
pump
maintaining
configuration the best
shall
compromise
For engines with thrust greater propellants other than hydrogen. single-entry
pump
would
deliver
the
among
required
the other
than 5000 lbf, use Consider double-entry
fall to the
right
of the
109
headrise system
and flowrate
requirements.
single-stage Centrifugal pumps for applications
peak-efficiency
while
specific
speed
pumps for in which a in figure
15.
Table
VII. - Guide for Screening Arrangements
Candidate
Turbopump
Recommended arrangement (fig. 17)
Engine Features Propellant type
Size
Turbine drive cycle Staged combustion & expander
Large
H2 fuel
(F > 25 000 lbf)
Dense oxidizer
Dense fuel
g
GG & tapoff
c,g
Staged combustion & expander
a,g
GG & tapoff
a,c,g
Dense oxidizer
Small
H2 fuel
(F < 25 000 lbf)
Dense oxidizer
Dense fuel
Staged combustion & expander
f,g
GG & tapoff
c,g,f
Staged combustion & expander
a,e,g
GG & tapoff
a,c,e,g
Dense oxidizer
Use shrouded impellers for liquid oxygen. shown in figure 19, and select the pump
For hydrogen, conduct an analysis similar to that type and number of stages by weighing the results
against the other system requirements. In general, use centrifugal pumps for hydrogen if the throttling ratio Qdes/Qmin is greater than 2. If the throttling ratio is very large, increase the number of centrifugal stages (figs. 13 and 19). If the ratio is less than 2, the axial pump also may be considered, particularly efficiency is very important. Do unshrouded titanium centrifugal
for missions with long operating durations where high not exceed approximate tip speed limits of 2800 ft/sec for impellers, 2000 (1700 to 2300 depending on design
specific speed, amount of sweepback, blade height, method of shroud attachment, etc.) ft/sec for shrouded titanium centrifugal impellers, and 1500 ft/sec for titanium axial rotors. For unshrouded centrifugal impellers, evalute the effects of tip clearance on efficiency (ref. 21) and the effects
of housing
weight
required
to maintain
thus
clearance.
Use the suction-performance relationships in section 2.2.1.1 to size the pump inlet to meet the NPSH requirements. Note that decreasing the inlet flow coefficient will not only decrease the NPSH requirement but will also decrease the pump efficiency (figs. 15 and 19). For
small
engines
are less than and (figs.
compare
400
(F < 5000 (figs.
other
lbf), initially
15 and
pump
types
analyze
18) or if impeller such
centrifugal tip
as partial-emission
18 and 20).
110
pumps.
diameters
If stage
are less than
centrifugal,
specific
speeds
1.5 in., analyze
Barske,
and
Rootes
3.1.2.6
TURBINE
The
CONFIGURATION
turbine
configuration
the best compromise Use axial turbines pressure-compounded
shall
among
the other
the
the required system
ratio
of the
do not turbine
pitchline
design'point
efficiency
curves
is unusually
critical,
velocity
Conduct
tradeoffs
between
similar
to the
(i.e.,
spouting
using
turbine
efficiency reference
more
than
two rotors.
cycles,
staged-combustion
consider
consider
temperature
and
weight
in which the the optimum
using
(secs.
Note,
2.2.1.4
partial-admission
reversing
the
the
ratio
however,
require additional and cooling. Figure
requirements
turbopumps to determine
21, use this velocity
to
turbine
and
turbines
so
that
that
supporting 23 may be
to determine
speed is combination
turbines in which the pitch diameter and, consequently, the pitchline must be reduced to keep a direct-driven, full-admission turbine
In
from
without causing axial thrust 4 for additional guidance.
that, in turn, for lubrication
flowrate
and inlet
in figure
For the
limit,
heights less 3 times the
(24)
speed,
minimum blade-height (figs. 17(e) and 17(0).
two-stage
velocity
diameter, small U/Co
ratio,
isentropic
to those
consider
diameter and, for centrifugal stress, pressure
rotors
1
large turbines require outboard bearings structure, an extra seal, and extra provisions used as a guide in selecting the type.
optimum rotor turbine-blade-root
achieving
requirements.
select the turbine type that will give the highest problems. Figure 23 may be used as a guide. Consult If efficiency
while
exceed 1800 ft/sec; do not use blade rotor diameters exceed approximately
/Co: From
horsepower
with, in general, no more than two and two-row (one-stage) velocity-compounded).
Use high pitchline velocities but than 0.15 in.; and do not let the pump impeller diameters. Calculate expression
deliver
the
limited by of rotor
2.2.3.1).
and
hot
velocity within
geared
gas
and the
turbines
from
the
precombustor can flow in through an annular axial inlet and, with the aid of a 180-degree annular elbow downstream of the turbine, the turbine discharge gas can flow back to the main combustion chamber through an annular axial discharge.
111
3:2
DETAIL
3.2.1
DESIGN
Limits
The and
turbopump low weight
curves
various
speed
2.1.2.2) funding
turbopump
limits
and
and
pump
configurational
is an example of this type with data relating turbopump
to compensate
turbine
efficiency,
3.2.1.1
not
exceed
required
efficiency
shall
remain
during the
free
from
the operational
inducer
suction
To
remove
preinducer optimization
information the
beyond
the
each limit.
example, assuming adequate 39 000 rpm is a reasonable
If the
curve
rotational
development selection for
speed
significantly
pump weight affects
on the plot.
cavitation speed
that
inducer
on inducer cavitation
speed
limits
upstream of and analysis
cavitation
overall
pump
profiles.
shown
turbopump a payload
in figure
24
at any
time
operation. optimization
If the pump inlet that considers
If long life (> 1 hr) is required, a be necessary. Reference 1 provides
limits.
on rotational
speed,
the main pump inlet. Use table to select the drive system for the
To pump propellants that are saturated inducer blade plus blade boundary-layer obtain the vapor-pumping capacity from
the
NPSH's near zero are desired, TSH limits expressed by equations (7) and
and tank weight. in figure 24 may
limit
impairs
and acceleration
specific
weight, pump efficiency, adjustment of the curves
detailed
to proceed
indicate
Use the information on the curves in and performance to vehicle payload (sec.
(8) (in combination with fig. 27) at any time during NPSH is not specified by the system, conduct
more
speed;
on figure 16; higher speeds would reduce bearings and seals, and would not reduce
two factors.
a turbine
vs rotational
the mission. For applications where operation should be considered. Do not exceed the NPSH
turbopump downward
performance limits of any
CAVITATION
inducer
performance
changes
speed. For requirements,
the other
include
INDUCER
The
for
efficiency
of plot. weight
turbopump optimization shown would require more complex
enough
during effects
weight the
shall provide for achieving good the mechanical or hydrodynamic
assembly.
as a basis for selecting pump and high vehicle-performance
the LOX efficiency,
Do
Speed
rotational speed without exceeding
in the turbopump
of
Figure 16 conjunction
INTEGRATION
to Rotational
component Plot
AND
place
either
a boost
pump
or a
VIII boost
as a guide in a system pump or the preinducer.
in the tank (zero tank NPSH), first determine the blockage (normally 20 to 40 percent) and then either figure 29 or from equation (9). For a narrow \
112
Table
VIII.
-
Drive system
Comparison of Typical and Preinducers
type
Drive
Advantages
Systems
for Boost
Pumps
Disadvantages
....
Gear (fig. 28(a))
Through-flow hydraulic turbine (fig. 28(b))
i
Positive speed control
Complex
Most efficient
Close coupled pump
Simple
Lags main pump start
Efficient
Close coupled pump
Relatively few rotating-shaft seals provide increased reliability
Limited
,
Recirculated-flow hydraulic
to main
to main
headrise
i
Allows remote location
Lags main pump start
In some appfications, more efficient than
Relatively
turbine
(fig. 28(c))
low efficiency
gas turbine
Gas turbine
Relatively few rotating-shaft seals provide increased reliability
Recirculated propellant may vaporize at turbine discharge
Allows remote
Complex
location
Allows pre-start
Requires source
separate
gas
In general, more efficient than recirculated-flow draulic turbine h
Electric
,,
motor
hy-
,
Allows
remote
location
Heavy electric
motor
(fig. 28(d)) Allows pre-start
Requires electrical energy source
Efficient
Limited
With a propellant.cooled motor, relatively few rotating.shaft seals provide increased reliability
113
headrise
range
of zero-NPSH
an (i//3)L
of 0.7;
operation,
use a value
for a wider
range,
vapor fraction and minimum allowable
the minimum inducer inlet
these
hydrogen
relations
pumping,
3.2.1.2
for
consult
references
BEARING
DN
bearings
shall
The
use
for (i//3)L of approximately
an (i//3)L
zero-NPSH annulus area.
and
oxygen.
operate
at DN
more
detailed
never
exceed
use the corresponding
tank-saturation pressure For zero inlet-line losses,
For
17, 23, 24, and
0.6 and
of 0.3 to 0.4. Then
to determine the figure 37 displays
information
on
two-phrase
25.
levels
that
will
permit
them
to meet
the life
requirements. Do
not
exceed
required
(10)
bearing
life including
a downward If the
the
preliminary
adjustment
turbine may
DN limits
be placed
(refs.
26 and
3.2.1.3
between
27) may
SEAL
the
calibration,
and
flight
the operating
operation
life. If
exceeds
1 hour,
size the shaft
the
speed
limit
on critical-speed for this
type
basis
(ref.
8). Equation
Of arrangement.
To permit
turbine bearing outboard (fig. 30(b)). Equation (11) speed limit. To permit even higher speeds, the pump
inducer
and
the
impeller
(fig.
30(c)),
maybe bearing
or a hydrostatic
bearing
be considered.
RUBBING
The seals shall
6 at any time during
may be necessary.
