m
APTD-1290
AUTOMOTIVE GAS
TURBINE OPTIMUM
CONFIGURATION STUDY
U.S. ENVIRONMENTAL PROTECTION AGENCY
Office of Air and Water Programs
Office of Mobile Source Air Pollution Control
Advanced Automotive Power Systems Development Division
Ann Arbor, Michigan 48105
-------
APTD-1290
AUTOMOTIVE GAS TURBINE
OPTIMUM CONFIGURATION
STUDY
Prepared by
E. S. Wright, L. E. Greenwald, R. R. Titus
United Aircraft Research Laboratories
East Hartford, Connecticut 06108
Contract No. 68-04-0013
Project Officers:
T. M. Sebestyen (EPA), A. W. Nice (NASA)
Prepared for
U.S. EWIRONMENTAL PROTECTION AGENCY
Office of Air and Water Programs
Office of Mobile Source Air Pollution Control
Advanced Automotive Power Systems Development Division
Ann Arbor, Michigan 48105
May 1972
-------
[FINAL REPORT TITLE PAGE REVERSE SIDE FORMAT]
The APTD (Air Pollution Technical Data) series of reports is issued by
the Office of Air Quality Planning and Standards, Office of Air and
Water Programs, Environmental Protection 'Agency, to report technical
data of interest to a limited number of readers. Copies of APTD reports
are available free of charge to Federal employees, current contractors
and grantees, and non-profit organizations - as supplies permit - from
the Air Pollution Technical Information Center, Environmental Protection
Agency, Research Triangle Park, North Carolina 27711 or may be obtained,
for a nominal cost, from the National Technical Information Service,
5285 Port Royal Road, Springfield, Virginia 22151.
This report was furnished to the U.S. Environmental Protection Agency
by United Aircraft Research Laboratories in fulfillment of Contract
No. 68-04-0013 and has been reviewed and approved for publication by
the Environmental Protection Agency. Approval does not signify that
the contents necessarily reflect the views and policies of the agency.
The material presented in this report may be based on an extrapolation
of the "State-of-the-art". Each assumption must be carefully analyzed
by the reader to assure that it is acceptable for his purpose.
Results and conclusions should be viewed correspondingly. Mention
of trade names or commercial products does not constitute endorsement or
recommendation for use.
Publication No. APTD - 1290
ii
-------
FOREWORD
This report describes a comprehensive analysis of gas turbine engines
designed for automotive propulsion. The study was performed by the United Aircraft
Research Laboratories under contract to the Environmental Protection Agency, Office
of Air Programs, Department of Advanced Automotive Power Systems Development
(Contract Ho. 68-0^-0013).
The Contractor's study team involved the participation of the United Aircraft
Research Laboratories (UARL), Pratt and Whitney Aircraft Division (P&WA),
United Aircraft of Canada Limited (UACL), and Hamilton Standard Division (BSD),
all of United Aircraft Corporation. Additionally, several outside organizations
representing potential vendors of selected critical components assisted in
the study on a voluntary no-cost basis. Program management, economic analysis,
and mission analysis were provided by UARL. Engine cycle analysis was provided
by P&WA, base-line technology and preliminary design by UACL, and fuel control
design by HSD.
Within UARL, the program was managed by Edward S. Wright'. Vehicle
performance and mission analysis were performed by Dr. Larry E. Greenwald,
transmission analysis and vehicle integration by Richard R. Titus, and
manufacturing cost studies by W. Richard Davison.
The program monitor for EPA was Mr. Thomas Sebestyen, and technical
direction was provided by Mr. Arno W. Nice, of the NASA-Lewis Research Center.
Ill
-------
L-9712U9-7
Report L-9712H9-7
Automotive Gas Turbine
Optimum Configuration Study
TABLE OF CONTENTS
Page
SUMMARY 1
RESULTS AND CONCLUSIONS 2
RECOMMENDATIONS U
INTRODUCTION 5
ESTABLISHMENT OF BASE-LINE TECHNOLOGY 9
General 9
Compressors 11
Turbines IT
Stress Data 25
Turbine Blade Cooling 29
Engine Loss Data 33
Heat Exchangers 33
Transmissions 38
CYCLE ANALYSIS AND PRELIMINARY CANDIDATE SELECTION 63
Engine Cycles Evaluated 63
Selection Criteria 67
DESIGN 77
Engines 77
Power Trains 88
Installation Concepts 95
Fuel Control 110
IV
-------
TABLE OF CONTENTS (cont'd)
Page
ENGINE PERFORMANCE 139
Program Description 139
Output 139
Selection of Engine Operating Line li+5
Emissions Characteristics
VEHICLE EVALUATION 153
Transmission Modeling 153
Engine Sizing 159
Acceleration Performance 163
Fuel Economy and Emissions l8l
ECONOMIC ANALYSIS 205
Direct Manufacturing Cost Estimates 205
Cost of Capital 213
Cost of Fuel 2lU
Cost of Repairs and Maintenance 215
Results 219
RECOMMENDED ENGINE SELECTION 221
Emissions 221
Initial Price and Production Cost Uncertainties 223
Fuel Economy 227
Size and Weight 230
Reliability and Maintainability 230
Technological Risk 231
Development Time and Costs 231
Other Factors 232
DEVELOPMENT PROGRAM 235
Demonstration Program 235
Future Production 238
-------
TABLE OF CONTENTS (cont'd)
Page
REFERENCES 2^1
APPENDIX I - VEHICLE DYNAMICS WITH GE HYDROMECHAHICAL TRANSMISSION 2^5
APPENDIX II - VEHICLE DYNAMICS WITH BORG-WARNER AUTOMATIC 255
APPENDIX III - COMPUTER PROGRAM DESCRIPTION 265
APPENDIX IV - HISTORIC ENGINE-RELATED OWNERSHIP COSTS 271
APPENDIX V - DESIGN OPTIONS FOR IMPROVED FUEL ECONOMY 279
-------
Report L-9712^9-7
Automotive Gas Turbine_
Optimum Configuration Study
SUMMARY
A wide variety of candidate S&s turbine cycles was evaluated for application
to automobile propulsion. Simple cycles, regenerated and recuperated cycles,
cycles including intercooling and reheat, single-shaft and free-turbine
engines, and a combined gas turbine and Raiakine cycle were evaluated over a, range
of pressure ratios and turbine inlet temperatures for an initial total of 60
separate candidate cycles. Following a preliminary evaluation on the basis of total
lifetime costs of all designs, three leading candidates were chosen for
more detailed analysis — a simple cycle, a regenerated cycle, and a recuperated
cycle, all of which were single-shaft designs.
The selected engines were subjected to a more detailed preliminary design
analysis which included a. comprehensive study of fuel control, transmission,
and installation considerations. The installed engines were analyzed on the
basis of vehicle performance over a number of driving cycles for two candidate
transmission types. In these analyses, the acceleration performance, fuel
economy, and emissions characteristics were calculated. A direct manufacturing
cost estimate was made as part of an overall economic analysis in which total
lifetime engine-related automobile ownership costs were estimated. Also,
estimates were made of the probability of achieving 1976 federal emissions
standards, probable automobile first cost, and probable vehicle fuel economy.
These estimates were compared with estimated values for both present and future
emissions-treated Otto-cycle engines.
A broad outline of a development plan was generated for one of the selected
engines, the single-shaft, simple-cycle configuration, and some of the problems
associated with achieving a demonstration of this engine by 1975 are discussed.
-------
L-97121J9-7
RESULTS AMD CONCLUSIONS
1. In an engine selection process which considered 60 potential gas turbine
cycles, three engines were selected for detailed analysis. These were:
(l) A single-shaft simple-cycle gas turbine (designated SSS-10)
(2) A single-shaft regenerated gas turbine (designated RGSS-6)
(3) A single-shaft recuperated gas turbine (designated RCSS-8)
2. The SSS-10 has the best potential of meeting the 1976 Federal HOX emission
standards but careful combustor development is required to establish the fact. Heat
exchanger engines inherently pose a more difficult problem from the IOX emission
standpoint and the combustor development program would be correspondingly greater.
Federal standards for carbon monoxide (CO) and unburned hydrocarbons (UHC) are
expected to be met by all three of the selected engines.
3. The single-shaft engines recommended require an advance in transmission
capability to provide adequate driving characteristics. Several potentially eligible
concepts are under study. It is important to ensure that engine and transmission
developments proceed concurrently.
U. The gas turbine engines analyzed can be produced at a cost which would
result in an automobile selling price on the same order as today's cars, and con-
siderably lower than for a 1976 automobile powered with an emissions-treated Otto-
cycle engine (OC-76). The SSS-10-powered automobile is expected to have the
lowest cost.
5. Total engine-related lifetime cost predictions for 105,000 miles of opera-
tion are approximately the same for all three selected gas turbine engines. The
lower capital and maintenance costs associated with the simple-cycle engine
compensate for its higher fuel cost when compared with heat-exchanger engines.
6. The heat-exchanger engines exhibit superior fuel economy when compared
with either the simple-cycle gas turbine engine or emissions-treated Otto-cycle
engines. While the simple-cycle engine is predicted to fall short of the EPA goal
of 10 miles per gallon over the Federal Driving Cycle, it may compare favorably with
the emission treated OC-76. It is reasonable to postulate that future refinements
of the first generation turbine-transmission combination will effect substantial
improvements in fuel economy.
7. All three selected engines are compatible, in terms of size and weight,
with the installation requirements of standard-sized automobiles. The simple-cycle
engine has the greatest flexibility, due to its relatively small size and weight.
-------
L-9712^9-7
8. The design of the selected engines is based on advanced, but demonstrated,
technology. The resulting high specific output (power-to-airflow ratio) leads to
small size of critical components and consequently requires a minimum of high-cost
materials. 'The high engine shaft speed associated with these high-work components
is consistent with demonstrated bearing technology.
9. Of the.three gas turbine cycles selected for detailed study, the SSS-10
offers the highest probability of meeting the Office of Air Programs goals for
emissions and performance, and capable of being developed by 1975- The base-line
engine is a simple-cycle, single-shaft engine rated at 130 hp, under standard
conditions, utilizing a high-efficiency single-stage centrifugal compressor capable
of a pressure ratio of 10:1, and a single-stage radial inflow turbine. Its per-
formance is based on state-of-the-art technology and, while improvement in emissions
is required relative to estimated levels based on current state of the art, it is
highly probable that these can be achieved by 1975 if adequate programs are imple-
mented.
10. A meaningful demonstration of the SSS-10 could be accomplished by 1975.
The necessary concurrent transmission program requires further definition.
11. The "normal" period for introducing a novel concept into large-scale
automotive production is on the order of 10 years. With sufficient impetus and
support this time could probably be materially shortened. Again, the definition
of such a program requires additional study.
-------
L-971249-7
RECOMMENDATIONS
1. It is strongly recommended that a high-priority program be initiated
for the development and demonstration of the simple-cycle, single-shaft engine
(SSS-10) defined herein. This recommendation is based on the prime result of
this study that the SSS-10 is clearly the best solution for a low-emissions
automobile powerplant, this engine offering the highest probability of meeting
all systems requirements, including low cost,'low pollution, and satisfactory
fuel economy.
2, Concurrently with the engine program, a parallel study, with a subsequent
development program, is recommended to define and demonstrate the optimum
transmission(s) for the single-shaft engines defined herein.
3. It is recommended that a parallel engine development program be initiated
for the development and demonstration of the ECSS-8 recuperated gas turbine,
defined herein, to gain the advantage of superior fuel economy for those
applications (such as trucks and buses) in which this characteristic is of
prime importance, and also to allow for the possibility that an extraordinary future
increase in fuel price may add significance to fuel economy. This program
can have a somewhat lower priority than that for the recommended automobile
engine because of the fewer numbers of applicable vehicles and consequent
lesser pollution problems » as well as greater technological risk.
-------
INTRODUCTION
Gas turbines for automotive applications have been considered for many
years but, up until recently, they have been rejected as either too costly
to produce or too expensive to operate. However, with increased emphasis on
air pollution, their favorable emissions characteristics have stimulated more
intensive investigation.
One such study was reported in Eef. 2s which was intended to determine if the
application of new technology could result in a cost-competitive solution as
compared with the cleaned-up Otto-cycle engine. In that study, a simple-cycle
gas turbine was found capable of low emissions, superior performance (i.e., lower
weight and volume for a given power output), adequate fuel economy, and competitive
initial and lifetime operating costs. The significant technological innovation
which contributed to this result was a, highly efficient centrifugal compressor
which, in a single stage, could produce pressure ratios high enough to achieve
a cycle efficiency sufficient to avoid the necessity of regeneration, a feature
which had been largely responsible for the high cost associated with the traditional
approach to gas turbine design.
The simple-cycle design approach resulted in a small engine, relatively
insensitive to design tolerances, which offered the possibility of low material
and labor costs. While adequate substantiation was offered in Ref. 1 to
lend confidence in the results, the limited scope of that study did not permit
extensive optimization of the simple-cycle engine or a comprehensive comparison
with all of the many gas turbine cycle options available. Consequently, a
more detailed optimization of gas turbine alternatives, represented by the present
study, was needed.
Accordingly, the objectives of this study were (l) to define the optimum
gas turbine engine(s) capable of meeting the 1976 federal emissions standards and
of being developed by the year 1975, and (2) to outline the research and development
programs necessary to develop and demonstrate a selected optimum engine(s) by
1975 and mass-produce vehicles powered by such a low-emission engine at the
earliest possbile date after 1975-
The study was organized into three sequential phases as summarized in Fig. 1.
Within each phase, specific tasks are noted, but these were not necessarily time
sequenced as shown, there being a strong requirement for iteration to account for
interdependencies among the tasks. Phase I required the establishment of state-
of-the-art technology to permit the estimation of realistic performance for all of
the 60 gas turbine cycles initially considered. Out of a preliminary evalutation
of these cycles emerged three candidates for detailed analysis.
*References are located beginning on page 2kl.
-------
L-971249-7 FIG. 1
STUDY SCOPE
PHASE I - CANDIDATE SELECTION
• ESTABLISHMENT OF BASE-LINE TECHNOLOGY
• PARAMETRIC INVESTIGATION OF MANY GAS TURBINE CYCLES
• PRELIMINARY EVALUATION OF PROPULSION SYSTEMS
• SELECTION OF CANDIDATE SYSTEMS FOR DETAILED STUDY
PHASE H- CANDIDATE OPTIMIZATION
• PERFORMANCE ANALYSIS OF SELECTED ENGINES
• MISSION ANALYSIS
• VEHICLE PERFORMANCE, EMISSIONS, AND FUEL ECONOMY
• ENGINE INSTALLATION INVESTIGATION
• DEFINITION OF OPTIMUM CONFIGURATIONS
• PRELIMINARY DESIGN OF OPTIMUM CONFIGURATIONS
• ECONOMIC ANALYSIS
• MANUFACTURING COST ESTIMATION
• LIFETIME COST PREDICTION
• FINAL ENGINE SELECTION
PHASE HI - PROGRAM PLANS
•OUTLINE RECOMMENDED DEVELOPMENT PROGRAM
• DISCUSS IMPLICATIONS FOR ENGINE PRODUCTION
-------
Phase II involved the optimization of the three engines selected from Phase I
for the detailed requirements of low-polluting automobiles as established by the
Office of Air Programs. This phase involved very close coordination among the
noted tasks since the optimization process was essentially iterative in nature.
The optimization program included design, performance, and economic analyses
of the selected configurations. Finally, Phase III involved the preparation of
an outline of a development program for the most attractive of the gas turbine
engines considered.
This report consists of a description and results of the tasks performed
under each phase, and is organized in very much the same order as the above
tasks are described.
-------
-------
L-9712U9-7
ESTABLISHMENT OF BASE-LINE TECHNOLOGY
General
Early in the study of automotive gas turbines at United Aircraft Eesearch
Laboratories, it became evident that, to be competitive, a gas turbine engine
must be based on advanced-technology components. Advanced technology in this
regard is defined as the best demonstrated level of sophisticated aerodynamic,
thermodynamic, and mechanical design capable of large-volume, low-cost production.
The failure of a successful automobile engine to emerge in some 20 years of previous
effort by various, highly competent individuals and organizations can be attributed
primarily to the level of technology represented by the components used in these
efforts. Low component efficiencies and low stage work components coupled with
low achievable peak pressure ratios have led to only one path of development for
automobile engines, i.e., the examination of large heat exchanger areas in an
effort to compensate for deficiencies in the flow path efficiency in seeking
reasonable fuel consumption characteristics. Eef. 1 documents the high costs
associated with this approach.
Since the objective of this study was to define the best engine for future
OAP development programs, the engines studied herein were based on the best
applicable advanced technology. A previous study had identified the advantages
of this approach for both a simple-cycle and regenerated engine (SSS-12 and
RSS-7 of Eef. l)5 but it was far from clear whether or not an advanced regenerated,
a recuperated, or a simple-cycle engine might be the optimum choice of advanced-
technology engines. The method of approach for this study used comparable levels
of component technology, supportable by test results obtained within UAC wherever
possible.
Another significant constraint which affected the results of the study was
that regarding the 1975 demonstration requirement. In order to minimize the
technical risk, several compromises in the performance previously reported in
Eef. 1 were made. Specifically, compressor test data were used directly from
test results with little or no improvements assumed for 1975 performance. In
addition, specific speeds of the engines were reduced, thereby resulting in
reduced rotational speeds and specifications for stresses and bearings which more
closely adhere to the current state of the art. The pressure' ratio was also
reduced for the simple-cycle engine, and a reduction of up to 100 degrees in
maximum uncooled turbine inlet temperature (to 1900) was assumed. These compromises
were made at the expense of increased size of the engine and resulted in a
technologically less advanced engine than that postulated in Ref. 1. Consequently,
there is no doubt that the engines described in this report represent state-of-the-art
-------
L-971249-7 F(G. 2
TYPICAL HIGH-PRESSURE RATIO COMPRESSOR PERFORMANCE PREDICTIONS
NOMINAL EXIT MACH NO. 0.12
COMBINED 30°, 0° INLET SWIRL MAP
(30° SWIRL - 100% SPEED INTERPOLATED)
0° INLET SWIRL
30° INLET SWIRL
11.0
10.0
9.0
8.0
7.0
6.0
5.0
4.0
3,0
2.0
1.0
80.0
78.0
76.0
74.0
72.0
70.0
100.0
85.5
20.0 30.0 40.0 50.0 60.0 70.0 80.0 90.0 100.0 110.0
10
-------
L-9712U9-7
technology with no breakthroughs required to achieve' the performance estimated,
provided that a reasonable and thorough development program is followed.
A discussion of the base-line technology used for evaluation follows.
Some design features examined for the base-line engines (such as internally
cooled turbine blades) were eliminated in the Phase I review and, hence, did not
enter into the detailed Phase II evaluation.
Compressors
The most critical component of a gas turbine engine (since it is generally
the most difficult component for which to obtain advanced performance) is the
compressor. United Aircraft Corporation has achieved a significant breakthrough
in the area of small centrifugal compressor performance largely as a result of
research conducted at UACL. Both rig demonstration and actual engine application
of advanced centrifugal compressors have confirmed that these compressors, which
feature UACL's patented pipe diffuser, are the most advanced in the world from
the standpoint of efficiency, pressure ratio, flow capability, handling ease,
and predictability.
An example of the performance believed to be achievable from these 'compressors
is shown in Fig. 2. This compressor map is an estimate previously reported in
Ref. 1, and is representative of those used for performance calculations in
this study.
Single-Stage Design Point Performance
Figure 3 shows the relationship between stage efficiency (riij_s) sn& stage
total-to-static pressure ratio (PRrp_g) which was used for this study, at design poii
The curve shows the predicted level available for engine demonstration in 1975
following a suitable research program. Because the time period for this research
is short, UACL considered it advisable to assume only modest improvements over
currently available technology. It is for this reason that this curve has an
irregular form between 8 and 9-50 PR. It has been constructed from experimental
data from series of tests on two different types of impellers, identified as
the "F" and the "K" impellers. The "F" impeller incorporates prewhirl and the
inducer is designed to be subsonic below about 5-5;l pressure ratio, which also
has the erfect of reducing impeller loading. However, for a given pressure ratio,
this approach results in higher diffuser entry Mach number, and as press'ire
ratio increases, this Mach number increases to the point where losses become
significant. Above 6.0:1 the inducer also becomes supersonic, resulting in a
11
-------
PREDICTED DESIGN POINT COMPRESSOR STAGE EFFICIENCY AND PRESSURE RATIO
•o
I
-J
0.85
0.80
t-
K-
0.75
0.70
0.65 —
1975 DEMONSTRATION
ENGINE
8 10
P&T-S
12
14
16
130
120
110
100
90
80
70
60
50
40
1/9
\
CO
D£
O
a.
Of
o
•
CO
-------
L-9712^9-7
fairly steep gradient in efficiency with pressure ratio. These effects are
discussed in greater detail in Ref. 5 which includes test data on this class of
impeller up to 1^.0:1 pressure ratio. Substantial improvements are possible
at high pressure ratio as the result of the development of supersonic inducers
which give efficient impellers and reduced diffuser inlet Mach numbers. The
potential performance of this class of impeller is represented by the right-
hand branch of the curve, above 8.0:1 PR.
As a result of this approach, i.e., essentially assuming that demonstrated
performance levels could be improved by roughly the same amount at all pressure
ratios, the predicted efficiency levels reflect the variation experienced with
current compressors. To remove the apparent "dip" in this curve between 8
and 9.5 PR would mean assuming that the efficiency at 8.0:1 PR could be increased
by a further k points. It was considered unwise to assume that this could be
done for a 1975 demonstrator engine.
Two-Stage Design Point Performance
The staging of two such compressors provides an excellent level of technology
with which to investigate multispool engines. The point of maximum efficiency
for each overall pressure ratio was found (Fig. U) and the corresponding values
of first-stage pressure ratio are shown in Fig. 5- As can be seen from Fig. 5j
the optimum first-stage pressure ratio for a two-spool machine is approximately
the square root of the overall pressure ratio.
In a single-shaft compressor, equal specific speeds would give a high first-
stage pressure ratio, while equal pressure ratios would give a large difference
in specific speed. Therefore, the shape of the efficiency vs. pressure ratio
characteristic will establish two regimes. At low pressure ratios this curve
is flat and the effect of specific speed will dominate (i.e., pressure ratio
across the first stage, PRj, will be high).
When the first-stage pressure ratio exceeds U:l, the optimum will start
to shift toward equal stage pressure ratios, and when both stage pressure ratios
exceed 5:1 the effect of pressure ratio dominates and PRj = 'v/PRoverallT' *n
Fig. 5 the two-spool curve is a simple transform of the single-stage characteristic
while the single-shaft curve shows the effect of the further constraint on the
stage specific speeds.
13
-------
L-971249-7
FIG. 4
OPTIMUM TWO-STAGE CENTRIFUGAL COMPRESSORS-OVERALL EFFICIENCY
VS. OVERALL PRESSURE RATIO
0.84
0.82
0.80
> 0.78
o
(J
uu
U 0.76
0.74
0.72
0.70
TWIN SPOOL
J L
J L
10 14 18 22 26
OVERALL PRESSURE RATIO, PROV(T_S)
30
-------
FIRST STAGE PRESSURE RATIO VS OVERALL PRESSURE RATIO
OPTIMUM CENTRIFUGAL COMPRESSORS
O TWO SPOOL
D SINGLE SHAFT
23
6 8 10
OVERALL PRESSURE RATIO, PR0v
-------
L-971249-7
FIG.6
AXIAL TURBINE FULL-POWER EFFICIENCY
QIN = 0.5
Q|N=1.2
3-
U
z
UJ
y
u.
U.
IU
0.9
0.8
0,7
0.6
0.5
1.0
T|N --= 1984 F
2563 F
1984 F
3200 F
2563 f
3200 F
2.0
3.0 4.0
PRESSURE RATIO
5.0
6.0
16
-------
L-9712U9-7
Turbines
The aerodynamic design of turbines at UACL is largely based on a fundamental
analysis of fluid flow within the turbine flow passages. This approach was
used for the design of the turbines for the original PT6 engines, and this sophisticated
technique has been progressively refined since then with a series of uprated
PT6 versions, the JT15D turbofan engine, research turbines, and advanced engine
design studies. This emphasis on analytical design requires a continuing effort
to refine design and analysis techniques. UACL has, since 1963, been engaged
in a series of aerodynamic turbine research programs on both axial and radial
turbines. Current axial turbine research at UACL is funded jointly by the DRB*
and UACL, and is concentrated on high-work stages. A stage of 3.6 total-to-total
pressure ratio is being tested along with a number of variations to establish
design criteria for high-work regimes. The effects of cooling compromises such
as I6w aspect ratio and thick trailing edges are being tested. Extended stage
loadings were chosen for the present research program due to the potential for
performance improvement in the high-work regime of small, cooled axial turbines.
Aerodynamic research on radial turbines has been carried out at UACL since
1963. A design point efficiency of 88.7% at 5.9:1 pressure ratio, and a peak
efficiency of 90% at 7.15:1 pressure ratio were achieved on the first build of
a test turbine early in.this program. Results of turbine test work have been
partially reported in Refs. 6 through 9-
The base-line technology data were selected to adequately cover the required
turbine inlet temperature (TIT) range. Three levels were selected such that the
first TIT would require no airfoil cooling; the second would require conventional
amounts of convection cooling; and the last would require some advanced form of
cooling. Two different sets of the flow parameter, W V"T/P, were selected: two
for axial turbines and another two for radial turbines. The philosophy followed
in this investigation was to design the turbines aerodynamically for a part-
load design point and then correct the information to maximum power conditions.
Axial Turbines
Preliminary designs and efficiency estimates were made for several single-
stage axial turbines with different pressure ratios, turbine inlet temperatures,
and inlet nondimensional flows. The results ai*e shown in Figs. 6, 7, 8 and 9.
From the inlet conditions and assumptions, turbine annuli were constructed
and the airfoil numbers and chords were estimated. Stage efficiency estimates
followed, including cooling air effects. For the lowest-TIT case, no airfoil
*Canadian Defence Research Board.
17
-------
L-971249-7
FIG. 7
AXIAL TURBINE GAS TEMPERATURE RELATIVE TO BLADE MEAN SPAN
3000
2800
2600
UJ
a.
UJ
O_
<
_l
UJ
Q£
2400
2200
2000
1800
1600
2563 F•
1984 F
3 4
PRESSURE RATIO
18
-------
L-971249-7
FIG. 8
AXIAL TURBINE BLADE STRESS PARAMETER
o
X
CM
z
z
< 5
in
Q
10 3
T|N =3200 F
I
3 4
PRESSURE RATIO
19
-------
L-971249-7
FIG. 9
RADIAL TURBINE EFFICIENCY AT FULL POWER
Q|N = 0.3
Q=
0.90
0.85
u
g
u
ui
z
m
u.
3
0.80
2563
3200 F
0.75
0.7
5 7
PRESSURE RATIO
2u
-------
'1-9712^9-7
cooling was required. For the intermediate TIT, airfoil cooling was assumed
with 1.5$ of vane cooling air and 2% blade cooling air. For the highest-TIT
oases, efficiency decrements typical of transpiration cooling were used.
The. procedure for two-stage axial turbine designs was similar to the one
described above. For the intermediate TIT, cooling of only the first stage
was assumed, with cooling flows similar to those given above. For the case of
the highest turbine inlet temperature, transpiration cooling of airfoils of both
stages was assumed and the accuracy of information for the two-stage axial
designs was of the same order as for the single-stage cases. The results are
not shown in this report, since none of the designs which were selected for
cycle analysis included the use of two-stage axial turbines.
Radial Turbines
For the small sizes under consideration in this study, radial turbine efficiency
(Fig. 9) tends to be higher than that of an axial turbine designed to the same
conditions. Because of the high cycle pressure ratios under consideration, the
radial turbine was also an obvious choice since it is capable of higher expansion
ratios. For this reason, major effort was applied to the radial turbine preliminary
designs and efficiency estimations. The pressure ratio range investigated was
from h to 12. In the process of this investigation, the efficiency estimation
technique had to be revised to cover the cases with very low inlet nondimensional
flows, very high pressure ratios, and very high inlet temperatures. The results
are shown in Figs. 10 through 12.
Radial turbine off-design performance was predicted using Fig. 12. In order
to use this figure, the following steps are used:
a. Calculate U/C0' for the required off-design point, noting that Ah1
is the isentropic enthalpy drop
b. The upper curve is entered to find
c. The lower curve is entered for the correction to ri/1"iTyEgjQji
the selected pressure ratio
d. Calculate n from the design point
21
-------
RADIAL TURBINE TIP SPEED
ro
ro
3200
2000 -
f
10 12
PRESSURE RATIO
14
18
20 L
-------
RADIAL TURBINE GAS TEMPERATURE RELATIVE TO BLADE TIP
2800 -
2600
I 2400
< 2200
2000
1800
1600
Q IN =0.6
Q IN =0.3
T,N=3200 V
= 2563 F
11 13
PRESSURE RATIO
15
17
= 1984 'F
19
P
-------
L-971249-7
RADIAL TURBINE OFF-DESIGN PERFORMANCE
CORRECTIONS
FIG. 12
1.00
UJ
Q
C-
PRE5SURE RATIOS ARE TOTAL TO TOTAL
U/C,
UJ
Q
c-
-0.02
-0.04
I
0.2 0.4 0.6 0.8
P.R./P.R. DES.
1.0
1.2
-------
L-9T12H9-T
Stress Data.
Axial TurMne Stress Limits
Stress limits for integrally cast axial turbines were developed to serve
as a preliminary guide in the selection of feasible rotors. Figures 13 and 1 it-
show the maximum allowable, relative gas temperature at the blade mean section
and the maximum rotational speed limits for various A^N2 blade stress parameter
values, where A^ is the mean gas annulus area in square inches, and H is rotational
speed in revolutions per minute. The material shown in Fig. 13 is IN-100, while
Fig. lit- shows similar data for INCO-713 material.
If the rotors are solid with no bore, the rotational speed limits are
indicated by burst requirements or rim low-cycle fatigue requirements. The
data presented have been based on the following assumptions:
1. Design limits are for the maximum power conditions.
2. Eotors have a constant blade mean peripheral speed of 1350 ft/sec.
3. The blades are unshrouded. ,
U. The blade creep life is 150 hr.
5. The ratio of disc bore radius to rim radius is approximately 0.3.
6. The rotor burst speed margin is 1.2.
7. The disc fatigue life is 25,000 to 30,000 start-stop cycles.
8. The thermal stresses were assumed constant.
Radial Turbine Stress Limits
The maximum allowable relative gas temperature at the blade tip for the
radial turbines is given in terms of the tip peripheral speed for various rotational
speeds, as shown in Fig. 15•
The data presented in Fig. 15 are valid for the following design conditions:
1. Design limits are for the maximum power condition.
2. Material is forged Udimet 700 (PWA-689), which is believed to be
the best for this application. Certain engine cycles may allow
the use of cheaper materials.
.25
-------
L-971249-7
FiG. 13
140,000
120,000
100,000
80,000
60,000
40,000
AXIAL TURBINE STRESS LIMITS
MATERIAL IN-100
MAX LIMIT FOR BURST REQUIREMENT
MAX LIMIT FOR BLADE CREEP LIFE
1800
IL!
> U-
< I
1700 -J K
J1"
1600 to §
*$
1500 ^ I
X UJ
< l-
1400
10
AAN2HN.2 RPM2 X 10"10
26
-------
L-971249-7
FIG. 14
AXIAL TURBINE STRESS LIMITS
MATERIAL INCO-713
100,000
80,000
60,000
40,000
20,000
MAX LIMIT FOR BORED DISC BORE LCF
MAX LIMIT FOR BLADE CREEP LIFE
I
1300
1700
U.
1600 j,.
1500
1400
AAN IN. RPM X 10
6 7
-10
10
-------
STRESS-RELATED TEMPERATURE LIMITS FOR RADIAL TURBINES
UDIMET 700
SOLID HUB
TOTAL INLET TEMPERATURE
40,000
2000
1900
no
CO
Of.
Ill
Q.
Ill
1800
1700
1600
1500
GAS RELATIVE TEMPERATURE AT TIP
120,000
2000 —
1900
1800
1700
1600 —
1500
BORED HUB
1800
2000
2200 2400 2600 2800
TIP PERIPHERAL SPEED ,
TOTAL INLET TEMPERATURE
40,000 RPM
1800 2000
,,Tlp_ FT/SEC
2200
2400
-------
L-9712^9-7
3. The blade creep life is 150 hr at design limit.
b. The hub ratio is selected to give maximum temperature capability,
but was limited to a maximum value of 0.55-
5. The bore ratio for the bored hub is 0,2.
6. The maximum allowable theoretical strip analysis stress was kept
at 125 ksi for the bore, 75 ksi for the rim, and 90 ksi for the blade
root; presently these limits are believed to permit a fatigue life
of 25,000 start, max. power, stop cycles.
7. The effect of "blade tolerance is considered in the blade definition
(tolerance +_ 0.005 in. wavelength = 20% of tip radius). Nominal blade
weight only is considered for disc stress calculation.
Turbine Blade Cooling
Because cooling air mass flow is a function of stress as well as temperature
level, it has been correlated with the primary stress criteria. Other limitations
due to high stress or high temperature must be established since there are major
areas of concern regarding the combinations of high temperatures and high speeds ,
erosion of blade tip coating at high temperatures, requirements for relatively
large cooling holes and small turbine rotors, and small-size effects on material
properties and component strength. The limitations due to the low-cost requirement
are difficult to establish. However, it was felt that this problem would be
faced following the selection of engine cycles should the engines with cooling
requirements appear to be particularly attractive. However, in no case was an
internally cooled turbine selected for detailed analysis.
The cooling air mass flow was obtained from a cooling effectiveness correlation
shown in Fig. l6 for which the material application and stress parameter have to
be known. Materials which were considered for the automotive engine include:
WX-188 and WI-52 for turbine vanes, forged Udimet 700 and cast IH-100 for radial
turbine rotors, and cast IH-100 and IMCO-713 for axial turbine, rotors. The
allowable temperature (T^-^) to be used in the cooling effectiveness calculation
is given here for typical designs. In the case of rotors, T^j-^ has been correlated
with the following stress parameters:
2
axial: AN , where: A is gas path annulus area, N is rotational speed;
units are in. KPM
29
-------
Lr971249-7
FIG. T6
COOLING AIR FLOW VS COOLING EFFECTIVENESS
0.4
u 0.3
c-
a
<•>
in
U.