(fig. 30(a)),
to approximate
higher design speeds, place the used to estimate this rotational may
testing,
of the limits
is overhung
be used
in reference
operate
SPEED at rubbing
velocities
that will permit
them
to meet
time
during
the life
requirements. Do
not
exceed
the
seal speed
limits
given
in reference
7 at any
the
operating
life. If required life exceeds 1 hour, a downward adjustment of the limits may be necessary. Consider the use of static liftoff seals, which will eliminate dynamic rubbing seals. If the
turbine
8 provides are used,
is overhung,
detailed equation
equation
information (13) may
(12)
may
be used to estimate
the speed
limit.
Reference
on shaft sizing. If outboard bearings (figs. 30(b) and 30(c)) be used to estimate the limit. To permit operation at even
higher speed, advanced configurations are necessary (refs. 7, 26, and 28).
such
as liftoff,
114
hydrodynamic,
and
hydrostatic
seals
450
Ii00 11
45
10
5 40
400
i000
6
45
4
40
35 900
--
350
;
8oo -
o
300 700 20 <
-25
<
250
600,-2O
15
k
200 .,..q
500
--
400
--
Choked (max.
Ln
*/A)
Choked
0
150
(max.
0
O/A)
0
Flow
._
process
-
equilibrium
Tank conditions Solution method
lOO
:-.]
Official Subscript SO
isentropic
= saturated not critical
300
-Flow
liquid 200
minus
--
100
Line
vapor
20
30
volume
fraction, (a)
40
50
--
Po-PI
al'
percent
6O
0
refers
inlet
l i0 Line
l 20 vapor
40 fraction,
(b)
37.
-
Zero-NPSH for
oxygen.
pumping
capability
requirements
for
hydrogen
inlet
Oxygen
and
total
minus
pressure
l
30 volume
to
static
l
Hydrogen
Figure
liquid integration
Offical oxygen properties Subscript 0 , tank conditions 1 = inlet line conditions
0 t 10
= equilibriumisentropic
process
Tank conditions = saturated Solution method = numerical of momentum equation
Oo
hydrogen properties 0 = tank conditions 1 = inlet line conditions
Po-PI refers to inlet total inlet static pressure
©
l 50 al"
l 60 percent
l 70
3.'2.1.4
TURBINE-BLADE
The
turbine
which Do
not
blades
failure
exceed
CENTRIFUGAL shall
the turbine
N 2 Aa limits
given
a downward
adjustment
To optimize the turbopump that consider turbine inlet type,
pump
efficiency.
turbopump An iteration rotational
turbine
blade
Figure
speed for speed.
materials,
hot
that are below
the
4 at any time
of the limit may
turbopump flow charts
A pressure-ratio
in these
gas on
in reference
weight,
the levels at
which
is the
outside
can
cause
the
allowable
N2Aa
rotational
speed,
and
for determining
basic types of drive cycles. efficiency is a function of
result
of pressure ratios, is partially illustrated the centrifugal stress on the upstream
rotors
the operating
limit, conduct tradeoffs turbine pressure ratio,
procedures
limits for four because pump
optimization,
during
be necessary.
turbopump
of the analytical
and weight at the blade stress pump efficiency is indicated
compensate for the smaller limit somewhat. However, by
stresses
at the turbine-blade centrifugal-stress temperature, turbine inlet pressure,
38 presents
figure 38(a) over a range turbines are used, check temperature
at centrifugal
will occur.
life. If long life is required,
turbine
operate
STRESS
of applying
the
logic
in
in figure 31. If multi-rotor rotors, because the higher
to drop
enough
to more
than
annulus area. Turbine blade cooling can be utilized to relieve this before using turbine cooling, consider the thermal stresses caused and
cold
cryogenic
propellant
on the
inside;
for repeated
rapid
starts, these stresses can cause cracking due to thermal fatigue. If repeated starts cause thermal fatigue even without cooling, consider hollow blades with hot gas ducted to the hollow center. Hollow tapered blades can be used to increase the allowable N2Aa limit without
affecting
the
outer
contour
of the
blade
and, therefore,
without
penalizing
turbine
performance.
3.2.1.5
GEAR
The gears
PITCHLINE shall operate
levels at which In designing tooth
bending
a geared stress,
failure
VELOCITY at pitchline
velocities
and tooth
stresses
that are below
the
will occur.
turbopump, and tooth
observe compressive
the
recommendations
stress
116
given in reference
on gear 5.
pitchline
velocity,
TORBINE INPUT VELOCITY TURBINE
INPUT
PITCHLINE
VELOCITY,
INLET PRESSURE, INLET
RATIO,
O/C o
TURBINE EFFICIENCY, U
_T
INLET TEMPERATURE, (To)T1 TURBINE TYPE
(Po)TI
TENPERATURE, (To)T1
PRESSURE RATIO,
PR
TURBINE TYPE N2Aa LIMIT
"VELOCITY " TURBINE N2Aa
RATIO,
U/C o
EFFICIENCY,
_T PI'I_HLINEVELI_ITY,
LIMIT FOR TURBINE Pt_P
INLET PRESSURE,
INPUT
PRESSURE HEADRISE, TURBINE
FLO_ATE,
WT
EFFICIENCY, ANNULUS
ROTATIONAL
AREA,
SPEED,
TURBOPUMP TURBOPUMP WEIGHT, ROTATIONAL WTpSPEED,
TURBINE ANNULUS AREA, (Aa)T
I
ROTATIONAL
!
SPEED,
,1
NT
TORBOPOI_ TURBOPUMP ROTATIONAL WEIGHT, WTpSPEED,
NTp
(b) Staged-combustion (a) Gas-generator
and
tapoff
HI)
FLO_ATE, _p I Pt_P INPUT EFFICIENCY, _p
_p
I
NTp I
HEADRISE,
FR
Wp
1
(Aa) T
NT
RATIO,
U' (Po)T1
Hp
FLO_ATE,
"-4 TURBINE
FOR _BINE
cycles
Figure 38. - Flow charts for determining turbopump speed and weight at the turbine-blade centrifugal-stress limit.
i
and
expander
cycles
3.2.2
Pump Design
3.2.2.1
INDUCER
INLET
FLOW COEFFICIENT
The inducer inlet flow good pump efficiency. Do
not
design
at inlet
coefficient
tip flow
inlet flow coefficient, generate 19. Reference 1 provides more
3.2.2.2
shall provide
coefficients curves detailed
for good
much
below
suction
0.07.
To
performance
determine
and
the
impact
at the design point that are similar to those information on inducer suction performance.
of
in figure
STABILITY
The pump operating
shall
be stable
and shall
have
predictable
performance
over
the
entire
range.
For axial pump rotors and stators, do not exceed a blade diffusion factor DR of 0.70 mean blade diameter (or station) at any point within the design operating range:
W2
DB,ms
Wl
u
--
W2
at the
u
= 1 ----+----
(25)
W 1
2O
W 1
where D n,ms = blade wl
= inlet
diffusion
factor
fluid velocity
w2 = discharge
fluid
at blade
relative
velocity
to the
relative
station
blade
to the blade
wl u = tangential
component
of inlet relative
w 2 u = tangential
component
of discharge
= blade
solidity
(ratio
of chord
For centrifugal pumps, design such that zero-slope points on the performance-map pump
mean
always
For more references
operates
in the
detailed information 2 and 3.
negative
relative
length the
slope
velocity velocity
to spacing)
entire speed
operating lines or,
region
of the pump
on centrifugal
pump
118
stability
range falls to the right of tile in other words, such that the speed and
line (figs.
axial
pump
12 and stall,
13).
consult
3.212.3
TIP SPEED
The pump range
impellers
of rotational
shall
maintain
mechanical
over
the entire
operating
speed.
Design pump impellers with an adequate and the tip speed at which burst occurs; pump impellers. capability, the
integrity
For forged approximate
margin between the maximum operating tip speed reference 2 may be used as a guide for centrifugal
titanium, tip speed
which limits
is the material are 2800 ft/sec
with the highest tip-speed for unshrouded centrifugal
pump impellers, 2000 ft/sec for shrouded centrifugal pump impellers (1700 to 2300 ft/sec depending on design specific speed, amount of sweepback, blade height, method of shroud attachment, hole in the center, etc.) and 1500 ft/sec for inducers and axial pump rotors. Note
that
impellers references
3.2.3
3.2.3.1
the
tip-speed
limit
decreases
with
degree
of blade
sweepback
and that hydrogen pumps are the only pumps that approach 1, 2, and 3 for more detailed information on tip speed limits.
Turbine
for
these
centrifugal
limits.
Consult
Design
PERFORMANCE
OPTIMIZATION
The turbine shall provide for low weight and high efficiency necessary to maintain mechanical integrity. For a tip-speed-limited turbine design (i.e., effect of inlet temperature on allowable temperature on various design 39(a). If the turbine diameter
a design pitchline
within
the limitations
limited by rotor stress), velocity. Consider the
determine the effect of inlet
factors and conduct an optimization as indicated in figure becomes disproportionately large (greater than 3 times the
pump impeller diameter) or the moment of inertia becomes too large for rapid starts, it may be necessary to back off from the tip-speed limit. This condition often occurs with the oxidizer
turbopump
in a hydrogen-fueled
application.
For designs limited by turbine-blade centrifugal stress, determine the effect of inlet temperature and pressure ratio on the allowable rotational speed (sec. 2.2.1.4). Consider the effects of inlet temperature on various design factors and conduct an optimization as indicated
in figure
hydrogen
turbopumps.
To
determine
the
39(b).
number
Note
that
of stages,
the
use
blade
stress
equivalent
optimum tradeoff weight and performance. Unless engine, do not use so many stages that an outboard reference 4 for more detailed information on turbine
119
limit
generally
weight
(sec.
is encountered
2.1.2.2)
only
to determine
the payoff is highly significant turbine bearing is necessary. rotor stress limitations.
in
the
to the Consult
Rotational Pressure
speed ratio
N = constant" PR = constant
o .4 0 o
Pitthllne
o
l_wbine
q_
$i o
veZecity
U • coas_t
ratio
presmare
PR = (Po)T1/PT2
!
k
hO 0
I
-,4 O
| _0
O
o
Turbine
inlet
o
temperature------_
Turbine
(a) At the turblne-blade
tip-speed
Figure
39.