O
O
O
u
02
U.i
0.1
VANES
1%
2%
3%
COOLING AIR MASS FLOW -PERCENT
OF ENGINE MASS FLOW
30
-------
radial: UT,
Blade tip peripheral speed, (ft/sec)
Material
Application
Allowable Temperature
T
ALL»
NX-188
WI-52
UDIMET 700
IN-100
IN-100
INCO-713
Turbine vane
Turbine vane
Radial turbine rotor
Radial turbine rotor
Axial turbine rotor
Axial turbine rotor
1780
1750
1710
1770
1810
1730
2
- 57 (UT/1000)
- 57 (U?/1000)
- U5 AH
- U5 AN2
The cooling effectiveness used in the cooling air mass flow determination
was defined as follows:
for vanes
TIT
n =
TIT
(UT)2
TRTIP - 32 ——- - TALL
for radial turbines nc = 1.25 \±uuu)
TRIIP " 32
TCA
for axial turbines n =1.25
TRMEAN - TALL
T-,
RMEAN -
where : TIT = turbine temperature , F
TALL = allowable temperature, F (obtained from Table l)
= co°linS air temperature, F (which is assumed equal to
compressor delivery total temperature plus 100 degrees)
31
-------
L-97I21+9-7
TABLE I
ENGINE CYCLE LOSSES
SSS-10 * RGSS-6 * RCSS-8
1. Intake pressure loss 1.3
2. Primary burner pressure loss 3%
3. Duct pressure loss between turbine exhaust,
and heat exchanger 0
i|. Duct loss from turbine or heat exchanger
to nozzle 3.3
5. Nozzle loss k.J
6. Nozzle dump loss 1.2
7. Reduction gearbox loss (ASHP) ^.2
*For definition of engine designationa, see p. 73
8. Exchanger cold side Ap/p ° 0.0125 0.0057
9. Exchanger hot side Ap/p ° 0.0213 0.0231
10, Carryover flow loss 0 1.5% 0
11. Exchanger leakage flow loss 0 4.5$ 0
12. Turbine leakage flow 1.0$ 1.0? 1.0$
13. Bearing seal leakage flow loss 0.5$ 0.5$ 0.5$
32
-------
= total relative temperature at blade tip, F
= total relative temperature at blade mean section, F
Cooling mass flows are obtained from correlation with the cooling effectiveness
as plotted in Fig. 16 for static vanes and rotor blades. This correlation is
applicable to simple convection cooling of small-size blades and vanes.
Engine Loss Data
The major losses which were factored into the engine performance program
were estimated based on experience and detailed analysis of installation
requirements. Table I shows the cycle losses which were included for the optimization
portion of this study.
Heat Exchangers
Recuperator
Two candidate recuperator types were evaluated for the automotive engine
application — plate-fin and tubular — and the tubular type was chosen. The
main factors influencing the choice of an axisymmetric, tubular type were as
follows:
1. Engine Layout. The axisymmetric heat exchanger possible with the tubular
type allows for more compact, low-loss ducting, particularly in the turbine
exit diffuser. This duct will be much more tolerant of the large swirl
variations which it will have to accept. The plate-fin exchanger does not lend
itself readily to this configuration, and would require the exhaust stream to be
split into discrete flows feeding rectangular heat exchangers. These split
ducts would have larger losses and bulk than the design chosen, thereby tending
to minimize the advantage in compactness normally associated with plate-fin
heat exchanger matrices.
2. Heat Exchanger Durability. Cyclic thermal stresses affect all
heat exchangers. Experience at Hamilton Standard Division (HSD) with the plate-
fin type has shown tendencies for progressive increases in leakage as the closure
joints crack open under this cycling. This problem is aggravated by the additional
restraints at the corners of a rectangular-section heat exchanger. The best
33
-------
L-971249-7
FIG, 17
AXISYMMETRICAL TUBULAR RECUPERATOR
SINGLE PASS - HOT
DOUBLE. PASS - COLD
2140 TUBES, 0.012 IN. O.D.
70
60
UJ
Z
HI
-------
L-9712^9-7
solution to this problem at present is to use material with good cyclic fatigue
properties. The tubular type appears to suffer much less from these difficulties.
The braze joints will not tend to open up under the pressure loads; thus, even
if cracked this should not exhibit much leakage. Furthermore, the annular
headers can be arranged to float radially inside the engine casing, thus
eliminating a large source of thermal stress. This feature opens up the
prospect of a cheaper material for the tubular exchanger.
3. Manufacturing Cost. The potential manufacturing cost of a plate-fin
exchanger is apparently relatively low. However, experience with actual
production at HSD shows that it is extremely difficult to avoid leakage at the
closure joints on newly brazed parts. This difficulty must be resolved at low
cost for the plate-fin design to be acceptable. The major difficulty for the
tubular heat exchanger is the cost of the tubes themselves. The solution to
this problem may be some form of helically wound strip welded or brazed into tubes
and drawn, if necessary, to the finished size. The other fabrication problems
do not appear to be too difficult. No clear distinction can be drawn between
the fabrication cost of these two types, but the possibility of less expensive
materials for the tubular type may provide some cost savings.
The disadvantage of the axisymmetric tubular recuperator is its somewhat
larger matrix size when compared to a compact plate-fin type. This disadvantage
is somewhat offset when the volume of the entire recuperator, -including headers
and ducting, is considered; nevertheless the practical effectiveness of the tubular
recuperator is probably limited, for the automobile application, to values of
less than 70%. The variation of effectiveness with recuperator length is presented
in Fig. 17- It is noted that, for the recuperator considered, it is necessary
to double the length (from 8.5 in. to 17-5 in.) to increase effectiveness from
60 to 70%. In the optimization portion of the study, it was initially attempted
to design the recuperator for an effectiveness of 70%; however, it soon became
apparent that packaging and installation problems, and the increased manufacturing
cost, could not justify the relatively small gain in fuel economy (approximately
1 mile per gallon) for the higher value of effectiveness.
Regenerator
The heat exchanger of the rotary type considered (regenerative) was considered
to be of the CEECOE® ceramic configuration. Data entered into the program are
shown in Fig. 18 and Table II. These data were supplied by the manufacturer,
Corning Glass.
Although metallic-matrix rotary regenerator cores have been used for
truck-type engines, their material cost precludes serious consideration in
35
-------
L-971249-7
0.200
f AND j FACTORS FOR CERCOR MATRIX HEAT EXCHANGER
FIG. 18
0.100
0.090
0.080
0.070
0.060
0.050
0.040
0.030
0.020
0.010
0.009
0.008
0.007
0.006
0.005
\
\
-13.33/Np
a= 1548 FT2/FT3
P = 63.7%
4 r = 1.645 X 10 ~3FT
f = FANNING FRICTION FACTOR
J = COBURN MODULUS
__--. THEORY FOR EQUILATERAL
TRIANGULAR TUBES
J I I LJ_
_L
60 70 80 90 >00
200
300
400 500 600 700
36
-------
L-9712U9-7
TABLE II
HEAT EXCHANGER CORE GEOMETRY FOR DATA REDUCTION
(Single-Blow Transient Technique)
•Core Designation
Core Mass
Core Dimensions
Core Volume
Solidity
Porosity, p
Area Density, a
Heat Transfer Area, A
Frontal Area, A^j.
Flow Area, PAfr = Ac
Hydraulic Diameter, Urh
L/rh
Specific Heat, c
Density, solid, p
Thermal Conductivity, k
Conduction Area, Ak
Humber of Cells per in. , N
Cell Height/Cell Width, d/c
507 (Corning 22-02*)
0.8961** (U06.6 grams)
width 3.312 in.
height 3.296 in.
length 2.830 in.
30.892+ in.3
0.3633
0.6367
15U8 ft2/ft3
27.67 ft2
10.92 in.2
6.95 in.2
0.0016^5 ft
575.2
0.200* Btu/db-F)
138** rbm/ft3
0.^2 Btu/(hr ft-F)
3.97 in.2
10l4l.6+
0.853+
Date: 6/U/68
* Manufacturer's specifications
+ Inputs for p, a, and rjj evaluations
37
-------
L-9712U9-7
automobiles. Another ceramic material for this application is provided (>y
Owens-Illinois, tut its properties are very similar to those of the Corning Glass
material for engine cycle analysis purposes.
Transmissions
The transmission is the power-train component which proTides the torque
flexibility required for proper operation of an automobile. The selection of
this unit is primarily influenced by the engine torque/speed characteristics.
An internal combustion (1C) automobile engine can provide high torque over most
of the engine speed range and therefore requires gear-ratio changes (torque
multiplication) only for starting from a standstill. The free-turbine gas turbine
engine also has relatively high low-speed torque. Consequently, a standard three-
speed automatic transmission can be used, with or without a torque converter,
or a manual shift could be employed.
The single-shaft gas turbine has torque characteristics which make the
selection of a transmission an acute problem. The full-load torque decreases
rapidly as speed is reduced until, at 50% rated speed, the engine produces no
appreciable torque at all. Therefore, the transmission must provide considerable
torque multiplication over the entire speed range to make the single-shaft gas
turbine suitable for an automobile.
Previous studies had identified the extreme importance of the transmission
in determining the engine characteristics and the output of optimization
studies. Therefore, considerable emphasis was placed on the selection of
proper base-line technology transmissions. The transmission analysis for this
study is directed, mainly toward, identifying and determining the operating
characteristics of those units which may be suitable for a single-shaft engine.
Geared systems provide torque multiplication via a set of gears which are
alternatively meshed or demeshed to provide discretely stepped gear ratios .
The simplest mechanical device is a multigeared stepped transmission, such
as an 8-speed unit suggested by Borg-Warner and examined for single-shaft automotive
gas turbines in Ref. 1. This unit incorporates a controlled, mechanical, slipping
clutch which is fully engaged and disengaged within a relatively narrow engine
38
-------
speed range. The clutch permits transmission of a constant engine torque (determined
by clutch input speed) at any clutch output speed up to and including input
speed. The shifts would occur automatically, governed by throttle setting and
road speed, in a fashion similar to that employed in current automotive transmissions.
The rate of deceleration of the slipping clutch during upshifts is assumed to be
controlled at a rate sufficient to assure passenger comfort. Figure 19 illustrates
a steady-state horsepower/speed profile for the 8-speed gearbox with a 150-hp
simple-cycle gas turbine. Similar data are shown for an equal-output spark-
ignition (SI) engine with a three-speed torque converter automotive transmission.
Considering the shift points from a steady-state standpoint s for both systems ,
there is a large torque (horsepower) mismatch which needs to be smoothed. From
steady-state considerations, the average power for the 8-speed transmission up
to 60 mph is about 106 hp, but transient analyses indicate the availability
of a substantially higher average power, as shown in Fig. 20. Here, the
effects of engine inertial energy are seen to increase the average power to
about 120 hp for the conditions analyzed.
As presently conceived by Borg-Warner, the shifting control will be a
function of engine speed and vehicle speed, and not dependent on road load as
in current transmission practice. Thus, the shifts take place at the same
speeds, regardless of throttle setting.
The slipping clutch must have sufficient life to avoid costly and frequent
repairs, and it is the Borg-Warner belief that this durability is current state
of the art. However, the torsional vibration associated with a pure mechanical
system may not be damped to any significant degree. The cost of an 8-speed
transmission consisting of a standard four-speed automatic gearbox preceded by
a two-speed planetary splitter and a controlled slipping clutch may not be higher
than present automatic transmissions with torque converters. In addition,
efficiencies should be high for this completely mechanical system. The weight
of such a unit is still somewhat uncertain since detailed designs have not been
executed, but is estimated to be less than 200 Ib for a 150-hp engine.
Fluid-Coup11ng_S v_s tern
An alternative to the slipping clutch for this transmission may be a fluid
coupling. The first major advantage would be extremely long lifetimes with
little maintenance. Secondly, a fluid coupling is an excellent "shock absorber"
and greatly reduces torsional vibration transfer to the passengers.
If the fluid coupling were located uetween the turbine and the transmission,
a variable-fill system could be incorporated for starting from a stop. At idle,
39
-------
L-97-1249-7
FIG. 19
POWER VS SPEED
8-SPEED MECHANICAL (SSGT)
. 3-SPEED TORQUE
CONVERTER (SI )
-------
PERFORMANCE OF SINGLE-SHAFT GAS TURBINE WITH BORG-WARNER 8-SPEED TRANSMISSION
k\N ENGINE INERTIA
STEADY STATE
200
o
a.
UJ
10
20
30
40 50
ROAD SPEED ~MPH
60
70
80
-------
L-971249-7
FIG. 21
ELECTRIC-DRIVE SCHEMATIC
-------
L-9712^9-7
the coupling would be sufficiently empty to avoid automobile "creep." As the
accelerator is depressed, the coupling would fill at a rate which would vary with
throttle setting and road load. Thus, wheel, spin could be avoided from a standing
start at full throttle. At part throttle, the coupling would fill rapidly to
provide maximum available torque to the transmission. At high slip ratio, the
power loss through the coupling is converted to heat which tends to break down
the coupling fluid. However, the variable-fill system would incorporate an
external oil reservoir, which would also act as an oil cooler. Partially-
filled fluid couplings have a tendency to "foam" under high slip conditions,
thus greatly reducing efficiency.. Therefore, the coupling should be as near
full as possible.
If the coupling were located behind the transmission, it would experience the
combined speed range of the engine and transmission, thus possibly eliminating
the need to reduce the fluid level at idle since the torque transmitted varies
as the square of the input speed.
Since the fluid coupling transmits torque at a maximum of 9&% efficiency,
its.overall efficiency would approach that of a mechanical clutch. Also, the
coupling would probably be cheaper than the clutch, but the added oil reservoir/
cooler unit and pumping system would result in an overall higher cost.
Electric Drive
Rather than to rely on high gear ratios to achieve high torque multiplication
at low speeds, other systems might be incorporated which have the desired
characteristics. One such system is electric drive. Here, the gas turbine
drives a generator (or alternator) which produces electrical power to drive
traction motors which are connected to the vehicle's wheels (Fig. 21). The
controlled-speed traction motors allow the engine to operate near peak efficiency
at all times, since engine speed is not a function of road load. Electric-
drive systems have been successfully used for off-road equipment where great
flexibility is required, and where cost and weight considerations are secondary.
In these applications, four basic types of electric motors have been used —
shunt and series-field d-c motors, a-c induction motors, and a-c synchronous
motors. D-c motors generally have good speed control but are speed-limited
due to the brushes and commutator. A-c motors can operate at higher speeds and
are more rugged than d-c motors but are more difficult in speed control.
The costs of electric-drive systems used to date have normally been quite
high, due primarily to low production rates. Also, the components have been
heavy and bulky to a degree which has made an electric-drive system appear infeasible
-------
L-971249-7
FIG. 22
A-C INDUCTION MOTOR WEIGHT CORRELATION
1000
CO
_l
I
I-
X
o
o 100
10
0.001
I I I
I I
I
I
j I III
0.01
HORSEPOWER/RPM
0.10
-------
L-9712^9-7
for an automobile. As an example, assume two electric traction motors are used,
each rated at 50 hp and 3000 rpm. Figure 22 indicates that each would weigh
about 300 Ib and have a retail cost of about $1+50 at low-volume production rates.
A conventionally-sized 100-hp generator, which is driven by the gas turbine,
may weigh 600 to TOO Ib and cost about $1000. While these prices could be
greatly reduced if millions were produced annually specifically for the automotive
market, a redirected technology to design lightweight units for automotive
use would appear to be a necessary first step.
Recent developments in electric-drive systems indicate that significant
advances could be made in system weight and performance. Delco-Remy Division
of General Motors has tested a high-speed a-c drive system in a 2 1/2-ton
truck (Ref. 10). The traction motors are brushless, self-synchronous units which
have torque and speed control characteristics much like a d-c motor. They are
electronically commutated as a function of rotor position, thus eliminating rotor
windings, the rotor commutatorj and brushes. The motor can be operated-at
high speeds (17,000 rpm) and also provide high continuous torque (essentially
constant hp). These high-speed units should result in significant weight
reductions. It is estimated that a 50-hp motor would weigh less than 100 Ib
and have a retail cost of about $200 at smaller production rates than is normal
for automobiles. Likewise, the alternator can be similarly designed and would be
estimated to weigh under 200 Ib and cost $500.
i
i
Similar developments have been advanced by the Pratt & Whitney Division of
United Aircraft for d-c systems (Ref. 11). Rather than eliminate the brushes
and commutator, the design of these elements was improved to the point where
rotational speeds of 15,000 rpm were successfully run on a 30-kw prototype
unit. This unit weighed only 112 Ib. It is estimated that a 50-hp mass-produced
unit could be designed to weigh about the same.
The Solar Division of International Harvester has developed a small 10-kw
turbo-alternator on a common shaft which operates at 100,000 rpm. Similar
technology would allow a single-shaft gas turbine and alternator to be used
. for an electric-drive automobile.
'It appears that significant advances could be made in electric-drive
technology for automotive applications if a directed effort were to be made.
Possibly the high-speed alternator (or generator) would weigh no more than
today's transmissions, and the traction motors and control systems would weigh
no more than the drive shaft, differential, and axle they would replace.
The control systems have been estimated to cost about $20/kw on a high-production
basis and may approach $2/kw for automobile production rates.
-------
L-971249-7
FIG. 23
EFFICIENCY OF GENERAL ELECTRIC HYDROMECHANICAL TRANSMISSION
I- 0.80
li-d-p)12] [s
I 11
2p
0.366
0.366
0.8 I.2
K --= OUTPUT RPM/RATED RPM
,6
-------
L-9712U9-7
Hydrostatic Transmissions
Pure hydrostatic transmissions have been tested in automobiles for years,
but the systems have suffered from hydraulic noise, low overall efficiency
(~80$), and appear to "be subject to speed limitations (approximately 5000 rpm
maximum). They have also been heavy and somewhat bulky.
The pure hydrostatic system can be greatly improved by combining it with
an epicyclic gear system to produce a transmission which has higher efficiency
and greater versatility than the basic system. These hydromechanical transmissions
(HMT) are sometimes called "split-torque" transmissions in which the input torque
is divided between the hydraulic system and the mechanical gear train. One
hydraulic unit is attached to the input of the transmission system; the other
is geared (or attached) to the remaining planetary shaft which does not connect
to the output. Depending on load conditions, either hydraulic unit may act as
a pump or motor. .' ;
Another variation is the "split-speed" system in which one hydraulic unit
is attached to the output and the other is attached to the floating member of
the planetary gear system.
A third type employs two planetary gear systems where each hydraulic pump
is attached to the floating member of one of the .planetaries. From Ref. 12,
it appears that the split-speed and double-planetary systems can provide efficiencies
of 9Q% over a wide speed range. However, the split-torque configuration would
probably be the cheapest because of simplicity. :
| All HMT's allow a smooth, infinitely variable change in gear ratio and
allow the engine to operate near peak efficiency.
' The literature suggests that the gear systems coupled with the hydraulic
pumps and controls will be quite expensive, even on a mass-produced basis.
The hydraulic noise problem must also be solved, along with weight reduction.
However, conversations with General Electric personnel indicate that a hydro-
mechanical transmission can be produced as cheaply as a modern torque-converter
transmission and would probably weigh less. The potential noise problem
remains to be resolved. However, GE has isolated the source of the major
contributors to noise in its HMT and has redesigned its units, although absolute
noise levels have yet to be measured. Figure 23 presents efficiency characteristics
for a General Electric HMT with'an 8.7-cu in. capacity. The equation shown on
the figure very closely represents test data derived from this unit.
-------
L-971249-7
FIG. 24
TORQUE CONVERTER CHARACTERISTICS
i.o
o.
X
I-
o.
z
\
(X
x
Q.
K
o
0.2
0.4 0.6
OUTPUT RPM/INPUT RPM, N
0.8
UJ
Q.
Z
a.
I-
o
1.0
-------
L-9T12U9-T
Sundstrand Corporation has also built hydromechanical transmissions for large
trucks and is currently preparing to market its unit for the trucking industry.
This transmission is a dual-mode configuration (DMT) which incorporates two
internal clutches. These clutches allow the DMT to operate in a pure hydrostatic
mode at low speeds and then shift to a hydromechanical mode through a planetary
gear system. This system would allow higher part-load efficiencies than the
HMT and better full-load performance at somewhat greater"expense. This unit,
as now marketed for the trucking industry, is too expensive for automobiles and
weighs 675 lb.
Hydrokinetic Transmissions
A pure hydrokinetic coupling (fluid coupling) does not provide torque
multiplication but merely transfers torque. However, a torque converter (TC)
similar to those used with today's automatic transmissions can multiply torque
as much as 5-"to-6:l at stall. Thus, the torque converter may be adaptable to
the single-shaft gas turbine. If so,' its low cost due to fully developed technology
and manufacturing techniques, plus its high reliability and relatively good
efficiency, would make it a good Candidate. Several authors (see Ref. 13,
for example) have suggested the torque converter for use with single-shaft engines
and have shown preliminary data .to support their positions. However, these
authors considered only the torque-multiplication ratio and efficiency variations
as a function of output speed (or slip ratio). For various slip ratios, the
input torque was multiplied by the torque multiplication factor to obtain output
torque. This approach in essence assumes that the torque converter is infinitely
small; i.e., there is no retarding torque which tends to reduce input speed.
Simultaneously, it assumes the torque converter is infinitely large and can
absorb whatever torque is applied to it.
Actually, one must examine the torque absorption characteristics of the
converter and match these with the'engine output torque over the entire speed
range. The converter .must be sized and matched to a specific engine pow^r
level. Converters are often characterized by a set of curves of torque multiplication
factor (Q) and efficiency (n) vs. output/input speed ratio (N), as shown in
Fig. 2h. Since efficiency (n) is defined as output/input horsepower, it can
also be defined as the product of N and Q, i.e.,
hp2 (torque)2 (output speed)
(tbrque)-[_ (input" speed)
-------
L-971249-7
TORQUE CONVERTER-ENGINE CHARACTERISTICS
FIG. 25
ENGINE CHARACTERISTICS
600 r-
500 -
200
400 600
OUTPUT RPM
50
800
1000
-------
L-9712U9-7
Thus5 a curve of Q vs. H could define a converter. However, these characteristic
curves are not sufficient for determining hov it will perform when mated to a
specific engine. Figure 25 presents a set of torque absorption curves for a
spark-ignition (Si) engine. The intersections of the SI torque curve with the
converter curves represent a locus of operating points. If this particular
converter were mated to the single-shaft gas turbine (SSGT), one would again
superimpose the engine torque curve as shown by the dashed line in Fig. 25.
The resulting locus of operating points indicates that over most of the input
speed range, the output speed will be about constant. Also, there is no stall
condition, i.e., zero output speed. Therefore, this converter and the SSGT
are incompatible.
Another means of determining compatibility between engine and converter
is to examine a curve of torque vs. engine speed upon which is superimposed
a typical parabolic converter stall curve which is matched to the maximum engine
torque (Fig. 26). The torque of the SSGT is below the converter stall requirement
over a large percentage of the operating range. In this region, the converter
will tend to reduce engine speed until a torque match can be found. However,
the more the rpm is reduced, the greater the mismatch and eventually the engine
will stall. A smaller converter could be used, but a governor would be required
to control maximum rpm and much less torque would be delivered.
Torque converters can be made with many different torque absorption
characteristics and torque multiplication ratios. One type has torque characteristics
which appear to be more compatible with the SSGT. The primary torque of this
converter at a given input rpm decreases as the output rpm approaches stall
over a range of about O.hO to 0 (output rpm/input rpm). By combining a properly
sized converter and the proper gear reduction ratio for the SSGT engine, a match
between engine and converter can be achieved as shown in Fig. 27. The engine
would not be operated to full rated rpm to circumvent the "curl" at the top of
the curve. Therefore, 10 to 15 hp would be lost at the top end of the operating
range.
Idle conditions must also be examined. With the converter and SSGT engine
defined above, the engine would stall at idle speed when the power train is
engaged. To prevent stall, one could automatically increase the throttle
setting when the shift lever is moved from neutral. Another means of preventing
stall at idle speed is to change the ratio of the reduction gear, thus changing
the engine idle torque to a high enough level to avoid stall. This approach,
exemplified by Fig. 28, leads to a split transmission in which a gear change can
be made before and after the converter. Thus, good idle conditions and full-
power performance would be attainable.
51
-------
L-971249-7
FIG.26
TORQUE CONVERTER STALL CHARACTERISTICS
100 -
Of
o;
O
I-
o
z
U4
O_
(J
o;
LLl
a.
T.C; TpO LARGE
FOR ENGINE
40 60
PERCENT PEAKING SPEED
80
100
-------
L-9712.49-7
FIG. 27
TORQUE ABSORPTION CHARACTERISTICS FOR SINGLE-SHAFT
GAS TURBINE APPLICATION
o
Of
o
OUTPUT RPM
53
-------
L-971249-7
FIG. 28
TORQUE CONVERTER-ENGINE MATCHING
400
360
320
i
to
a
Q-.
o
280
240
200
160 -
120
1200
1400
-IDLE SPEED
1600-
1800
2000
2200
RPM
-------
L-9712U9-7
Another means of providing satisfactory idle conditions may be to use'a
variable-fill converter in which fluid is pumped in and out according to some
schedule. By reducing the mass of the working fluid, the converter characteristics
would be changed to allow idle without engine stall. However, partially filled TC's
tend to "foam" and can rapidly lose efficiency. Thus, one would have to carefully
study this problem and determine to what practical degree the converter could
be emptied.
Traction Drives
Traction drives (TD) have been in use for many years and have been incorporated
in systems ranging from small toys to large industrial machinery. The principle
involves rolling friction, rather than gear teeth, to transfer power. For
small torque levels, dry mating surfaces can be employed. However, as the torque
to be transferred increases, the compressive force between the components
must be increased to prevent slippage, thus producing large amounts of heat
(lost power), which destroys the contact surfaces, and excessive repetitive
Herzian stresses, which lead to fatigue (shelling) failures. The use of lubricants
allows the transfer of high torque without excessive heat build-up. Recent
development in lubricants by Monsanto are reported to have resulted in efficiencies
of about 91%, and can greatly reduce surface stresses by effectively enlarging
the total contact area (Ref. lU).
Transmissions using traction-drive principles have been built on a small
scale and have proven very successful from the standpoints of both efficiency
and endurance. One basic design problem is that of achieving stable zero output
speed on a device which does not also include a clutch and a separate reversing
device.
It appears that traction-drive transmissions would require precision contact
surfaces to ensure smoothness and roundness. This precision is generally
expensive, as compared with a torque converter, for instance, which is made from
stampings and automatically assembled.
A multitude of traction transmission designs have been suggested. The
New Departure Division of General Motors has developed a nonratio-changing gear
reduction unit which automatically increases the compressive forces between
the rolling elements as the torque load increases. Ho data are available
as to costs but efficiencies have been quoted as being over 95$.
The General Motors Hydramatic Division Toroidal transmission was tested
extensively in several automobiles, and is reported to have high efficiencies, long
life, and competitive manufacturing cost. It offers little advantage when coupled
55
-------
L-9712U9-7
TABLE III
TRANSMISSION RATINGS
(Each Transmission Weighted by Category - Max. = 7)
Category
Efficiency
Weight
Smoothness
Durability
Temp. Sensitivity
Eng. Braking
Woise
R&D Costs
Manuf. Costs
Availability
Type of Transmission
Hydroklnetic
S
5
6
5
5
5
6
5
7
5
Hydromechanical
k
6
1
5
k
6
k
1
6
1
Traction
Drive
6
k
1
3**
5
7
6
k
k
h
Mechanical Electric
7 6
k 5*
3 7
it 7
7 7
7 6
5 6
6 2
5 1*
6 2
* Considering elimination of drive shaft, differential, and gear box
** Highly questionable, dependent on R&D effort
-------
L-9712^9-7
with a standard V-8, but might be adapted for a single-shaft gas turbine.
Additional devices are "being tested by TRACOR and AirResearch. Confirmation of
reported efficiency figures for test devices in practical automotive units would
lead to significant improvements for automotive transmissions, regardless of
powerplant type used.
Transmission Ratings
A general comparison of the basic transmissions is shown in Table III.
For each category, the transmissions are rated from 1 to 7, where the higher
numbers reflect in favor of the system. The ratings shown were derived from
both quantitative and qualitative sources, including literature surveys, manufacturer
contacts, and computations. It should be emphasized that these ratings are
very interdependent and could change during development and/or manufacture.
Research and development status is a dominant factor, not only because of the
contract requirement to demonstrate the optimum engine by 1975, hut since progress
can greatly affect the ratings of efficiency, weight, smoothness, durability,
etc. It is currently judged that these ratings should be representative
for reasonable research" and development costs.
Each category will also carry its own ratings as to overall importance
in a program to develop an automotive transmission. When considering the ultimate
goal of customer acceptance of the system, one must surely place high emphasis
on research, development,and manufacturing costs, each of which directly affect
the cost to the customer. Probably as important is the vehicle performance
delivered by the transmission system. A measure of performance would include
smoothness, engine braking capability, temperature sensitivity, and to a lesser
degree, efficiency. Although the engineer, designer, and salesman stress
operating efficiency, rarely does a driver show concern for gas mileage variations
in a given vehicle, as long as the vehicle is running smoothly and starts
reliably. "The driver is seldom aware that worn spark plugs and pitted distributor
points can cost him a 10$-to-30% increase in fuel consumption. The same is
true of low fluid levels and/or clogged filters in a torque converter. Variations
in driving habits can cause significant variation in gas mileage as evidenced
by average driving versus the Mobil mileage runs. Therefore, it is felt that
theoretical operating efficiency is not as important as consistency, smoothness,
reliability, and initial cost.
Similarly, the transmission weight is important if significant weight
savings are attainable which could measurably affect vehicle performance,
vehicle cost, and gas mileage. However, from a fuel economy standpoint,
a 3000-lb automobile with a single-shaft engine will show only a 1 mpg increase
in fuel economy over a 5000-lb automobile.
57
-------
Transmission noise must be related to engine noise rather than absolute
values, since driver awareness to noise is relative. As an example, an HMT,
which is thought to have high pumping noise, is barely audible when installed
in a heavy truck or track-laying vehicle, but might "be the dominant source of
noise in an automobile.
The "availability" category reflects the timetable set by the Federal
Government for meeting the 1976 emissions requirements. If it is assumed
that these requirements will be strictly enforced, then the transmission
development must start almost immediately to allow lead time for "bench" testing
and prototype vehicles. Again, availability is a function of applied development
effort, where a four-year crash program with unlimited funds could probably
develop any transmission considered here by 1976. However, in the initial phase,
it is obviously prudent to use an available unit (such as the HMT, rated "7")
rather than one requiring much R&D effort (such as the electric system, rated
"2").
An importance (or weighting) factor, ranging from 1 to 5, was also assigned
to each of the ten categories as follows:
Category Weighting Factor
Efficiency
Weight
Smoothness
R&D Costs
Manufacturing Costs
Durability
Braking
2
1
1+
5
5
k
3
Temperature Sensi-
tivity
Availability
Noise
58
-------
L-9712^9-7
These weighting factors were multiplied by the rating factors assigned in
Table III to obtain the modified relative ratings which are shown in Table Ilia
along with the total rating for each transmission.
This analysis favors the HMT mainly because of availability, smoothness of
operation, and because units have been built under other programs which
substantially reduces the required development costs and technological risk
for a 1975^demonstration.
Differential
For the transmission systems studied (except electric), a mechanical differential
system is employed to split the total torque between the driving wheels and to also
provide a means of compensating for speed variations between the wheels, such
as when turning a c orner.
The system employed in this study is an epicyclic gear system (planetary)
(Fig. 29), where the ring gear is driven by a silent-running chain connected
to an output pinion from the transmission. This system is similar to that
currently used in General Motors' front-wheel drive cars, such as the Oldsmobile
Toronado. One axle is connected to the sun gear and the opposite axle is
connected to the carrier which contains the planet gears. To provide proper
rotational direction, the planet gears also have a corresponding set of idlers.
This epicyclic system eliminates bevel gearing and the associated thrust loads
on the bearings.
In a standard planetary system, the rotation speed relationship among the
planet, carrier, and ring gear is
usrs = uc(rr + rs)
where ur s c = speed of ring, sun, and carrier,orpm
rr s = radius of ring and sun gears, in.
The chain drive serves three major functions. First, a large distance
between centers can be traversed with low reduction ratios (=3:1) for a small
weight penalty. By incorporating an idler in the chain system, the differential
reduction ratio can be easily changed by changing the transmission output pinion
diameter. The idler is then used to compensate for chain tension.
59
-------
L-9712U9-7
TABLE Ilia
TRANSMISSION RATINGS*
(Each Category Weighted by Importance - Max. = 5)
Category
Efficiency
Weight
Smoothness
Durability
Temp . Sens-
itivity
Eng. Braking
Noise
R&D Costs
Manuf. Costs
Availability
TOTALS **
Rating
2
1
k
k
2
3
It
5
5
h
Type of Transmission
Hydrokinetic
10
5
2h
20
10
15
' 2h
25
35
20 '
188
(7950
Hydromechanical
8
6
28
20
8
18
J-6
35
30
28'
197
(82.8«
Traction
Drive
12
h
28
12
10
21
2h
20
20
16
167
(70.2$)
Mechanical
1U
It
12
16
lU
21
20
30
25
21*
180
(75.6*)
Electric
12
5
21
28
ik
18
24
10
20
8 i
160
(67.2?)
*Each transmission rating = (category rating) x (importance rating)
maximum = 7 x 5 = 35
**100% = 238
60
-------
PLANETARY DIFFERENTIAL
TRANSMISSION-
CHAIN
OUTER PINION
RING GEAR
TO LEFT
WHEEL
DIFFERENTIAL
RING GEAR
INNER PINION
TO RIGHT
" WHEEL
PINION CARRIER
INNER PINION
OUTER PINION
SECTION A-A
TRANSMISSION PINION
Cl
-------
L-9712U9-T
62
-------
L-9712^9-7
CYCLE ANALYSIS AND PRELIMINARY CANDIDATE SELECTION
Task 2 of Phase 1 (see Fig. l) involved a parametric design-point cycle
study of a large number of potential candidate engines for the gas turbine application,
Task 3 comprised the selection of the three most attractive candidates for further
optimization studies in Phase 2.
Engine Cycles Evaluated
A total of 60 engine cycles was evaluated during the parametric design-
point cycle study. These were as follows:
1. Regenerative with large-size regenerator/recuperator (e = 0.90)
2. Regenerative vith medium-size regenerator/recuperator (e = O.TO)
3. Regenerative with small-size regenerator/recuperator (e = 0.50)
U. Simple cycle with single shaft
5. Intercooled with dual shaft
6. Reheated version of free-turbine engine
7. Simple-cycle gas turbine combined with Rankine cycle (COGAS)
8. Simple cycle with variable inlet turbine nozzle area and free turbine
9. Simple cycle with free turbine
10. Simple cycle, single shaft with water injection for power augmentation
11. Simple cycle with power-transfer free turbine
Cycle combinations 1 through 7 above were each evaluated for two levels
of turbine inlet temperature and four levels of cycle pressure ratio, thereby
making a total of 56 cycles among these types. The turbine inlet temperature level
selected represented an uncoole'd version at 1900 F and a cooled version at 2300 F.