-
inlet
temperature
Turbine
-----e"
(b) At the turblne-blade
limit
Effect in
inlet temperature
of
turbine
turbine
inlet
optimization
temperature (gas-generator
on
various cycle).
design
factors
Turbine
inlet te_erature_
centrifugal-stre_s
limit
3.2.3.2
EXHAUST
The
turbine
performance To
avoid
cycles, choke
PRESSURE shall
exhaust
and sufficient
unpredictable
engine
design for a turbine between the turbine
engine turbine
level
for the engine
to meet
performance
3.2.4
into the
thrust
Turbopump
When
solving
particular
3.2.4.1 The
bearing
simplest
flareback
in GG and
tapoff
engine
for sufficient plumbing and
Integration
design
shall
components
reflect
problem,
evaluation
also
consider
in the turbopump
location
shall
without
provide
exceeding
turbopump
design
the
maximum
of" component
the
effects
of
that
system.
individual with
no
support
bearing
more
than
inboard of the turbine if the resulting bearing DN DN is not within limits or the turbine has more
the
thrust.
PLACEMENT
bearing outboard of the component development arranging
engine
pressure great enough to permit the gas flow to the end of the exhaust duct. To meet design
mechanical-design
on all other
manner
a single-shaft
bearing bearing
mechanical
BEARING
its design
predictable
chamber.
a particular
solution
for
and expander engine cycles, design flow to pass through the downstream
the
Mechanical
The turbopump in teraction.
sufficient
or exhaust
discharge static discharge and
thrust for staged-combustion discharge pressure to permit
the injector
For
at a pressure
turbine
turbine. (pump
relative
To simplify future alone and turbine
for
the
rotor
in the
limits. two
turbine
rotors,
place
the
aft
is within limits. If the resulting than two rotors, place the aft
development alone prior
effort, consider individual to combining them) when
to the pump.
For a single-stage pump, place the forward bearing inboard of the pump impeller: If the pump has two or more stages, place the forward bearing between stages. If the pump has a separate inducer stage, place the pump and use the inducer stators the pump have more own sets of bearings geared
turbopump
than three and Seals.
forward bearing to support the
between the inducer and the rest of the bearing housing. If both the turbine and
rotors, consider designing each as a separate unit with its For engines with thrust less than 10 000 lbf, consider the
arrangement.
121
_'.2.4.2
TURBINE
The turbine effects of concentricity, be simple;
ROTOR
ASSEMBLY
and shall be easy
Concentric
pilots
to assemble
may
various components, the pilots. (2)
Disk mounting coefficient that
(3)
Even
disk
from
not
and disassemble. thermal
growth,
be dependable
or if thermal
consider
if different
growth
and
conditions,
clamping
Disks will deflect the other.
axially
the
disks
bolts
must
if the
temperature
may
adapt
operate
to the
If the rotor tip speeds are approaching the stress limits, designing without a center hole. To permit the two sides
following
when
are used
for the
force
act together
a thermal
at different
resulting
on one
the
materials
centrifugal
(or clamp) bolts may loosen if they have is much different from that of the disks.
at steady-state
If so, the growth. (4)
ATTACHMENT
rotor assembly and its attachment to the main shaft shall adapt to the centrifugal stress and thermal growth on deflections, normality, alignment, and clearances, shall transmit the torque reliably; shall
To minimize problems resulting designing the turbine rotor assembly: (1)
AND
on
expansion
temperatures.
differences
side is different
in radial
from
that
on
minimize centrifugal stresses by of such disks to be symmetrical,
use a drive coupling such as a curvic coupling to attach the disks to the shaft and to each other. For curvic couplings, place the clamping bolts at the coupling mean diameter in order to avoid disk deflection. In general, use a disk diameter that is less than 4 times the coupling diameter; if a greater diameter disk stability and a reasonable
ratio is desired, conduct a rotor unit loading for the coupling (ref.
dynamics analysis 8). If the stresses
to ensure are within
the material limits, the torque may be transmitted by torque pins (a minimum of three), the clamping bolts, or a spline (with radial pilots) on the extended hub of the disk. To attach the disks to the main shaft, use clamping bolts if the disks do not have a hole in their centers. If they have a center hole, mount the disks on clamping bolts or a center clamping nut on the shaft.
a shaft
and secure
them
with
either
To attach the turbine rotor to the main drive shaft of a single-shaft turbopump, coupling in combination with an involute spline alone, a curvic coupling alone, torque-pin joint. The main considerations are normality and concentricity.
use a curvic or a bolt and For a geared
turbine,
requirements
consider
gear
separation
cause problems, consider using pump are separate units, use deviation.
and
allowances
for
misalignment.
If these
a quill shaft arrangement (fig. 4). If the turbine and the a coupling that is adaptable to misalignment and axial
122
If the turbine rotor consistsof individual disks,use an idiot pin or asymmetricalbolt holes to prevent backward mounting. To facilitate handling and positioning, install the disks vertically (vertical turbopump shaft). If the aft bearing is outboard, consider stackup tolerances,concentricity tolerances,runouts,and thermaldeflectionsduring assembly. If the pump has a thrust balancepiston, position the turbine rotor axially relative to the main driveshaft to avoid rubbing of the rotors andstators. Consultreference8 for detailedinformation on designof shaftsandcouplings.
3.2.4.3
TURBOPUMP
The
turbopump
pump
and
permit
housing
the
components To
HOUSING
turbine
and avoiding
a single-shaft
the cold pump and use an arrangement without
losing
thermal
than
growth
If the turbine avoid contact struts
of
the
to small
and have
to adapt
to the
large
to the
the pump housing.
the
the
various
pump.
If the
difference
for the turbopump casing can slide cylindrical
should remain This is the point
between
housing or unrestrained
housing
is used,
large
torque
turbopumps
are
for the radial
by
passing
them
more
subject
to unsymmetrical
pins to withstand. sides of the rotor, design manifold or by shielding through
hollow
vanes;
the casing to the support use
wherever possible. This design maintains a constant relative position between bearing supports and ensures alignment between the nozzles and the rotor blades. deflection is staill excessive, use a self-centering linkage arrangement to connect to the bearing To
avoid
regions.
high-temperature
insulation the two If thermal the casing
carrier.
cracking To
the
ambient and, therefore, the at which the cylinder should
casing. Use a structural cone to attach the hot turbine The radial pin method is less applicable to large
because
bearing
between
between
temperature
either use a long cylinder on which the hot turbine
is a separate unit with bearings on both with the hot gas by using a discharge discharge
differences
alignment
failure.
relation
a larger
temperature proper
along the cylinder will remain constant.
and bolted onto to a cylindrical
turbopumps
structural
turbopump
its concentric
to the
maintaining
the hot turbine, of radial pins
temperature somewhere diameter at that point be flanged manifold
shall adapt while
of
avoid
the sealing
housing,
minimize
problems
as
material well
regions.
123
as
cross cracking,
sections do
in high-temperature not
use
flanges
in
3.2.4.4
BEARINGS
The
bearings,
AND
SEALS
seals, and
requirements, shall minimize
bearing
Provide lubrication to all bearings; If the pumping fluid is hypergolic lubricant To
from
avoid
a separate
early
with a slight so that more
system
bearing
lubrication is more or if the turbine
that
difficult if the aft bearing is a separate unit, supply
failure,
prevent
sliding
by installing
the
bearing
the
bearing
seal
that
DN is within
will
limits,
permit
consider
the
seal
the effects
The thrust overloading
to absorb centrifugal
races
is outboard. the bearing
on the
to operate
shafts
bearing, Obtain
effectively,
of the various
rubbing velocity is within limits. These limits are a function and type of lubricant. See sections 2.2.1.2 and 2.2.1.3.
AXIAL
axial
and
operating
and radial and axial .loads. In general, use is a separate unit. Consult reference 6 for
In some configurations, particularly large ones, use a flexible bearing housing to the turbine manifold. If critical speed or shaft try changing the bearing preload so as to change the bearing spring
If the
life
of the rotor,
hot turbine gas from entering the bearing, install a seal between the bearing and source. In general, use a shaft-riding seal and a face-contact seal. Check to ensure
that the seal drop, material,
3.2.4.5
the turbopump
interference fit. If axial or radial loads are severe, consider a split-race balls may be inserted. Use a seal material compatible with the lubricant.
conditions such as temperature, kind of lubricant, ball bearings to absorb axial loads if the turbine more information on selection of bearing type. To prevent the hot-gas
shall meet
source.
the pressure balance across the particularly if the seal is dynamic. To ensure
lubrication
shall satisfy the axial and radial load requirements rotor dynamic problems.
thrust
THRUST
it. If it is not, pumps), which,
diaphragm to attach the whirl becomes a problem, rate.
BALANCE
balance system the bearings. is within
of seal pressure
shall
resist
the load-carrying
the
axial
ability
thrust
of
the
of ball bearings,
rotor
without
use ball bearings
consider balance ribs on the back faces of the impellers by causing the propellant to spin with the impeller, decrease
net axial thrust on the rotating assembly. If the axial again, if the pump is centrifugal, consider a labyrinth the impeller back face in combination and, therefore, the axial thrust.
with
a cavity
124
thrust seal vent
is too great arrangement
to reduce
alone (for the
for balance ribs and, to form a cavity on
the pressure
in the cavity
If either the variation in axial thrust between operating points or a reasonblemargin for error in predicted thrust are too great to be carried by the bearings,use bearingsthat are free floating axially in combination with a thrust balancepiston (refs. 2 and 3). To avoid balance-pistonrubbing, use the bearingsaslimit stopsor, if the bearingDN's are too high to carry the loads,userubbing stops.For a double-ringorifice-type balancepiston in hydrogen (fig. 33), the orifice rings can be usedas rubbing stops.However,in oxygen,a limit stop with a lower rubbing speedthat is located elsewherein the turbopump may be necessary. Locate the limit stops close to the balancepiston (axially) to minimize the influence of thermal and pressuregradients on the axial distancebetween the stop and the balance piston. To minimize axial length and thereforeturbopump weight and to minimize leakagelosses, use balancepistonsthat areintegralwith the pump impellersor rotors. Analyzehousingand impeller deflections during designbecauseintegral balancepistonsaresensitiveto clearance variations. Separatebalancepiston mounted elsewhereon the turbopump shaftmay beused for caseswhich havedeflection programsor where axial length is not a prime concern. To obtain more predictable forces and minimize shaft flexing, use shroudedimpellerson centrifugal pumps. If impellersare unshrouded,momentarydifferencesin the flow patterns in the flow passages will generateadditional unbalancedforcesthat mustbe accommodated in the design.