The levels of cycle pressure ratio selected were dependent on the cycle and ranged
from U:l to as high as 30:1. The final four combinations 8 through 11, were
evaluated for a single most attractive pressure ratio and turbine inlet temperature,
as determined from the preceding 56 evaluations, for a total of four cycles.
63
-------
TABLE IV
SAMPLE ENGINE iPERFORMAMCE COMPUTER PRINTOUT
________ SI_N@LS_ SPOOL
FNGINF (SSR>> cpyNTERFi_ow_
DESIGN POINT INFORMATION
ALTITUOI
A I RFLOW
T.I.T.
E
1.
0. FT "1
?30 LBS/SEC C
SMgiENTT!
>ELTA"~AW
;MP. 59.0 F '
IENT TEMR -.8 F
1900, F BURNER INLET TEMP. 102B.4 F
PRES. RATIO 8. CO BUHNER INI
DESIGN
RPM 95000. RPM FUEL FLOW
EXH. AREA 8.73 SO. IN. E
TRANS.
.QFF-OES
0/0 HP
. .. 1 ,000
.900
.800
.700
.600
.500
.TOO
.200
.100
.050
.020
.800
.700
.600
.500
.400
,->oo
.200
.050
.400
,30O
.200
.106
.050
.020
.200
.150
.lOO
.050
.0?0
.IOC
.075
,o«o """
AREA 391.0 SO. FT. t
IGN DATA
0/0 Nl
1 . 000
!«poo i .
1.000
... U ooc .
1 .000 -
1.000
tjOOO
1.000
1,000
1.000
1.000
I .000
.980
,950
.950
.950
.990
.950
195.0
. 95ft...
.aoo
.800
.800
.floo
,aSo
,800
,650
i650
.650
.650
.65^
.555"
.500
.506
HP Nl
(HP) £RPM)
1 1 6,9 95OOO,
103.9 95000.
90.9 95000,
78.0 9^000.
65.0 9500O.
39.0 9500O.
26,0 95000.
13.0 950CO.
6.5 95000.
i 2.6 95000.
103.9 9025O.
90.9 90250,
78. O 90250.
* 65. O 90250,
52.0 90250.
30.0 902BO.
26. O 90250.
13.0 9025O.
6.5 9025O.
2.6 9P250.
52.0 76000.
39.0 760OO.
26.0 76000.
13.6 76000.-
6.5 76000.
2.6 76OOO.
26,0 61750.
19,5 61750.
&•= 61750.
13. n 47500.
;~ 9.7 47500.
?',f. 47500.
>ESIGN HOI
EFFECT tvEi
_ET PRES. 1 15.2
62.4
5SEPOWER 13O.O
•1ESS .75
PSIA
LBS/HR
HP
WF BURNER T.I.T. TURB.EXH
.63
34.92
32.22
30.98
3O.07
22.72
20.16
17.60
15.02
13.65
12.62
1025.4 ~ 1900.3
982.5 1823.9
939.8 1747.4
897.9 1670.7
B55.5 1593,4
B15.7 1518.3
778.8. 1445.5
740,8 1372.8
703.6 1299.6
668.2 1227.6
651.4 1192.6
640.5 1 170.6
921.4 1702.0
871.6- 1615. R
823.9 1530. B
.778.3 1447.9
734.9 1367.1
693. n 1288.4
653.2 1209.2
614.4 1131.7
595.8 1093.5
584.2 1O70.O
833.3 1364.3
6§2 .6 11 42 . 9
522.6 884, a
TEMP.
1162.1
1 IO6.3
105O.9
995.7
940.1
887.2
836. 3
787.1
737.1
688.7
665.7
650.9
1037.7
974.2
912.1
852.2
794.8
740.1
685.6
633.3
607.9
592.1
1 147.0
924.5
748.9
619.6
557.7
519.7
EXH. AIRFLOW
TEMP. JLB/SEC)
770.1 1.2304
754.7 1.2317
739.1 1.2322
723. O 1.2318
707.2 1.2316
691.6 1.2310
676.2 1.2300
662.9 1.231O
648.7 1.2311
634.7 1.231!
628.1 1.2311
623.9 1.2311
...
692.5 1.1648
676.5 1.1702
659.8 1.1741
643.5 1.1758
627.7 1.1763
612.5 1.1756
597.3 1.1 765
582.6 1.1763
575.4 1.1762 '
571.0 1.1762
467.6 .4936
4
-------
L-9712^9-7
For each combination, the end output item consisted of performance data at
full- and part-load operating points, including specific fuel consumption and
output shaft rpm vs. horsepower. The tabular design point output included
airflow, tur"bine inlet temperature, pressure ratio, burner inlet temperature,
burner inlet pressure, fuel flow, exhaust area, heat exchanger surface area, and
heat exchanger effectiveness. A sample output is shown in Table IV.
. The rationale for the, selection of the cycles is described in the following
paragraphs. ;
Turbine Inlet,Temperature - .•;•..
Central1 to the problem of automotive gas turbines is the question of maximum
allowable turbine inlet temperature which involves a discussion as to whether
turbine blade- cooling technology, should be advanced to the stage where highly
copied turbines are a practical proposition or whether only uncooled turbines
should: be considered. Of the two turbine inlet temperatures selected for study
(1900 and 2300 F), the 1900-deg level represents current state of the art for ;
the gas turbines under consideration, where this maximum value is attained only
at the full power point at any operating speed and does not represent a continuous
operating line. • The higher turbine inlet temperature, which is'typical of the
current state of the art for internally cooled blades of aircraft gas turbines,
was not necessarily believed to be cost effective for an automobile engine, but
was selected to 'quantify its relative advantages or disadvantages.
•i '
Cycle Pressure Ratios
• All engines employing heat exchangers were evaluated at cycle pressure
ratios of k-, 6, 8, and 10, whereas simple-cycle engines Were evaluated at
pressure ratios of 6, 10, 15,' and 30. Single-stage compressors were assumed
for all pressure ratios up to and including 10:1, and two-stage compressors were
considered for all other pressure ratios. Single-stage radial turbines were
considered for the first stage of expansion in all engines with pressure ratios
up to 10:1. Free-turbine engines.consisted of a radial turbine driving the gas
generator shaft, and an axial-stage power turbine. ,j
Description of Cycles .Evaluated ' , ' .
The first three cycles above (regenerative) were chosen to evaluate the heat
exchanger engine. It was determined that, for the purpose of the preliminary
evaluation, there would be'no,;need to differentiate between recuperated and
regenerated engines; therefore, only one set of data, representative of both,
-------
L-9712l*9-7
was run. The large-size regenerator had an effectiveness of 90%; the medium
size regenerator had an effectiveness of 70%; while the small regenerator
had an effectiveness of 50%. The fourth engine type (simple-cycle single-shaft)
was selected as being similar to an engine in a previous study (Ref. l) which
provided acceptable fuel consumption characteristics at a minimum manufacturing
cost. The fifth engine cycle (intercooled with dual shaft) was selected to see
if there were advantages in efficiency to be gained by intercooling the flow
between stages of a two-stage compressor. The sixth cycle, which was a reheat
version of the free-turbine engine, was selected since it is assumed that full-
power requirements will size the basic engine, whereas fuel consumption
and emission requirements at partial load will determine its other characteristics.
The reheated engine, with its characteristic high specific power, would thus have
a smaller basic engine size and a smaller compressor and, consequently, would
be working at a higher percentage of full (no reheat) power during the typical
mission, with a corresponding improvement in specific fuel consumption at the
average mission power. The associated high specific fuel consumption at the
design point due to adding heat at a point in the cycle where the pressure is
low is likely to be of little consequence in normal automobile applications ,
where operation at full power is of very limited duration.
The seventh cycle, a simple-cycle engine combined with a Rankine-cycle
engine operating from exhaust heat, was included in the evaluation for the following
reasons. It is the system with the greatest assurance of meeting 1976 emissions
standards and offers additional advantages, when compared with the simple-cycle
engine, with regard to exhaust cooling, silencing, torque-speed characteristics,
and heat for the car interior- When compared with the regenerated engines, it
would offer greater power per pound of airflow and greatly improved emissions
for a comparable level of fuel economy. At the same time, when compared with
a straight Rankine cycle, the system would be smaller, less costly, lighter,
and permit startup without undue complexity.
The last four cycles were evaluated at a 10:1 pressure ratio at 1900 degrees
turbine inlet temperature. They included a, free-turbine engine with and without
variable inlet turbine nozzles, and a power transfer device to transfer power
in scheduled amounts from the free turbine to the single-shaft engine as has
been demonstrated in the General Motors GT309 engine. One of the engines
evaluated considered using water injection for power augmentation. This engine
shares, with the reheat engine, a similar advantage of producing a higher
specific output and therefore a smaller basic engine size than the simple-cycle
engine, thereby resulting in expected improvements in cost and fuel economy.
The problem of obtaining pure water for injection was not considered in the
preliminary evaluation since it is a problem which would have to be faced only
if the cycle showed sufficient advantage for further design study.
66
-------
L-9712U9-7
Selection Criteria
The selection criteria which were applied to all 60 engine cycles and used
to select the three engines for further study in Phase II of the program are
described "below. The final rankings were expressed in dollars, representing the
total engine-related lifetime costs (TLC) for seven years and 105,000 miles of
automobile life. These costs represent approximately one-fourth of the total
lifetime cost of the vehicle, as reported in Ref. 15, appropriately adjusted for
the cost of money. Therefore, an engine-related cost ranking which was below
lUO$ of present experience represents compliance with the contract requirement
that overall cost of operation not exceed 110$ of the cost of operation of
an equivalent 1970 automobile, if it is assumed that the other 75$ of the expenses
do not vary with engine type. The lifetime cost calculations were derived as
follows. First, it was assumed that the total lifetime engine-related costs
were: all fuel consumed by the engine, and engine-related costs of depreciation,
maintenance, cost of money, and taxes. The fuel costs were derived, as detailed
below, from calculations from the mission analysis program using the fuel
consumption figures which were the output of the engine cycle analysis program
of Task 2. The remaining items are as derived from the following table which
resulted in a total cost of 235$ of the basic engine cost, using an average of
the high and low tabular values:
Item Related to Lifetime Costs as a Percentage
Price of Engine of Engine Cost
Depreciation
Maintenance 70$* 100$
Sales Tax
Property Taxes 0$ 15$**
Excise Tax
Interest Lost/Finance
Totals
Average 235$
as per Ref. 1 *** interest lost at 6$
5$ of annual value of engine **** finance charge of 12$ for
3 years, plus 6$ interest lost
on equity; refinanced once
67
-------
L-9712U9-T
TABLE V
DERIVATION OF EHGIME COMPLEXITY FACTOBS
Direct Manufacturing Costs (DMC)
SSS-12 Engine with Direct Labor at
$5.50/hour (9.15^/min)
Compressor Rotor
Hot Sheet, etc.
Turbine
Gears, Bearings,
Seals
Assembly, Test,
Miscellaneous
Parts
Diffuser Case
Air Inlet Housing
Accessory Drive Cover
Pressure Vessel
Indueer, Impeller
Turbine Shroud
Scroll
Fuel Nozzle
Burner Liner
Turbine Shroud
Compressor Shroud
Star Exducer
Nozzle
All Gearing
Compressor Shaft
Output Shaft
Bearings-
Seals
/ _1 Others
DKC
CF
U6.81
9-38
38.71
23.t
16.18
0.29
0.06
0.20
0.20
0.15
0.10
CFg Total 1.00
-------
L-9712U9-7
The basic engine cost was derived by, first, estimating the manufacturing
cost of the engine. This value vas obtained by multiplying the cost of the
simple-cycle engine, derived carefully in Ref. 1, by a complexity factor
estimated as a function of the needed components of each engine relative to the
configuration.of the simple-cycle engine, as detailed below. A retail price
of each engine vas then derived by multiplying its manufacturing cost times 2.7
to obtain an average retail price of that portion of the vehicle believed due
to the engine. This price vas finally multiplied by 235?> the estimated lifetime
cost factor, as derived above.
The total lifetime cost was then derived by summing the costs of the engine,
the cost of its control, and the cost of fuel.
The fuel cost derivations were made by assuming that mileage driven was
one-half urban and one-half rural as per national averages reported in Ref. l6.
It was assumed that the urban mileage was represented by the Federal Driving
Cycle and that rural mileage was approximately equal to the UARL country cycle
of Eef. 1 which is approximately equivalent, for fuel economy, to driving at
70 mph. An adjustment was made to reduce the mileage reported in Ref. 15
to that more representative of national averages as reported in Ref. 16.
Turbine fuel was assumed to weigh 7 Ib/gal and was costed at 33^/gal. Gasoline,
for Otto-cycle engines, was assumed to weigh 6.2 Ib/gal and was costed at 35^/gal.
The engine complexity cost factors were derived in the following manner: From
Ref. 1 the direct manufacturing costs of the simple-cycle single-shaft engine
(SSS-12) were normalized to the value 1.0. This derivation is shown in Table V.
The derivation of complexity factors for all engines was then derived from this
base of data by adjusting for airflow, heat exchanger surface area, and design
changes required.. In general, differences in airflow were accounted for by
adjusting costs by the square root of the ratio of the airflow of the candidate
engine to that of the original SSS-12 of Ref. 1. Heat exchanger areas were
compared by contrasting effectiveness to that of the RSS-7 of Ref. 1 and then
adjusting for the area of the heat exchangers by the cube root of the surface
area of the heat exchangers. Free-turbine engines were costed on the basis of
comparisons between the SSS-12 and SFT-12 of Ref. 1. Controls were compared on
a complexity basis according to estimates furnished by Hamilton Standard. The
tabulation of engine complexity factors for all 60 engines is shown in Table VI.
The engine costs as derived from the complexity factors, plus fuel and
control costs, are presented in Table VII along with the values of total lifetime
costs for all 60 engines plus the 1970 Otto-cycle engine (OC-70). The total
lifetime costs are also shown in Fig. 30 as a function of pressure ratio.
-------
L-971249-7
TABLE VI
HIGIHE COMPLEXITY FACTOBS
Complexity Factors
Series Designation
1. HSS-4 (50?)
RSS-6
RSS-8
RSS-10
RSS-4C
RSS-6C
RSS-8C
RSS-J.OC
2. RSS-4 (75?)
RSS-6
RSS-8
RSS-10
RSS-4C
RSS-6C
HSS-8C
RSS-10C
3. RSS-4 (90%)
RSS-6
RSS-8
RSS-10
RSS-4C
RSS-6C
RSS-8C
BSS-10C
4. sss-6
SSS-10
ESS-15
SSS-30
SSS-6C
SSS-10C
SSS-15C
SSS-30C
5. SFT-6I
SFT-10I
SFT-15I
SFT-30I
SFT-6IC
SFT-10IC
SFT-15IC
SFT-30IC
6. SFT-6R
SFT-10B
SFT-15H
SFT-6RC
SFT-10FC
SFT-15KC
SFT-30BC
7. WSS-6
WBS-10
WSS-15
WSS-20
wss-6c
WSS-10C
wss-isc
WSS-20C
8. SFT-10VG
9. srr-io
10 . SSS-10W
11. SFT-10PT
VAHEX
1.6/106
1.4/92
1.45/96
1.51/102
1.38/73
1.17/62
1.19/64
1.21/65
1.67/324
1.4/275
1.47/285
1.52/297
1.44/204
1.19/173
1.21/175
1.23/179
1.85/1008
1. 47/786
1.50/807
1.54/870
1.6/557
1.24/1(28
1.24/435
1.25/454
1.1
1.22
1.35
2.65
0.93
0.98
1.03
1.51
2.23«
2.0*
1.0
1.33
0.91
0.81
0.78
0.88
1.06
1.01
0.99
0.96
0.91
0.87
1.39
0.71
0.82
0.92
1.15
0.52
0.57
0.61
0.68
1.22
1.21
1.22
1.22
Gears , Bearings
Casings
0.38
0.34
0.35
0.36
0.31
0.27
0.28
0.28
0.56
0.49
0.51
0.52
0.45
0.39
0.39
o.4o
0.87
0.71
0.73
0.75
0.66
0.53
0.54
0.55
0.28
0.30
0.39
0.57
0.26
0.27
0.34-
0.41
0.53
0.50
0.5
0.57
0.48
0.45
0.44
0.47
0.56
0.55
0.60
0.53
0.52
0.57
0.72
0.23
0.25
0.32
0.36
0.20
0.21
0.26
0.28
0.48
0.48
0.48
0.48
Compressor
0.04
0.06
0.07
0.07
0.04
0.05
0.06
0.06
0.05
0.06
0.07
0.07
0.05
0.05
0.06
0.06
0.05
0.06
0.07
0.07
0.04
0.05
0.06
O.OT
0.05
0.06
0.14
0.20
0.05
0.06
0.12
0.14
0.10
0.12
0.12
0,13
0.09
O.oO
0.10
0.11
0.05
0.06
0.12
0.05
0.06
0.11
o.i4
0.04
0.05
0.11
0.12
0.03
0.04
0.09
0.10
0.06
0.06
0.06
0.06
Hot Sheet
0.26
0.24
0.24
0.25
0.24
0.22
0.22
0.22
0.26
0.24
0.25
0.25
0.24
0.22
0.22
0.22
0.27
0.25
0.25
0.25
0.26
0.23
0.23
0.23
0.20
0.21
0.23
0.33
0.19
0.19
0.20
0.24
0.21
0.19
0.19
0.22
0.19
0.17
0.17
0.16
0.40
0.39
0.39
0.38
0.37
0.36
0.46
0.16
0.18
0.19
0.21
0.14
0.15
0.15
0.16
0.21
0.21
0.34
0.21
Turbines
0.23
0.22
0.22
0.23
0.62
0.57
0.57
0.58
0.24
0.22
0.22
0.23
0.63
0.57
0.58
0.58
0.25
0.22
0.23
0.23
0.66
0.58
0.58
0.59
0.19
0.20
0.21
0.31
0.51
0.52
0.53
0.64
0.46
0.44
0.44
0.50
0.74
0.70
0.69
0.73
0.65
0.63
0.63
0.95
0.93
0.91
1.15
0.16
0.17
0.18
0.20
0.38
o.4o
o.4i
0.43
1.07
0.48
0.20
0.48
and Seals
0.18
O.lff
0.18
0.18
0.18
0.18
0.18
0.18
0.18
0.18
0.18
0.18
0.18
0.18
0.18
0.18
0.18
0.18
0.18
0.18
0.18
0.18
0.18
0.18
0.18
0.18
0.18
0.18
0.18
0.16
0.18
0.18
0.54
0.54
0.54
0.54
0.54
0.54
0.54
0.54
0.36
0.36
0.36
0.36
0.36
0.36
0.36
0.18
0.18
0.18
0.18
0.18
0.18
0.18
0.18
0.36
0.36
O.lB
0.69
HEX
0.76
0.68
0.70
0.73
0.62
0.54
0.55
0.56
1.13
0.98
1.01
1.04
0.90
0.77
0.78
0.79
1.73
1.42
1.46
1.50
1.32
1.07
1.07
1.09
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.1
0.09
0.09
0.10
0.09
0.08
0.08
0.08
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.6
0.6
0.6
0.6
0.6
0.6
0.6
0.6
0.00
0.00
0.00
0.00
Ml
0.10
0.1
0.1
0.1
0.12
0.12
0.12
0.12
0.12
0.11
0.11
0.11
o.i4
0.13
0.13
0.13
o.i4
0.13
0.13
0.13
0.15
0.14
0.14
0.14
0.08
0.08
0.10
0.13
0.10
0.10
0.11
0.13
0.16
0.15
0.15
0.17
0.17
0.17
0.17
0.17
0.16
0.16
0.17
0.18
0.18
0.18
0.22
0.08
0.08
0.09
0.10
0.09
0.09
0.10
0.11
0.17
0.13
0.11
0.15
WE
TMAL
1.95
1.62
1.86
1.92
2.13
1.95
1.98
2.0
2.54
2.28
2.35
2.40
2.59
2.31
2.34
2.36
3.49
2.97
3.05
3.11
3.27
2.78
2.80
2.85
0.98
1.03
1.25
1.72
1.29
1.32
1.48
1.74
2.1
2.03
2.03
2.23
2.30
2.21
2.19
2.28
2.18
2.15
2.27
2.45
2.42
2.49
3.05
1.45
1. 51
1.67
1.77
1.62
1.77
1.79
1.86
2.35
1.72
1.37
2.07
-------
Engine
0. OC-TO
1. RSS-li (50X)
RSS-6
RSS-8
RSS-10
RSS-4C
RSS-6C
RSS-6C
RSS-10C
2. RSS-4 (75?)
RSS-6
RSS-8
ass- 10
RS3-4C
H35-6C
BSS-8C
R3S-10C
3. RSS-4" (90*)
RSS-6
RSS-8
RSS-10
RS3-4C
ES3-6C
RES-8C
3SS-10C
4 . 3SS-6
SSS-10
SSS-15
SSS-30
S3S-6C
3SS-10C
SSS-15C
SS3-30C
5 . SFT-6I
3PT-10I
SFT-151
SFT-30I
SFT-6IC
siT-ioic
SFT-15IC
SFT-30IC
6. SFT-6R
SFT-10R
SFT-15R
SFT-6RC
SFT-10RC
3FT-15RC
W2S-6
HEE-1G
USS-15
W3S-20
WSS-6C
wr.s-ion
K3S-15C
'033- 20 f!
5ES-10 V.G.
3FT-10
fss-iow
9.5
17.2
17.6
9.1
15.9
15.6
20.1
10.9
18.6
13.2
18.1
18.1
•18.1
17-5
20.0
20.0
22.0
12.6
17.7
19.3
22.5
7.3
9.2
10.3
6.7
7.6.
8.1
10.0
9-8
5.E
11.7
13.14
7.8
5.8
9.2
10.fi
9.6
5.7
7.1
8.0
7.1
8.1
10.2
16. 4
I6.li
16. li
10.6
17. li
17.2
17.8
15.3
9.2
9.2
9.2
9.2
TABLE VII
TOTAL LIFETIME COSTS
Rural MPB
12.2
13.6
16.9
17.5
17.2
12.0
15.1
.l&J,
16.7
17.9
20.2
18.5
18.5
16.1
18. It
18. li
18.3
20.5
22.8
21.5
20.6
15.2
23. 4
18.8
18,8
12.2
13.5
lit. 3
12.1
11.2
12.7
13.7
14.0
10.1
13.8
15. 4
12.9
10.2
13.0
15.1
.lit. 6
11.1
-42.7
13.2
10.9
12. It
14.5
3.5
19.2
19.2
19.2
16.1
20. ll
21. 3
21.9
21.0
13.5
13.5
13.5
13.5
CFj.
l.Oll
0.87
0.70
0.65
0.65
1.07
0.80
B.J-3
0.70
0.69
0.56
0.61
0.61
0.77
0.62
0.62
0.62
0.60
0.53
0.55
0.53
0.82
0.60
0.60
0.55
1.17
1.00
0.92
1.21
1.21
1.09
0.96
0.95
-1.1,3
Q.89
0.79
1.10
1.142
1.02
0.88
0.94
1.35
1.15
1.07
1.26
1.11
0.92
1.65
0.64
o.6ii
O.filt
0.85
0.60
0.59
0.57
0.63
1.00
1.00
1.00
.1.00
Fuel
Cost HI
.2652
2218
1785
1658
1658
2728
20140
1862
1785
1760
1428
1556
1556
196k
1581
1581
1581
1530
1352
1402
1352
2091
1530
1530
ll|D2
2981i
2550
23!46
3086
3086
2780
21*8
2422
3646
227.0
2014
2805
3621
2601
224 ll
2397
31*2
2932
2728
3213
2830
23*6
Ii208
1632
1632
1632
2168
1530
1501
Ilt5lt
1606
2550
2550
2550
2550
OTj
0.90
1.9k
1.82
1.86
1.92
2.13
1.95
1.98
2.00
2.54
2.28
2.35
2.40
2.59
2.31
2.3lt
2.36
3.1i9
2.97
3.05
3.11
3.27
2.78
2.80
2. aii
0.98
1.03
1.25
1.72
1.29
1.32
l.ltS
1.7ll
2.10
2.03
2.03
2.23
2.30
2.21
2.19
2.28
2.18
2.15
2.27
2.145
2.112
2.149
3.05
1.145
1.51
1.67
1.77
1.62
1.77
1.79
1.86
2.35
1.72
1.37
2.07
Engine Control
Cost (t) CFC Cost (t)
91)5 0.50 235
20lt8 1.50 705
1911
1953
2016
2237
20148
2079
2100
2667 1.50 705
239lf
?lt68
2520
2720
2^26
2457
2478
366k 1.50 705
3118
3202
3266
3434
2919
2940
2992
1029 1.00 470
1082
1312
1806
1354
1366
1554
1827
2205 1.08 508
2132
2132
2342
2415
2320
2300
2394
2289 2.05 964
2258
2384
2572
254l
2614
3202
1522 1.50 705
1586
1754
1859
1701
1859
1880
1953
2468 1.45 682
1806 1.28 602
1428 1.15 540
2174 1.85 870
Total
Lifetime
Cost .($)
3832
4971
4401
4316
4379
5669
4793
46!|6
4590
5132
4527
4729
47.81
5389
4712
4743
4764
5899
5175
5309
5323
6230
5154
5175
5099
4483
4102
4128
5362
4910
4636
4472
4719
6359
4910
4654
5655
6544
5429
5052
5299
6695
6154
6076
6749
6335
5924
8374
3859
3923
4091
4732
3936
4066
4039
4264
5700
4958
4528
5594
% of
OC-70
100
130
115
113
114
148
125
121
120
134
118
123
125
141
123
124
124
15k
135
139
139
163
134
135
133
117
107
108
140
128
121
117
123
166
128
121
148
171
142
132
138
175
161
159
176
165
155
219
101
10?
107
123
'103
106
105
111
149
129
118
146
71
-------
L-971249-7
ENGINE-RELATED TOTAL LIFETIME COSTS
(UNCOOLED TURBINE, 1900-F MAX. INLET TEMP.
ENGINE DESIGNATIONS
SERIES
1
2
3
4
5
6
7
8
9
10
II
REGENERATED,
REGENERATED,
REGENERATED,
SIMPLE CYCLE,
SIMPLE CYCLE,
SIMPLE CYCLE,
COG AS,
SIMPLE CYCLE,
SIMPLE CYCLE,
SIMPLE CYCLE,
SIMPLE CYCLE,
f = 50%, SINGLE SHAFT
( = 75*,SINGLE SHAFT
( = 90%, SINGLE SHAFT
SINGLE SHAFT
INTERCOOLED, FREE TURBINE
REHEAT, FREE TURBINE
VARIABLE TURBINE
FREE TURBINE
WATER INJECTED
POWER TRANSFER
O
u
ui
u.
7000
6000
5000
O 4000
3000
• SELECTED PHASE II
STUDY ENGINES
10 15
PRESSURE RATIO
20
25
30
72
-------
L-9712^9-7
Preliminary engine selections for Phase II study are noted on Fig. 30. However,
slight modifications, principally with respect to the low-effectiveness regenerated
engine (Series l) were made at the request of the Office of Air Programs, and
the selected configurations are as follows:
RGSS-6 - Regenerated, S_ingle-S_haft engine, (5:1 pressure ratio,
75$ regenerator effectiveness
RCSS-8 - Recuperated, Single-Shaft engine, 8_:1 pressure ratio,
60% regenerator effectiveness
SSS-10 - Simple-Cycle, Swingle-Shaft , 10:1 pressure ratio
Additional Iterations
While some of the engines shown in Fig. 30 appear to show better potential
than the ones chosen (notably Series 7, the COGAS cycle), the final selection
was made on the basis of additional iterations in the selection process as
requested by the Office of Air Programs. Those iterations included an elimination
of the COGAS system on the basis of technical and economic uncertainties. They
also included further consideration of free-turbine engines, taking into account
transmission effectiveness and costs required. The final selection also considered
the use of the OAP-furnished driving cycle which is described in more detail
in a subsequent section. In the case of the fuel cost comparisons, the effect
of the OAP-supplied driving cycle (which consists of equal times over the FDC ,
the suburban driving cycle, and the rural -driving cycle) is to increase overall
fuel consumption and to favor more strongly the part-load operation of the engine.
It was found that the fuel costs for the Series h SSS-10 engine increased by
12%, and those of the Series 1 RSS-8 and the Series 2 RSS-6 increased by 6%.
The total lifetime cost comparisons are then as follows.
Series Engine Original TLC Revised TLC Selected Engine
It SSS-10 $tt!02 $ltlt02 SSS-10
1 RSS-8 U316 I*ltl3 RCSS-8
2 RSS-6 It527 ^619 RGSS-6
Since it was obvious that changes of this magnitude would not affect
the choice of engines to be studied in Phase II, no further revisions to the original
calculations were made. The OAP-supplied driving cycle was, however, used for
the economic analysis and the optimizations which were performed in Phase II.
73
-------
L-97121*9-7
Comparison of Free-Turbine and Single-Shaft Engines
Comparisons relative to the transmission complexity required for free-
turbine and single-shaft engines were developed subsequent to the above work.
These comparisons confirmthe elimination of the free-turbine version as described
below.
The direct manufacturing cost (DMC) estimates for presently used automatic
transmissions for UOOO-lb automobiles are as follows:
1. Three-speed with torque converter - $62
2. Four-speed with torque converter - $75
3. Torque converter alone - $15
Therefore, a four-speed automatic for a free-turbine engine would have a
DMC of approximately $60 if the torque converter were eliminated. This transmission
would provide fuel economy approximately comparable with the single-shaft engine
coupled to an infinitely variable transmission which allows the engine to
follow its optimum fuel flow schedule, since the four-speed transmission coupled
with the free turbine would not allow the free-turbine engine to follow its
optimum fuel control schedule, and the resulting fuel consumption degradation
of from 8 to 10/5 negates the inherently greater efficiency of the geared
transmission. The three-speed transmissions without torque converter coupled
to the free-turbine engine would result in an additional fuel economy degradation
of approximately k to 5$ for the free-turbine engine.
Either of the above resulting free-turbine propulsion systems is still
unsatisfactory for automobile use since engine braking (negative torque) is
not provided. The lowest-cost approach to this problem would probably be to
add a hydraulic retarder to the transmission. This fluid coupling, plus piping,
radiator, and control would have a DMC of at least $10. If this braking method
is unacceptable from a safety standpoint, i.e., if negative torque must be
supplied by the engine itself, the best alternative is to use variable geometry
in the power turbine nozzles at an estimated DMC of $32.50 plus an additional
engine control DMC of $8. In addition, the variable geometry results in an
estimated fuel consumption penalty of 3 to h%. Since these higher costs are
clearly unacceptable, the most likely transmission for the free-turbine engine
appears to be a four-speed automatic without the torque converter but with a
hydraulic retarder with a consequent estimated DMC of $70. On a basis consistent
with the previous calculations, this DMC would amount to a complexity factor for
the transmission of CFT = 70/165 = 0.1*2.
71*
-------
L-9712^9-7
In the case of the single-shaft engine, the transmission assumed is the
hydromechanical transmission (HMT) designed by General Electric. It is possible
that its DMC will not exceed that of the present transmission and, based on an
in-house manufacturing survey, it appears that the $90 DMC assumed is conservative,
The CFT = 90/165 = 0.55 for that assumption, and the fuel economy for this
propulsion system, as mentioned above, is comparable with the free turbine with
the four-speed transmission selected above.
Using the general methods previously derived, an overall complexity factor
may be assigned to both single-shaft and free-turbine versions of the engine by
adding the engine complexity factor from Table VI to that of the transmission
required as described above. This procedure is followed below for the Series
2 RSS-8 engine and a free-turbine version of this engine, designated the RFT-8.
VAHEX
RSS-8 1.1+7/^85
RFT-8 same
Casings
0.51
0.66
Compr .
0.07
same
Hot Sheet
0.25
same
Turbitif
0.22
0.50
L_
GB&S
0.18
0.36
HEX
1.01
same
A&T
0.11
0.12
CF
..35
2.98
CI't
Trans
0.55
0.1+2
c*,
2'
3
From the above it can be seen that the cost of the transmission assumed for
the single-shaft engine would have to almost double for its complexity factor
to equal that of the free-turbine engine. The resulting DMC of $172 for the
transmission appears to be completely unrealistic.
As a result of this analysis, it is concluded that the free-turbine version
cannot be competitive with the single-shaft version even when the most favorable
economic assumptions are made for the free-turbine and the least favorable are
made for the single-shaft engine. Therefore, the selection of a free-turbine
engine for Phase 2 optimization was not justifiable on either a technological
or economic basis.
75
-------
L-9712^9-7
76
-------
L-9712U9-7
DESIGN
Engines
The three engines designed for the Phase 2 optimization comparisons are
quite similar in all respects except for the inclusion of the heat exchanger.
All engines are based on a single-shaft design with single-stage centrifugal
impellers and single-stage radial turbines. All are designed to be robust, low-
cost in construction consistent with high performance, and simple in concept and
execution. All three designs make the minimum use of variable geometry. In
each case, the only variable geometry is that for the compressor inlet guide vanes.
These inlet guide vanes might possibly be eliminated during the development
program as a result of further compressor development improvements, or they might
be included in high-performance versions of a basic engine only. Their addition,
including control complexity, raises the cost of the engine less than 10% and
they provide a powerful means of optimizing engine operation during at least the
development phase without the cost, performance, and mechanical reliability
penalties of variable-geometry hot parts, such as power turbine nozzles. It is
possible that some variable geometry might be associated in the combustor design,
particularly for the regenerated and recuperated engines. For current lack of
definition, this possibility was not included in the present design studies.
Specific comments on engines follow.
Simple-Cycle Engine (SSS-10)
The simple-cycle engine is a single-shaft machine with all radial turbo-
machinery as are all Phase 2 study engines. It achieves a pressure ratio of
10:1 and a design mass flow of 1.15 Ib/sec with a rotor speed of 106,000 rpm.
Design and off-design point operational characteristics are shown in Table VIII,
together with comparable detailed information for the other two study engines.
The layout drawing of the SSS-10 engine is shown in Fig. 31 in cross section.
The end view of the engine installed transversely with the GE HMT is shown in
Fig. 32. An identical drawing numbered and keyed to a parts list is shown later,
in Fig. 109-
The air intake is configured radially to achieve minimum installed length,
and includes 18 variable inlet guide vanes. The inlet guide vanes are produced
from plastic or aluminum strip stock and their actuation is automatically
controlled by the fuel control unit. Structural support across the intake is
provided by a series of axial struts. This configuration provides maximum .flexi-
bility for intake ducting.