3.2.4.6
THERMAL
Thermal
barriers
temperature simpler
BARRIERS between
rise during
and more
the
pump
shutdown
and
the
turbine
so as to make
shall
minimize
subsequent
restart
the
pump
procedures
efficient.
For applications in which cryogenic turbopumps are restarted, isolate the hot turbine rotor from the pump rotor by connecting them with an involute spline together with a quill shaft, or with a ball spline coupling, or, if stress permits, with a coupling made out of a different material (e.g., plastic). To prevent heat flow from a hot turbine housing to a cold pump housing, material which
isolate between cryogenic
the housing by using the the two housings, or using fluid may
be recirculated
radial pin arrangement, a thermal barrier such
after
125
shutdown.
clamping insulating as a manifold through
3.2.4.7
ASSEMBLY
The turbopump design features assembly shall be compatible. If
the
turbopump
is designed
and
for
long
the procedures
service
life
and facilities
and
overhaul
access to all points. To avoid mistakes during reassembly, noninterchangeable fasteners. Make provisions for extracting, must
be separated
for
turbopump
capability,
provide
index components without damage,
easy
and use parts that
by force.
If extremely close clearances between the rotor and the housing must be maintained, consider using matched assemblies. Consider fabrication and maintenance when selecting the turbopump
materials.
use materials Assemble class-100 rocket
that
For
require
example, annealing
the turbopump 000, condition-C engine
for parts
in clean
turbopumps.
that
and reheat
are likely
treatment
after
a temperatureand room is recommended
Class
100 000
means
that
to require
weld
repair,
do not
welding.
dust-controlled environment. for the assembly of all types the
room
must
satisfy
one
A of
of two
criteria: (1) no more than 100 000 particles per cubic foot if they are 0.5 micron (#) and larger, or (2) no more than 700 particles per cubic foot if they are 5/_ and larger. Condition C means that the-temperature is 72 -+ 5 ° F. During assembly, check seal leakage rates, rotor axial and radial clearances, and bearing drag. Use locking devices to secure all bolts, nuts, and other fasteners. To avoid damage due to broken pieces, never twice. After assembly,
use lockwire within the turbopump and never use a locking device check the torque to ensure that there is no interference between the
rotor
and the housing
and that
3.2.5
System
Interfaces
3.2.5.1
PUMP INLET
The
pump
inlet
and
the bearing
upstream
and
seal drags
ducting
shall
are within
specifications.
enhance
the
pump
suction
performance. Minimize
losses
by minimizing
bends
and changes
in duct
cross section.
Use turning
vanes
in
elbows that are too close to the pump (within approximately 15 to 20 pump inlet diameters upstream of the pump). Consider tangential pump inlets if there is sufficient NPSH and if the overall result is a more compact unit. Use straightening vanes at the tank discharge to minimize
fluid rotation
reference
1 for more
at the pump detailed
inlet
information
and to prevent
fluid vortexing
on flow-distortion
126
effects
in the tank. and pump
Consult
inlet design.
3_2.5.2
PUMP DISCHARGE
The pumt_ shall meet
discharge connections and downstream ducting shall be leak tight and the requirements of light weight, minimum pressure drop, and ease of
servicing. Use a bolted O-rings for
flange to connect the pump discharge noncryogenic propellants and with
to the discharge pressure-actuated
line. Seal the joint with seals for cryogenic
propellants. If the flanges become disproportionately large and heavy (this may occur for small line sizes), consider welding the discharge line to the pump discharge. Use a diffuser at the pump discharge to minimize losses drop against line weight to obtain' the diffuser area more detailed
3.2.5.3
ratio to match pump information on pump
TURBOPUMP
The
turbopump
in the downstream optimum line size
discharge discharge
expansion large pad turbopumps
3.2.5.4
use close and with
mounts
shall support
coupled,
GAS-GENERATOR
The gas-generator
consider
and
weight
reference
10 for
and all loads applied
at one end of the turbopump other end. expansion
mounting
CONNECTION
connections
expansion
expansion
Consult
without causing unacceptable or the thrust chamber systems.
rigid pads
AND
shall be leak
deflections
and,
and
to accommodate
For cast structures, consider one and contraction. For very small
at the pump
discharge
flange.
MOUNTING tight
and shall adapt
to a high degree
and contraction:
Since the gas connection to the turbine turbine manifold and select its material thermal
the turbopump
contraction, ball joints at the integral keys to accommodate
with cast volutes,
of thermal
velocities.
MOUNTING
to the turbopump during operation distortions in either the turbopump In general,
and line lines.
plumbing. Trade line pressure and, therefore, the appropriate
contraction.
manifold also supports to withstand the added For
ease
of assembly
the gas generator, design the loads, including those due to and
disassembly,
use
bolted
flanges with pressure-actuated metallic seals at the connection point. To obtain greater reliability and lower cost, consider welded connections if the production rate is to be high. Use X-ray and penetrant detection methods to check the quality of the welds.
127
3.2.5.5
TURBOPUMP
The turbopump without removal Provide provide
for for
ON THE
ENGINE
system design shall provide for routine from the engine or major disassembly.
replacement of the seals without disturbing instrumentation replacement, leak checks,
possible,
provide
bearings, of rotors
rubbing, and and seals.
3.2.5.6
SERVICE
for
TURBOPUMP
The turbopump
hand
turning
worn
seals.
the
turbopump
Consider
the
servicing
the bearing inspections,
to permit
assembly. In addition, and torque checks. If
early
use of boroscope
and inspection
detection
ports
of damaged
to allow
inspection
OVERHAUL system
design
shall provide
for easy and inexpensive
overhaul.
Because overhaul
the turbopump replacement requires expensive engine recalibration, plan for rather than replacement. Plan for easy replacement of life-limited parts. When
deciding facilitate
whether repair,
to integrate parts to facilitate original manufacture or to separate parts to weigh the cost of original manufacture against the probability of failure. To
permit replacement of either unit without the turbine rotor assemblies to be balanced
3.2.6
The
the other,
try to design
the pump
and
Start Systems
start
system
shall
repeatability,-response, and turbine exhaust integrity. Consider leading
disturbing individually.
meet
the
vehicle
requirements
of start
time,
restart
time,
light weight, and simplicity," shall match the turbine type disposal method," and shall not endanger vehicle structural
main propellant tank-head and candidates to meet the combined
solid-propellant engine and
detailed trade study to make a final selection below in sections 3.2.6.1 through 3.2.6.4.
128
of a start
start-cartridge start vehicle requirements. system,
using
systems Conduct
the methods
as a
given
3.2.6.1
MAIN-PROPELLANT-TANK
The start start
system
transient
Consider
employing
without
this
start
propellant-expander
HEAD main-propellant-tank
undue system
drive
suitable
cycles.
Conduct
of a computerized mathematical control system requirements for simple ope.n-loop enough to cause transient occur,
tank pressure
model a main
head shall provide
(weight)
a satisfactory
penalty.
for
gas-generator,
a detailed
analysis
staged-combustion, of the engine
start
and
with
the aid
to determine the most satisfactory sequence and propellant tank-head start. For a GG cycle, use a
valve sequence system unless the propellant inlet pressure range is large wide variations in the engine start transient. If wide variations in the start provide closed-loop control of the GG oxidizer flow (controlling GG
temperature). For staged-combustion and expander turbine drive cycles, conduct a detailed computer analysis to determine if closed-loop control is required during the start transient. The start of a staged-combustion-cycle engine is more sensitive to transients than that of a GG-cycle
system,
3.2.6.2
and the need
PRESSURIZED-GAS
The pressurized provide
START
gas in the start
satisfactory
starts
Consider pressurized-gas start low-pressure-ratio (high-flow) Conduct determine
for closed-loop
under
control
should
tanks
shall be sufficient
all operating
tanks a suitable turbines used
mathematical
considered.
TANKS
model
in energy
and
quantity
starting method for GG cycles but on staged-combustion or expander
of the
engine
system mission.
system
is required
for repressurizing the start tank when long orbital Compensate for pressure variation with temperature
that, under normal environmental a relief valve or regulator that pressure-regulator Minimize
the
type number
possible; specify Class and visual inspection). transients.
of device of leak I welds Insulate
conditions, is set higher
temperature than the
or a narrow-band
paths
by combining
relief valve
not for cycles.
all operating requirements, to possible leakage of the stored for this analysis.
Provide the capability for increasing the gas storage pressure to a higher value than in the analysis in the event that testing indicates that additional energy is required. backup of the
to
conditions.
a detailed analysis of the start system, considering the required stored-gas energy. Account for any
gas. A computerized
be carefully
(and normal
coasts are a requirement by storing the gas such pressure) operating
valve for the functions.
that used Provide a
relief
will increase. Use level. Specify a system.
Use welded
joints
where
(i.e., welds that require radiographic, penetrant, dimensional, the tank and utilize reflective coatings to minimize thermal
129
3.2.6.3
LIQUID-PROPELLANT
The
liquid-propellant
unacceptable
start-tank
TANKS
system
shall
provide
satisfactory
for
systems
start
without
overpressures.
Liquid-propellant complexity,
START
start
tank
probably
are
systems
not
can
suitable
be used
for new
engine
GG
systems.
but,
because
However,
of system
if liquid-propellant
start tanks are suitable, control the start-tank pressurization rate to prevent overpressures in the GG feed system. A start-tank pressurization time of 3 seconds is satisfactory for most GG systems. rate. Conduct
Provide
a detailed
orificing
capability
heat-transfer
in the
system
and flow analysis
for
adjustments
of the start-tank
of pressurization
system
during
chilldown
to determine required flowrate through the system; size the fill and vent system accordingly. Under steady-state liquid conditions during fill, maintain some pressure (5 to 15 psig) in the start tank by restricting restriction experimentally. this purpose. Provide
a bleed
hole
the
in one
vent Design
system. If chilldown time is critical, the vent system oversize and include
of the
start-tank
isolation
check
valves.
determine the exact orifice provisions for
Determine
the hole
by experimentation, since the required size depends on volume of trapped propellant, of propellant, and environmental conditions. For ground-mounted start-tank systems, the bleed hole in the check valve in the GG supply line from the main feed system°
3.2.6.4
SOLID-PROPELLANT
The solid-propellant provide a satisfactory Consider
the
START
start-cartridge start transient
solid-propellant
start
thrust-chamber-tapoff cycles. Conduct considering all operating requirements,
size type place
CARTRIDGE combustion rate and propellant under all operating conditions. cartridge
a suitable
a detailed to determine
starting
analysis of the required
grain size shall
method
for
GG
and
the engine start system, solid-cartridge burn time
and flowrate. A computerized mathematical model of the engine system is required for this analysis. Control the grain temperature so that the start transient is within the safe operating range
of the
engine
and control installation.
of vehicle Conduct
temperature system less
variation sensitive
system.
Minimize
grain
temperature
variation
boattail temperature and by conditioning a detailed heat-transfer analysis to define
by
insulation,
coatings,
the start cartridge prior to the design for minimum
and to determine the range of grain temperatures. Make the turbine drive to grain temperature effects by utilizing a low-energy start cartridge
(providing acceleration to the 50 +- 10 percent level), thus allowing overlap of the power sources; then either provide closed-loop control to sense start-cartridge burnout and then turn on the GG, or use a check valve to isolate the start-cartridge turbine drive system from
130
the mainstage turbine drive system, thereby providing an automatic transition from start-cartridgepower to mainstagepower. If start-time repeatability is critical, utilize an electric heaterblanket to maintain the propellant grainat someconstanttemperaturehigher than the maximumexpectedenvironmentaltemperature. To avoid damagingdetonationsin the oxidizer feed system,avoid contamination by using (1) an inert-gaspurge through the oxidizer systemwhile the solid-propellantstart cartridge is burning, (2) a poppet-type injector designdiscussedin reference9,or (3) a check valve isolating the liquid-propellant GG from the solid-propellant start cartridge. To prevent thrust overshoot for cases in which the solid-propellant start cartridge ignites the liquid-propellant GG, provide pyrotechnic igniters in conjunction with the start-cartridge gasesto minimize the overlap and use the start-cartridgegasesonly for ignition of the igniters. Maintain the ignition sourcefor t to 2 seconds to ensure reliable ignition of the liquid
propellants.
3.3
DESIGN
3.3.1
3.3.1.1
EVALUATION
Engine-System
DESIGN-POINT
Characteristics
SYSTEM
The power requirements shall be consistent with engine system.
BALANCE
and output flow the requirements
characteristics of the turbopump system imposed during operation of the rocket
Determine the pump headrise, pump flowrate, and turbine performing a design-point balance of the turbopump system rocket engine system. Include in the system balance detailed engine
components.
pumps fulfilled
and by
for the
Write the power turbopump
steady-state available system
equations from during
for
the turbine. steady-state
available from the turbine is equal to the horsepower to be used when performing the power calculations engine thrust at the desired mixture ratio balance can be found in reference 35. The
design-point
system
engine-system requirements. turbopump-system design.
balance If
power The basic operation
whether
engine-system
131
and
flow
power is that
required
the
A detailed
example
turbopump
requirements
are
system not
by the
condition to be the horsepower
required by the pump. The is that required to produce
of the propellants.
determines the
the
flowrate requirements by when it is operating in the characteristics of all of the
met,
pump flow the design of a system
meets modify
the the
3.3.1.2
OFF-DESIGN
SYSTEM
The
turbopump
system
over
the entire
operating
Obtain
the
range
BALANCE
shall operate
reliably
range.
of planned
Determine
turbopump-system
the
operating
In order to assess properly the requirements, analytically determine other turbopump parameters. examples of which are shown and has
the 95 or 99.5
requirements
operation
from
the
engine-system
balance
tolerances from existing engines to determine components at the extremes of the planned
envelope
within
which
95 percent
effects of off-design operation the simultaneous changes in pump
of the engines
on system design head and flow and
Present this information in the form of operating ellipses, two in figure 40. Each figure shows three ellipses, marked 50, 95,
99.5 percent. These percentages of falling within these ellipses.
either
performance
range.
(sec. 2.3.1.1) and then use known component the effects of off-design engine and turbopump operating will fall.
and meet
percent
are the probabilities A turbopump system
that any normally
normal operating point should be designed to
limits.
Percent
58
x103
/
72
99.5
Percent
xlO 2
99.5 37
,_
36
_
5s
._
70 68
_
66
Q
62-
I
!
I
I
"85
87
89
91
Pump
(a)
Pump
flovrate,
head
vs
Figure
point
operating
Nominal point
53
5 operating
-Nominal
64
xlO 2
I
I
I
I
I
26.5
27
27.5
28
28.5
gpm
pump
40.
Turbine
flowrate
-
Probability
(b)
ellipses
132
for
turbopump
Turbine
operation.
speed,
power
xlO 3
rpm
vs
turbine
speed
Compute the ellipses by normal probability techniques. Treat : each of the performance-determining parameters of the components as an independent, randomly'distributed variable with a mean equal to its nominal value, and a standard deviationdeterminedfrom its tolerance.The standarddeviationsof eachengine-dependent variable should be calculated from a table of enginelinear influence coefficients and'the toleranceassignedto the engineindependentvariable.The bivariateprobability distribution shouldbe calculatedby firgt computing the correlationcoefficient betweenthe turbopump variablesbeing considered(ref. 36). Finally, the effectsof the predictablyvarying operating conditions (e.g.,the variation of pumpinlet pressureduring flight)should be determinedfor each dependent variable and added algebraically to the 95-percent range to yield the 95-percentlimiting valuesshown. If the turbopump designdoesnot meet the reliability andperformancerequirementsat the extremesof this envelope,modify the designandrecheckit. 3.3.1.3 The
CONTROL turbopump
engine,
CONSTRAINTS system
shall be simple,
Use system balances (sec. effects of control system
controls
shall
and shall cause
meet
the
minimal
2.3.1. l) and system types and locations
response
performance
requirements and weight
dynamic models (sec. 2.3.2) on pressure drops, control
of the penalties.
to determine the ranges, transient
responses, and stability. In addition, determine the corresponding effects of any variations in pump pressure requirements on turbopump weight and performance. Conduct tradeoffs of simplicity against resultant performance and weight penalties and, finally, select the optimum control requirements are
system. illustrated
staged-combustion
engine
3.3.2
3.3.2.1 The
System
Typical effects in figure 34
of complexity for a range of
on pump discharge-pressure engine mixture ratios for
a
cycle.
Dynamic
Analysis
START turbopump
support
system
the attainment
moment
of inertia
of the engine-system
To improve turbopump system reducing the turbopump-system
transient moment
and
low-speed
start-transient
torque
133
shall
objectives.
response, reduce the time of inertia and increasing
torque. To reduce the turbopump moment of inertia, with the limitations discussed in section 2.2.1. Also
capability
constant (eq. the low-speed
design at the highest select materials that
(15)) by turbine
speed consistent, have the highest
Strength-to-densityratios andareadequatefor the load andlife requirements.Increasingthe low-speed turbine torque will also reduce the turbopump time constant. To provide increasedturbine torque at low speedsfor GG rocket engine systemsthat start with availablepropellant tank head pressure,use a hot-gasvalvein seriesbeforethe turbine. This valvewill be open during enginestart to provideaddedturbine inlet pressure.At mainstage conditions, the hot-gasvalvewill be closedto achievethe steady-stateoperatingconditions of the engine. For systemshaving separateturbopumpsfor the oxidizer andfuel, the fuel turbopump time constant should never be greaterthan the oxidizer turbopump time constant.This helps preventtemperaturespikesduring enginestart. Checkthe steepnessof the head/flow curveof the pump.A steepcurvewill tend to stabilize the system.Sweepbackof the bladesin a centrifugalpumpwill producea steeperhead/flow curve. Use analytical digital or analog engine start models to verify stable turbopump operation during transient conditions. These models are generally for a specific engine configuration and caution should be used in trying to generalizeresults. Evaluate the structural adequacyof the turbopump designto withstand transient overloadsthat may occur during the start.
3.3.2.2
THROTTLING
Turbopump
throttling
control
systems
shall
meet
the
required
engine-system
response. Consider (thrust before
control
of the
turbine
working
fluid
for systems
requiring
moderate
response
rates
change up to 20 percent per second). Assess hot-gas valve technology carefully using hot-gas valves. Use throttling of the main liquid propellant valves when faster
ramp rates are needed (thrust change greater than 20 percent per second). Also reduce the turbopump system time constant (sec. 3.3.2.1) for better response to all commands. If the pump inlet flow coefficient becomes low enough during throttling to cause suction-performance
3.3.2.3
problems,
prewhirl
(ref.