The compressor is a centrifugal impeller of forged titanium alloy with 18
full vanes and 18 partial vanes, or splitters. Titanium is specified because of
77
-------
TABLE VIII
ENGINE OPERATIONAL CHARACTERISTICS
- Standard Day
Includes All Installation Losses
ro
fr
I
Engine
RGSS-6
RCSS-8
SSS-10
Power
(hp).
2.6
19.5
130.0
2.6
19-5
130.0
2.6
19.5
130.0
H
(Krpm)
1*1.0
57.1*
82.0
vr.5
66.5
95.0
10.0
7l*. 2
106.0
Wf
(Ib/hr)
1*.51
15.79
69.00
it. 88
17.10
73.59
9.8?
22.1*6
82.16
Burner
Inlet Temp.
(F)
761
723
1038
620
636
950
211*
3W
692
TIT
(F)
987
1130
1900
877
1098
1900
959
1266
1900
Airflow
(Ib/sec)
0.1*2
0.79
1.1*9
0.38
0.71
1.36
0,27
O.UY
1 . 21
Fuel /Air
Ratio
0.0033
0.0060
0.0139
0.0036
0.0068
0.0152
0.0105
0.0131*
0.0191
Equiv.
Ratio
e
0.92
0.81*
0.75
0.82
0.72
0.60
~
Coripr ,
P.R.
1.70
2.76
6.00
1.97
3.50
8.00
1.99
3.1*5
10.00
-------
SSS-10, HMT LAYOUT
REDUCTION GEAR BOX
COMPRESSOR
-------
SSS-10, HUT POWER SYSTEM
HOOD U«E__..-
FUEL INLET
SPARK PLUG
CONNECTION
CO
o
, ACCESSORIES
I DRIVE SHEAVE
' OIL PUMP
£
«SP
I
-------
L-9712^9-7
its superior fatigue strength in comparison -with current steels. The particular
grade of titanium (6 Al-2 Zn-9 Zr-6 Mo) was selected over the more common 6 Al-Uv
(used in the RGSS-6 and the RCSS-8) because of the higher temperatures encountered
in the 10:l-pressure-ratio engine. The compressor, mounted on the turbine shaft,
is driven by a friction joint provided by the axial lockup of components on the
turbine shaft.
The diffuser passages are integrally machined in the ductile cast iron
diffuser case with leading edges machined in a cast-in stainless steel ring to
Improve strength. The diffuser passages consist of straight and conical sections5
which intersect tangentially, providing a curved leading edge to the flow of
gases leaving the impeller tip.
The combustor is a single-can design selected for reliability and low poten-
tial exhaust emissions. The gases are ducted from the combustor to the radial
turbine nozzle by a toroidal scroll. The length of the combustor provides for
a lengthened secondary combustion zone to thoroughly combust carbon monoxide and
unburned hydrocarbons, Both the combustor and the turbine entry duct may be
manufactured from either Hastelloy-X or INCO-625, depending on production economics.
The radial turbine nozzle is a one-piece investment casting with 15 airfoils.
There are several alloy options for this component, including MAR-M-509 with MDC-9
coating, IHCO-738 with PWA-73 coating, or WI-52 with MDC-9 coating. The final alloy
selection would be based upon production economics and performance. The front and
rear turbine shrouds may be constructed from the same material as the turbine
nozzle or consideration may be given to using reaction-bonded silicon nitride
ceramic as a potentially lower-cost material.
The turbine rotor would be manufactured from forged Udimet-700 and inertial
welded to the AMS-U3^0 shaft. Among the high-production volume forging techniques
which might be developed for this application, consideration should be given to
superplasticity forging.
Exhaust gases are ducted away from the engine by an annular exhaust duct,
part of which is shown in the cross section (Fig. 3l). A straight diffusing length
of approximately 9 inches is required. This duct may be manufactured from
aluminized mild sheet steel.
The gearbox shown in Fig. 31 is configured to suit the particular installation
as is discussed both in the previous section on base-line technology and in the
following sections on installations. However, it is representative of the
components required for a wide variety of installations. If a large offset were
not required between the engine centerline and the output shaft, one gear and
shaft could be eliminated from the gearbox. The gear train is shown as a double-
reduction unit with an idler gear giving an output shaft speed of 3200 rpm.
81
-------
RGSS-6, HMT LAYOUT
SECTION A-A
COMBUSTOR
REDUCTION
GEAR BOX
(SEE FIG.Tf)
-------
-. 971249-7
Another possibility is to replace the idler gear with a chain drive such as is now
in use on such front-end-drive vehicles as the Oldsmobile Toronado. The initial
reduction gear mesh requires carburized and ground gear teeth. Because of their
lower pitch line velocities, the last two gear meshes may possibly make use of
carburized, unground gear teeth.
All shafts are supported on plain bearings for low cost with the exception
of the main engine shaft itself. Rolling-element bearings were retained for
the input shaft since power losses in plain bearings at this speed become excessive.
The specification of three main shaft bearings as shown in the layout, Fig. 31,
could hopefully be reduced to two bearings.
Plain bearings were specified for low cost, except for the main shaft, where
power losses would otherwise be excessive. Further engine optimization is
expected to reduce the three main shaft bearings shown in Fig. 31 to two angular
contact bearings, since designs for these have been satisfactorily demonstrated.
For example, similar bearings designed for the Army's Mobile Equipment Research
and Development Center 10 Kw Turbo-alternator have run without failure for more
hours (U636 vs. 3500 design) at higher speeds (93,500 rpm vs. 7^,000 rpin average).
Accessories are mounted on or are near the gearbox and are shown as belt-
driven (Fig. 32). Sheaves for belt drives are provided on the 3200-rpm output
shaft. In addition to driving such items as the generator and hydraulic pumps
and perhaps air conditioning drives, a belt-driven starter is assumed feasible for
the engine since starting torques are relatively low and cranking speeds relatively
high. The gear casing contains an integral oil tank of the wet sump type. The
oil pump is mounted on the casing and is directly driven by the idler shaft. A
centrifugal breather impeller and hose connection is provided on the first
intermediate shaft. The oil filter, oil cooler, and breather filter will be
externally mounted and appropriate connections completed by hose. Exhaust from
the breather filter will be connected back into the engine intake to further reduce
emissions and solid oil separated from the air, and the breather filter will be
drained back to the gearbox sump.
All major casings are shown as cast iron for low cost. Hot-end casings are
of ductile .cast iron, while the intake and gearbox casings are of gray cast
iron. Aluminum could be substituted for some of the intake casings should further
weight reductions be desired. The weight for the entire engine, less its accesso-
ries, is estimated to be 2^0 Ib.
The 'performance of the SSS-10 engine is discussed in a subsequent section of
this report.
Regenerative Engine (RGSS-6)
The regenerative-cycle engine is a single-shaft machine with all radial turbo-
machinery. Its layout is shown in Figs. 33 and 31*. Its basic parameters are:
a pressure ratio of 6:1, a mass flow of 1.1*9 Ib/sec, a rotor speed of 82,000 rpm,
and a design regenerator effectiveness of 75% using a regenerator comprising a
single ceramic disc. Structural construction is quite similar to the SSS-10,
except for the addition of the regenerator and its associated ducting passages.
The material for the compressor is specified as titanium for long fatigue life and
resistance to corrosion and is of alloy AMS 1*928 (6 A1-1*V), A conventional single-
83
-------
RGSS-6, HMT POWER SYSTEM
I
-------
L-9712^9-7
can combustor with one fuel nozzle and ignitor is shown for the regenerative
engine. As will fee pointed out later in the study, it is uncertain at the moment
•whether this type of a simple combustor can provide the required emissions
reductions to meet OAP goals. It is most probable that a more complicated
device involving variable geometry and perhaps multiple nozzles and multiple
casings may be required to provide promise of meeting these requirements.
The exhaust gas from the turbine is diffused to reduce its Mach number and
is passed through the regenerator matrix where it gives up heat to the ceramic
core. It is then dumped into a plenum and makes its way to the single exit port.
The' regenerator has a disc 13 inches in diameter and 2,8 inches thick and
is composed of CERCOE material mounted in a steel drum. The regenerator disc
rotates at approximately 20 rpm and is driven by a gearing system consisting of
a worm gear set in the reduction gearbox, a cross shafts and a chain drive to
a spur gear pinion driving a ring gear on the outer drum of the regenerator core.
Other candidate drive methods considered were an electric or hydraulic motors
but these were rejected as being higher-cost and less-reliable drives.
The gearbox is essentially the same as that shown for the simple-cycle engine
except for (a) an increased distance between the engine centerline and the output
shaft centerline to accommodate the larger engine diameter, (b) the addition of a
worm gear set and cross shaft for the regenerator drive, and (c) a slight
difference in reduction ratios required to achieve the 3200 rpm caused by the
somewhat lower shaft speed. Design and off-design point information are shown
for the RGSS-6 in Table VIII. The weight of the engine is estimated to be
520 Ib.
Recuperated Engine (RCSS-8)
The recuperated engine is a single-shaft machine with all radial turbomachinery.
The recuperator is a tube type which is annular in design. Its basic parameters
encompass a, pressure ratio of 8:1, a mass flow of 1.36 Ib/sec, a rotor speed of
955000 rpm, and a heat exchanger design effectiveness of 60%. The layout of the
RCSS-8 engine and cross section is shown in Fig. 35 and an end layout is shown in
Fig. 36.
The turbomachinery design of the RCSS-8 is very similar to that of the
SSS-10 and^RGSS-6 engines. The titanium alloy centrifugal impeller of AMS
k92Q (6 Al-Vv) has been selected for long fatigue life and resistaace to corrosion.
Compressor air is delivered via 20 diffuser passages machined into the
diffuser case to a cast passage which ducts the diffused air to !:,he recuperator.
The recuperator is a double-pass air, single-pass gas type eouFtructed of 2280
U-tubes of 0.125 in. outer diameter and 0.010 in. thickness which are brazed
to a header plate. The recuperator has a cylindrical shape with airflow axial
through the tubes and gas flow radial. The material to be us«J. in the recuperator
is specified as a 300-series stainless steel.
85
-------
RCSS-8, HMT LAYOUT
I-
CO
o\
REDUCTION GEARBOX
(SEE FIG. 31)
COMBU5TOR
SECTION A~A
•TUBE RECUPERATOR
-------
RCSS-8, HMT POWER SYSTEM
SPARK PLUG
CONNECTION
FUEL INLET
-------
L-9712l*9-T
The combustor is a single-can design with a single fuel nozzle and igniter
for good combustion efficiency. In this engine, however, the combustor is buried
beneath the diffuser-to-recuperator passage in order to minimize flow distortion
at the entrance to the recuperator. Its size, therefore, directly influences the
engine outside diameter. In order to minimize the engine diameter, the dilution
portion of the combustor is curved on a radius about the engine axis. The primary
zone remains cylindrical to avoid sacrificing combustion efficiency. The igniter
and fuel uczzle are fitter the flame tube through bosses cast into the diffuser
passage and are accessible from outside the engine.
Exhaust ducting is constructed of ductile cast iron and may be configured
either as a two-port or a single-port configuration as the installation dictates.
The gearbox is essentially the same as that shown for the SSS-10 engine except
that the distance between the engine centerline and the output shaft center has
been increased to accommodate the larger engine diameter. As in the previous
engines5 all major casings are cast iron for low cost, with hot-end casings of
ductile or heat-resistant cast iron, and intake and gearbox casings of gray cast
iron. The weight for the entire engine, less its accessories, is estimated to be
U80 It). .
Power Trains
.• j
The transfer of power from a power source to the traction (or reaction)
members of a powered vehicle should ideally be performed with high efficiency
(small power loss), quietness, and smoothness,, (small time rates of change of
acceleration). The power must also be applied in the proper relationships of
speed and torque'to meet the wide range of conditions", associated with an automobile's
operation. Therefore,.the power output of the engine must be controlled by engine
throttle for varying loads, while the final reaction torque delivered to the wheels
is controlled by a transmission system which is capable of changing the input
torque to provide the required reaction force.
The power train for a gas turbine automobile essentially consists of a gear-
box to reduce the engine speed to the rated speed.of'the transmission, a trans-
mission which can provide varying gear ratios, a differential which provides a
final overall gear ratio and a means of differentially splitting the power for a
two-wheel drive, and, finally, an axle system to deliver the power to the wheels.
The design of these power-transfer components, depends on the type of trans-
mission employed, the installation configuration,-the speed and torque to be carried,
and the gear ratio(s) required.
-------
L-9712^9-7
Gear Seduction Box
Since the gas turbines considered in this study are rated at approximately
100,000 rpm, a large reduction in speed is'required before the power can be input
to the transmission. This reduction ratio is dependent on the type of transmission
employed. The rated speed limits of various transmissions are assumed as follows:
Mechanical - 6000 rpm
Hydromechanical - 3200 rpm
Hydrokinetic - 6000 rpm
In the case of electric transmissions, where the gas turbine drives a generator
(alternator), the reduction ratio is much smaller since generators of advanced
design have been operated at speeds up to 100,000 rpm.
The design of the gear system can take many forms, such as a planetary system,
offset gearing, or double-reduction gearing. While one would tend toward the
lightest-weight configuration, the actual installation characteristics and the
mating of the engine to the transmission through the gearbox must be considered,
as well as total cost.
For this study, a very compact propulsion system configuration can be
achieved by mounting the engine and transmission transversely to the automobile
centerline and side by side. The distance, d, between the engine output shaft
and transmission input shaft varies between IT in. and 20 in., depending on which
of the three candidate engines is employed. The rated engine shaft speed among
the three engines varies from 82,000 rpm to 106,000 rpm. For the specific engines,
the overall reduction ratio for each engine-transmission combination is as follows:
Transmission
Mechanical
17.7
13.7
15.8
HMT Hydrokinetic (TC)
33.1
25.6
29.7
17-7
13.7
15.8
SSS-10
RGSS-6
RCSS-8
Since the rotor of each engine has a high level of rotational energy (inertia),
it is desirable to minimize the added inertia of the reduction gearbox to help
maximize the transient response of the engine to throttle variations.
The kinetic energy, E, of each gear is
E = £ I w2 (1)
-------
where "I" is the moment of inertia of the gear about its rotational axis and "u"
is its rotational velocity. Assuming each gear is a solid disc, the moment of
inertia of each is
(2)
where: m = mass density of gear, slugs/in.
r = pitch radius of gear, in.
t = gear thickness, in.
Combining Eqs. (l) and (2) yields:
(3)
Simple offset gearing for large reduction ratios involves a large final gear
which tends to be heavy and is volumetrically inefficient; therefore, double-
reduction gearing is employed.
The radius of the pinion driven by the engine must be sufficiently small to
avoid supersonic tip speed (creating a noise-producing shock wave and making it
difficult to use spray lubrication) and yet large enough to carry the engine-
produced torque. A one-inch radius pinion rotating at 106,000 rpm would have a
tooth tip speed of 922 fps, which is nearly supersonic. Therefore, the maximum
radius will be limited to about one inch.
The total rotational energy of the four gears in a double reduction system
can be written as
E =
assuming the gears are made of the same material and are the same thickness. Then,
an energy parameter, e, can be defined as:
rl + ~~+^2 (5)
J2 «Q
where: RQ = overall reduction ratio
rl 2 3 4 = pitch radii of four gears, in.
90
-------
For known values of r\ and TZ, the radii of the remaining tvo gears are
.,.
In Eq. (5), the gear face width (thickness), t, was assumed to be constant.
However, the face width is a function of the power to be transferred and the rota-
tional speed. This relationship is expressed as
, 31,500 hp (R+l)3 , v
D2t = k u, R > (7)
where D = distance between gear center lines, in.
t = face width, in.
= input horsepower
R = reduction ratio
u = pinion speed, rpm
= surface durability factor, lb/in.2
Also:
D = rj + r2 (8)
where rj = pinion radius and r2 = radius of mating gear. Equation (8) can also
be written as
D = rj(l + R). (9)
Substituting Eq. (0) into Eq. (7) yields
. _ 31.500 hp|/:
-------
L-971249-7 FIG. 37
INERTIAL ENERGY PARAMETER FOR DOUBLE-REDUCTION GEARING
92
-------
L-9712U9-7
A face width parameter, T, can be defined as
(11)
The second set of reduction gears can be analyzed in a similar manner, yielding
(12)
By applying Eq. (ll) to the first two terms of Eq. (5), and Eq. (12) to the last
two terms, Eq. (5), modified for gear thickness, becomes
(r§ H- rg)
(13)
Substituting Eq. (6) into Eq. (13) yields an expression for e in terms of the
characteristics of the pinion and mating gears.
Figure, 37 presents the variation of the gear system inertial energy
parameter, e, with pinion gear and mating gear size. With a 0.5-in. radius pinion,
the total inertial energy is about 30% less than that for the 1.0 in. radius
pinion, for the minimum-e cases. The 0.5-in. pinion requires a h-in, mating gear.
The corresponding third and fourth gears will have radii of 3.26 in. and 12.23 in.,
respectively.
Since the gearbox weight will be a small fraction of total vehicle weight,
its weight effect on vehicle performance will be small. However, the gears' inertia
directly affects the response of the, engine to throttle changes and is therefore
the dominant factor to be considered.
Figure 38 presents the radii of the third and'fourth gears in the double
reduction system. The radius of the final gear"is rather large and contributes a
significant amo1mt to the total gear system inertial energy. The basic problem
with using single-stage double reduction gearing is the relatively large span
93
-------
GEAR SIZE RELATIONSHIPS
16
d = 20 IN.
RO =30
FIG. 38
a
a:
14
12
10
MINIMUM ENERGY
n = 0.5
\
2 4 6
GEAR RADIUS, '2 ~ IN.
-------
L-9712U9-7
between the input and output shafts. If this distance were smaller, the final
gear size could be reduced. The same effect can be achieved by adding a fifth
gear the same size as the final gear, as an idler. A schematic is shown in Fig. 3C
With the idler in the system, the energy parameter, e, is defined as
l2
(15)
The gear radii r3 and r^ are:
r3 =
1 + 3 —
(16}
The inertia energy parameter, corrected for varying gear thickness becomes
a-ri-r2
3 -L
1 +
(I?!
The inertial energy for this gear train system is about 50% lower than for
the standard double reduction system, as seen by comparing Figs. 37 and 39- Also,
the gear sizes are significantly reduced, as evidenced in Fig. Ho. Mow the largest
gear is only 5 in. radius. Consequently, the double reduction system with an
idler was chosen for the reduction gearbox in this study.
Installation Concepts
The installation of any power/transmission system in an automobile should
provide for easy maintenance, minimal susceptibility to impact damage, and should
not encumber the driver or passengers. The selection of a single-shaft gas turbine
offers great flexibility as to installation in a full-size six-passenger sedan,
because of its small size and light weight. The HMT was chosen as a candidate
transmission for purposes of examining the installation problem. The engine and
transmission can be mounted transversely and neatly connected with a double
95
-------
L-971249-7 FIG-
INERTIAL ENERGY PARAMETER DOUBLE-REDUCTION GEARING WITH IDLER
70
UJ 60
K-
iu
2
ce
< en
n_ w
o
m
z
111
_l
i 40
i-
Q£
UJ
Z
30
UJ
O
20
d = 20 IN.
Ro =30
96
-------
L-971249-7
FIG. 40
z
?
l/>
3
Of
UJ
O
DOUBLE REDUCTION GEARING
GEAR SIZE RELATIONSHIPS
d = 20 IN.
Ro =30
r , = 0.5
GEAR RADIUS, r2~IN.
97
-------
INSFALLAriQN SKtfCH, SINGLE SHAFT GAS TURBINE WITH HMI", .'.iONF AliLE.. lif- L
XN
vo
oo
-------
L-9T121+9-T
reduction gearbox incorporating an idler. The final drive would be through the
epicyclic differential. This arrangement is so compact that a front-wheel drive
(FWD) configuration could easily be developed. Such a system is shown in Figs.
Ul, U2, and h3 for a simple-cycle engine. Drive-line flexibility can be provided
with Rzeppa joints. Figure Ul shows the entire unit neatly mounted between the
frame members.
The air intake is shown only schematically; an actual system would incorporate
a filter system and moisture trap. The exhaust system is conceived to be a straight-
through system which is muffled by \long-strand fiber glass, contained by inner
walls perforated according to the noise profile over the length of the system.
The 6-in. exhaust pipe will flatten beneath .the vehicle to provide road clearance.
The accessory drive is taken from the idler gear of the reduction gearbox in the
form of a series of pulley belts. ..Because of the low engine torque, a belt-drive
starter can be incorporated.
The installation shown would allow a large, flat floor space for the front-seat
occupants, and ir.ore importantly provides a large amount of empty space, between
the engine and grill of the vehicle, which is vital]in avoiding serious and
expensive damage in light, to moderate, collisions. It. is felt that the insurance
companies will watch closely the development of the lo*-emissipn systems and would
most certainly impose high, rates on any system which is excessively vulnerable to
light collisions. An example would be a collision in which a standard vehicle
sustains a broken grill, a smashed radiator, and fan. A new radiator core costs
about $60 and the fan about $25. This same collision would only bend the intake
ducting for the single-shaft gas turbine-powered ^ehicle. Conversely, a Rankine-
cycle system would probably sustain damage to th£ condenser, vapor generator,
regenerator, condensor fans, and considerable plumbing. If such components were
repairable at all, the total labor costs would be very high, and replacement costs
would be similarly severe.
The same basic unit would also make an attractive rear-engine installation,
as shown in Figs, kh and h5. The air intake would be through;openings near the
rear of the car, using either protruding scoops or flush grilles, depending on the
pressure distribution and airflow in that particular area, as well as jstyling
features.
The exhaust would be serpentined to provide sufficient length for muffling
and exhaust-gas cooling. Note that the rear-engine system also provides substantial
crush-structure between the rear bumper and the engine.
The regenerated and recuperated single-shaft configurations are somewhat
larger than the simple-cycle unit, but can still be installed quite nicely in the
same basic configurations. Figures h6 and kl illustrate the front-drive system
using the recuperated engine. Wote that the engine and transmission are reversed
to allow hood clearance over the engine. The rear-drive installation for the
recuperated engine is shown in Figs. h8 and 1*9.
99
-------
INSTALLATION SKETCH, SSS-10 WITH HMT, FRONT-WHEEL DRIVE
o
o
p
Jk.
to
-------
INSTALLATION SKETCH, SSS-10 WITH HMT, FRONT-WHEEL DRIVE
10
*>•
-------
H SKETCH, S?NGL;-'-i,-'.Mr- GAS TL'RHi^L "^TH riftJTv PEAR-WH;-!. DRIVE
o
ro
\
O
-------
INSTALLATION SKETCH, SSS-10 WITH HMT, REAR-ENGINE DRIVE
5
•a
H
O
U)
-------
INSTALLATION SKETCH, RCSS-8 WITH HMT, FRONT-WHEEL DRIVE
MO
I
Tl
Cl
-------
o
Ul
INSTALLATION SKETCH, RCSS-8 WITH HMT, FRONT-WHEEL DRIVE
REDUCTION
GEAR BOX
•o
—I
-------
INSTALLATION SKETCH, RCSS-a WITH HMT, REAR-ENGINE DRIVE
*
r
v EXHAUST .
-------
INSTALLATION SKETCH, RCSS-8 WITH HMT, REAR-ENGINE DRIVE
>o
••J
NJ
-------
O
CD
INSTALLATION SKETCH, RGSS-5WITH HMT, FRONT-WHEEL DRIVE
• N
i-
i
f*
u<
-------
INSTALLATION SKETCH, RGSS-6 WITH HMT, FRONT-WHEEL DRIVE
REDUCTION GEAR BOX
-------
The large regenerated, engine is shown in Figs. 5-'; afld '-'> "or "fb" front-wheel
'irive vehicle. The ieai-.-engine syst.riu is very simila? to bL-: reeix; , vfced engine
installation.
Fue.1 Control
This section summarizes the study program conducted by the United Aircraft
Hamilton Standard Division for a fuel control for o.r.i automotive gas turbine. Tli-j
program was undertaken to conceptually design fuel control ^yst^as for three
different engines; namely, single-shaft simple-cycle, single-shaft recuperative,
and single-shaft regenerative engines. The fuel control system discussed in this
report satisfies the requirements for all three engines when usec in conjunction
with the General Electric Infinitely variable HMT transmission.
The selected mode of control for this appli:p.tion consists of a fuel flow/
compressor discharge pressure (Wf/P3J schedule with exhaust gas temperature (EGT)
limiting. A detailed discussion of the reasons for this mode selection is pre-
sented later, In addition, a schematic diagram and a control packaging concept
have be-iii generated and are discussed at length.
The control evolved during this study program is basically a hydromechanics!
unit with an integral fuel pump, a remote inlet guide v?.rte (IGV) actuator and a
remote EGT sensor Specific design features-, of the control include the following:
1. Control of starting, acceleration, and deceleration fuel flow
2, Automatic start sequencing as a function of pump (and engine) speed
3. Exhaust gas temperature limiting
h. Automatic IGV actuation as a function of pump (and engine) speed
5. Foot pedal bias of fuel flow
6. Electrical shut-off
7. Automatic altitude compensation froi. sea level to 10,000 feet
Although His; control which was 'X/acaptually designed for this application is
hydromechanical, electronic iTn.pl'v:n:jivtat,ion was also considered, Similar control
logic was used for sn electronic control, and a logic dia.gram and wiring schematic
were prepared, Both methods of control irs-plemsnt-ation were then costed, and costs
110
-------
to manufacture either the hydroraechanical or electronic version compared
favorably. However, it was decided that the elect ?circ implementation presented
a somewhat larger uncertainty than the hydroraechanical version IE both coft and
performance. Therefore, a program decision wan reached to implement tiie cout'col
hydroinechanically . However, foy future programs of this natui-ea it is recommended
that an electronic control be eiven consid^j-ation, Electronic implementation me./
prove beneficial if the automotive engine requirements bacvome somewhat more sophi-
sticated.
Control Mode Determination
General
The mode of control presented is the reriu.lt of discussions witii UARL arid UACL
relating to implementation of a low-coot, fuel control for the single-shaft
regenerative and recuperative engines (RGLn-6 and. RCSS-8) and the single-shaft
simple-cycle engine (SSS-10). Fuel c-.oatro.J. requirements for these engines were
considered in conjunction with the General Bterrric IVT-870 infinitely variable
hydromechanical transmission, and althojgb engine tempera cure, pressure, and fuel
flows differ for the three engines 3 the operating characteristics allow a common
mode of control for all three eu^iiiss. Thus, only the detailed requirements of
the SSS-10 have been considered in sizing control components, with the assumption
that differences in the temperature-, Tn-T'ssure, and flow requirements of the other
two engines would have a negligible effect on cost, weight, and complexity.
Reviewf
.'. typical engine performance Map, shi.vn in Fig, 52, indir.ahes engine fuel
requirements for steady-state and acceleration operation of the engine. The
maximum acceleration fuel flow is limited by turbine inlet temperature considera-
tions.
The steady-state operating tins is a result of the infinitely variable
transmission controlling engine speed (as a function of foot pedal position) by
varying transmission drive ratio to hold engine speed constant when the foot
pedal is in a fixed position. Thus, the fuel control is not required to provide
engine speed governing (as is required on aircraft engines) because of the combina-
tion of a fixed-shaft engine and an infinitely variable transmission. Therefore,
fuel control requirements are reduced to providing acceleration and deceleration
fuel limits and auxiliary functions such as IGV, starting, and shut-off functions.
Traditional aircraft engine control methods of controlling turbine inlet
temperature during acceleration are by open-loop scheduling of fuel flow as a
function of speed, inlet temperature, arid compressor discharge pressure (P3)
l-:-/el. These techniques would result in excessive control cost and complexity foi:
111
-------
US71249 -7
SSS-10 TORQUE CHARACTERISTICS
DESIGN SPEED = 106,000 RPM
DESIGN POWER = 130 >IP
T.I.T. = 1900 F
AMBIENT TEMP= 59 F
FIG. 52
§ 0.6 U
o
DESIGN POINT FUEL
FLOW = 82.2 LB/HR
FUEL FLOW - LB/HR = 75
CONSTANT POWER, HP/HP MAX=
DESIGN SPEED - W/W
MAX
112
-------
TURBINE EXHAUST GAS TEMPERATURE (EGT) VS TURBINE INLET TEMPERATURE (TIT)
EGT
STEADY-STATE LINE
50% N
I
-O
IO
Jk
I
MIN FUEL FLOW
-TIT
o
-------
L-971249-7
FIG. 54
FUEL FLOW PARAMETER
SSS-10
0.80
Ot
a.
0.70 -
0.60 -
0.50 0.60 0.70 0.80
SPEED RATIO, ~ N/NMAX
0.90
Q MAX
Wf/P3
111+
-------
L-971249-7
FIG. 55
FUEL FLOW PARAMETER
RGSS-6
Of
as
o
0.70
0.60
0.50
0.40
0.30
0.20
0.10
0.5
MAX
'Wf/P3
0.6 0.7 0.8
SPEED RATIO,~ N/NMAX
0.9
MIN
WF/P3
1.0
115
-------
AUTOMOTIVE FUEL CONTROL LOGIC DIAGRAM
-o
r
MAX WF/P3
FOOT PEDALI
POSITION
I
r
i_
EGT REF."
i_
SENSED
EXHAUST-
GAS TEMP
P3
.METERING
UNIT
SENSOR
*- IGV POSITION
Tl
O
Ul
-------
L-9712U9-7
an automotive fuel control, and would "be further complicated by considerations for
regenerative engines since ambient temperature, engine speed, and P3 do not uniquely
define fuel flow for a given turbine inlet temperature for such engines.
Acceleration-limiting based on turbine exhaust gas temperature (EOT) presents
a simplifying alternative and allows a common mode of control for the three engines
under consideration. Figure 53 shows turbine exhaust gas temperature plotted
vs turbine inlet temperature (TIT) as a function of engine speed and fuel flow.
The curve indicates that holding EGT constant will result in a lower TIT at idle
speed than at maximum speed. This feature is desirable, to a certain extent since
a constant inlet gas temperature would result in higher, turbine temperatures at
idle than at maximum speed. Acceleration at constant-EOT would be relatively
slower due to the lower TIT imposed by the EOT limit. Acceleration transients
at constant EGT imply an EOT sensor with no time lags. Thus, proper selection of
the sensor time constant can result in an acceptable overshoot in EGT which will
allow higher transient turbine inlet temperature during acceleration throughout
the operating range. Since the time constant of the.temperature sensor varies with
airflow, the EGT overshoot will be the greatest at low .speeds, where the TIT
margin is the greatest, and will be lowest at maximum speed.
I
Either fuel flow (Wf) or Wf/P3 could be used as the manipulated variable to
provide temperature-limiting; however, .Wf/P3 was chosen for several reasons which
are illustrated in Fig. 5** which shows Wf/P3 vs engine speed for steady-state,
constant EGT, and constant TIT. A maximum limit on Wf/P3 can be chosen so the
TIT limit is not exceeded, even though a fixed EGT reference may be exceeded
during transient operation. Also,, the Wf/P3 mode provides for altitude bias,
thereby permitting the EGT limiter to be implemented as a proportional control.
Use of a Wf mode would require an EGT control that would allow fuel flow to integrate
as a function of EGT error to hold the EGT limit regardless of altitude. Failure
of the integrator would allow the engine to exceed the TIT limit; however, failure
of the P3 sensor would be in a safe direction since it would result in underfueling
the engine.
The use of Wf/P3 suggests the possibility of simplifying the control for the
SSS-10 by eliminating the EGT sensor through an appropriate choice of the maximum
Wf/P3 limit. It can be seen on Fig. 51* that a level of Wf/P3 can be picked (as
a result of compromising acceleration potential), without exceeding the turbine
inlet temperature limit. Elimination of the EGT sensor for the regenerative and
recuperative engines is not feasible because, as shown in Fig. 55, a fixed Wf/P3
limit would cause excessive turbine inlet temperature. Thus, Wf/P3 was chosen as
the manipulated parameter in the interest of reduced mechanical complexity and
fail-safe considerations.
P_rop£sed__M£d.e_ of_Contr£l_Lo,gi_c_
The logic for the proposed control mode is illustrated in Fig. 56. Wf/P3
level is scheduled between minimum and maximum limits as a function of foot pedal
117
-------
AUTOMOTIVE FUEL CONTROL SCHEMATIC
/=V£L
TO
ADJ
//// ///////
-------
L-9712U9-7
position. As long as the preselected exhaust gas temperature is not exceeded,
the fuel flow to the engine will "be proportional to foot pedal position and
compressor discharge pressure level only, since Wf = (Wf/P3) x P3.
If the exhaust gas temperature limit is exceeded, the P3 level sensed by
the control is reduced by the EGT sensor, thus making Wf = (Wf/P3) x K x P3,
where K is a function of EGT error. Minimum Wf/P3 is scheduled at idle foot pedal
position, resulting in (Wf) decel = (Wf/P3 min) x (P3), since the engine will be
decelerating at a temperature below the EGT limit. The inlet guide vane control
schedule is a linear combination of engine speed and inlet temperature and will
be discussed in detail in the section treating the functional description of components.
It is possible that the requirement for IGV's may be eliminated on production engines,
particularly the RGSS-6 and RCSS-8 which have lower pressure ratios than the SSS-10.
The proposed mode of control, scheduling Wf/P3 as a function of foot pedal
position with a proportional exhaust gas ' temperature bias, results in a control of
minimum complexity suitable for the RGSS-6, RCSS-8, or SSS-10. . It should be noted
that a portion of this simplicity is due to the utilization of the infinitely
variable hydromechanical transmission to provide engine speed control. Further
potential simplification might result from elimination of the EGT sensor or IGV
actuator as discussed above.
Functional Description
GenerajL
A hydromechanical implementation of the proposed mode of control is illustrated
schematically in Fig. 57 . This system provides for acceleration, and deceleration
fuel flow scheduling as a function of foot pedal position, exhaust gas temperature,
and compressor discharge pressure (P3) level. The system consists of a metering
unit, a remote IGV actuator, and a remote exhaust gas temperature sensor. The
metering unit contains an engine-driven centrifugal fuel pump, a throttle valve
and pneumatic multiplier, a metering head regulator (biased by foot pedal position),
a start valve, and a shut-off solenoid. A description of these components and their
performance characteristics are presented below.
A forced-vortex centrifugal fuel pump has been chosen because its charac-
teristics make it particularly suited to automotive gas turbine requirements. The
forced-vortex pump shown schematically in Fig. 57 is an open-impeller type repre-
senting extreme simplicity in design. The pump design is based on the assumption
that the liquid rotates in the housing like a solid body, along with the impeller,
to form a perfect forced vortex, and that the quantity of liquid discharged is
small relative to the rotating volume. This implies that rotational velocities are
119
-------
L-9712U9-7
predominant and that other components of the absolute liquid velocity are negligible.