37) or simple
flow recirculation:
SHUTDOWN
The pump
structural
design
Compute surge pressures at engine-system configuration, geometry.
consider
Design
the inlet
shall accommodate
the pump inlet, using valve characteristics,
to withstand
these surge
134
shutdown
surge
pressures.
mathematical models fluid characteristics, pressures.
that reflect the and suction-line
3.3.2.4
scope
INSTABILITY
LoW-frequency Investigate
To
verify
the
volume
system
change
per unit
shall be defined. pressure
change)
compliance
and
the
dynamic
of the turbopump system. Test the turbopump the operating fluid system with a measured
and
pressure) 33.
gain,
conduct
with a flight-type pulse of varying
/
System
3.3.3.1
TURBOPUMP
The
Development
turbopump
system
and
anticipated
extremes
Conduct
structural
testing. confirm
the
testing
of ac_ual
testing
the
that
are required
ensure
that
demonstrate
assembly engine
of certain
turbopump-system system,
shall
turbopump
Conduct hydrostatic the adequacy of the disks
Testing
SYSTEM
components
engine
(the
pump-cavitation
3.3.3
For
compliance
of the turbopump
partial derivative of discharge pressure with respect to inlet design following the methods recommended in reference
tests exciting
frequency.
turbine
characteristics
pump-cavitation
frequency-resp°nse inlet duct by
(POGO)
response
the dynamic gain (the for the turbopump-system
_e pl
dowl
SYSTEM
under
operating
testing design.
to operate
components of pump Conduct
near material
testing
conducted
the
instrumentation
adequacy that
of
critical
simulate
the
conditions.
turbopump
pressure structural
the conditions
before
volutes spinpit
prior
to turbopump-system
and turbine manifolds testing of pump rotors
to and
limits. integration
is adequate
of the turbopump to verify
pump
into and
the
turbine
performance. To do this, hold a formal design review of the entire turbopump system to ensure that the instrumentation requirements are met before making a hardware release. Critical measurements include fluid flowrates, pressures, and temperatures, and rotor speeds, accelerations, To
avoid
positions, engine-system
turbopump. range potential
and
To
failures
do this,
of engine-system
torques.
calculate operation.
due the
to turbopump extremes
Then
test
operation,
of turbopump the
turbopump
apply
limits
operation at these
testing
from
the
extremes
to the planned
to expose
problems.
_lianc_
If pump tests are conducted prior to the turbopump-system tests, consider variable-speed electric motor drive for low horsepower units. For high-horsepower may be necessary to use the same turbine as will be used on the engine. tests, use the suitable pump fluids from table VII so as to reduce propellant pump
tests,
use the design
fluid.
135
using pumps,
a it
For early pump costs. For later
Term
Definition
or Symbol
cryogenic
fluids
or conditions
at low temperatures,
usually
at or below
-150 ° C
•(222 ° R) D
diameter,
DB
blade
ft or in.
diffusion
factor DHIA
D
S
DN
specific
diameter,
fts/4/gpm
bearing
speed-capability
mm and rotational
drag pump
index,
the product
of bearing
bore
size (D) in
speed (N) in rpm
pump whose rotor consists of a disk with many short radial blades. The flow enters radially and is carried within the blade passages around the disk and is discharged
radially
EW
equivalent
weight,
Ibm
EWF
equivalent
weight
factor,
F
engine
GG
gas generator
thrust,
acceleration
g
1/2 , Ds = -Q,Iz
a port.
lbm/(lbm/sec)
lbf
due to gravity,
H
head
or headrise,
HP
shaft horsepower,
hp
h
specific
Btu/lbm
ZXhv
latent
I
rotating
mass moment
Is
specific
impulse,
i
fluid incidence
i/_
ratio
enthalpy, heat
through
ft/sec 2
ft or ft-blf/lbm
of vaporization,
Btu/lbm
of inertia
lbf-sec/lbm angle, deg
of fluid incidence
140
angle to impeller
blade
angle
Definition
Term or Symbol
inertance
the impeding effect fluid-filled conduit
J
converstion
k
thermal
KD N
empirically
derived
coefficient
for bearings
Kss
empirically
derived
coefficient
for seals
L/D
length-to-diameter
MR
mixture
N
rotational
NPSH
net positive
factor,
of fluid inertia
of oscillations
in a
J = 778 ft-lbf/Btu
conductivity,
ratio:
on transmission
Btu/(hr-ft-°R)
ratio ratio
speed,
of oxidizer
to fuel, MR = _Vo/gZf
rpm
Po - Pv suction
head,
ft or ft-lbf/lbm,
NPSH -
at pump
inlet
P
M S
specific
o/f
ratio
P
pressure,
psi
Pc
chamber
pressure,
Pv
vapor
Po
total
Pitot
pump
speed,
rpm-gpmV2/ft
of oxidizer
to fuel
pressure, pressure,
3A, N s -
psi
psi psi
pump in which a rotating liquid ring is created inside pressurized fluid is scooped from this ring by stationary ducted to the outside.
PL
stage payload,
PR
pressure
Q
volume
lbm
ratio flowrate,
gpm
141
a rotating drum; Pitot heads and
TermorSymbol R Rootes
Definition
reaction pump
rotary
pump
produce S
consisting
of
two
a positive-displacement
clearance
between
rotor
intermeshing
pumping
and casing,
cam-like
rotors
that
action
in.
NQ V2 a S
s;
suction
specific
speed, rpm-gpml/2/ft
3_, S s -
(NPSH) 3_
as
corrected
suction
specific
speed,
S's -
(1 - u2) _
Seq
torsional subject
stress shaft,
in a solid
shaft
of the
same
outside
diameter
as the
psi
SS
seal speed,
SSME
space shuttle
s
specific
stall
loss of pumping capability surface of the blades
T
temperature,
To
total
temperature,
T/C
thrust
chamber
Tq
torque,
in.-lbf
Tesla pump
pump similar to a centrifugal configuration, which consists energy to the flow by friction
TSH
thermodynamic
time constant
time for a variable
ft/sec main
entropy,
engine Btu/(lbnr-°R) as a result
of flow separation
on the suction
°R °R
suppression to reach
142
pump with the exception of the rotor of many closely spaced disks that add
head,
ft
63% of its final value
Term
or symbol
Definition
U
turbine
U/Co
isentropic
velocity
u
tangential
velocity,
ut
blade
VC
velocity
pitchline
velocity,
ft/sec J
vane pump
tangential
pump
velocity
consisting
(tip speed),
ft/sec
of a rotor
with
sliding
vanes
that
is mounted
housing
W
weight,
;v
weight
w
fluid velocity
Z
number
lbm flowrate,
lbm/sec relative
to blade
of blades
vapor volume factor
percent
at the pumping
for thermodynamic
gas specific incremental ratio
ft/sec
compounded
eccentric
A
ratio
of inlet
I?
efficiency
%)T.s
turbine
P
ratio
P
density,
o
blade
limit
suppression
head; blade
to discharge
tip diameter,
heat ratio change tip diameter
total-to-static
to inlet tip diameter,
lbm/ft 3 (ratio
143
Dtl 6 =-Dt2
efficiency
of inlet hub diameter
solidity
angle
of chord
length
to spacing)
Dh 1 u =---Dtl
in an
Term
or symbol
Definition
turbopump
time constant,
I _- = --TqN
Gin
_o
flow coefficient,
ref. to blade
tip speed,
_0= -tl t
stage head coefficient,
gH ref. to blade tip, ff =--
ut
SUBSCRIPTS act
actual
des
design
E
engine
f
fuel
h
hub
I
inducer
id
ideal
isen
isentropic
L
liquid
m
meridional
min
minimum
ms
mean station
o
oxidizer
P
pump
si (i = 1,2,3,etc)
specific
component
value for entropy
144
(fig. 35)
stmsc urrs stall
stall
T
turbine
T/C
thrust
TP
turbopump
t
blade
u
tangential
v
vapor
0
total
1
inlet
2
discharge;
chamber
tip component
exhaust
Identification
MateriaP
A-50
50/50
Armalon
trade
Bearium
B-10
blend
of N2I-I _ and UDMH
name of E. I. duPont
trade name bronze
of
CRES
corrosion-resistant
Freon
trade
GN:_
gaseous
Hastelloy
name
Bearium
Metals
Corp.
Teflon
(Rochester,
N.Y.)
for
leaded
steel
of E.I. duPont
nitrogen
Haynes-Stellite
C
Co. for glass-filled
Co. for a family
of fluorinated
hydrocarbons
per MIL-P-27401 Corp.
designation
for
a nickel-base
high-temperature
alloy hydrazine
N2H4,
1Additional
information
on
Plaza,
York,
in
Defense, Society
New
N.Y.;
Washington,D.C., for
Metals
(Metals
metallic
materials
MIL-HDBK-5B, Sept.
Park,
1971 OH),
; and
propellant herein
Metallic in
Metals
can
grade per MIL-P-26536 be
found
Materials Handbook
1961.
145
in
and (8th
the
1972
Elements ed.),
SAE for
Vol.
Handbook,
Aerospace
1 : Properties
SAE, Vehicle and
Two
Pennsylvania
Structures,
Selection
Dept.
of Metals,
of Am.
Material
Inconel
Identification
trade
625,718
names
IRFNA
inhibited
Kel-F
trade
of International
red fuming name
of
Nickel
nitric
Co. for austenitic
acid, propellant
3M Corp.
for
grade
nickel-base
alloys
per MIL-P-7254
'
a high-molecular-weight
polymer
of
chlor otrifluorethylene "K"
monel
LH2 LOX,
LO 2
trade
name
alloy
containing
liquid
hydrogen,
liquid
oxygen,
N2H4
hydrazine,
N204
nitrogen
RJ-1
ram grade
of International
propellant
tetroxide,
jet
copper,
propellant
propellant
fuel,
grade
per MIL-P-27201
per MIL-P-25508
grade
a high-density,
per MIL-P-26539
kerosene-like
high-energy
name
of General
Teflon
trade
name
of E.I. duPont
UDMH
unsymmetrical
321,347
austenitic
431
martensitic
2219
wrought
6061
wrought
41
age-hardenable
per MIL-P-26536
propellant
trade
Rene
for a wrought
and aluminum
grade
grade
Co.
hydrocarbon,
propellant
per MIL-F-25558
kerosene-base MIL-P-25576
RP-1
nickel,
Nickel
hydrocarbon
Electric
Co. for a polymer
nickel-chromium-iron
aluminum
grade
nickel-base
per
alloy
of tetrafluorethylene
propellant
grade
per MIL-P-25604
steels
steel alloy with copper
aluminum
propellant
Co. for an austenitic
dimethylhydrazine,
stainless
fuel,
alloy
with
as principal
magnesium
and
alloying silicon
element as principal
alloyil_g elements
Vehicles
Identification
and Engines
J
Agena
YLR81-BA-
11
engine for Agena upper stage; manufactured by Bell Aerospace
146
15 000 lbf thrust; uses IRFNA/UDMH; Company, Division of Textron
Vehicles
and Engines
Identification
Atlas
launch
A-7
engine system for Redstone; manufactured by Rocketdyne
Centaur
high-energy
vehicle
using MA-5 engine
upper
stage
for
system
78 000 Division, Atlas
lbf thrust; used North American
and
Titan;
uses
LOX/alcohol; Aviation, Inc.
two
RL10A-3-3
engines F-1
engine
for S-IC; 1 500 000 lbf thrust;
Rocketdyne H-1
engine
Division,
Rockwell
for S-IB; 205
uses LOX/RP-1
; manufactured
by
uses LOX/RP-1;
manufactured
by
uses LOX/LH2;
manufactured
by
designed
and
International
000 lbf thrust;
Corp.
Rocketdyne J-2
engine for S-II; Rocketdyne
J-2S
uprated
J-2;
developed
230 000
265000
lbf
lbf
thrust;
thrust;
uses
LOX/LH2;
by Rocketdyne
Aerojet engines for first stage of the Titan • the -3 uses LOX/RP-1, and develops • the -5, -7, -9 uses N204/A-50
vehicles 150 000 lbf thrust
, and develop
215 000 lbf thrust
LR-91-A J-3, -5, -7, -9
Aerojet engines for the second stage of the Titan vehicles • the -3 uses LOX/RP-1, and develops 80 000 lbf thrust • the -5, -7, -9 use N204/A-50 , and develop 100 000 lbf thrust
MA-5
five-engine 1 sustainer
system for Atlas vehicle containing 2 booster, 2 vernier, and engines; boosters provide 330 000 to 370 000 lbf thrust;
sustainer, 57 000 by Rocketdyne MB-3
engine
system
manufactured
to 60 000
for Thor
lbf thrust;
vehicle;
uses LOX/RP-1
170 000
lbf
thrust;
; manufactured
uses
LOX/RP-1;
by Rocketdyne
Redstone
early
RL I 0A-3-3
engine for Centaur; 15 000 lbf thrust; uses LOX/LH 2 ; manufactured Pratt & Whitney Aircraft Division of United Aircraft Corp.
Saturn
launch
V
launch
vehicle
vehicle
using A-7 engine
for Apollo
147
manned
system
mission
to the moon
by
Vehicles andEngines
Identification
S-IB
booster
S:IC
first
using a cluster stage
(booster)
of eight of
the
H-1 engines Apollo
Saturn
V vehicle;
uses
five
F-1
engines S-II
second
stage
of the Apollo
Saturn
V vehicle;
uses a cluster
of five J-2
engines S-IVB
third
Thor
launch
Titan
X-8
I, II, III
stage of the Apollo vehicle
Saturn
using MB-3 engine
family
of launch
rocket
engines
vehicles
developed
using
V vehicle; system the
by Aerojet
experimental throttleable rocket LOX/LH2 ; developed by Rocketdyne
148
uses a single J-2 engine
LR-87-AJ Liquid engine;
and LR-91-AJ
Rocket 90000
series of
Co. lbf
thrust;
uses
REFERENCES .
Anon.:
Liquid
SP-8052, 2.
Anon.:
Anon.:
Engine
Turbopump
Inducers.
Space Vehicle
Design
Criteria
Monograph,
NASA
Design
Criteria
1971.
Liquid
Monograph, 3.
Rocket
May
Rocket
NASA
Liquid
Engine
SP-8109,
Rocket
Centrifugal
December
Engine
Axial
Flow
Turbopumps.
Space
Vehicle
1973. Flow Turbopumps.
Space Vehicle
Design
Criteria
Monograph
(to
be published). 4.
Anon.: January
Liquid 1974.
Rocket
Engine
Turbines.
5.
Anon.:
Liquid
Rocket
Engine
Turbopump
SP-8100, 6.
Anon.:
March
Rocket
March
Gears.
Engine
Turbopump
Turbopump
8.
Anon.:
Turbopump
10.
* 11.
Anon.: March
Liquid
Liquid 1972.
13.
Propellant
September
Balje,
O.E.:
Performance
A Study
Flexible
Discharge
American
A. J.: Centrifugal
Pump
Klassen,
Vehicle
Monograph,
Design
NASA
SP-8110,
Criteria
Monograph,
NASA
Space Vehicle
Design
Criteria
Monograph,
NASA
on Design and Matching
Criteria
Shafts
and
Space Vehicle
Tubing,
Flow
Aviation,
and Axial
Rotating-Shaft
Seals.
Couplings.
Space
Vehicle
Design
Criteria
Space
Vehicle
Design
Criteria
1972.
Gas Generagors.
Pump
Div., North
Stepanoff,
84, 1962, 14.
Engine
SP-8101,
A. E.: Induced
Rocketdyne 12.
Rocket
NASA
Anon.: Liquid Rocket Lines, Monograph (to be published). Marks,
Criteria
1971.
Anon.: Liquid Rocket Engine Monograph (to be published).
9.
Design
Space
Bearings.
7.
Monograph,
Vehicle
1974.
Liquid
SP-8048,
Space
Design
Bellows,
Oscillations
and
Filters.
and Matching
NASA
Vehicle
SP-8081,
Design
H-Q Operation.
Criteria
LAP 67-100,
1967.
John Wiley & Sons,
of Turbocomponents.
MOnograph,
Space
Due to Position
Inc., unpublished,
Flow Pumps.
Criteria
Inc.,
of Turbomachines: J. Eng. Power,
1957.
Part Trans.
B-
Compressor
ASME,
Series
and A, vol.
pp. 103-114. H. A.:
Mean-Diameter
Cold-Air
Investigation
Single-Stage
Axial
of Effects
Flow Turbine.
of Partial NASA
Admission
TN D-4700,
on Performance
August
*Dossier for design criteria monograph "Turbopump Systems for Liquid Rocket Engines." Unpublished. material available for inspection at NASA Lewis Research Center, Cleveland, Ohio.
149
of 3.75-Inch
1968. Collected source
v
15.
Stenning,
A. H.: Design
79, Dynamic
of Turbines
Anal. and Control
for High-Energy-Fuel,
Lab.,
Mass. Inst.
Tech.,
Low-Power-Output September
Applications.
1953.
16.
Balje, O.E.: A Study on Design Criteria and Matching of Turbomachines: Relations and Design Criteria of Turbines. J. Eng. Power, Trans. ASME, Series 83-102.
17.
Wong, G. Turbopumps-
S.;
and Second
Wagner, Interim
W. R.: Thermodynamic Report. R-7585, Rocketdyne
Report
Improvements Div., North
Part A- Similarity A, vol. 84, 1962, pp.
in Liquid Hydrogen American Aviation, Inc.,
July 1967. 18.
"19.
Huzel,
D. K.;
SP-125,
1971.
Dinges,
R. A.:
and
Huang,
Mark
Inc., unpublished,
D. W.: Design
19 Final
August
Report.
MacGregor, C. A.; and Csomor, Rocket Engines. NASA CR-72965,
21.
Hoshide,
R. K.; and Nielson,
Report. 1972.
NASA
CR-120815,
Gelder,
T.
Ruggeri,
Measured D-3509, 23.
Bissell,
F.;
Pressure July
TDR-3117-3011,
A.: Rotating Rocketdyne
C. E.: Study
R.
and
Propellant
Rocket
Rocketdyne
Engines.
ed., NASA
American
Aviation,
Div., North
Second
1963.
20.
22.
of Liquid
and Positive Displacement Div., Rockwell International
of Blade
R-8806,
Rocketdyne
S.;
Moore,
and
Temperature
Clearance Div.,
R. D.:
Depressions
Effects
North
Cavitation
for Low-Thrust February 1974.
on Centrifugal
American
in Cavitated
Pumps Corp.,
Rockwell
Similarity Regions
Pumps Corp.,
Considerations of
- Final
November
Based
Freon-114.
on
NASA
TN
Pumps
for
1966.
W. R.; Wong,
O2/H 2 Rocket
G. S.; and Winstead,
Engines.
J. Spacecraft
T. W.: Analysis
Rockets,
24.
King, North
J. A.: Design of Inducers for Two-Phase American Rockwell Corp., July 1970.
25.
King,
J. A.:
North
American
Design
of Inducers
Rockwell
Corp.,
Operation
for Two-Phase January
of Two-Phase
vol. 7, no. 6, June
Oxygen
- Final
- Final
Flow
in
LH2
pp. 707-713.
Report.
R-8283,
Rocketdyne
Div.,
Report.
R-8832,
Rocketdyne
Div.,
1972.
26.
Wong, G. S.: High Pressure Pumping Technology Aviation, Inc., October 1964. (Confidential)
(U).
27.
Young,
of Hydrostatic
W. E.; and Reddecliff,
1970,
J. M.: Investigation
Cryogenic TurbopumpsFinal Report (U). United Aircraft Corp., May 1967. (Confidential)
AFRPL
R-5833,
TR-67-130,
Rocketdyne
Bearings Pratt
Div., North
for Use in High Pressure & Whitney
*Dossier for design criteria monograph "Turbopump Systems for Liquid Rocket Engines." Unpublished. material available for inspection at NASA Lewis Research Center, Cleveland, Ohio.
150
American
Aircraft
Div.,
Collected source
q
28. Anon.:
Final Program R-5652-4P, Rocketdyne
29. Lieblein, Limiting
30. Anon.:
S.; Schwenk, Blade
31. Anon.: (Pogo).
32. Anon.: Analysis
33. Anon.:
H-1 Engine R-6813,
34. Anon.: Criteria
Diffusion
Design
Corp., System
System
Investigation
Broderick,
R.
in Axial-Flow-Compressor
Rocketdyne
Program.