The noticeable distinctions from a conventional centrifugal pump are^a straight^
radial open-bladed impeller, a circular housing concentric with the impeller (with
ample clearances), and a wider cross section than is normally employed in conventional
centrifugal pumps. A simple discharge nozzle is located tangentially to the peri-
phery of the impeller bore. The open impeller offers the advantage of developing no
end thrust since it is hydraulically balanced, which also results in essentially no
pressure drop across the shaft seal.
Pump pressure and flow characteristics, as shown below, illustrate a flat
pressure characteristic over a wide flow range at constant speed which is parti-
cularly advantageous to the IGV implementation. The sharp cutoff at high flows is
typical of this type of pump and is influenced by discharge nozzle size, pump inlet
pressure, and pump speed. Since the altitude range of the automotive control is
from sea level to 10,000 feet, the discharge nozzle is easily sized to allow maximum
flow at these extremes, Pump cutoff flow is essentially linear with speed, whereas
the engine fuel requirements are reduced more sharply; therefore, sizing the nozzle
for maximum speed and flow conditions results in ample flow margin at reduced speed.
100%
n = 100%
PRESSURE
n = K/N*
100%
FLOW
Pump discharge pressure level requirements are a function of compressor
discharge pressure, burner fuel nozzle'pressure drop, and fuel control metering
valve and flow passage pressure drops. Pump sizing for the SSS-10 engine is based
on 500 psi discharge pressure at maximum speed (N/H* = 100?). The allocation of
pressure drops and level is tabulated as follows:
120
-------
L-97121*9-7
Compressor discharge pressure lU8 psi at max speed
Burner fuel nozzle AP 200 psi at max flow-
Throttle valve metering head AP 1*2 psi at max Wf/P3
Start valve AP (at 1*0$ K*) 80 psi for start flow-
Metering head regulator nozzle
AP and line losses 30 psi at max flow
Required Pump Discharge Pressure 500 psi at 100$ H*
In the interest of minimizing pump size and complexity, it is desirable to
operate the pump by a direct drive from the engine. The current design for the
SSS-10 (Fig. 31) indicates a three-stage reduction between engine output speed
and transmission input speed with three speeds for consideration: 3200 rpm,
15,000 rpm or 105,000 rpm at maximum speed.
Considering the fuel pump horsepower requirements, pump efficiency might not
be expected to be critical since the theoretical power requirements for a 100$
efficient pump delivering 78 Ib/hr at 500 psi is 0.06 horsepower. However, pump
impeller diameter and speed determine disc friction loss (or "-windage" loss), and
this loss is independent of flow capacity of the pump. For a given discharge
pressure requirement at a fixed speed, the impeller tip diameter is fixed since the
output pressure is directly proportional to the square of the tip speed (U<* TO).
The friction force on the disc is dependent on the fluid density, the centrifugal
energy (-which is proportional to U2), the area of the disc surface (which is
proportional to D2) and an empirical friction factor f'. Multiplying the product
of these factors by U to obtain the power absorbed (P^jr) gives: PDp = f p U2D2U =
f p U3D2, and substituting WD<* U, Pep11 f p N3D5. This indicates that to
minimize friction horsepower losses for a given pump delivery pressure, the impeller
3 / 9
diameter should be minimized since U= p3'^ and is fixed by the pressure require-
ments. A more detailed consideration of the pump disc losses for the forced-vortex
pump leads to the following equation:
PDF = 7.62 x 10-6 x S x v2 x (l^)2'8 x (d2*-6 + 1.38 d^
where: PDF = disc friction, hp
S = fluid specific gravity
v = fluid viscosity, centistokes
d = impeller tip diameter, in.
d = impeller inlet diameter, in.
N = shaft speed, rpm
A pump sized for 500 psi at 15,000 rpm results in an impeller tip diameter of
It inches and the resultant disc friction is 7.0 horsepower. A pump sized for 500 psi
121
-------
INLET GUIDE VANE SCHEDULE
•o
-J
50
40
ro
ro
O
UJ
I
01
o
z
o
30
APPROXIMATE SCHEDULE
POSITIONING IGV AS
FUNCTION OF CENTRIFUGAL'
FUEL PUMP PRESSURE
20
10
50
60
70 80
SPEED RATIO, ~ N/NMAX
90
100
110
•n
O
-------
L-9712^9-7
at 105,000 rpm results in an impeller tip diameter of 0.?U inches and the resultant
disc friction is 0.12 horsepower, resulting in approximately 33% overall efficiency.
Clearly, the operation of a pump sized for 500 psi at 15,000 rprn requiring 7 horse-
power is not suitable for an engine required to develop 130 net horsepower. TIius,
it is recommended that the fuel pump be designed to run at the maximum engine
speed of 105,000 rpm.
Fuel pumps, seals, and bearing cartridges have been designed and developed by
Hamilton Standard for turbopump applications in this speed range, although not at
this low flow capacity. A carbon face seal backed up with a static "0"-ring and
wave spring similar to "Cartriseal" 1-530 assembly should prove suitable for this
application since the open impeller pump design results in low sealing force levels.
It should also be noted that the engine is designed for 25 hours of operation at
105,000 rpm, and that the normal maximum operating speed is 80%' of this or 8U,000
rpm which leads to more favorable conditions for long seal life.
Inlet. Guide. Vane_(IGV)_Sy_stem
A typical IGV schedule of blade angle (g) vs. speed for hot-day, standard-day,
and cold-day conditions is shown in Fig. 58. The remote IGV actuator (shown ,
schematically in Fig. 57) is spring loaded so that the actuator does not move until
the fuel pump discharge pressure is sufficient to overcome the spring load. Since
the pump discharge pressure is a function of speed squared, the resulting actuator
motion closely approximates the desired schedule.
Several options exist regarding temperature bias of the IGV schedule. The
first, in the interest of simplicity, is to eliminate the need for temperature
bias by choosing a compromise schedule that could give adequate engine performance.
Changes of schedule slope and position can be accomplished by choosing different
spring rates and preloads. If such a compromise schedule proves unsatisfactory,
two methods of providing temperature bias are suggested. The first method would
bias the preload on the piston with a bimetallic disc so that, on hot days, the
force, and hence the speed, required to initiate stator vane motion would be
increased. The temperature bias of the point at which the actuator begins to move
is at a constant W//9" with this method, and can be illustrated by considering the
following equilibrium force balance:
Actuator pressure force = KjN2 = Spring force = K2 x Temp
then N2/T <* constant
and W//8" = constant
123
-------
THROTTLE VALVE AND PRESSURE MULTIPLIER CHARACTERISTIC
SSS-10
-o
-J
100
MAX
80
ce
x
CO
60
40
IDLE
20
HR-PSIA
I
I
I
20
40
60 80 100
SENSED PRESSURE-PSIA
120
140
160
-------
L-9T12U9-7
Thus, this method can provide 3 vs. N//?; however, the bimetallic disc would have to
be sensitive to compressor inlet temperature. Placing the actuator in the compressor
inlet air path is not feasible due to the actuator size required to provide 100 in-
Ib torque through a 50-degree arc. A second method of biasing-stator vane position
with compressor inlet temperature would be to use bimetallic elements in the stator
vane positioning linkage. This method will provide a more effective temperature
bias since the IGV linkage is inherently in the compressor inlet airflow path.
The proposed IGV system which positions the stator vanes as a function of fuel
pump pressure is recommended only in the interest of low cost and simplicity. If
temperature bias should become a requirement, it is recommended that it be
accomplished by incorporating bimetal elements in the stator vane positioning linkage
in the compressor.
TJir_o^tl_e_Va_lve_an_d_Pnieumati_c_Mult^ip_li^er_
The throttle valve and pneumatic multiplier is a flapper valve positioned by
an evacuated bellows. This system provides the compressor discharge pressure (P3)
bias of fuel flow and the interface for the exhaust gas temperature sensor such that
fuel flow is a linear function of sensed P3 pressure. This is illustrated by con-
sidering the following metering system equations:
WF = K A/AP~,
where Wf = fuel flow, pph
A = throttle valve metering area, in.2
AP = throttle valve metering head, psi
K = constant
Setting /AP = (Wf/P3), and A « sensed pressure, then
Wf <* (Wf/P3) x Sensed Pressure,
where sensed pressure = KxP3 = f( P3, EOT), and Wf/P3 = f(foot pedal position). A
typical fuel flow characteristic of this device is shown in Fig. 59 for pressure
and foot pedal extremes. Absolute limits on fuel flow can be established by posi-
tioning the sensing bellows in conjunction with the foot pedal input limits.
Met_ering_E[ead_ R.e£U.lator_
The metering head regulator shown schematically in Fig. 57 regulates the
metering head, or AP across the throttle valve by adjusting the flapper nozzle
opening in response to force changes on the sensing diaphragm. The desired metering
head is set by varying the torsion spring force applied to the diaphragm as a func-
125
-------
L-971249-7
FIG. 60
MAXIMUM FUEL FLOW PARAMETER
7i 0.6 -
111
Of
ss
0.7
SSS-IO
MAX
FOOT PEDAL POSITION
126
-------
1-9712U9-7
tion of foot pedal position. A simplified analysis of this force balance shows that
AP = F/A, where A is the diaphragm sensing area. The metering logic requires that
Wf/P3 Toe proportional to the square root of the throttle valve metering head.
Therefore, the Wf/P3 schedule vs. foot pedal position will be of the form Wf/P3 a
/foot pedal position. The ratio of maximum to minimum, metering head is equal to the
square of the max.-to-min. ratio of Wf/P3. Thus, for the SSS-10 engine with a range
of Wf/P3 from 0.2 to 0.53 pph/psia, a max.-to-min. metering head ratio of 7.0 is
required. Preliminary metering system sizing indicates that a range of 6 to \2
psia results in reasonable component sizes.
Figure 60 illustrates the Wf/P3 vs foot pedal characteristic for the SSS-10
engine. Idle and max Wf/P3 limits can be varied by the adjustable stops on the
foot pedal input lever.
Exhaust_ Gas_ Tjsmpejrature. _(EGT)_ Sensor_
The temperature sensing system selected to provide exhaust gas temperature
biasing of acceleration fuel flow is shown schematically in Fig. 57- This system
makes use of a bimaterial probe to sense turbine exhaust gas temperature by using
the thermal differential expansion between the high-expansion element in the air
stream and the low-expansion element shielded by the high-expansion element. The
differential motion is used to position a spool valve which provides a pneumatic
signal to the throttle valve and pneumatic multiplier. The spool valve is supplied
with compressor discharge pressure (P3) and is essentially closed to ambient
pressure (Pamb,) and full open to P3 at temperatures below the EGT limit. As the
EGT limit is approached, the spool valve bleeds a small amount of air from the P3
sensing line and reduces the signal to the throttle valve sensing bellows because
of the series orifice pressure drop characteristic shown below.
SENSED
SENSED
1.0
PSENSED /P3
A2/A]
SPOOL VALVE SERIES
ORIFICE CHARACTERISTICS
EGT
TEMPERATURE
CHARACTERISTIC
127
-------
The rate of change of the reduction in pressure with EGT is a function of the tempera-
ture coefficients of the high- and low-expansion elements, the length of the probe,
and the amount of spool valve underlap. A valve with O.OOU in. underlap would be
capable of reducing fuel flow from effectively max. to min. Wf/P3 with an exhaust
gas temperature change of 100 F when used with a probe 3.3 in. long, consisting
of IHconel-600 for the high-expansion element and lithium aluminum silicate (zero-X)
for the low-expansion element. Zero-X material was chosen because of its low-
expansion characteristic, high strength, and excellent thermal shock resistance.
Because the low-expansion element has a temperature coefficient that is
negligible compared with that of INconel-600, the time response of the probe can
be calculated by considering only the heat transfer between the Inconel sleeve and
the exhaust gas. Thus, the time constant can be represented by the following
equation :
Kt Y C
T =
h
where K = constant
T = wall thickness
Y = density of wall
C = specific heat of probe
h = film coefficient
A simplified expression for the film coefficient for a tube in air based on Reynolds
number and Prandtl number variations is given by:
h = KG0'58^0'1*2
where G = mass velocity, pps/ft2
D = tube diameter.
Then the analytical representation of the sensor time constant becomes
128
-------
L-9712U9-7
Variation of the sensor time constant with airflow density is shown in Fig.
6l. The speed reference points indicated are based on the airflow-vs-speed data
for the SSS-10 engine and assume the sensing probe is mounted in a 3.5-in. I.D.
turbine exhaust passage, thus resulting in time constant variations from 3.5
seconds at idle to 1.5 seconds at max. speed.
Finalization of the EGT sensor dynamic and static operating characteristics
will require further engine and control study to optimize engine acceleration per-
formance .
Start, Valve. and_ £hutj^Oj[f_S£lenoid.
The start valve, as shown schematically in Fig.57» is in the metered flow
path. The ball valve is responsive to forces provided by the fuel pump and the
solenoid.
During the starting sequence, the solenoid plunger would retract from the ball
valve when the ignition or engine starter is energized. The spring on the ball
valve will hold the valve closed until the engine speed, and hence the fuel pump
pressure, is at a level that will assure that the engine will start. To shut off
fuel flow to stop the engine, the solenoid is de-energized and the spring on the
solenoid will close the ball valve.
Estimated starting fuel flow characteristics are illustrated in Fig. 62 for
the SSS-10 and assume a nominal speed for starting fuel flow initiation of ^0%.
The spring load on the ball valve is set to open when the difference between
compressor discharge pressure and pump discharge pressure is 65 psia. Thus, starting
flow will be initiated at lower speeds at higher altitudes since the compressor
discharge pressure level will be lower. The two-slope characteristic of the
starting fuel schedule at a given altitude and foot pedal position is a result of
the pressure rise characteristic of the fuel pump and the pressure drop charac-
teristics of the metering unit. Although flow is initiated at nominally kO%
speed, a further increase in speed is required to develop the pressure necessary
to overcome the pressure drops needed to provide fuel flow at the levels associated
with various foot pedal, positions. In other words, the steep portion of the
starting schedule is due to pump pressure limitations, and the shallow portion of
the schedule is achieved when the pump produces sufficient pressure to allow the
control to schedule fuel flow at the Wf/P3 level set by the foot pedal position.
129
-------
CALCULATED EXHAUST GAS TEMPERATURE SENSOR TIME CONSTANT VS AIR FLOW
I
•o
to
*-
I
U)
o
1/1
0
Z
o
u
111
o
u
PROBE: INCONEL 600
0.500 O.D. X 0.009 WALL
SPEEDS BASED ON SSS-10
AIR FLOW WITH PROBE MOUNTED IN 3.5 IN. TURBINE
EXHAUST PATH.
2 —
1 —
8 9 10 11 12 13
AIR FLOW DENSITY-(LBM/SEC/FT2)
14
15
16
17
18
f*
Ov
-------
L-971249-7
FIG. 62
ESTIMATED STARTING FUEL FLOW
SSS-10
ASSUMPTIONS: (1) START VALVE SET (3 90% NMAX
(2) W,/P3 MAX = 0.65
(3) Wf/P3 MIN = 0.20
(4) FUEL BURNER NOZZLE
A PN = K Wf 2
25
CO
_l
I
20
15
10
Ol
U.30
SEA LEVEL
X 10,000 FT
SEA LEVEL
10,000 FT
MAX
FOOT PEDAL
MIN
FOOT PEDAL
0.40
0.50
0.60
SPEED RATIO ~ N/N
MAX
131
-------
AUTOMOTIVE FUEL CONTROL DETAIL
•o
c.T'/r-t-/.. J-A
-------
Physical Description
The selected fuel control has "been studied carefully for determination of
weight and cost while maintaining high standards of reliability, durability, flexi-
bility, and maintainability. Several arrangements were examined prior to selection
of the installation shown in Pigs. 63 and 6k. Realistic volumes and weights are
obtained by preparing a detailed arrangement in this manner.
The fuel control package consists of three separate units connected by fuel
and pneumatic lines. The fuel control unit is positioned on the engine gearbox
to utilize the high-speed drive required for the fuel pump. The IGV actuator is
mounted adjacent to the engine IGV linkage mechanism, thus insuring minimum linkage
weight. The EOT sensor is mounted at a convenient location in the turbine exhaust
gas stream. Standard bolt retentions are used for each unit, and special tools are
not required for maintenance .
The housing is of die-cast aluminum for low cost and weight. It encloses the
high-speed centrifugal pump, pressure regulating system and foot pedal linkage,
shut-off solenoid and start valve, and the P3 sensor and throttle valve. The
metering elements are closely grouped to minimize pressure drops within the
control. Close spacing assures minimum volume and weight.
The unit is accessible for maintenance and adjustment after installation. Idle
and maximum speed adjustments are externally accessible at the foot pedal input
lever location. P3 bellows external adjustment is performed with a screwdriver.
All internal subassemblies are accessible for scrutiny by removing the control
cover bolts and control cover. Since no operating parts are attached to the
cover, all internal parts remain undisturbed during inspection.
The fuel control features a metal gasket with integral, continuous packing for
positive sealing. "Printed circuit" flow lines are incorporated in this gasket to
simplify the housing casting. The cover is of die-cast aluminum for low cost and
weight. Provisions in the cover are made for upstream and downstream throttle
valve pressure taps which are also useful for purging air after fuel control
installation. Hydraulic 'and pneumatic connections are made by use of standard
open-end wrenches. Positive sealing is assured by utilizing "0"~seal packings in
aircraft-type threaded fittings. All housing mounting bolts are standard, and
likewise are readily installed by use of standard socket or open-end wrenches.
133
-------
FUEL CONTROL-EXTERNAL VIEW
i
-------
L-9T12U9-T
The high-speed centrifugal pump is a cartridge-type assembly, and is located
low in the housing to minimize suction losses. The cartridge is restrained by a
snap ring, and may be readily removed without disturbing the control calibration.
Pump interface to the engine gearbox is a continuous surface to eliminate contamina-
tion to the drive shaft and bearing seal, A standard, face-type carbon seal mates
with the anti-drive end of the pump shaft, thus allowing the high attendant surface
speeds while effectively preventing fuel-to-air leakage.
The shut-off solenoid and integral start valve are located at the top of the
control housing. Electrical terminals are oriented upward so that waterproof
protectors can be effectively used. Metered fuel flow is upward from the control
through the start valve, and then out to the engine. Any entrapped air is purged
automatically by virtue of this arrangement.
The P3 sensor and throttle valve subassembly is located low within the main
housing to allow the P3 connection to enter from the bottom; also, any condensate
that forms in the bellows cavity drains out by gravity. An externally accessible
throttle valve position adjustment screw is located at the side of the main housing
for field adjustments, if required.
The pressure regulating system and foot pedal linkage shaft terminates at the
side of the control opposite the engine, allowing flexibility of design of foot
pedal linkage. Adjustment of idle and maximum-stop screws is possible with a
standard Allen wrench. Self-locking inserts prevent screw rotation due to
vibration.
The IGV actuator is a simple plunger-cylinder design mounted adjacent to the
engine inlet guide vanes, and remote from the main control housing. The actuator
housing is a permanent-mold aluminum casting for low cost and weight. An adjustable
rod is provided at the end of the IGV piston shaft for exact actuator positioning
with respect to the IGV linkage. Aircraft-type connections insure leak-proof
sealing.
Exhaust_ Gas. Temper ature_ _(_EGT_)_ £>en_sor_
The exhaust gas temperature sensor is mounted in the turbine exhaust gas
stream at the rear of the engine, remote from the main control housing. The
sensor assembly consists of a cast iron housing and a slide valve actuated by a
high-temperature bimetallic probe inserted in the gas stream. Orientation of the
sensor is not critical, thus allowing flexibility in mounting location.
135
-------
L-9712^9-7
Cpmponent_ Weights,
The estimated dry weights of the fuel control units, excluding electrical
wiring and hydraulic and pneumatic tubing provisions and mounting bolts, are as
follows:
Fuel control and pump 5-03 lb
IGV actuator 2.29 lb
EGT sensor 2.01 lb
Total estimated system dry weight 9-3 lb
The fuel control and pump weight is broken down as follows:
Housing 2.50 lb
Housing cover 0.57 lb
Housing cover bolts 0.24 lb
Gasket plate and packing O.lU lb
Fuel pump and drive 0.37 lb
P3 sensor and throttle valve O.l6 lb
Shut-off solenoid and start valve 0.75 lb
Pressure regulating system and foot pedal linkage 0.30 lb
Total estimated dry weight 5.03 lb
Externa;l_Conn_e£tipn_s_R_eqiiiredi
Fuel control
Hydraulic:
- fuel line to pump
- fuel line to engine
- fuel pressure line to IGV actuator
Pneumatic:
- signal line from EGT sensor to P3 sensing bellows cavity
Electrical:
- wiring to shut-off solenoid
Mechanical:
- foot pedal input lever linkage
Inlet guide vane actuator
Hydraulic:
- fuel pressure line to pump suction line
- fuel pressure line to pump discharge line
136
-------
L-9712U9-7
Mechanical:
- IGV linkage
Exhaust ftas temperature sensor
Pneumatic :
- P3 line from engine compressor to EGT sensor
- Signal line from EGT sensor to fuel control
Summary of Fuel Control Analysis
This analysis indicates that an engine control mode based on scheduling Wf/P3
as a function of foot pedal position with a proportional turbine exhaust gas
temperature bias results in a control of minimum complexity suitable for the RGSS-6,
RCSS-8, or SSS-10 engines. However, additional study of the engine control system
will be required to determine control design requirements .
It is recommended that engine, control, and transmission -dynamics be studied
in a comprehensive analysis of overall system performance in 'both acceleration and
steady-state speed governing modes of operation. Results of this ''analysis will
establish. requirements such as: rate of change of "gear ratio" for the IVT as a
function of engine speed error, exhaust gas temperature sensor time constant , and
rate of change of P3 reduction with temperature. In addition, hot- and cold-day
performance analysis of the acceleration mode can determine the feasibility of
eliminating the need for an exhaust gas temperature sensor by establishing an
appropriate maximum Wf/P3 limit for the SSS-10 engine. -
Evaluation of engine performance at hot- and cold-day extremes without
temperature bias of the IGV schedule is recommended to determine the feasibility
of simplifying the control system by eliminating the requirement for temperature
bias of the IGV. In addition, a determination of IGV system forces due to friction
and aerodynamic loading will be required to establish actuator positioning
accuracy. ••'-:.*.• •
Finally, it is recommended that further studies be conducted on the pumping
and burner fuel nozzle system. These studies would be aimed at reducing fuel system
pressure requirements, thereby enabling lower pump speeds to be utilized to assure
long pump life. Perhaps a combination of pump and injector such as a "slinger
nozzle" could be incorporated into the engine to further simplify the fuel control
system.
137
-------
L-9712l*9-T
138
-------
L-9712^9-7
ENGINE PERFORMANCE
Program Description
All engine cycle analyses were performed by Pratt & Whitney Aircraft (P&WA)
using its proprietary State-of-the-Art Performance Program (SOAPP). The SOAP
program is an engine cycle analysis system which easily organizes any type of
engine configuration using any combination of components. It is particularly
applicable for this optimization study because of the contract requirement to
study a large number of alternative engine cycle parameters and configurations.
The engine, within SOAPP, is made up of components such as inlets, compressors,
burners, turbines, heat exchangers, and reheat burners. The computations involving
each of these components are made in subprograms called modules. These modules
contain the latest calculation techniques and are continually updated for this
purpose. The analyst may select any desired engine configuration which can have
any of a number of compressor and turbine types and any pressure ratio, turbine
inlet temperature, and value for other parameters of interest.
The SOAP program locates the modules in the machine library and links them
together in a running engine calculation. The actual order of calculation is
automatically determined by the program through a preprocessor program. The para-
meters, such as temperatures and pressures, are passed along downstream in the
calculation and build up an array of values at each station for all of the para-
meters as the calculation flows toward the exit nozzle. There are four basic arrays
that bind the engine deck into a running system. There is a continuity array which
passes the values of parameters from module to module, the spool array which links
the turbine with its compressor, the stream array which contains parameters that
have only one value per stream, and the constants array which are the values
associated with specific components.
The input conforms to preselected designations and runs directly from the input
block to the modules. When a design point is run, the scaling constants for any
compressor or turbine version are calculated so that the off-design points that
follow are properly scaled.
Output
The output computed by the SOAPP for this study included the following items:
1. Percentage horsepower
2. Percentage speed
3. Horsepower
139
-------
L-971249-7
FIG. 65
DESIGN TORQUE AND SPEED CHARACTERISTICS
SINGLE - SHAFT GAS TURBINES
1.2
CONSTANT POWER, HP/HPM,X =
0.2 0.4 0.6 0.8
DESIGN SPEED ~ W/WMAX
-------
L-9712U9-7
U. Speed
5. Fuel flow
6. Burner inlet temperature
7. Turbine inlet temperature
8. Turbine exhaust temperature
9. Engine exhaust temperature
10. Airflow
11. Compressor pressure ratio
12. Engine exhaust flow
13. Fuel-air ratio
lU. Heat exchanger effectiveness
15. Engine torque
16. Percentage torque
Each computer output sheet contains printouts of these values for ;fixed horsepower
and speed commands. Horsepower values were selected from 2.% to 100% at the design
speed, and appropriate values of horsepower were computed for design speeds
ranging down to 50%. Eight different engine speeds were selected and approximately
ko points run for each engine condition.
Performance was generated for the following five ambient conditions:
1. Sea level, standard day
2. 5000-ft standard day
3. Sea level, 32 F
U. Sea level, 85 F
5. Sea level, 105 F
Therefore, the complete set of output data for Phase II consisted of five.conditions
for three engines, and approximately ho points per engine, for a total of approxi-
mately 600 calculated engine operating points.
Although complete sets of data were generated for all three engines, the final
changes which occurred during the optimization process were not necessarily included
in the final output for each engine. Trends for various ambient conditions, however.
were firmly established and it was judged not essential to rerun the entire 200
points for each final engine.
The final data consisted of the sea level, standard-day calculations for the
simple-cycle engine, the regenerated engine, and the recuperated engine. The data
were then .plotted for each engine in order to determine the maximum torque character-
istics, the fuel flow lines, temperature lines, etc. Lines for each operating point
were determined by the maximum allowable turbine inlet temperature. For all three
engines, this was 1900 degrees, corresponding to the use of uncooled turbines.
Figure 65 shows the torque-speed characteristics for all three engines for standard-
day conditions. Fuel flow is superimposed on the torque-speed curve for the RGSS-6
11*1
-------
L-971249-7
FIG. 66
RGSS-6 TORQUE CHARACTERISTICS
DESIGN SPEED = 82,OOORPM
DESIGN POWER = 130 HP
T.I.T. = 1900 F
AMBIENT TEMP = 59 F
eCS)6N POINT FUEL FLOW = 68.0 LB/HR
FUEL FLOW - LB/HR =
CONSTANT POWER,HP/HPMAX=
0.4 0.6
DESIGN SPEED - W/WMAx
-------
L-971249-7
FIC. 67
RCSS-8 TORQUE CHARACTERISTICS
DESIGN SPEED = 95,000 RPM
DESIGN POWER = 130 HP
T. I.T. = 1900 F
AMBIENT TEMP = 59 F
DESIGN POINT FUEL
FLOW = 73.6 LB/HR
CONSTANT POWER, HP/HPUAV
MAX
0.2
0.4 0.6
DESIGN SPEED- W/WMAX
0.8
1.0
-------
L-971249-7
SSS-10 TORQUE CHARACTERISTICS
DESIGN SPEED = 106,000 RPM
DESIGN POWER = 130 HP
T.I.T, = 1900 F
AMBIENT TEMP= 59 F
FIG. 68
DESIGN POINT FUEL
FLOW = 82.2 LB/HR
CONSTANT POWER, HP/HP MAX= O.J
0
DESIGN SPEED - W/W
MAX
Ikk
-------
L-9712^9-7
on Fig. 66, while fuel flows for the RCSS-8 and SSS-10, are plotted on Figs. 67
and 68, respectively.
Selection of Engine Operating Line
Engine operating lines were selected to provide common modes of operation for
all three engines. Two considerations entered' into the selection of these lines.
These were: vehicle acceleration response and engine operating life.
With regard to engine acceleration response, there have been reservations
relative to the capability of a gas turbine engine to provide response similar to
that currently experienced by the owners of automobiles. The problem of response
is many-fold. It includes a ^consideration not only of the engine itself and its
steady-state torque capabilities, but of the engine inertia, the transmission
characteristics, and the transient characteristics of the engine.
It is well known that the single-shaft engine suffers from much poorer torque-
speed characteristics than its free-turMne counterpart. Thus, the single-shaft
engine requires a fairly sophisticated transmission in order to provide suitable
propulsion system characteristics for vehicles. On the other hand, the character-
istics o'f the single-shaft; engine are such, that its maximum torque at any operating
engine speed is instantly available. That is to say, an increment in fuel flow to
the burner is immediately felt as a greater torque output through the turbine which
is connected directly to the operating shaft. Free-turbine engines have a "built-in
transient lag since the additional energy added in the combustor is extracted first
through the compressor turbine and is felt secondarily by the power turMne. Because
of this response lag, many automotive gas turbine engines with a free-turbine con-
figuration include variable geometry on the power turMne in order to extract more
energy from the expanding gases for propulsion purposes.
f -.,,
Since a single-shaft engine requires a sophisticated transmission in order to
serve as a practical propulsion system,, the transmission itself can be used to solve
response problems which'remain. Specifically, the control system of the engine,
transmission and accelerator, pedal can be integrated so that the vehicle operator
feels an instantaneous response to his pedal command, while the engine can
simultaneously be allowed to accelerate to speeds where more torque is available,
provided the engine's steady-state operating line is selected at some combination
of speed and output which provides an acceleration margin at that speed. Figure 69
shows six potential operating, lines, ,. The first operating line (AA) is the maximum
torque characteristic of "the engine at , any given speed. It is obvious that an
engine operating on. this*. line, to deliver steady-state power requirements, would
be unable by itself to accelerate, either the 'vehicle or itself without provisions
in the transmission to drop the 'load;' But this dropping of load would aggravate
the response characteristics and be unacceptable to the motorist; thus line AA
-------
L-971249-7
FIG. 69
ENGINE OPERATING LINES
RCSS-8
DESI GN SPEED = 95,000 RPM
DESIGN POWER = 130 HP
T.I.T. = 1900 F
AMBIENT TEMP = 59 F
1.2
1.0
0.8
a
Of
o
o
0.6
0.4
0.2
DESIGN POINT FUEL.
FLOW = 73.6 LB/HR-
FUEL FLOW - LB/HR =
CONSTANT POWER, HP/HP MAX= fl.3
DESIGN SPEED - W/W
MAX
1U6
-------
L-9712^9-7
is not acceptable as an operating line. At the other extreme, operating line DD
can provide an engine with 100$ power response at any power requirement, but its
fuel consumption would be unacceptable since it is always operating at maximum
airflow.
The operating lines designated as A, B, C and D represent reasonable compromises
between the extremes of fuel economy and response characteristics as represented by
lines AA and DD. At this point in the development cycle, it cannot be stated with
certainty which operating line is optimum or that any one would be optimum for all
drivers. The infinitely variable transmission makes it feasible to offer several
different operating lines to suit the driver's desires. For example, a driver-
selected option can be visualized for performance with a fuel economy penalty, or
vice versa. Considering the lines in sequence, line A provides the best fuel
economy for the heat-exchanger engines but the least amount of response for all
engines. Line B provides the best fuel economy for the simple-cycle engine, a
somewhat poorer fuel economy for the heat-exchanger engines, but provides a more
adequate response margin. Line C provides the greatest response margin of the
three lines which start at the idle speed of 50%, but at the expense of fuel economy
in all three engines. Line D represents an even greater margin of response, which
is achieved by fixing idle speed at 60%. For the purposes of the present study,
it is judged that line B represents the best overall combination of fuel economy
and response for all three engines.
A second consideration in the establishment of the operating line is that of
engine life. With reference to Figs. 66 and 67, it can be seen that the optimum
fuel flow line for the heat-exchanger engines is coincident with the maximum torque
line. This is because the greatest efficiency is achieved at the highest turbine
inlet temperature for the engines of this configuration. But the stresses asso-
ciated with high temperatures are such that the maximum turbine inlet temperature
should not be used on a steady-state basis for more than 100 hours of engine
operation in order to achieve the desired design life. Thus, the optimum fuel flow
line for the heat-exchanger engines does not represent a practical operating line.
On the other hand, transient excursions beyond the maximum turbine inlet temperature
are allowable and acceptable. The simple rule of thumb of greatest validity is
that, for steady-state operation, the turbine exhaust temperature at the design
point should not be exceeded since the turbine exhaust temperature is closely
related to the turbine metal temperature which is the determining factor in engine
life calculations. The fuel control selected, which has been described previously,
permits transient overtemperature subject to time-constant design parameters as a
function of exhaust gas temperature. It assures that steady-state operation will
not exceed allowable steady-state exhaust gas temperatures.
11*7
-------
GAS TURBINE EXHAUST EMISSIONS
NITROGEN OXIDES
r
•o
200
180
160
140
2
o.
± 120
UJ
Q
X
O
o
o
100
80
60
40
20 -
TF-33
0.05
(S 130 HP
PT-6
-56
IS/69
_L
0.10 0.15
EQUIVALENCE RATIO,
0.20
0.25
0.30
-------
L-9712H9-7
Emissions Characteristics
Emissions characteristics are based on state-of-the-art data as reported in
various survey documents. In the case of the heat-exchanger engines, emissions
measurements for state-of-the-art regenerated engines have been reported (Ref. 17).
In the case of the simple-cycle engine, the data are developed from Ref. 18 which
presents a survey of various simple-cycle aircraft engines. From these data,
emissions index curves have been prepared suitable for input to the mission analysis
program. These emissions index data are consistent with the operating lines discussed
above, with emissions characteristics developed from fuel-air ratios (converted to
equivalence ratios) which were reported as the output from the engine cycle analysis
program.
Figure 70 presents NOX exhaust emission data for a number of simple-cycle gas
turbine engines from Ref. 18. The upper and lower values of NOX are shown for three
equivalence ratios corresponding to the operating line for the SSS-10 at powers of
2.6 hp, 19«5 hp and full horsepower. Figure 71 shows the same information for
carbon monoxide. The bars shown are those believed to be typical of state-of-the-
art, high-low values for simple-cycle gas turbine engines. It will be noted that
the Contractor-produced PT6 engine is eliminated from this analysis since it is
known that it is not typical, in carbon monoxide emissions, of low-emission com-
bustors. Figure 72 presents these same data for unburned hydrocarbons. It will be
noted that the extreme scatter of the UHC data renders it highly questionable and
that there is practically no correlation between hydrocarbon output and carbon
monoxide output. However, it is believed that hydrocarbon output is one of the
most easily controlled emissions characteristics in the gas turbine engine.
Several emissions cases were run for both simple-cycle and heat-exchanger
engines. In the case of the regenerated engine (RGSS-6), the emissions indices were
derived from data reported on the General Motors GT-309 in Ref. 17, as was"done in
Ref. 1. The emissions indices for the RCSS-8 engine were slightly modified from
those for the RGSS-6 to account for the lowered combustor inlet air temperature
predictions as derived from the engine performance program. For the SSS-10, the
emissions indices were derived from the T-56 engine (Allison Division of General
Motors) as shown in Figs. 70, 71, and 72 for HOX, CO, and UHC, respectively,
correlated to the equivalence ratios calculated in the engine cycle analysis
predictions. The corresponding equivalence ratios at 2.6, 19-5, and 130 hp are
shown on these figures.
In addition, for all engines, emissions indices were approximated to indicate
the levels required to meet Federal standards.
-------
\r\
o
550
500 -
450
400
350
300
o_
0.
8 250
200
150
100
50
GAS TURBINE EXHAUST EMISSIONS
CARBON MONOXIDE
JT3C-4
T76-G-12
GT-309
I
0.05
0.10 0.15
EQUIVALENCE RATIO,
0.20
0.25
0.30
-------
GAS TURBINE EXHAUST EMISSIONS
HYDROCARBONS
>o
••J
0.05
0.10 0.15
EQUIVALENCE RATIO,
0.20
0.25
0.30
-------
The emissions index data resulting from these selected curves for all three
engines are shown in Figs. 97 through 103 and are discussed in the Emissions
portion of the following section. These emissions index data serve as input in the
vehicle mission analysis program for purposes of deriving emissions estimates.
152
-------
L-9712^9-7
VEHICLE EVALUATION
Transmission Modeling
The engines chosen for final optimization restricted the choice of available
automatic transmissions to those compatible with the characteristically poor low-
speed torque characteristics and narrow operating speed range of the single-shaft
gas turbine. Further restrictions on the transmission are that it should be light-
weight, durable, and have (assuming there are no off-the-shelf units available) the
potential for mass producibility and low cost. Of the five basic types of transmis-
sion described in an earlier section (mechanical, hydromechanics!, hydrokinetic,
electrical and traction-drive), the GE HMT was chosen for evaluation. In addition,
the Borg-Warner 8-speed mechanical transmission was chosen as an alternate for per-
formance calculations.
Transmission Description - GE HMT
The GE HMT is an infinitely variable hydromechanical transmission designed to
provide a stepless and continuously variable ratio change from full reverse to full
forward and overdrive. The basic rotating elements are two hydraulic ball-pumps
and a simple planetary gear set. The support items are a suitable control and two
stroking actuators together with a small fixed-displacement charge pump. The input
shaft from the engine is coupled directly to the sun gear of a planetary gear system
and also drives the input hydraulic element. The output hydraulic element, which
is hydraulically coupled to the input element, drives the ring gear. The planet
gears, through their carrier, drive the output shaft. Ratio change is accomplished
by varying ring gear speed and direction. This is .done by changing the displacement
of the input hydraulic element in response to signals from the ratio controller.
The transmission incorporates a ratio control system which is programmed to reflect
the desired torque for any particular engine speed. Approximately 60% of the torque
is carried hydraulically, and hO% is carried mechanically through the sun gear.
The transmission weighs 165 lb dry and will use 20 Ib of hydraulic fluid (SAE
10-30 motor oil is recommended). Rated and maximum input speeds are 3200 and ^-200
rpm, respectively. However, these speeds will not be reached since 100% engine
output speed is about 2700 rpm following the primary gear reduction. Nominal rated
transmission output torque is 700 Ib-ft (equivalent to wheel-spin torque for a
itOOO-lb vehicle with 50-50 weight distribution and a 2.86 rear-axle ratio).
The transmission, efficiency data on which all HMT vehicle performance calcu-
lations were based are shown in Fig. 23. The analytical expression for transmission
efficiency at the top of Fig. 23 fits the efficiency data over the range of input
powers shown on the figure. This equation was used to extrapolate transmission
efficiency when engine input power ratios (engine input power * maximum power) were
below 10$, since no data were available in that range. However, for the mission
153
-------
L-971249-7
HYDROMECHANICAL TRANSMISSION INERTIA
GE HMT 870 (8.7 CD. IN./REV)
FIG. 73
0.09
0.2 0.4 0.6 0.8 1.0 1.2
TRANSMISSION RATIO (OUTPUT SPEED/INPUT SPEED)
1.6
-------
analysis work done in this study, the vehicles spend very little time at such low
power levels. The rotational inertia of the 8.7 cu-in HMT, as a function of trans-
mission ratio, is shown in Fig. 73.
Rati£ £ontrol
The transmission ratio is automatically controlled. For any road load con-
dition, the control will select ratios to operate the engine according to a pre-
determined speed-load schedule. The speed-load schedule is established "by the
relationship of fuel control setting to controlled speed for each accelerator
pedal position.
The optimum fuel flow line may "be selected as the speed-load schedule, or
some trade-offs may be desirable in the interest of overall vehicle performance
to accommodate such factors as emissions, engine durability, high-speed power
requirements and shape of the transmission efficiency curve.
The transmission control receives inputs of accelerator position and engine
speed. If the accelerator is depressed, the transmission ratio controller calls
for more engine speed. This signal is simultaneously fed by direct mechanical
linkage to the engine fuel control. Since the engine speed cannot change as fast •
as the mechanical input from the accelerator, a speed difference signal is produced
in the controller. This differential moves the controller output linkage to in-
crease transmission ratio, unloading the engine and allowing it to accelerate. As
engine speed increases, the speed error in the control approaches zero and trans-
mission ratio stabilizes at a new value consistent with the engine torque for the
particular engine speed selected by the operator and the horsepower load at the
wheels. Viewed another way, if the accelerator and fuel control are maintained in
a fixed position with the vehicle moving over varying terrain, the transmission
will continually change ratio to maintain the desired engine torque value for the
engine speed. Engine speed under this condition would be essentially constant.
Transmission Description - Borg-Warner 8-Speed Mechanical
The Borg-Warner,8-speed mechanical transmission consists of a four-speed
planetary automatic gearbox preceded by a two-stage splitter which, in effect,
doubles the number of ratios in the gearbox. The controlled slipping clutch can
transmit constant torque to the wheels at any engine speed (below the clutch
lock-up point) without multiplication. Its advantage is that the stall speed can
be adjusted in the design process so that the engine can be completely idled or be
completely engaged within a fairly narrow range. In addition, when the clutch is
locked up, power is transmitted with a higher efficiency than associated with a
hydraulic slipping device. The overall mechanical efficiency of the 8-speed gear-
box is about 85% (not including clutch losses).
155
-------
L-971249-7
CONTROLLED SLIPPING-CLUTCH CHARACTERISTICS
FIG. 74
O
K
O
MAXIMUM TORQUE
TRANSMITTED BY CLUTC
MAXIMUM TORQUE
AVAILABLE FROM ENGINE
.b
ENGINE SPEED
PRESSURE FORCE
FROM PUMP
ENGINE SPEED
156
-------
L-9T12U9-7
The clutch is a multiplate, oil-bath, disc-type design, about 9 in- in
diameter. An oil cooler is required to dissipate excess heat. The torque trans-
mitted by the clutch increases with increasing engine speed. This is done by
applying increasing hydraulic pressure to the clutch pressure plate (s) against the
action of a negative bias pressure from mechanical springs. The clutch system is
designed so that engagement begins to occur at just above idle speed.
The action of the controlled slipping clutch can be more easily understood
with the aid of Fig. 7^*. In Fig. 7^A, the output pressure of the pump is shown
increasing with the square of the engine speed. This pressure acts on the clutch
pressure plate against the action of the negative spring bias (an opposing spring
force). The sum of the pump and spring pressures is negative (that is, no pressure
is applied to the clutch disc) until the engine speed increases to the value
indicated by w^.. At engine speeds above o^, the clutch begins to transmit torque
because the pump pressure is greater than the negative spring bias, and a positive
pressure, which increases with engine speed, is applied to the clutch disc.
Figure 7^B shows the torque transmitted by the clutch, as a function of engine
speed, superimposed on a typical single-shaft gas turbine torque curve. The value
"ij_ is the engine speed at which the clutch begins to transmit torque. The torque
transmitted by the clutch increases as the square of the engino speed (because the
torque transmitted is proportional to clutch pressure) until the speed w. is
reached. Beyond this speed, the maximum torque of the engine is less than the
torque capacity of the clutch and the clutch will be locked. If an engine torque
is commanded which is greater than the maximum torque that the clutch can transmit
(say point a), the torque difference between the commanded and clutch torques vill
cause the engine to accelerate. If an engine torque is commanded which is less
than the maximum clutch torque (point b, for example), the clutch will be locked.
The speed wj_ would normally be selected as an idle speed and the speed u)» would be
designated as the clutch lock-up speed. The values «^ and u can be selected by
the proper choice of clutch design parameters such as spring bias and pump charac-
teristics.
Conversations with Borg-Warner personnel indicate that it will be possible to
incorporate a device, known as an inertia- valve , with the clutch system. The valve,
which is sensitive to angular acceleration, would be used to limit the rate of
change of engine speed during gear changes. The valve is held by a spring whose
stiffness is such that the valve can bypass oil to the clutch (to allow slippage)
for angular accelerations greater than the design value. The effect of the valve
is to produce smoother, less jerky acceleration.
The inertia valve is not suitable for use with engines w. '*':-*> torque output
is not a smooth function of time. For example, the discretp i twer strokes of the
Otto-cycle engine produce transient accelerations of the craw -haft which would
adversely affect the action of the valve. The continuous ami f-mooth power produc-
tion «f the gas turbine, however, would render it suitable for use with the device,
157
-------
L-9712U9-7
TABLE IX
SELECTED PHYSICAL CONSTANTS AND BASE-LINE VEHICLE
Acceleration of gravity
Air density (55 F)
Fuel density (gas turbine fuel)
Heat content
Rolling resistance coefficients
Wheel radius
Product of frontal area and drag coefficient
Weight distribution
Rear axle efficiency
Road-tire adhesion coefficient
Accessory power
Fuel economy runs
Acceleration performance runs
Vehicle test weights (Wt) (Appendix V):
RGSS-6
RCSS-8
SSS-10
Clutch inertia (8-speed transmission)
Transmission gear ratios
(8-speed mechanical transmission):
32.2 ft/sec2
0.0023769 slug/ft3
7.0. Ib/gal (0.1^3 gal/Ib)
18.UOO Btu/lb
r0 = 0.015
rj = 2.15 x 10-5 ib/lb-fps
r2 = 1.85 x 10-f Ib/lb-fps2
1.1 ft
12 ft2
0.95
1.0
1.3 hp
H.O hp
3970 Ib
•1+020 Ib
3850 Ib
200 lbm-in2
Gear
Ratio
1st
2nd
3rd
Uth
5th
6th
7th
8th
50
30
56
06
69
1*0
18
1.00
158
-------
L-9712U9-7
Engine Sizing
The purpose of this task was to define the probable engine size (maximum
power) and idle speed in order that the required off-design engine data could "be
generated for further engine optimization. To accomplish this, a computer program
was written which predicts the maximum acceleration performance of a vehicle
equipped with the GE-HMT transmission. Other results from this program, and two
other programs written for this project, are discussed in later sections. The
derivations for these programs are presented in Appendices I and II. The programs
themselves, including sample printouts are described in Appendix III.
The criteria used to select the engine size and idle speed were the OAF
vehicle acceleration performance goals of Ref. 19. These requirements are listed
below. The parameter Wt is the vehicle test weight which is given in Table IX using
standard OAP calculations. Table IX contains constants and base-line parameters used
for all Vehicle Evaluations.
Acceleration from a standing^start:
The minimum distance to be covered in 10.0 sec is UUO ft. The .maximum time
to reach a velocity of 60 mph is 13.5 sec. Ambient conditions are lU.7 psia,
85 F. Vehicle weight is Wt, and acceleration is on a level grade and is ini-
tiated with the engine at the normal idle condition.
Acceleration in merging traffic:
The maximum time to accelerate from a constant velocity of 25 mph to a velocity
of 70 mph is 15.0 sec. Time starts when the throttle is depressed., Ambient
conditions are lU.7 psia, 85 F. Vehicle weight is Wt, and acceleration is on
level grade.
Acceleration, DOT High-Speed Pass Maneuver:
The maximum time and maximum distance to go from an initial velocity of 50
mph with the front of the automobile (l8-foot length assumed) 100 feet
behind the back of a 55-foot truck traveling at a constant 50 mph to a
position where the back of the automobile is 100 feet ahead of the 55-foot truck
is 15 sec and 1^00 ft. The entire maneuver takes place in a traffic lane
adjacent to the lane in which the truck is operated. The vehicle will be
accelerated until the maneuver is completed or until a maximum speed of
80 mph is attained, whichever occurs first. Vehicle acceleration ceases
when a speed of 80 mph is attained, the maneuver then being completed at
a constant 80 mph. (This does not imply a design requirement limiting
the maximum vehicle speed to 80 mph.). Time starts when the throttle is
depressed. Ambient conditions are lU.7 psia, 85F. Vehicle weight is W+,
and acceleration is on level grade.
159
-------
L-971249-7
FIG. 75
HOT-DAY ACCELERATE PERFORMANCE
GE HMT
SSS-10
W - 3H50 LB
IDLE SPEED RATIO - W/WMAX:
O 0.5
V • 0.6
3 1.7
ACCESSORY POWER- 1.3 HP
111
O
i
o
u
UJ
2
I
o
HOT-DAY LIMIT
>0.95 X 440 = 418 FT
HOT-DAY LIMIT
1.05 X 15 = 15.75 SEC
HOT-DAY LIMIT
1.05 X 13.5 - 14.18 SEC
''20 130 140 150 160
STANDARD-DAY RATED POWER (HP)
170
160
-------
L-9712^9-7
Reference 19 also states that the vehicle must be capable of providing the
above acceleration performance within 5$ of the stated values when operated at
ambient temperatures from -20 to 105 F. For convenience, the GAP acceleration
criteria are presented below in tabular form, for both the 85° and 105° days:
105° Day
Performance Requirement 85° Day (5% performance degradation)
0-60 mph acceleration time (sec) 13.5 lU.2
Minimum distance covered in 10 sec (ft) UUO 1+18
25i-70 mph acceleration time (sec) .15.0 15.8
DOT High-Speed Pass Maneuver:
Minimum time (sec) • 15.0 15.8
Minimum distance (ft) lUOO 1^70
Because of the relatively severe degradation of performance with increasing
temperature' '(about 15$ loss of power between 59 to 105 F. ambient temperature),
the engines were sized on the basis of the hot-day (105°) requirements rather
than the 85° ambient criterion.
Engine sizing .was performed by inputting a matrix of ( standard-day) engine
powers and idle speeds into the HMT vehicle performance computer program to obtain
a corresponding matrix of vehicle performance data. (The program corrects for the
hot-day performance degradation.) The computed performance data consisted of the
requirements for each of the OAP acceleration criteria (0-60 mph acceleration
time, etc.). The performance data were then plotted to obtain the different
combinations of engine size (maximum power) and idle speed which would just meet
the OAP hot-day requirements. This process is illustrated,in Fig. 75. The
various combinations are formed by the intersections of the 105° day performance
requirements (horizontal dashed lines) and the constant-horsepower lines on
Fig. 75. The resulting engine power and idle speed pairs were then plotted as
shown in Fig. 76. The locus of these points, for each acceleration criterion,
divides the power and idle speed map of Fig. j6 into two regions.. The region
above and to the right of the locus contains power and idle speed combinations
which will exceed the particular acceleration criterion* Conversely, all power
and idle speed pairs below and to the left of the locus will not meet the accelera-
tion requirements. Thus, the smallest engine size which will just meet all the
OAP requirements will -lie on the rightmost locus (of the three shown) for a given
idle speed. Similarly, the lowest acceptable idle speed will lie1 on the uppermost
locus for a given engine size. For the conditions shown in Fig. 76, the required
size is 111 hp at an idle speed ratio of 0.60. The nominal unmodified standard-day
engine rating to meet this requirement is 130 hp, assuming that 85? of standard-day
torque is available at 105 F. Therefore, all three engines were specifiedto be
rated at 130 hp at 59 F. A normal 50$ idle speed was chosen because it is certain
161
-------
GAP VEHICLE DESIGN GOAL PERFORMANCE ENVELOPE
• ENGINE SIZED FOR HOT-DAY (ios F> PERFORMANCE
• SSS-10GAS TURBINE
• GE HMT
• VEHICLE WEIGHT = 3850 LB
• ACCESSORY POWER = 1.3 HP
o\
ro
0.8
0.7
g
h-
g 0.6
Q_
*/>
LU
O
0.5
0.4
25-70 MPH IN 15 SEC
100
440 FT IN 10 SEC
110 120 130 140 150 160 170
STANDARD-DAY RATED POWER - HP
I | I I I
180
110 120 130
MAXIMUM HOT-DAY POWER-HP
140
!SO
-------
L-9T12I49-T
For known values of ri and TZ, the radii of the remaining two gears are
d-rrr2
r3 = , r, .=
In Eq. (5)5 the gear face width (thickness), t, was assumed to he constant.
However, the face width is a function of the power to he transferred and the rota-
tional speed. This relationship is expressed as
2 31,500 hp (R+l)3
~ k to R ' ^''
where D = distance tetween gear center lines, in.
t = face width, in.
= input horsepower
R = reduction ratio
a) = pinion speed, rpm
= surface durability factor, lh/in.2
Also:
D = ra + r2 (8)
where rj = pinion radius and r2 = radius of mating gear. Equation (8) can also
"be written as
D = r:(l + R). (9)
Substituting Eq. (0) into Eq. (T) yields
t = 31,500 hp [feU x| (10)
91
-------
L-971249-7 FIG. 37
INERTIAL ENERGY PARAMETER FOR DOUBLE-REDUCTION GEARING
92
-------
L-9712U9-7
A face width parameter, T, can be defined as
Wn
(11)
The second set of reduction gears can be analyzed in a similar manner, yielding
"lr3
(12)
By applying Eq. (ll) to the first two terms of Eq. (5), and Eq. (12) to the last
two terms, Eq. (5), modified for gear thickness, becomes
-
f! * 1)
(13)
Substituting Eq, (6) into Eq. (13) yields an expression for e in terms of the
characteristics of the pinion and mating gears.
+ (d-ri-r2)2 U
i +&) Rn 1R°
Figure 37 presents the variation of the gear system inertial energy
parameter, e, with pinion gear and mating gear size. With a 0.5-in. radius pinion,
the total inertial energy is about 30% less than that for the 1.0 in. radius
pinion, for the minimum-e cases. The 0.5-in. pinion requires a !+-in. mating gear.
The corresponding third and fourth gears will have radii of 3.26 in. and 12.23 in.,
respectively.
Since the gearbox weight will be a small fraction of total vehicle weight,
its weight effect on vehicle performance will be small. However, the gears' inertia
directly affects the response of the engine to throttle changes and is therefore
the dominant factor to be considered.
Figure 38 presents the radii of the third and fourth gears in the double
reduction system. The radius of the final gear is rather large and contributes a
significant amo1mt to the total gear system inertial energy. The basic problem
with using single-stage double reduction gearing is the relatively large span
93
-------
L-971249-7
GEAR SIZE RELATIONSHIPS
16
d = 20 IN.
R0 =30
FIG. 38
14
12
10
I
«/>
3
a
Q!
111
o
MINIMUM ENERGY
r, =0.5
_L
_L
_L
246
GEAR RADIUS, r2~ IN.
9U
-------
L-9712U9-7
between the input and output shafts. If this distance were smaller, the final
gear size could be reduced. The same effect can be achieved by adding a fifth
gear the same size as the final gear, as an idler. A schematic is shown in Fig. 39-
With the idler in the system, the energy parameter, e, is defined as
(15)
The gear radii r3 and r^ are:
(16)
The inertia energy parameter, corrected for varying gear thickness becomes
I"1*®1
(IT)
The inertial energy for this gear train system is about 50% lower than for
the standard double reduction system, as seen by comparing Figs. 37 and 39- Also,
the gear sizes are significantly reduced, as evidenced in Fig. Uo. Now the largest
gear is only 5 in. radius. Consequently, the double reduction system with an
idler was chosen for the reduction gearbox in this study.
Installation Concepts
The installation of any power/transmission system in an automobile should
provide for easy maintenance, minimal susceptibility to impact damage, and should
not encumber the driver or passengers. The selection of a single-shaft gas turbine
offers great flexibility as to installation in a full-size six-passenger sedan,
because of its small size and light weight. The HMT was chosen as a candidate
transmission for purposes of examining the installation problem. The engine and
transmission can be mounted transversely and neatly connected with a double
95
-------
L-971249-7
FIG. 39
INERTIAL ENERGY PARAMETER DOUBLE-REDUCTION GEARING WITH IDLER
d= 20 IN.
R0 =30
96
-------
L-971249-7
FIG. 40
DOUBLE REDUCTION GEARING
GEAR SIZE RELATIONSHIPS
d = 20 IN.
R0 =30
z
I
a
o
MINIMUM ENERGY
= 0.5
4 6
GEAR RADIUS, r2~IN.
= 1.0
= 0.5
10
97
-------
INSTALLATION SKETCH, SINGLE-SHAFT GAS TURBINE WITH HMF, r:;ONT WHEEL. D,'« • L
\
o
-------
L-9712^9-7
reduction gearbox incorporating an idler. The final drive would be through the
epicyclic differential. This arrangement is so compact that a front-wheel drive
(FWD) configuration could easily be developed. Such a system is shown in Figs.
Ul, 1*2, and ^3 for a simple-cycle engine. Drive-line flexibility can be provided
with Rzeppa joints. Figure Itl shows the entire unit neatly mounted between the
frame members.
The air intake is shown only schematically; an actual system would incorporate
a filter system and moisture trap. The exhaust system is conceived to be a straight-
through system which is muffled by long-strand fiber glass, contained by inner
walls perforated according to the noise profile over the length of the system.
The 6-in. exhaust pipe will flatten beneath the vehicle to provide road clearance.
The accessory drive is taken from the idler gear of the reduction gearbox in the
form of a series of pulley belts. Because of the low engine torque, a belt-drive
starter can be incorporated.
The installation shown would allow a large, flat floor space for the front-seat
occupants, and ir.ore importantly provides a large amount of empty space between
the engine and. grill of the vehicle, which is vital in avoiding serious and
expensive damage in light, to moderate, collisions. It is felt that the insurance
companies will watch closely the development of the low-emission systems and would
most certainly impose high rates on any system which is excessively vulnerable to
light collisions. An example would be a collision in which a standard vehicle
sustains a broken grill, a smashed radiator, and fan. A new radiator core costs
about $60 and the fan about $25. This same collision would only bend the intake
ducting for the single-shaft gas turbine-powered vehicle. Conversely, a Rankine-
cycle system would probably sustain damage to the condenser, vapor generator,
regenerator, condenser fans, and considerable plumbing. If such components were
repairable at all, the total labor costs would be very high, and replacement costs
would be similarly severe.
The same basic unit would also make an attractive rear-engine installation,
as shown in Figs. W and k$. The air intake would be through openings near the
rear of the car, using either protruding scoops or flush grilles, depending on the
pressure distribution and airflow in that particular area, as well as styling
features.
The exhaust would be serpentined to provide sufficient length for muffling
and exhaust-gas cooling. Hote that the rear-engine system also provides substantial
crush-structure between the rear bumper and the engine.
The regenerated and recuperated single-shaft configurations are somewhat
larger than the simple-cycle unit, but can still be installed quite nicely in the
same basic configurations. Figures 1+6 and k7 illustrate the front-drive system
using the recuperated engine. Hote that the engine and transmission are reversed
to allow hood clearance over the engine. The rear-drive installation for the
recuperated engine is shown in Figs. hQ and 1*9-
-------
INSTALLATION SKETCH, SSS-10 WITH HMT, FRONT-WHEEL DRIVE
V
o
p
*>.
K>
-------
INSTALLATION SKETCH, SSS-10 WITH HMT, FRONT-WHEEL DRIVE
-------
INSTALLATION SKETCH, SINGLE-SHAFT GAS TURBINE WITH HMT, REAR-WHEEL DRIVE
o
ro
-------
INSTALLATION SKETCH, SSS-10 WITH HMT, REAR-ENGINE DRIVE
o
oo
-------
INSTALLATION SKETCH, RCSS-8 WITH HMT, FRONT-WHEEL DRIVE
I
•o
-------
INSTALLATION SKETCH, RCSS-8 WITH HMT, FRONT-WHEEL DRIVE
-------
INSTALLATION SKETCH, RCSS-8 WITH HMT, REAR-ENGINE DRIVE
<0
I
0
fe
-------
INSTALLATION SKETCH, RCSS-8 WITH HMT, REAR-ENGINE DRIVE
o
—5
hO
fc
I
-------
INSTALLATION SKETCH, RGSS-6WITH HMT, FRONT-WHEEL DRIVE
o
Ca
O
8
-------
INSTALLATION SKETCH, RGSS-6 WITH HMT , FRONT-WHEEL DRIVE
io
REDUCTION GEAR BOX
O
-------
L-9712U9-7
The large regenerated engine is shown in Figs. 50 and 51 for the front-wheel
drive vehicle. The rear-engine system is very similar to the recuperated engine
installation.
Fuel Control
Introduction
This section summarizes the study program conducted by the United Aircraft
Hamilton Standard Division for a fuel control for an automotive gas turbine. The
program was undertaken to conceptually design fuel control systems for three
different engines; namely, single-shaft simple-cycle, single-shaft recuperative,
and single-shaft regenerative engines. The fuel control system discussed in this
report satisfies the requirements for all three engines when used in conjunction
with the General Electric infinitely variable HMT transmission.
The selected mode of control for this application consists of a fuel flow/
compressor discharge pressure (Wf/P3) schedule with exhaust gas temperature (EGT)
limiting. A detailed discussion of the reasons for this mode selection is pre-
sented later. In addition, a schematic diagram and a control packaging concept
have been generated and are discussed at length.
The control evolved during this study program is basically a hydromechanical
unit with an integral fuel pump., a remote inlet guide vane (IGV) actuator and a
remote EGT sensor. Specific design features of the control include the following:
1. Control of starting, acceleration, and deceleration fuel flow
2. Automatic start sequencing as a function of pump (and engine) speed
3. Exhaust gas temperature limiting
k. Automatic IGV actuation as a function of pump (and engine) speed
5. Foot pedal bias of fuel flow
6. Electrical shut-off
T. Automatic altitude compensation from sea level to 10,000 feet
Although the control which was conceptually designed for this application is
hydromechanical, electronic implementation was also considered. Similar control. •
logic was used for an electronic control, and a logic diagram and wiring schematic
were prepared. Both methods of control implementation were then costed, and costs
110
-------
L-9712^9-7
to manufacture either the hydromechanical or electronic version compared
favorably. However, it was decided that the electronic implementation presented
a somewhat larger uncertainty than the hydromechanical version in both cost and
performance. Therefore, a program decision was reached to implement the control
hydromechanically . However, for future programs of this nature, it is recommended
that an electronic control be given consideration. Electronic implementation may
prove beneficial if the automotive engine requirements become somewhat more sophi-
sticated.
Control Mode Determination
The mode of control presented is the result of discussions with UAEL and UACL
relating to implementation of a low-cost fuel control for the single-shaft
regenerative and recuperative engines (RGSS-6 and RCSS-8) and the single-shaft
simple-cycle engine (SSS-10). Fuel control requirements for these engines were
considered in conjunction with the General Electric IVT-8TO infinitely variable
hydromechanical transmission, and although engine temperature, pressure, and fuel
flows differ for the three engines, the operating characteristics allow a common
mode of control for all three engines. Thus, only the detailed requirements of
the SSS-10 have been considered in sizing control components, with the assumption
that differences in the temperature, pressure, and flow requirements of the other
two engines would have a negligible effect on cost, weight, and complexity.
Revi3wof
A typical engine performance map, shown in Fig. 52, indicates engine fuel
requirements for steady-state and acceleration operation of the engine. The
maximum acceleration fuel flow is limited by turbine inlet temperature considera-
tions. .
The steady-state operating line is a result of the infinitely variable
transmission controlling engine speed (as a function of foot pedal position) by
varying transmission drive ratio to hold engine speed constant when the foot
pedal is in a fixed position. Thus, the fuel control is not required to provide
engine speed governing (as is required on aircraft engines) because of the combina-
tion of a fixed-shaft engine and an infinitely variable transmission. Therefore,
fuel control requirements are reduced to providing acceleration and deceleration
fuel limits and auxiliary functions such as IGV, starting, and shut-off functions.
Traditional aircraft engine control methods of controlling turbine inlet
temperature during acceleration are by open-loop scheduling of fuel flow as a
function of speed, inlet temperature, and compressor discharge pressure (P3)
level. These techniques would result in excessive control cost and complexity for
111
-------
L-971249-7
SSS-10 TORQUE CHARACTERISTICS
DESIGN SPEED = 106,000 RPM
DESIGN POWER = 130 'HP
T.I.T. = 1900 F
AMBIENT TEMP- 59 F
FIG. 52
DESIGN POINT FUEL
FLOW = 82.2 LB/HR
CONSTANT POWER, HP/HP MAX
0.6
0.5
DESIGN SPEED - W/W
MAX
112
-------
TURBINE EXHAUST GAS TEMPERATURE (EGT) VS TURBINE INLET TEMPERATURE (TIT)
EGT
STEADY-STATE LINE
50% N
65% N
MAX FUEL FLOW
LIMIT
«o
I
MIN FUEL FLOW
•TIT
-------
L-971249-7
FIG. 54
FUEL FLOW PARAMETER
SSS-10
0.80
0.70 -
0.60 -
D.
0.50 -
0.40 -
0.50 0.60 0.70 0.80 0.90
SPEED RATIO, ~ N/NMAX
•Q MAX
Wf/P3
MIN
Wf/P3
1.0
-------
L-971249-7
Fl C. 94
EFFECT OF OUTPUT SHAFT SPEED ON
SSS-10 FUEL ECONOMY
n
10
o
a.
I
o
u
UJ
iu 8
3
2000
AMBIENT TEMPERATURE = 59 F
VEHICLE WEIGHT = 3850 LB
AVERAGE DRIVING
FDC
I
I
2500 3000 3500
MAXIMUM OUTPUT SHAFT SPEED -RPM
4000
187
-------
L-971249-7
EFFECT OF GAS GENERATOR INERTIA ON
SSS-10 FUEL ECONOMY
AMBIENT TEMPERATURE - 59 F
VEHICLE WEIGHT =3850 LB
FIG. 95
11
10
o
a.
o
o
u
FDC
1 2 3
GAS GENERATOR INERTIA- LBM-IN.2
188
-------
L-971249-7
Fi6.fi
EFFECT OF ACCESSORY POWER ON
SSS-10 FUEL ECONOMY
AMBIENT TEMPERATURE
59 F
n
10
O
Q_
I
O
u
UJ
uj 8
, - AVERAGE DRIVING
FDC
I
2 3
ACCESSORY' POWER - HP
189
-------
L-9712U9-7
TABLE XI
EFFECT OF CHOICE OF OPERATING LINE
ON FUEL ECONOMY
Engine
RGSS-6
RCSS-8
SSS-10
Fuel Flow Line
A
B (base-line)
C
D
A
B (base-line)
C
D
A !
B (base-line )
C
D :
Fuel Economy (mpg)
FDC
13.0U
12.16
9.85
8.77
12.07
11.20
9.00
8.00
7A8
7.51
7.11
6.82
Suburban Route
'19.26
17.52
13. 11*
12.19
17.80
<16.15
12.00
11.15
10.98
11.01
10. 2k
9.90
Country Route
• 17-29
16.13
13. ^1
13.M
16.29
15.06
12. 5k
12. 51*
12 . 00
12.02
11.1*9
.11/U9
Composite
16.08
lU/90
11.88
11.10
ll*. 97
13. -79
10.95
10.20
9.75
9.77
9.21
8.97
190
-------
L-9T12U9-7
effect on average mpg is slight, the change being due to the change in FDC economy.
The latter is adversely affected by increases in engine inertia because of the
continuous variations in vehicle (and therefore engine) speed.
The effect on fuel economy of accessory.power is shown in Fig. 96 for the
SSS-10. Both average and FDC economy figures are adversely affected by increases
in the accessory load. Decreases of 13!? and 16% for the average and FDC fuel
economies,.respectively, are incurred for a change in accessory load from 0 to
5 hp.
The fuel economy results for the various engine operating lines described in
the Engine Performance Section (Fig. 69) are sho'wn In Table XI. Operating line A
shows somewhat better fuel economy than*line B (the base-line operating schedule)
for the two regenerated engines, and approximately fthe same fuel economy;for the
SSS-10 engine. This result is due to the character of the respective fuel flow
maps.,„For,the heat-exchanger engines, the lines of constant fuel flow are fairly-
straight and parallel, with negative slopes. Thus, line A, being above and to
the left of line B, produces better fuel fe;conomy for the RGSS-6 and RCSS-r8 (see - [
the fuel flow maps of Figs. 66 'and 67). , The constant fuel flow lines for the SSS-10,
however, are flatter in the region of lines A and-B, and the fuel economy of this ;.
engine is therefore less sensitive to the selection of the operating line than '
either the RGSS-6 or the RCSS-8. The fuel flow;line of the simple-cycle engine ,
(Fig. 68), in fact, became tangent to the lines of constant power (for low-to- ;i
medium power settings) and there exis'ts an optimum fuel flow line which comprises
the locus of the points of tangency. :
\ <* • ! •'
j Line C reduces the fuel economy for^all three>engines, but the reduction is
less severe for the SSS-10 by the above reasoning. Line D, which utilizes a 60%
idle speed ratio, produces further penalties in:fuel economy,-with decreases of
28%, 29% and 10$ on the FDC for the RGSS-6, RCSS-8 and SSS-10 engines, respectively.
Emissions Results
For the purposes of calculating vehicle emissions over the Federal Driving
Cycle, basic data were expressed in terms of an emissions index (ib of pollutant
produced per 1000 Ib fuel consumed) as a function of engine, power setting for
each of the important constituents. Figure 97 shows the emissions index (El) for
oxides of nitrogen (WOX)_ for the RGSS-6 engine. Line A, representing existing
data for automotive engines, was taken directly from Ref. 1. Line B is the
estimated variation of El required to just meet the Federal standards for NOX.
Figure 98 shows similar data for the RCSS-8. The NOX El data for the SSS-10 engine
are shown in Fig. 99. Line A is representative of data for the T-56 engine as
described earlier, and line B is an estimate of the El variation which would just
meet the Federal standards for UOX emissions.
191
-------
L-971249-7
FIG. 97
EMISSIONS INDEX FOR NITROGEN OXIDES (NOX)
(EXPRESSED AS N02)
RGSS-6
2
x
x
Ul
O
f>
—
Ill
NORMALIZED POWER - HP/HP.