Rocketdyne
J. C.; and
Engine
Aircraft
Engine
Engineering, Task I, Turbomachinery Inc., May 1965. (Confidential)
Loadings
Advanced
Div., United
Progress Report for Product Div., North American Aviation,
July
Functions
Div., North
Prevention
of
NASA
Coupled
Loop
Rockwell
for Estimating
NACA
RM-E53D01,
FR-2488
Structure-
SP-8055,
October
Inc., December
of S-V Vehicle
American
Instability Corp.,
August
Propulsion
Losses
and
1953.
,
Pratt & Whitney
Aircraft
Longitudinal
Stability
1966.
Longitudinal
Aviation,
Stability
Inc., March
on S-II Stage
(Pogoa
1967.
of Saturn
V. R-7970,
1969. Instability
(Pogo).
Space
Vehicle
Design
1970.
35. Wieseneck, Model
Factor
of S-IVB Vehicle
Aviation,
for Support
Div., North
17 Hz Closed American
Monograph,
PWA Report
for Support
American
Functions
Rocketdyne
of
Div., North
Bell (AEB).
Elements.
1967.
Transfer
Transfer
R-6929,
Study,
L.: Blade
(U).
H. C.: RSS-8314-1, Space Shuttle Main Program. R-8892P (Enclosure 1), Rocketdyne
Engine, Phase B, Engine System Div., North American Rockwell
A.F2
McGraw-Hill
Performance Corp., April
1971. Mood,
Introduction
Wong, G. S.; Macgregor, Throttled Turbopumps.
to the Theory
of Statistics.
C. A.; and Hoshide, J. Spacecraft Rockets,
Book
Co. (New
York),
1950.
R.K.: Suppression of Cavitation and Nonstable vol. 2, no. 1, Jam-Feb. 1965, pp. 73-80.
151
Flow
in
C-I
NASA SPACE VEHICLE DESIGN CRITERIA MONOGRAPHS ISSUED TO DATE
ENVIRONMENT SP-8005
Solar Electromagnetic
SP-8010
Models
of Mars Atmosphere
SP-8011
Models
of Venus
SP-8013
Meteoroid March
Radiation,
(1967),
Atmosphere
Environment
Revised
May 1971
May 1968
(1972),
Revised
Model-1969
September
(Near
Earth
1972
to Lunar
Surface),
1969
SP-8017
Magnetic
Fields-Earth
and Extraterrestrial,
SP-8020
Mars Surface
SP-8021
Models
SP-8023
Lunar
SP-8037
Assessment
and Control
of Spacecraft
SP-8038
Meteoroid Environment October 1970
Model-1970
SP-8049
The Earth's
SP-8067
Earth
SP-8069
The Planet
SP-8084
Surface Revised
SP-8085
The Planet
Mercury
SP-8091
The Planet
Saturn
SP-8092
Assessment June 1972
and
SP-8103
The Planets
Uranus,
SP-8105
Spacecraft
Models
of Earth's Surface
(1968),
May 1969
Atmosphere
(90
Models,
March
and Emitted
Jupiter
(1970),
Atmospheric June 1974
km),
Revised
March
1973
July
December
1971
Control,
153
(Launch
March June
September
1970
and Planetary),
1971
and
Transportation
Areas),
1972
1972
of Spacecraft
Neptune,
Fields,
(Interplanetary
Radiation,
(1971),
Control
Magnetic
1971
Extremes
(1970),
Thermal
to 2500
1969
May 1969
Ionosphere,
Albedo
March
and Pluto May 1973
Electromagnetic
(1971),
November
Interference,
1972
STRUCTURES SP-8001
BuffetingDuringAtmospheric Ascent,Revised November 1970
SP-8002
Flight-Loads Measurements DuringLaunchandExit, December 19_4
SP-8003
Flutter,Buzz,andDivergence, July1964
SP-8004
PanelFlutter,Revised June1972
SP-8006
LocalSteadyAerodynamic LoadsDuringLaunchandExit, May1965
SP-8007
BucklingofThin-Walled CircularCylinders, Revised August1968
SP-8008
Prelaunch GroundWindLoads,November 1965
SP-8009
Propellant SloshLoads,August1968
SP-8012
NaturalVibrationModalAnalysis, September 1968
SP-8014
EntryThermal Protection, August1968
SP-8019
BucklingofThin-Walled Truncated Cones, September 1968
SP-8022
Staging Loads,February1969
SP-8029
Aerodynamic andRocket-Exhaust HeatingDuringLaunchandAscent May1969
SP-8030
Transient LoadsFromThrustExcitation,February1969
SP-8031
SloshSuppression, May1969
SP-8032
Buckli:,gof Thin-Walled DoublyCurvedShells, August1969
SP-8035
WindLoadsDuringAscent,June1970
SP-8040
FractureControlofMetallicPressure Vessels, May1970
SP-8042
Meteoroid Damage Assessment, May1970
SP-8043
Design-Development Testing, May1970
SP-8044
Qualification Testing,May1970
SP-8045
Acceptance Testing,April 1970
SP-8046
LandingImpactAttenuationfor Non-Surface-Planing Landers,April 1970 154
SP-8050
StructuralVibrationPrediction, June1970
SP-8053
NuclearandSpace Radiation EffectsonMaterials, June1970
SP-8054
Space Radiation Protection, June1970
SP-8055
Prevention of CoupledStructure-Propulsion Instability(Pogo),October 1970
SP-8056
FlightSeparation Mechanisms, October1970
SP-8057
StructuralDesignCriteriaApplicable to aSpace Shuttle,Revised March 1972
SP-8060
Compartment Venting,November 1970
SP-8061
Interaction withUmbilicals andLaunchStand,August1970
SP-8062
EntryGasdynamic Heating, January1971
SP-8063
Lubrication, Friction,andWear.June 1971
SP-8066
Deployable Aerodynamic Deceleration Systems, June1971
SP-8068
Buckling Strengthof Structural Plates,June1971
SP-8072
AcousticLoadsGenerated by thePropulsion System, June1971
SP-8077
Transportation andHandling Loads,September 1971
SP-8079
Structural Interaction withControlSystems, November 1971
SP-8082
Stress-Corrosion Cracking in Metals, August1971
SP-8083
Discontinuity Stresses inMetallicPressure Vessels, November 1971
SP-8095
PreliminaryCriteria for the FractureControl of SpaceShuttle Structures, June1971
SP-8099
Combining AscentLoads,May1972
SP-8104
StructuralInteractionWith Transportation andHandlingSystems, January1973
GUIDANCE ANDCONTROL SP-8015
Guidance andNavigation forEntryVehicles, November 1968
155
SP-8016
Effectsof StructuralFlexibilityonSpacecraft ControlSystems, April 1969
SP-8018
Spacecraft Magnetic Torques, March1969
SP-8024
Spacecraft Gravitational Torques, May1969
SP-8026
Spacecraft StarTrackers, July1970
SP-8027
Spacecraft RadiationTorques, October1969
SP-8028
EntryVehicleControl,November 1969
SP-8033
Spacecraft EarthHorizonSensors, December 1969
SP-8034
Spacecraft MassExpulsion Torques, December 1969
SP-8036
Effectsof StructuralFlexibilityon LaunchVehicleControlSystems, February1970
SP-8047
Spacecraft SunSensors, June1970
SP-8058
Spacecraft Aerodynamic Torques, January1971
SP-8059
Spacecraft Attitude ControlDuringThrustingManeuvers, February 1971
SP-8065
TubularSpacecraft Booms(Extendible, ReelStored),February 1971
SP-8070
Spaceborne DigitalComputer Systems, March1971
SP-8071
Passive Gravity-Gradient LibrationDampers, February 1971
SP-8074
Spacecraft SolarCellArrays,May1971
SP-8078
Spaceborne Electronic ImagingSystems, June1971
SP-8086
Space VehicleDisplays Design Criteria,March1972
SP-8096
Space VehicleGyroscope Sensor Applications, October1972
SP-8098
Effectsof StructuralFlexibilityon Entry VehicleControlSystems, June1972
SP-8102
Space VehicleAccelerometer Applications, December 1972
156
CHEMICAL PROPULSION SP-8087
LiquidRocketEngineFluid-Cooled Combustion Chambers, April 1972
SP-8109
LiquidRocketEngineCentrifugal FlowTurbopumps, December1973
SP-8052
LiquidRocketEngineTurbopump Inducers, May1971
SP-8110
LiquidRocketEngineTurbines, January1974
SP-8081
LiquidPropellant GasGenerators, March1972
SP-8048
LiquidRocketEngineTurbopump Bearings, March1971
SP-8101
LiquidRocketEfigineTurbopumpShaftsandCouplings, September 1972
SP-8100
LiquidRocketEngineTurbopump Gears, March1974
SP-8088
LiquidRocketMetalTanksandTankComponents, May1974
SP-8094
LiquidRocketValveComponents, August1973
SP-8097
LiquidRocketValveAssemblies, November 1973
SP-8090
LiquidRocketActuatorsandOperators, May1973
SP-8080
LiquidRocketPressure' Regulators, ReliefValves,CheckValves,Burst Disks,andExplosive Valves, March1973
SP-8064
SolidPropellant Selection andCharacterization, June1971
SP-8075
SolidPropellantProcessing Factorsin RocketMotorDesign,October 1971
SP-8076
SolidPropellant GrainDesign andInternalBallistics, March1972
SP-8073
SolidPropellant GrainStructuralIntegrityAnalysis, June1973
SP-8039
SolidRocketMotorPerformance Analysis andPrediction, May1971
SP-8051
SolidRocketMotorIgniters,March1971
SP-8025 SP-8041
•SolidRocketMotorMetalCases, April 1970 Captive-Fired Testingof SolidRocketMotors,March1971 •It,U.S.
GOVERNMENT
157
PRINTING
OFFICE:
1975--635-049/73
Related Documents
Turbopump Systems For Liquid Rocket Engines
November 2019 8
Liquid Rocket Engine Turbopump Gears
November 2019 12
Liquid Rocket Engine Turbopump Inducers
November 2019 13
Liquid Rocket Engines Centrifugal Flow Turbo Pumps
November 2019 8
Russian Liquid Propellant Engines
November 2019 7