MAX
192
-------
L-971249-7
FIG.98
EMISSIONS INDEX FOR NITROGEN OXIDES (NOX)
(EXPRESSED AS N02)
RCSS-8
20
16
2
x
oa
x
UJ
Q
Z
2
Ul
12
0.2
0.4
0.6
0.8
1.0
< NORMALIZED POWER - HP/HP
MAX
193
-------
L-971249-7
FIG. 99
EMISSIONS INDEX FOR NITROGEN OXIDES (NOX)
(EXPRESSED AS N02)
SSS-10
10
£fl
x
ui
UJ
0.2
0.4
0.6
0.8
1.0
NORMALIZED POWER - HP/HP
MAX
-------
L-971249-7
HG. 100
EMISSIONS INDEX FOR CARBON MONOXIDE (CO)
RGSS-6
RCSS-8
28
NORMALIZED POWER - HP/HP,
MAX
195
-------
L-9712U9-7
TABLE XIT
FEDERAL DRIVING CYCLE EMISSIONS DATA
Baseline Vehicles
Ambient Temperature = 59 F
Engine Operating Line B (Fig.69 )
Emissions (gm/mi)
Engine
RGSS-6
RCSS-8
SSS-10
Emissions Index
Line & Source
Figs. 97 Through 103
A - GT 309
B - Typical Required
A - Modified GT309
B - Typical Required
A - T-56
B - Typical Required
CO
0.53
2.90
0.57
3.13
1.86
2.58
UHC
(as C6Hllt)
0.15
0.38
0.16
O.Ul
0.31
o.Uo
NOX
(as N02)
2.72
O.Ul
2.38
O.Ul
1.03
o.ui
Federal Standards
3.1+0
O.Ul
0.40
196
-------
L9712U9-7
The carbon monoxide (CO) El variations for the RGSS-6 and RCSS-8 engines are
shown in Fig. 100. Line A was taken from Ref. 1, and line B is the estimated CO
allowable to approximately meet the requirements. Figure 101 shows CO El variations
for the SSS-10 engine, where line A represents data for the T-56 engine, and line B
is an estimate of the CO El variation allowable to still approximately meet the
requirements.
The unburned hydrocarbons (UHC) El data for the RGSS-6 and RCSS-8 engines
are shown in Fig. 102. Lines A and B were derived as described above for the
heat-exchanger engines. Figure 103 shows the UHC estimates for the SSS-10 engine.
Line A is based on the T-56 data, and line B is the estimate for allowable
emissions.
The emissions results as calculated for the base-line vehicles operating over
the Federal Driving Cycle (FDC) are shown in Table XII along with the current
Federal Standards. All results were derived for the base-line vehicles using
engine operating line B of Fig. 69. The existing El data for the heat-exhanger
engines (line A) yield HOX emissions greater than the Federal Standards by a
factor of 6. The corresponding CO and UHC levels are well below the standards.
A reduction in the NOX El levels for the RGSS-6 and RCSS-8 (line B, Figs. 91 and
98, respectively) is required to meet HOX standards, but a substantial increase
in CO and UHC is also allowable.
The T-56 data used to derive curve A for the SSS-10 engine yield NOX emissions
of 1.03 gm/mi. This value is greater than the allowable standard by a factor of
2.5, but still well below the NOX levels of the RGSS-6 and RCSS-8. The El data
of curve B, for the SSS-10, yield O.Ul gm/mi of NOX, with CO and UHC emissions of
2.58 and O.ltO gm/mi, respectively.
The effect of vehicle weight on emissions is shown in Fig. 10l+ for the SSS-10
engine. Increasing vehicle weight causes the CO and NOX production to rise,
although the variation of CO with weight is slight. The UHC emissions decrease
with increasing vehicle weight because the negative slope of the UHC El curve of
Fig. 103 more than offsets the increases in average power and fuel consumed which
occur with increasing weight. Figure 105 shows the effect on emissions (for the
SSS-10) of varying the rear-axle ratio. The CO emissions are relatively unaffected,
but the NOX emissions attain a minimum at an axle ratio of about 2.75. The UHC
emissions experience a "weak" maximum at the same axle ratio. The effect of
accessory power on emissions for the SSS-10 engine is shown in Fig. 106. The trends
are similar to those indicated in Fig. 10!+, and the UHC emissions decrease with
increasing accessory power. The KOX emissions are strongly influenced by the ,
accessory power loading, with an increase of 22$ occurring as the accessory load
increases from 0 to 5 hp.
197
-------
L-971249-7
FIG.101
EMISSIONS INDEX FOR CARBON MONOXIDE (CO)
SSS-10
0.2
0.4 0.6
NORMALIZED POWER - HP/HP
0.8
1.0
MAX
198
-------
L-971249-7
FIG. 102
EMISSIONS INDEX FOR
UNBURNED HYDROCARBONS (UHC) (EXPRESSED AS C6H]4)
RCSS-fr
RCSS-8
0.2
0.4 0.6
NORMALIZED POWER - HP/HP
199
0.8
MAX
-------
L-971249-7
FIG. 103
EMISSIONS INDEX FOR
UNBURNED HYDROCARBONS (UHC) (EXPRESSED AS C6H14)
SSS-10
10
o
x
ca
CO
-J
I
X
Q
z.
o
0,4 0.6
NORMALIZED POWER - HP/HP
0.8
MAX
200
-------
L-971249-7
FIG. 104
EFFECT OF VEHICLE WEIGHT ON EMISSIONS
SSS-10 ENGINE
FEDERAL DRIVING CYCLE
AMBIENT TEMPERATURE = 59 F
HYDROMECHANICAL TRANSMISSION
MAXIMUM POWER = 130 HP (5y F)
EMISSIONS INDEX C-LINES
3.0
2.5
2.0
1.5
o
u
1.0
0.5
I
0.60
0.55
0.50 5
o
i
0.45
X
O
0.40
u
IE
3000 3500 4000 4500
VEHICLE WEIGHT - LB
0.35
0.30
5000
201
-------
L-971249-7
FIG. 105
EFFECT OF REAR-AXLE RATIO ON EMISSIONS
SSS-10 ENGINE
FEDERAL DRIVING CYCLE
AMBIENT TEMPERATURE-59 f
HYDROMECHANICAL TRANSMISSION
MAXIMUM POWER - 130 HP
EMISSIONS INDEX C-LINES
3.0,
CO
2.51
2.01
o
i
i/j
z
o
LU
O
u
1.5 I
NO
NOV
1.0
UHC
UHC
0.51
J_
0.60
0.55
0.50
s
o
1
0.45 %
5
in
X
o
z
u
0.40 =>
0.35
0.30
3 4
REAR-AXLE RATIO
202
-------
1-971249-7
FIG. 106
EFFECT OF ACCESSORY POWER ON EMISSIONS
SSS-10 ENGINE
FEDERAL DRIVING CYCLE
AMBIENT TEMPERATURE =59 F
HYDRO-MECHANICAL TRANSMISSION
MAXIMUM POWER - 130 HP
EMISSIONS INDEX C-LINES
3.0
2.5-
2.0
5
s
o
s
UJ
o
u
1.5
1.0
0.5
CO
0.60
0.55
0.50 2
^
o
i
0.45
0.40
o
ox
u"
X
UHC
•0.35
I
0.30
2 3
ACCESSORY POWER -HP
203
-------
L-97121*9-7
201;
-------
L-9712^9-7
ECONOMIC ANALYSIS
The economic analysis is concerned with the. estimated owning and operating
costs to the vehicle owner over the life of the vehicle. The vehicle life has been
specified by OAP to be 15,000 miles annually for seven years, or a total of 1059000
miles.
The figure of merit used in this economic analysis is the total lifetime cost.
(TLC) of the engine since the costs associated with the remainder of the vehicle
are assumed to be constant regardless of the engine type. In a later section, seme
of the effects of the propulsion system upon the vehicle itself are also considered,,
The total lifetime costs include the initial engine cost as derived froa the direct
manufacturing cost, the cost of capital, fuel, repairs, and maintenance. The method
of calculation employed is discussed below.
Direct Manufacturing Cost Estimates
Manufacturing cost estimates were prepared for each engine design. Figures
107, 108 and 109 show the parts content for the RGSS-6, RCSS-G and SSS-10,
respectively. The part numbers are keyed to Tables XIII, XIV and XV, respectively,
for the RGSS-6, RCSS-8 and SSS-10, which present the manufacturing cost estimates.
These estimates of direct manufacturing costs are defined as the sum of the
direct materials costs and the direct labor costs. It should be noted that the
most important value for each of these engines is the combined total of direct
materials plus direct labor (the latter value of which is converted from minutes
to dollars at an appropriate wage). This distribution is emphasized because in
those circumstances where a component part most likely would not be made in an
engine-maker's facility, estimates of both labor and materials could not be made.
As a result, the noted direct labor times correspond only to the estimated time
allocated to the engine in the engine-maker's facilities.
The specific component parts defined in each of the three tables constitute
over 91% of the total parts costs for each engine, and 100$ of the parts whose
direct cost is $1.00 or more. Each of the noted parts, as well as some included
under the miscellaneous category, was examined individually to determine its weight,
material, and machining requirements. The background in production methods used
for the study reported in Hef. 1 was relied upon extensively to support the analyses
of the present study. However, whereas manufacturing cost estimates of Bef. 1
included purchased part costs with an allowance for machinery isacrtization in
addition to the cost of the basic materials, the results pres ;,uted in Tables XIII
through XV are for basic materials only.
205
-------
RGSS-6 PARTS DETAIL
-------
L-9712l*9-7
TABLE XIII
RGSS-6 DIRECT MANUFACTURING COST ESTIMATES
Part No.*
Item
1* Compressor Shroud
5 Impeller
8 Diffuser Case
9 Pressure Vessel
10 Combustor Can
16 Combustor Scroll
17 Turbine Nozzle
18 Turbine Shroud-Front
20 Turbine Shroud-Rear
23 Turbine Rotor
2k Turbine Shaft
26,30 Shaft Seals (3)
27,31 Bearings, Antifriction (3)
36 Cover, Press. Vessel-Rear
37 Exhaust Diffuser
38 Cover, Regen. Disk Housing
39 Regenerator Disk
1*0 Regen. Support, Seals, Gears, etc.
1*2,kk Regen. Drive Pulley, etc.
1*5 Case, Reduction Gearbox
1*5 Reduction Gears
^5 Cover, reduction gearbox
Miscellaneous
Fuel Control
Oil Pump
Alternator, Starter, Voltage Regulator
Final Assembly & Test
Subtotal
Direct Labor § $5/hr
Total Direct Mfg. Cost
Material**
Ductile Cast Iron
AMS U928
Ductile Cast Iron
Ductile Cast Iron
Hastelloy -X (IN 625)
Hast ell oy -X IN 625)
WI-52 (IN 738)
WI-52 (IN 738)
WI-52 (IN 738)
U700 (IN 100)
SAE 1*31*0
Carbon & Cast Iron
52 CB
Heat Resist. Cast Iron
Heat Resist. Cast Iron
Ductile Cast Iron
CERCOR (Ceramic)
Ductile Cast Iron
Cast Iron
Gear Steel
Cast Iron
— «. _
_ _ _
or - - -
Direct
Material
Cost - $
3.85
18.55
7-85
26.60
3.20
18.10
2U.OO
15-95
U.35
20.30
1.50
2.50
10.50
3.05
3.35
6.70
16.30
80.50
1.50
5.70
13. UO
7-35
20.55
39-50
1*.50
1^.55
_ _ _
$37'*. 20
16.95
Direct
Labor -
Min.
2.8
9.8
18.1)
6.0
2.0
i*.o
9.8
2.5
2.1
9.5
5.8
incl.
incl.
2.0
3.0
6.0
incl.
19.0
3.0
3.5
U3.1
U.5
11.2
incl.
incl.
incl.
35.0
202.7
$391-15
•See Fig. 107
••Alternate materials for certain high-cost parts shown in parenthesis.
Final specification to depend on cost.
207
-------
RCSS-4 PARTS DETAIL
e~
3
ro
o
co
-------
L-9712U9-7
TABLE XIV
ECSS-8 ESTIMATED DIRECT MANUFACTURING COST
Part No.*
Item
1+ Compressor Shroud
5 Impeller
8 Diffuser Case
9 Diffuser-to-Recup. Casing
10 Combustor Can
13 Recuperator
16 Combustor Scroll
17 Turbine Nozzle
18 Turbine Shroud-Front
20 Turbine Shroud-Rear
23 Turbine Rotor
2l+ Turbine Shaft
26 Recuperator Casing
27,31 Bearings, Antifriction (3)
30 Shaft Seals (3) •
36 Case, Reduction Gearbox "
36 Reduction Gears
36 Cover, reduction gearbox
37 Exhaust Diffuser
Miscellaneous
Fuel Control
Oil Pump
Alt., Start., Volt. Reg.
Final Assembly & Test
Subtotal
Di/rect Labor @ $5/hr
Total Direct Mfg. Cost
Material**
Ductile Cast Iron
AMS 1*928
Ductile Cast Iron
Ductile Cast Iron
Has tell oy-X (IN 625)
Type 301+ Stainless
Hastelloy-X (IN 625)
WI-52 (IN 738)
VI -52 (IN 738)
WI-52 (IN 738)
U700 (IN 100)
SAE 1+31+0
Ductile Cast Iron
52 CB
Carbon & Cast Iron
Cast Iron
Gear Steel
Cast Iron
Heat Resist. Cast Iron
Direct
Material
Cost - $
1.90
13.70
3.50
12.50
3.50
77.!+5
15.80
20.30
13.10
U. 65
19.80
1.05
10.95
10.50
2.50
5.70
13. ^0
7-35
5-65
Direct
Labor -
Min.
2.7
6.0
ll+.O
3.0
2.5
1+3.0
l+.O
10.1
3.6
2.2
9.1
5.0
8.5
incl.
incl.
3.5
1+3.1
1+.5
3.0
16.85
39.50
1+.50
$333.1+5
17,35
$350.80
9-1
incl.
incl.
incl.
30.8
207.7
•See Fig. 108
••Alternate materials for certain high-cost parts shown in parenthesis.
Final specification to depend on cost.
209
-------
SSS-10 PARTS DETAIL
ro
H
o
-------
L-9712U9-7
TABLE XV
SSS-10 DIRECT MANUFACTURING COST ESTIMATES
Part No.*
U
5
Item
9
10
16
IT
18
20
23
2k
26,30
27,31,33
39
1+0
U2
Compressor Shroud
Impeller
Diffuser Case
Pressure Vessel
Combustor Can
Combustor Scroll
Turbine Nozzle
Turbine Shroud - Front
Turbine Shroud - Rear
Turbine Rotor
Turbine Shaft
Shaft Seals (3)
Bearings - Antifriction (3)
Cover, Pressure Vessel - Rear
Case, Reduction Gearbox
Cover, Reduction Gearbox
Gears, Reduction
Miscellaneous
Fuel Control
Oil Pump
Alt., Start., Volt. Reg.
Final Assembly and Test
Subtotal
Direct Labor & $5/hr
Total Direct Mfg. Cost
Material**
Ductile Cast Iron
6-2-1+-6 Titanium
Ductile Cast Iron
Ductile Cast Iron
Hastelloy-X (IN 625)
Hastelloy-X (IN 625)
WI-52 (IN 738)
WI-52 (IN 738)
WI-52 (IN 738)
U700 (IN 100)
SAE 1+3J+0
Carbon & Cast Iron
52 CB
Cast Iron
Cast Iron
Cast Iron
Gear Steel
Direct
Material
Cost - $
1.30
13.35
!+.95
3.00
2.25
16.05
16. Uo
8.10
5.15
19.60
1.30
2.50
10.50
2.80
5.70
7.35
13.1+0
Direct
Labor -
Min.
2.6
5.7
16.0
6.0
2.0
l+.O
10.7
2.2
2.3
8.6
5.5
incl.
incl.
1.9
3.5
^.5
1*3.1
15.1+0
9.5
39.50
4.50
1^.55
$207.65
12.1+5
incl .
incl.
incl.
21.0
11+9.1
$220.10
*See Fig. 109
**Alternate materials for certain high cost parts shown in parenthesis.
Final specification to depend on lovest cost.
211
-------
L-9712U9-7
The following assumptions were common to all three engine designs in this
analysis. All cast iron parts were estimated for a material cost of between 15 #
and 200 per pound, depending on whether or not a high-temperature version of this
material was required. For all components, material cost estimates were based on
the weights before machining and, as such, include a scrap allowance. However, in
most circumstances, this scrap allowance never exceeds 20%, a. practice which is
quite typical of the high-volume, price-conscious automotive industry. Materials
such as coated WI-52 or its alternate IK 718, both of which could be specified for
investment-cast parts, were assumed to cost slightly less than $3 per Ib.
Hastelloy-X, which was used in burner cans and high-temperature sheet metal parts,
was estimated at $ij.50 per Ib in sheet form. In all engines, both the inducer/
impellers and the turbine rotors were assumed to be produced by advanced forging
techniques. One approach to advanced forging is the superplastic method (Gatorizing )
which recently has been perfected at the Pratt & Whitney Aircraft Division of UAC.
Previous analyses and consultations with Corporate personnel familiar with this
process made it possible to say that in high-volume production the raw material
costs for superplactically forged parts will cost approximately $7 per Ib, a value
which is practically independent of the type of material specified. Furthermore,
it is believed that because of the nature of this production process, labor times
of approximately 3 min per Ib (for small parts) would be required in the forging,
trimming and finishing operations of each part.
As noted above, the direct costs of purchased parts include profit, overhead
and machinery depreciation allowances in addition to the costs of labor and
materials. Previous consultation with suppliers of these high-vo-lume-preduction
items has led to the discovery that in most cases, the O.E.M. prices quoted by
a manufacturer of a part are approximately 1.8 to 2.0 times the direct production
cost of that component. As a result, all purchased component price data were
divided by this price/manufacturing cost ratio to reflect the present best estimate
of direct production cost for each of these items. In general, items which were
included in the vendor-purchased category were seals, bearings, starters, alternators,
voltage regulators, oil pumps, fuel controls and heat exchangers (where necessary).
Prices for these items were obtained from specialist vendors who either supply
these items directly to the automotive industry or who had an intimate working
knowledge of the prices of the various parts.
In the case of the regenerator or recuperator of the EGSS-6 and RCSS-8 engines,
detailed analyses of the various heat exchanger components were made. The ceramic-
type regenerator actually represents i complicated mechanism incorporating a chain-
driven gear drive, a disk holder and :he disk sealing mechanism. A previous study
(Ref. l) indicated that the heat recovery approach to automotive gas turbine design
is expensive, and,from the results generated in this study, this conclusion is sub-
stantiated. Of the nearly $100 indirect material costs associated with the re-
generator, the ceramic disk and its support and drive mechanism constitute
212
-------
L-9712U9-7
approximately $HO, whereas the direct cost of the sealing mechanism is estimated
to cost over $50. Discussions with the major manufacturer of this type of dynamic
seal have revealed that unless severe compromises are made on the part of the engine
maker, there is little chance that direct production costs of the seals can be re-
duced, even in high-volume production.
The recuperative type of heat exchanger represents a different problem "because
a large tube area must be provided to attain a reasonable effectiveness for the
thermodynamic engine cycle. As a result, 21^0 tubes, each with a diameter of 0.125
in. and an average length of slightly more than 23 in., are required. With Type-304
stainless steel formed from strip stock 0.010 in. thick, the base price of the
tubes alone becomes $61;.50. The addition of end'headers increases this cost to
slightly more than $77 per unit. Total labor time is estimated at a conservative
k3 minutes per unit even with the use of highly automated equipment. As a result,
the estimated cost of the heat exchanger on the recuperative-design automotive gas
turbine engine is only slightly less than the heat exchanger in the regenerated'
RGSS-6 engine. It must be emphasized, however, that there is currently no assurance
that satisfactory recuperators can be manufactured by these techniques with these
materials, and that a large-scale development program would be necessary to develop
not only more accurate cost estimates, but a realistic assessment of the practical
problems involved. J
Although no specific analysis was made, it is believed that the accuracy of
the material cost estimates are within ilO to 15% based on a root-mean-square
component-by-component analysis. This estimate of accuracy is approximately the
same as that noted in Ref. 1 since all engine components were analyzed in a similar
manner. Estimates of the direct labor probably have a higher statistical error,
although even a larger error would not significantly affect the results noted
because labor constitutes such a small portion of the total direct cost. Within
a given organization, a rather large multiplier normally is applied to labor to
account for overhead items (such as equipment amortization) which are intrinsic
to an individual firm's operations. Such overheads have not been included in this
analysis because this multiplier is an accounting convenience, is not a consistent
value, and is not necessarily a strong function of direct labor hours. Conse-
quently, it is believed that the results presented in Tables XIII, XIV and XV
represent an estimate of the direct cha.rges associated with the production of each
engine which reasonably could be attained in the 1976-77 period.
Cost of Capital
The cost of capital includes those costs related to investments either by the
owner or his creditors and/or to the value of the vehicle. In order to apportion
these costs it is necessary to rta-se several assumptions. First, as mentioned
abovej it is assumed that the ncnpropulsion system portions of the vehicle behave
in an identical fashion for all engines and, consequently, will not enter into the
213
-------
L-9712^-9-7
evaluation of the candidate engines and their propulsion systems. Secondly, it is
assumed that depreciation rates, interest rates, tax rates, etc., are unaffected
by the type of engine and that the only differences among the systems will be those
determined by the relative initial prices. It is further assumed that there is
a direct relation between manufacturing cost and' retail price and that this relation
is represented by a factor of 2.7 in order to mark up the direct manufacturing cost
to the original discounted retail price as per Ref. 1. Finally it is assumed
that these costs will continue to be incurred in accordance with historical trends.
The capital costs involved are those resulting from depreciation, interest
lost, finance charges, taxes, and collision and comprehensive insurance. All of
these may be reflected back to the original purchase price of the engine in terms
of probable lifetime percentage values.
Depreciation over the lifetime of the vehicle can be defined as the total of
the retail sales price minus the total of the trade-in (wholesale) price or allowance.
On a percentage basis, this sum is divided by the original sales price. Statistics
of Ref. 20 indicate that the average car bought new is traded at 3.9 years, and
that the average car bought used is traded at an average age of 7 years. If it is
assumed that all the study vehicles depreciate and resell on the same historical
basis, which includes the dealer's profit and cost on each of the two resales, a
depreciation value of approximately 120$ relative to the original value of the
vehicle can be derived using.Ref. 21 data for standard-sized Ford and Chevrolet
six-passenger V-8 sedans.
Appendix IV describes the historical data used to establish the cost of
ownership of conventional automobiles. On this basis, it can be determined that
the overall cost of capital historically amounts to approximately 1.66% of the
original price of the vehicle. Thus, the cost of capital for the engines con-
sidered in this study may be derived by multiplying their direct manufacturing
cost by 2.7 (to account for average retail price) and further multiplying this
figure by 1.66 (to account for depreciation, interest, taxes, finance charges
and insurance costs).
Cost of Fuel
The cost of fuel to the owner can be .established by multiplying the 105,000
miles driven by the assumed fuel cost in dollars per gallon and dividing the product
by the miles per gallon for the prescribed OAP driving cycle for each candidate
engine. The difficulty in applying this figure, lies,in estimating the price of
the fuel. Retail: pricing in the petroleum industry is a complex subject and results in-
highly varying practices. The competitiveness of this business results in large
swings which bear little or no immediate relation to the actual cost of refining.
21k
-------
L-9712U9-7
In the long run, however, it does appear that fuel costs must be related to refining
costs and to supplies. The prime feature of the fuel for the gas turbine engines
considered in this report is the lower pump delivered price. There are no octane
requirements nor cetane requirements, and straight-run distillery fuels (either
gasoline or light oils) would provide satisfactory fuel for the engines. Since
refining costs for gasoline (and particularly high-octane gasolines and most
particularly gasolines without lead) suitable for operation in motor vehicles are
necessarily higher than for turbine fuels, it appears safe to assume that, at least
for the foreseeable future, average pump prices for turbine fuels should be lower
than those for gasoline. This situation would be true even if the entire vehicle
production were converted, immediately, from Otto-cycle to gas turbine engines,
since in any given year only 10% of the vehicle population is replaced, and pre-
sumably the majority of refinery output for motor vehicles would continue to be
gasoline.
The assumed gas turbine fuel cost in this study has been assumed to be 31^/gal.*
The cost could easily range from 25^ to 1+0^, particularly for varying localities
and times, and quite obviously the engine with the higher rate of fuel consumption
is more sensitive to increased fuel prices.
Cost of Repairs and Maintenance
The basic data for estimating the repair and maintenance costs are the his-
torical data described in Appendix IV. Maintenance and repair estimates for the
three candidate gas turbine engines are also compared on this use basis and
involve estimation of frequency of repair as well as estimates of mechanic labor
and direct manufacturing costs of the parts involved. In this case, the'retail
spare parts cost is arrived at by multiplying the direct manufacturing cost estimate
by 3.5. ' -"•' :
The maintenance estimate for the three candidate' engines is shown in Table XVI
and the lifetime repair estimate is shown in Table XVII. It will be noted in Table
XVI that periodic maintenance is reduced considerably over that required historically,
and tune-up requirements become a minor item in the cost of maintaining a vehicle.
Primarily as a result of this situation, the repair and maintenance costs of the
gas turbine engines are predicted to be quite attractive in comparison with con-
ventional engines. x
Quite naturally, it is most difficult to predict the cost of repairs and
maintenance for new systems. Such predictions must always be only approximate
prior to the accumulation of sufficient amounts of historical data. While it
cannot be stated that the repair and maintenance costs for any of the gas turbine
engines would fall within the range assumed, it is felt that> the estimates made
here are logical extrapolations based on present knowledge. ' *
*A fuel cost of 33<£/gas was assumed for the preliminary candidate selection process
described earlier; however, at the request of OAF, the value was changed to 31^/gal
for the final engine selection.
215
-------
L-9712^9-7
TABLE XVI
ESTIMATED GAS TURBINE MAINTENANCE COSTS
Annual: (15,000 mi)
Check and adjust belts
Replace ignitor
Replace fuel filter
Check and calibrate•control
Replace air filter
Biennial: (30,000 mi)
Replace belts
Replace ignitor
Replace fuel filter
Check and calibrate control
Replace air filter
Check-balance
Change oil (SSS-10 only)
Check, heat exchanger (RGSS-6;RCSS-f
only)
Total Maintenance Costs (h Annual, 3 Biennial);
RGSS-6 = $279
RCSS-8 = $279
SSS-10 = $219
$61
216
-------
L-9712^9-7
TABLE XVII
GAS TURBINE REPAIR ESTIMATES
Material
Later Total per Lifetime
Repair Item E:
Repair Casings
Replace and Repair:
Combustor
Fuel Nozzle
Ignitor
Scroll
Turbine Nozzle
Shrouds
Wheel
Seals
Bearings
Reduction Gearing
Regen . Core
Reg en. Seals
Recup. Seals
ngine Frequency
All
RGSS-6
RCSS-8
SSS-10
All
All
RGSS-6
RCSS-8
SSS-10
RGSS-6
RCSS-8
SSS-10
RGSS-6
RCSS-8
SSS-10
RGSS-6
RCSS-8
SSS-10
All
All
All
RGSS-6
RGSS-6
RCSS-8
Total Lifetime Repair Costs
0.1
0.6
0.6
0.6
1.2
0.2
0.2
0.2
0.6 .
0.6
0.6
0.3
6.3 '
0.3
0.2 -
0.2
0.2
0.6
0.6
0.3
0.8
1.2
0.8
- RGSS-6
RCSS-8
SSS-10
Cost ($) Hours Repair ($)
0
12.80 •'
ii*. oo -' ;"- >•
9.00
2.00 ' -••
: Included
72
1*8 --•
1*8-; ••
- ;98.;,..
81 - - ' -
65 - '••
80
72
52
• 81
80
•'" ' 80 •" ' :
' "•-• 20
in
52
6U
320
310
k
• 2
< U
2
1
in
i*.
'k
h
-•• 7
T
6'
8
8
7
8
8
7
7
7
k
2
2
2
UO
33
•' 5U
29
12
Maintenance
112
88
88
168
•; 151
125
160
152
122
• '•••• l6l ..
160
150
90
111
92
8U
3^0
330
Cost ($)
h
20
32
17
lU
22
18
18
100
91
75
5^
U6
37
32
32
30
5^
66
19
67
UlO
261*
$862
$6UO
$33U
(1)
Frequency of repair over total lifetime; where less than 1.0, only fraction of
vehicles would require repair; e.q., one in ten vehicles would require casing repair.
217
-------
L-9712>+9-7
TABLE XVIII
TOTAL LIFETIME COSTS FOR PROPULSION SYSTEM - (TLC,
Fuel Economy "->*+ ™r Cost of Cost of
Engine mpg* Cost** Maintenance Repairs DMC Capital*** TLC-nS
HGSS-6
RCSS-8
SSS-10
llj-,9
13.8
9.8
2180
2360
3330
279
279
219
862
6ho
33U
391
351
220
1760
1520
990
5081
It859
U8T3
* GAP driving cycle
** Fuel @ 31^/gal
*** U.5 X DMC
218
-------
L-9T12H9-7
Results
The results of the economic analysis are shown in Table XVIII. As can be
seen, the total lifetime costs are extremely close for the three candidate engines.
Certainly, they are too close to clearly favor one engine over the other in view
of the accuracy of the assumptions entered.
Originally it had been thought that the economic analysis would provide a basis
for recommending the selection of one engine over the others. However, it is
obvious that the figures shown in Table XVIII are not sufficiently different to
allow the selection of any one engine. Therefore, a different selection procedure
had to be especially devised to meet the primary objective of the study. This
procedure is described in the next section.
219
-------
L-9712U9-T
220
-------
L-9712U9-7
RECOMMENDED ENGIUE SELECTION
The available information and "best estimates' concerning the performance, cost,
and other relevant characteristics of the various types of engines considered
(including, for comparison, present and future Otto-cycle engines) are summarized
and discussed briefly in this section, and recommendations are made as to which
of these engines is most likely to satisfy the requirements for clean, economical
prime movers for automobiles.
Emissions
Nitrogen Oxides
Nitrogen oxides (WOX) are the most dangerous of the pollutants attributable
to the motor vehicle and have been consistently judged as representing the most
serious pollution problem. Independent sources" have estimated its harmfulness
as. much as 50 times higher than that for carbon monoxide. The problem of controlling
NOX is not only the most important of those facing the automobile industry; it is
also generally considered to be the most difficult.
The results of this study show the SSS-10 to have a major advantage with
regard to NOX production. It is believed to be virtually certain that combustors
can be developed at reasonable cost for engines of the SSS-10 type which will
permit them to meet the 1976 Federal Emissions Standards of O.h g/mi for NOX.
As noted in the previous section, this requirement represents approximately a 60%
reduction in what is considered the state of the art for NOX emissions of untreated
simple-cycle engines. There is some indication that, because of low values for
UHC and CO emissions, slight increases in typical UHC production and a doubling
of typical CO production may be allowable in striving for the NOX standard. Pre-
liminary results of UACL combustor testing programs* have been inconclusive in this
regard, and definitive results are not in hand at the time of publication of this
report.
In the case of the heat-exchanger engines, it is far less certain that the NOX
standards can be met. Unconventional combustor design is almost certain to be called
for and, even if the required NOX reduction is demonstrated in test stands, con-
siderable question exists as to the likelihood of accommodating the combustor design
in a feasible automotive engine design, and whether such a design can be manufactured
at competitive costs.
A combustor development program is under way at UACL, with UARL participation,
to demonstrate that the desired emissions levels can be achieved, under EPA
Contract 68-OU-0015. Both regenerated and simple-cycle combustors are being
developed, and other contractors are also working with EPA support.
221
-------
The Otto-cycle engine is very unlikely to achieve the required level of NOX
production by 1976. Reference 22, for example, describes a maximum-effort auto-
mobile which will produce HOX emissions still too high by a factor of two. It is
expected that Wankel engines will have similar problems in meeting the standards.
Therefore, not only does the SSS-10 have a significant advantage in meeting
HOX standards, it appears that it is the only practical engine with a high proba-
bility of achieving this requirement.
Unburned Hydrocarbons (UHC)
and Carbon Monoxide (CO)
The production of both UHC and CO is related primarily to the efficiency
of combustion. They are, therefore, more easily controllable than the production
of KOX, which is fundamentally related to the combustion process itself. In
addition, their importance from a health standpoint is considerably less than that
associated with NOX. All of the candidate gas turbine engines have inherently
very low production levels of CO and UHC, unlike the Otto-cycle engine. However,
even the Otto-cycle engine may meet the UHC and CO standards by using post-
combustion treatment, at least for the majority of vehicles when they are produced.
Both the regenerative and recuperative gas turbine engines are certain to
achieve, these standards, since compliance has already been demonstrated by similar
engines. Even in the case of the simple-cycle engine, whose basic rate of
production of CO and UHC is somewhat higher than the regenerative engines, there
is little doubt that the standards will be met.
Production Variability
Production variability has been noted as of extreme importance in the case
of production-line Otto-cycle engines, which will rely heavily on post-combustor
treatment devices (such as catalytic converters) to convert a highly noxious
exhaust stream into a relatively pure one. The natural state of emissions for
Otto-cycle engines is some ten' to one hundred times worse than that required to
meet the standards, and in many cases the natural state is made even worse in an
attempt to lower NOX. Consequently, if a converter with a design efficiency of
98$, for example, delivers only 96$ efficiency as a result of normal production-
line tolerance stackups, the emissions will obviously double. On the other hand,
the gas turbines are inherently clean-burning engines, so that production-line
tolerance stackups will not result in significantly altered emissions levels.
The only area of possible concern would be a complex combustor design, such as
may be required for the heat-exchanger engines. The possibility of emission-
performance degradation due to tolerance stackups will have to be addressed for
these burners, and appropriate quality-control measures will have to be used.
222
-------
L-9712^9-7
Deterioration in Service
A major element in the emissions evaluation is the likelihood of satisfactory-
emissions performance over the lifetime of the vehicle. In the case of the SSS-10,
its emissions performance is likely to be constant, varying "by no more than 20 to
25$ in the extreme, with such variations as do occur "being caused primarily "by
lifetime degradations of power and component efficiency. In the case of the heat-
exchanger gas turbine engines, a slightly lower degree of lifetime performance is
expected because of heat exchanger leakage. Also, unconventional combustors will
undoubtedly be required to meet the NOx requirements in the heat-exchanger engines,
which may break down with consequent increase in emissions. In the case of the
Otto-cycle engine, there is presently no assurance that the required add-on equip-
ment can be made to last even 50,000 mi, i.e., half of the lifetime of the car.
As a result, in-service inspections would be required to check deterioration at
considerable inconvenience and cost to the owner.
Emissions Summary
In summary, it is believed that the SSS-10 engine possesses distinct advan-
tages with regard to the all-important emissions area. It is the only practical
engine with a high assurance of meeting the HOX goals. It, together with the
heat-exchanger gas turbines, offers the only reasonable approach among the
engines considered here for low lifetime-average CO and UHC formation.
Initial Price and Production Cost Uncertainties
It is believed that the strongest measurable factor governing buyer appeal
for a given set of vehicle options is the initial price. Historically, most
prospective car buyers are far more concerned with the initial price of the
vehicle than with its operating or lifetime costs.
Table XIX presents an analysis of price estimates which are based on (a)
vehicle weight estimates and (b) engine manufacturing cost estimates. For
purposes of standardization, OAP evaluation procedures require that the non-
propulsion system portion of the standard six-passenger vehicle weight be fixed
at 2700 Ib. In actuality, the weight of such components as springs, brakes,
tires, wheels, etc., are affected strongly by propulsion system weights. Auto-
mobile manufacturers tend to expect to save about one additional pound in vehicle
weight for every pound saved in the engine weight. A somewhat more conservative
value for this ratio can be derived as follows, based on a component breakdown
of vehicle weights originally established by Hoffman (Ref. 23). Of the total
vehicle weight, the trim and glass (total of 690 Ib) are assumed to be unaffected
by propulsion system weight, but the rest of the vehicle is assumed to vary with
the ratio of the propulsion system weight saved to the gross weight. Thus,
223
-------
TABLE XIX
VEHICLE WEIGHT AMD PRICE ESTIMATES
Line
i
2
3
k
5
6
7
8
9
_10_,
__11__J
12
13
Ik
15
16
17
18
19
20
21
22
23
I — • • • — *
Item
Engine
Radiator
Starter
Generator
Engine with Accessories
Transmission
Rear Axle and Drive Line
Battery ,
-> I ,--1 T-rl- 1 T' fP X 2°°
(200-mi range fuel)
Exhaust
Propulsion Accessories
(Wp) Propulsion System.
Adjusted Gross Weight
Nonpropulsive Variable Items
Nonpropulsive Fixed Items
Vehicle -without Propulsion System
Curb Weight
Curb Weight vithout Engine
Price Less Engine
DMC Engine
Price of Engine
Automobile Retail Price
Change Relative to Base Line OC
Commutation
£ lines Q-OD
£ lines (6)- 63
©+©
~^ij+ 1690
0.5U8
© - © - 1690
S - 0.75 (?)- 1000
$0.90 x @
2.7 x jj$
L_§§,3-10
l"b 235
0
25
20
L 280
180
100
50
250
75
655
935
JAoo
1775
690
21*65
3210
2930
$26UO
220
$3235
$-155
^RCSS-8_
Ib U75
0
25
20
520
180
100
50
170
65
565 .
1085
U650
1875
690
2565
3525
3005
$2700
350
9^5
$361*5
$+21*5
RGSS-6
Ib 515
0
25
20
560
180
100
50
160
65
555 .j
1115
4700
1895
690
2585
3580
3020
$2715
390
$3770
$+380
Base-Line OC
600
60
20
20
700
160
170
^5
175
50
600
OC-76
800
70
20
20
9.10..
160
170
^5
215
50
61*0
1300 j 1550
5000
2010
690
2700
3870
3170
$2850
200
51*0
$3390
$ 0
5)420
i
2180
690
2870
1*260
31*50.
$3010
'41*
j.200
$1*210
$+820
-------
where Wg is the gross weight, 690 lb is the invariant portion of the body (trim
and glass),, and 1000 lb is the weight of passengers and luggage to be added to
the fully fueled curb weight in order to reach the gross weight, and where 2010 lb
is the variable portion of the 5000-lb gross weight of a conventional automobile.
This yields the following relationship for adjusted gross weight:
Wg = 1.672 Wp + 2826
which indicates that about an additional 2/3 lb is saved for each pound saved in
the propulsion system.
Table XIX shows the vehicle weight and price estimates using this system for
five selected vehicles. The first vehicle is based on the SSS-10 engine; the
second the RCSS-8; and the third the RGSS-6. The fourth column represents a
vehicle employing a base-line Otto-cycle engine and is designated "base-line OC."
This vehicle represents a conventional ^tOOO-lb six-passenger automobile with standard
V-8 engine, as described in Ref. 1, and is, nominally, a 1970 vehicle. The final
column represents the estimate for a vehicle employing a pollution-treated 1976
Otto-cycle engine*.
Line 1 contains the engine weight. The first-three columns are those derived
from the design portion of this study. The OC-76 weight is derived by increasing
the size of the base-line OC engine in order to partially compensate for per-
formance reductions caused by emissions treatment devices and by adding 100 lb
of various pollution control devices. The radiator for the OC-76 is increased
slightly to account for the greater inefficiency and larger size of the propul-
sion system. The transmission weights for the gas turbine engines are listed as
25 lb higher than the conventional three-speed automatic with torque converter.
This is a conservative estimate, since the weight could well be less depending
upon the transmission selected. The rear axle and drive-line weight for the gas
turbine engines is considerably less by virtue of the lighter front-wheel drive
selected . Fuel and ..tank weights are calculated on the basis of a 200-mi range
for the Federal Driving Cycle.
The data for weights in this study have been generated by UARL for a contract
for the Department of Transportation.
225
-------
L-97121+9-T
From the calculated propulsion system weights shown on line 12, the adjusted
gross weight is then derived using the formula shown above. From this weight,
the key figure which is obtained (i.e., the curb weight of the vehicle without its
engine) is shown on line 18. The price of the vehicle less its engine is shown
in line 19, using the relationship that the price equals the weight multiplied
by 900/lb. The direct manufacturing cost of the engine is shown in line 20 and
is converted into a price of the engine by multiplying it by 2.7, as per Ref. 1.
Finally, the automobile retail price is calculated by adding the price of the
engine to the price of the vehicle-less-engine. As can be seen, the SSS-10
has the lowest price of any of the vehicles, and the difference between it and that
for the future predicted Otto-cycle engine is almost $1000. Also, the SSS-10
vehicle costs $^00 less than the RCSS-8 vehicle and more than $500 less than the
RGSS-6 vehicle.-
While a preliminary analysis of this sort cannot be considered to be conclu-
sive, it strongly suggests that on the grounds of low weight and low manufacturing
cost the SSS-10.engine has a clear advantage not only over the other two candidate
gas turbine engines but with regard to future predicted Otto-cycle engines.
Production Cost Uncertainty
No gas turbine has as yet been successfully placed in large-scale mass pro-
duction. Therefore,, the manufacturing cost estimates presented in this report,
although the best that can be made at this time, are subject to several uncer-
tainties.
Manufacturing cost estimates for the SSS-10 engine will probably be quite
close to those predicted here, since parts for which a large degree of uncertainty
exists are not of particularly high value. The greatest production uncertainty
is associated with the RCSS-8 engine because enough is not known to determine,
with confidence, the specification of the material used, or manufacturing
techniques involved, in the recuperator.
In the case of the regenerated engine, the major uncertainties stem from
the fact that production costs of ceramic cores are apparently not achieving
target values, so that it may actually be cheaper to purchase stainless steel
regenerator cores than ceramic cores (Ref. 2*0. Considerable uncertainty also
exists regarding the eventual cost of satisfactory seals for this duty, so that
the manufacturing cost estimates for the RGSS-6 may be considerably higher than
those estimated here.
Considerable uncertainty also exists regarding the production costs of treated
Otto-cycle engines, primarily because of the difficulty in defining the required
system.
226
-------
L-9712U9-7
Fuel Economy
The national average fuel consumption for all passenger cars is less than 114-
mile s per gallon (Ref. 16); this figure includes all of the compacts and inter-
mediates as well as the heavier standard-sized cars studied herein. UARL calcula-
tion of 11.14 mpg for present standard cars operating on the OAP composite cycle
appears to be consistent with this figure. The base-line SSS-10-powered vehicle,
which is predicted to deliver 9-8 mpg for the OAP cycle, burns 15% more fuel
than this standard car (OC-TO), while the RGSS-6 at 11+.9 mpg burns 2k% less fuel.
With a 15% fuel cost differential (31^/gallon vs 35^/gallon), the present auto
and the future SSS-10-powered vehicle have almost equal fuel costs for the OAP
cycle, whereas the costs for the RGSS-6 would be about one-third lower. Thus, the
fuel economy of the SSS-10 is believed to be adequate.
This situation suggests that the regenerated-cycle gas turbine might be
offered as an extra-cost fuel-economy option, which might be attractive for light
commercial vehicles, economy-conscious consumers, etc. For passenger cars,
however, most of the buying public would probably not avail itself of this option*.
The OAP specification for fuel economy calls for 10 mpg over the Federal
Driving Cycle**. The heat-exchanger gas turbines considered herein are expected
to exceed this specification by delivering between 11 and 12 mpg.. However, the
base-line SSS-10, as configured herein, is predicted to deliver only 7.5 mpg ;
over the Federal Driving Cycle using.the GE hydromechanical transmission (,HMT)-.
* Several relevant precedents are the following: (l) The overdrive, which at
about $100 extra over the standard transmission, delivered a 10 to 20%
improvement in fuel economy, saved on wear and tear on the engine, and :
furnished an automatic passing gear, has disappeared from the market because
of disinterest by the public. (2) Economy engines, such as six cylinder vs
eight, four vs six, etc., low compression (regular fuel) vs high compression
(premium fuel), generally have had only a very'limited appeal. (3) Extra-
performance options (four carburetor barrels vs two) enjoy significant sales,
despite their poorer economy and higher first cost. (k) A 1971 study by
General Motors based on a nationwide sample of 15,630 new car buyers of
ten different makes of B-body sized cars (see Table XX) indicated that the "
largest single group of car buyers surveyed (Chevrolet owners) ranked gas
economy 21st out of the 26 reasons most often mentioned as factors important
in determining final selection; in fact, only 7.8% of the Chevrolet buyers '-
mentioned gas economy, while the average for buyers of all the ten makes sur-
veyed was 10.\% (ranging from a low of k.6% for the Pontiac to 23.6$ for the
Ambassador).
** It is questionable whether future Otto-cycle engines can achieve this figure;
a more likely achievement for these engines is only 8 to 9 mpg.
227
-------
L-97121*9-7
TABLE XX
REASONS FOR CHOOSING CHEVROLET
Previous Experience with Make
Exterior Styling
Future Resale Value
Dependability
Reputation of Car
Riding Comfort
Handling Ease
Final Cost - Deal Offered
Interior Styling
Wanted Larger Car
% Responding
61*. 1%
1*9.3
U6.9
37.9
36.1
35.8
35. U
33.0
28.0
25.9
Interior Roominess
Dealer Service
Dealer Location
Perf . Engine Response
Delivery Made When Wanted
Durability
Overall Operating Economy
Quality of Workmanship
Safety Features
Sales Effort by Salesman
GAS ECONOMY
All Others
Total
% Responding
25.2
2k. 9
zk.k
22.3
20.3
18.7
18.5
12.8
11.7
9.7
7.8
1*0.2
629.U
NOTE-. Figures total more than 100? because of multiple answers
SOURCE: Automotive Industries, April J, 1972
228
-------
L-9712^9-7
It should be pointed out that the SSS-10, as configured herein, is a first-
generation engine with considerable potential, not only for growth, but for modi-
fications -which could improve the performance of even the initial version. Accor-
dingly, further optimization studies were conducted to investigate the possibility
of meeting the 10-mpg goal with both demonstration and production SSS-10 powered
automobiles. Several vehicle and engine design options are available to achieve
this goal, and these options, together with reasonable assumptions of their
magnitude and effect, are detailed in Appendix V and discussed below.
Of the likely vehicle design options, the use of the Tracor transmission
(rather than the GE HMT, together with a reduced vehicle test weight achievable
with the lightweight SSS-10) and slightly lower aerodynamic and tire resistance
values, results in a predicted FDC fuel economy of S.U mpg and an overall fuel
economy of 11.6 mpg. When these vehicle design options are combined with potential
improvements resulting from engine reoptimizations aimed at improving low-speed
fuel economy, a value of 10.0 mpg for the FDC may be achievable for a 197^
demonstrator vehicle. Associated with these design options is an improvement in
fuel economy on the OAF composite driving cycle to 12.8 mpg, as compared with
the predicted value of 9-8 mpg for the base-line vehicle.
For production automobiles in the 1977 to post-1980 time periods, further
engine refinements can be foreseen which might raise fuel economy on the FDC
to 10.5 to 11 mpg and, on the GAP composite driving cycle, to values exceeding
Ik mpg. These values are in the range predicted for the 1974 base-line RGSS-6
and RCSS-8 powered autos* and are considerably better than predictions for future
Otto-cycle engines.
In summary, the heat exchanger gas turbines, offer superior fuel economy, but
it is believed that the SSS-10 demonstration engine offers adequate fuel
economy for most owners, especially when measured against the current performance
of comparable engines (and potentially superior fuel economy when compared with
likely future Otto-cycle engines). Sufficient opportunity exists to substantially
improve SSS-10 fuel economy for the demonstration if this is required, while in
the future, with anticipated improvements in gas turbine (and transmission) techno-
logy, even the simple-cycle gas turbine is likely to exceed current fuel economies,
and advanced heat-exchanger gas turbines may even exceed the fuel economy of
current automotive diesel engines.
Fuel Availability
Related to fuel economy (which is an owner problem) is the problem of conserva-
tion of petroleum resources (which is a national and, ultimately, global problem).
* Many of the same improvements could be considered for the heat-exchanger
engines also, thereby offering still better fuel economy for these types.
229
-------
In this connection, not only the mpg on an absolute basis but also the refinery
yields must be considered. Future gasolines which have been proposed for Otto-
cycle engines (Ref. 25) will not only be more costly to refine, but the process
will yield less gasoline per gallon of refinery stock than current gasolines. The
heat-exchanger engines are given the highest rating in this respect, the Otto-
cycle engine the lowest, while the SSS-10 would rate somewhere in between*.
Size and Weight
The size and weight of the engine is considered to be of prime importance to
the vehicle producer. It is obvious that a small, lightweight engine can be
utilized in a much larger variety of vehicles than one which must be tailored to
each installation. The increased size and weight of the treated Otto-cycle engine
is surely a great deterrent to the future flexibility of automobile design. The
same is true for the heat-exchanger gas turbines. Conversely, the small size
and light weight of the SSS-10 engine makes it adaptable to a wide variety of
vehicles from compacts, through intermediates, to standard automobiles . In
addition, the smaller size and weight of the gas turbines, and particularly the
simple-cycle SSS-10, allows a greater compatability with the devices which are
required by new Department of Transportation regulations concerning safety, all
of which will certainly add weight to the car.
Further advantages of reduced size and weight relate to the ability of
packaging the complete reduced-size propulsion unit in either the front or the
rear of the vehicle, as shown in Figs. Ul through k'j. Both a front installation
with front drive and a rear installation with rear drive permit improved traction,
eliminate a long driveline (thus improving interior space flexibility), and
permit additional overall savings in the weight and cost of the vehicle.
Reliability and Maintainability
All of the gas turbines studied should be more reliable and more maintainable
than the future piston engine, based on the- greater simplicity as well as on
past experience in numerous gas turbine installations in ground vehicles. This
* In the future, as petroleum resources may become scarcer and higher-priced, it
is quite likely that vehicle designs may turn to highly efficient forms of
motive power, such as fuel cells. This event, however, is considered to be at
least several vehicle generations in the future, and is not of particular concern
at this time.
230
-------
predicted reliability extends not only to a no-breakdown situation, but to a1 good
cold-start capability. Furthermore, the simplest form of gas turbine engine is
expected to have the highest reliability and maintainability of those studied.
Estimates for the repair and maintenance costs of future Otto-cycle engines
indicate increases of from 30 to 50% in order to keep the more complicated and
less reliable pollution control equipment operating. (As indicated previously,
the failure of these devices can result in pollution worse than that emitted by
the untreated Otto-cycle engine.)
On the basis of these considerations, the SSS-10 engine is judged to have
the best reliability and maintainability potential of all engines considered.
Technological Risk
For the gas turbine engines considered herein, the greatest technological
risk lies in the development of the RCSS-8 because of the largely unproven
design of its recuperator. The second largest technological risk is assigned
to the RGSS-6 because considerable question still exists regarding the lifetime
of acceptably priced seals and ceramic disks. On the other hand, the SSS-10
incorporates proven technology in a conservative design in which the major unknowns
relate primarily to eventual production cost rather than technological uncertainty.
Hence, the SSS-10 is considered to involve the least technological risk of all
engines considered. ,
To place these technological risks in perspective, it should be observed that,
although an Otto-cycle engine does not in itself represent any technological risk,
building one which can drastically reduce its naturally dirty emissions to .accep-
table limits involves a considerable technological risk, because there is at
present no known acceptable technological procedure to achieve the Federal
Standards by 1976 (Ref. 26). In this sense, a large technological risk must be
assigned to the Otto-cycle engine.
Development Time and Costs
The development time and costs are governed by the technological risk.
Therefore, because of its simplicity and its extrapolation from direct state-of-
the-art technology, the SSS-10 is expected to require the shortest development
time and probably the lowest development costs. The successful demonstration of
the heat -exchanger gas turbines could probably be accomplished at a cost not greater
than about 1-1/2 times that of the simple-cycle engine. However, although the
estimates on development time and costs for developing pollution-treated Otto-
cycle engines are uncertain at this time, they are likely to be considerably
higher than those which would be required to successfully demonstrate the SSS-10
engine.
231
-------
L-9T12U9-T
Noise
All engines should be equally capable of low-noise operation. The heat
exchanger acts as a noise suppressor and results in a quieter engine for the RCSS-
and RGSS-6. The SSS-10 is likely to require an exhaust noise silencer. Possible
solutions include fixed ceramic cores and fiberglass mufflers, and a cheap and
simple installation appears probable.
Exhaust Heat
The SSS-10 will run with hotter exhaust gases, and the localized heat flux at
the tailpipe is a problem. Possible solutions include finning of ducts, and
dilution with outside air. The exhaust from the regenerated engines, on the
other hand, is cooled, and is less likely to be a problem. There is also a
similar heat problem associated with the future Otto-cycle engine, since the
OC exhaust will be heated and kept warm by the combination of converters and
reactors acting in its exhaust system.
Critical Materials
With respect to the requirement for critical materials, the SSS-10 is
slightly ahead of the other gas turbine engines under consideration because the
critical materials used in the heat exchangers associated with regenerated or
recuperated engines are not required in the simple-cycle engine.
Drivability
With an appropriate transmission, the SSS-10 should be as responsive and
as capable of engine braking as the best of present engine-transmission combina-
tions. The responsiveness of future Otto-cycle engines may not be up to this
standard especially if they are not maintained perfectly, based on observations
of current "semi-cleaned-up" engines.
The heat retention characteristics of the heat exchangers used in the RGSS-6
and RCSS-8 will affect the performance felt by the driver somewhat (i.e., degrade
the responsiveness); should this prove to be a severe problem, it will require
control system sophistications which will add to the cost of these already expen-
sive engines.
232
-------
L-9T12l*9-7
Results of Engine Selection Processes
In view of all of these considerations, it is believed that the SSS-10 engine
is the most likely to meet the EPA goals because:
1. It has the potential for the lowest emissions of any of the engines
considered herein and has a high probability of achieving the
Federal Standards.
2. It leads to the lowest predicted automobile price of any of the
low-emission engines, particularly including the future Otto-cycle
engined vehicle.
3. Its predicted fuel economy is adequate based on conservative estimates,
and considerable potential exists for significant improvement over
present Otto-cycle engined vehicles.
U. Its simplicity, small size, proven technology, and potential low
manufacturing cost lead to high reliability, low technological risk,
low development costs, installation flexibility, good drivability,
and fewer demands on critical resources.
The heat-exchanger engines are also believed to be far better alternatives
for vehicular power plants than the expected future pollution-treated Otto-cycle
engine. They have the advantage over the SSS-10 of lower fuel consumption (with
the implied lesser demand on petroleum resources), at the expense of greater cost,
complexity, and somewhat lesser responsiveness (or greater control complexity)
in the single-shaft, fixed-turbine nozzle configuration considered here, and
hence may be desirable for commercial applications (taxis, buses, trucks) for
which fuel economy is crucial. Therefore, it is believed that development and
demonstration of all three of the gas turbine engines considered herein could
certainly be justified on the basis of the nation's future needs, since the de-
velopment costs are likely to be modest in terms of national needs and in terms
of expenditures being made currently to modify Otto-cycle engines.
Since it is believed that engines generally similar to the RGSS-6 are
being currently developed by at least one automobile company it would seem
prudent to additionally develop only the RCSS-8 as an alternative to the SSS-10
engine. Although its development involves a greater technological risk than
the SSS-10, a developed RCSS-8 would provide a backup engine which might prove
more cost effective if future fuel prices increased more rapidly than the
future costs of labor and materials.
233
-------
L-9712^9-7
234
-------
L-9712^9-7
DEVELOPMENT PROGRAM
Demonstration Program
A development program has been laid out aimed at a demonstration of the
performance and emissions characteristics of an SSS-10 powered automobile. The
goal of the program would "be to achieve the performance levels of the base-line
vehicle as reported herein, including emissions levels below those specified by
law for 1976 vehicles. A total of five pre-production prototype engines would be
built and installed in three demonstration vehicles, with the demonstration to
consist of approximately 100 hours of essentially trouble-free operation with each
automobile. It is estimated that thirty months would be required for this program
(see Fig. 110).
Functionally, the development program consists of three closely related
areas: the engine, the transmission, and the vehicle. From the standpoint of
time, the development program is divided into three phases. Phase 1 is 8 months
in length and consists primarily of the engine design effort and supporting
component tests; Phase 2, 16 months in length, consists primarily of test and
development of the engines, up to and including proof test; and Phase 3, 6 months
in length, comprises the installation of the engines and transmissions into
suitably designed demonstration vehicles, and culminates in the actual vehicle
demonstration for a three-month period.
In order to complete the demonstration by the end of 197^ > the program must
be started by July 1, 1972. Since the program is highly compressed, lower program
costs would probably result from a 3 to 6 month stretchout (from the nominal
30-month program, to 33 to 36 months). However, this stretchout may not be
permissible because of the urgency of the automotive pollution problem and the
possibility that interim emissions treatment to Otto-cycle engines may be either
ineffective or involve unacceptable performance or cost penalties.
Engine Development
The engine development encompasses a design and assembly program, an extensive
engine test program, a support test program, and a component test program. A
detailed design of 8 months comprises the major effort in Phase 1. Coincident with
the last five months of the detailed design pha^,e, long-lead-time hardware items
which can be clearly defined are procured. Thin procurement phase extends for
four months into Phase 2 for those items which have short lead times and those
items which are not completely defined until the end of the detailed design
exercise.
The engine test program involves tests in eleven areas; i.e., performance,
controls, stress and vibration, lubrication and gearing, bearings, emissions,
235
-------
FIG. HO
RECOMMENDED DEVELOPMENT AND DEMONSTRATION PROGRAM
YEAR
MONTH
PHASE
MILESTONE i
ENGINE DEVELOPMENT
UJ
Q
PRELIM.
DESIGN
DETAIL
DESIGN
PROCURE
HARDWARE
PERFORMANCE
CONTROLS
STRESS &
VIBRATION
LUB 8.
GEARING
BEARINGS
EMISSIONS
NOISE
TRANSIENTS
ENDURANCE
PROOF
ACCEPTANCE
CONTROLS RIC
VIBRATION
STRESS
STRUCTURAL
STAGES
COMBUSTOR
DUCTS
COMPRESSOR
TURBINE
BEARING
CONTAINMENT
CONTROLS
TRANS.
SELECTION
DESIGN &
DEVEL.
PROJECT
SUPPORT
DELIVER
DEMO.
HARDWARE
MISSION
ANALYSIS
DESIGN
PROTOTYPE
PROCURE
VEHICLE
TEST
PROTOTYPE
DESIGN DEMO.
NSTALL
PROPULSION
TEST
DEMONSTRA
TION
^ 1972 »-
JASON D
PHASE 1
k START
>
-* 1973 »-
J F
^
i
MAMJJASON D
PHASE 2 i
FIRST RUN
i A A
DESIGN DELIVER
COMPLETE TRANSMISSION
K- 1
f—
h->
1
*— i
I——/
PROTOTYPE ^..........M.H
L,—,.
~* 1974 -"-
J F M A M J
COMPLETE PROOF
VEHICLE A A
DESIGN TEST
COMPLETE
1
— 1
(-H
h— (
l—
— 1
h-H
DEMO. H— ^~-
'
JASON C
PHASE 3 -i*.
START
ALL DEMO.
A A 4
ENGINES FND ]
DELIVERED DEMO.
— M
236
-------
L-9712U9-7
noise, transients, endurance testing, proof testing, and acceptance testing.
The first engine run is predicted to occur on the first of August, 1973, thirteen
months following start. Two dynamometers and an outdoor test stand are required.
The first engine assembly is tested on the performance test stand using the
first dynamometer rig for two months, incurring some 1+0 hours of engine running.
The performance testing continues at two other separated intervals, all on the
same dynamometer, for a total of 160 hours of testing. Controls testing takes
place on the second dynamometer and occupies Uo test hours. Stress and vibration
testing is also accomplished on the second dynamometer and totals 20 hours. The
lubrication system and gearing tests are also run on the second dynamometer and
total 150 hours. Bearings and air'system tests are run on the first dynamometer
and total 50 hours. Emissions tests'are run on the first dynamometer and total
20 hours. Noise tests are accomplished on an outside noise rig constructed for : >
the purpose of the program and-total some 30 hours. Transient performance is run
on a test vehicle and requires 20 hours. Endurance testing is also accomplished
with a teso vehicle and, totals 100 hours. Proof testing is accomplished on either
available dynamometer and totals 70 hours. Finally, acceptance tests for the
engines for the demonstrator, vehicle are run on any available dynamometer and total
kO hours. The total number of engine hours projected during the development phase
of the program is 700.
Transmission Development
The transmission development program should be conducted in parallel with ••
Phases 1 and 2 of the engine development program. The transmission development
program must be preceded by a transmission selection program. Its outcome will be
to select optimum transmissions for use with the SSS-10 engine. The design and
development of at least two transmissions should be pursued. This duplication is
required since the engine and vehicle performance depends significantly on the
type of transmission used, and it is advisable not to make the success of the
demonstration dependent upon only one transmission. At least two transmissions
should be delivered for testing in the prototype vehicle. This delivery would
occur on the first of November, 1973. (if during the prototype testing part of
the engine test program one transmission proves to have outstanding characteristics,
it will be selected for delivery during the actual demonstration, and a second
transmission will not be required.)
Vehicle Development and Demonstration .
The vehicle development program is involved in the first two phases of the
development and demonstration program, and the vehicle demonstration culminates.
Phase 3 of the effort. ...
The first requirement is a mission analysis program in support of the engine
design effort; it will investigate various duty cycles for the vehicle, and will
237
-------
L-9712^9-7
serve to specify the engine and transmission characteristics in such a fashion as
to insure the optimum combination of design parameters.
The prototype vehicle will be designed and procured in the vehicle develop-
ment and demonstration program, and will be tested in the engine test program.
The vehicles will be "off-the-shelf" vehicles, so that the design and modification
discussed herein pertains only to the modification required to make it suitable
for accepting and demonstrating the SSS-10.
The delivered propulsion systems will be ii-stalled in these vehicles and two
months are provided for internal testing prior to delivery to the designated EPA
facility for actual demonstration purposes. The Demonstration is shown in the
program as consisting of three months, during which time 100 hours of demonstration
will be accomplished on each of the three vehicles provided. It is anticipated
that the demonstration will include emissions testing, performance, testing, fuel ,
mileage testing, maintenance testing, and various other demonstrations as prescribed
by EPA.
Future Production
The normal procedure for the introduction of new concepts for automobiles
is that, following successful completion of a demonstration of the, type described
above, field experience is gained. On the basis of this experience, a production
prototype is designed and tested prior to the initiation of limited production
(< 100,000 units annually). Thus, from hardware program start to full mass pro-
duction of over 80% of a given vehicle population, the introduction of a new
product might cover a minimum of 10 years. Within this period, 3 to h years
would be devoted to the design and_ demons t rat ion of the concept (similar
to the program described above), 2 to 3 years for design and development of the
production prototype, an additional 3 to U years for fleet testing, and production
planning, and 2 to k years (and sometimes up to 10 years depending on demand) for
production build-up. On the basis of this approach, the SSS-10 could not be
expected to reach large-scale production until after 1980. However, the normal
product phase-in considerations, which are largely governed by economics, are
superseded in this case by the urgent requirement for engines with low emissions.
A comprehensive study is required to establish what is possible in terms of
placing a simple engine such as the SSS-10 into production on a crash basis.
It is believed that an integrated program closely developed in concert with a
vehicle manufacturer could result in limited production (100,000 vehicles or less
per year) at a relatively early date, i.e., within 2 or 3 years of a successful
demonstration (1977-78), and possibly sooner (at still greater cost and risk).
The major requirements which are not met by the above recommended demonstration
program which would be required for this limited production include extensive fleet
238
-------
L-9712U9-7
testing of vehicles,, subsequent redesign of the engine for production, and the design,
manufacture, and: assembly of the required production line.
Since the general configuration and the general manufacturing requirements
of the designated engine are expected to be known fairly early in the program,
considerable progress can be made toward integrating the manufacturing facility
with the development program. Production machinery- can be generally specified
and, in many cases, procured long before the exact final form of the required
production tooling is known (i.e., the size and location of drills, the exact
program for numerically controlled machine, the exact dies, molds, etc., for
castings and forgings). • .
It is obvious that this concurrent development procedure would involve some
waste; nevertheless, the proposed limited production in 1977 is consistent with
the reported progress which has been made toward the production of Wankel engines.
The additional effort required would be well within the capabilities of the
present automobile companies, and a ready market could be expected to exist for
the SSS-10 vehicle even if its initial production costs were higher than those
estimated in this report.
On the other hand, it does not appear practical to envision the entire change-
over of an industry to this type of engine in any calendar year close to 1975.
Extensive owner and consumer testing such as might be,derived.from limited produc-
tion runs should certainly be required before a massive changeover of the present
industry takes place. If it is assumed that the engine does go into limited
production in 1977, several years of extensive consumer operation of these produc-
tion cars is desirable, and it is difficult to imagine that a complete industry
changeover could occur prior to 1980 without unprecedented effort. Nevertheless,
an extraordinary effort might be Justified on the basis of national impact, and
the general requirements for implementation, i.e., placing the engine into high-
volume production, should be the subject of additional study.
Costs
A costing effort was not conducted for the development program outlined
here. The costs would depend on the basic objectives of the program and other
factors, and would require far more extensive analysis than that permitted by
the scope of this contract. Nevertheless, it is possible to make a rough estimate
for comparison with costs which are currently being incurred in the search for a
low-emission vehicle.
It is quite probable that the entire recommended development and demonstra-
tion program discussed above and shown in Fig. 110 (excluding that required to
prepare for limited production) could be accomplished for less than $10 million.
This cost would include the procurement and development of five engine sets and
three complete demonstrator vehicles, including the running of the demonstration
239
-------
L-9712U9-7
for 300 vehicle-hours. It would also include transmission and control costs,
although any estimates for transmission development are necessarily speculative
at this time because of lack of definition of the optimum transmission(s).
2UO
-------
L-9712U9-7
REFERENCES
1. Wright, E. S., L. E. Greenwald, and W. R. Davison: Manufacturing Cost Study
of Selected Gas Turbine Automobile Engine Concepts. UARL Report K-911017-1+.
August 19T1.
2. Kenny, D. P.: A Novel Low-Cost Diffuser for High-Performance Centrifugal
Compressors. Journal of Engineering for Power. January 1969.
3. Groh, F. G., G. M. Wood, R. S. Kulp, and D. P. Kenny: Evaluation of a High
Hub/Tip Ratio Centrifugal Compressor. ASME Paper No. 69-WA/FE-28.
H. Kenny, D. P.: Supersonic Radial Diffusers. Lecture Series No. 39 on
Advanced Compressors. AGARD-LS-39-70.
5. Morris, R. E., and D. P. Kenny: High Pressure Ratio Centrifugal Compressors
for Small Gas Turbines. Prepared for the 31st Meeting of the Propulsion
and Energetics Panel of AGARD, "Helicopter Propulsion Systems," June 10-lU,
1968, Ottawa.
6. Morris, R. W.: High Pressure Ratio Radial Compressors and Turbines. Diesel
and Gas Turbine Progress, December 1970.
7. Okapuu, U., and G. S. Calvert: An Experimental Cooled Radial Turbine. High
Temperature Turbines, AGARD Conference Proceedings No. 73 (preprint),
September 21-25, 1970.
8. Okapuu, U., and G. S. Calvert: Cooled Radial Turbine for High Pover-to-
Weight Applications. AIM Paper Ho. 69-52^.
9- Langton, R. L. , and R. E. V. Westerhout: Aerodynamic Testing and Instrumentation
of Components for Small Gas Turbines. SAE 6709^1.
10. Slabiak, Walter: An A-C Individual Wheel Drive System for Land Vehicles.
SAE Transactions, Paper No. 66013^. 1967-
11. Sedlock, Edward: Advanced Electrical Machinery and Component Development.
P&WA Report 66-972.
-------
L-9712^9-7
REFERENCES (cont'd)
12. Kress, James, H.: Hydrostatic Power-Splitting Transmissions for Wheeled
Vehicles. SAE Transactions, 1968, Paper No. 6805^9.
13. McLean, A. F.: Case for the Single-Shaft Vehicular Gas Turbine Engine.
IME Symposium, Warwick, England April 9» 1969•
lU. Yeaple, F.: Metal-to-Metal Traction Drives. Product Engineering, Oct. 1971.
15. Cope, E. M., and C. L. Gauthier: Cost of Operating an Automobile. U.S.
Department of Transportation Report, February 1970.
16. Anon.: 1969 Automobile Facts and Figures. Automobile Manufacturers
Association, 1969.
17. Anon.: Vehicle Design Goals - Six Passenger Automobile. Office of Air
Programs Division of Advanced Automotive Power Systems Development,
Revision C, May 28, 1971.
18. Corenlius, W. and W. R. Wade: The Formation and Control of Hitric
Oxide in a Regenerative Gas Turbine Burner, SAE.
19. Bujac, J. N. Jr., and R. Mantler: Impact of National Environmental Policy
Act of 1969 on the Advanced Technology Turbine Engine. Proceedings of
the National Conference on Environmental Effects on Aircraft and Propulsion
Systems, May 10, 1971.
20. Katona, G.: 1969 Survey of Consumer Finances, U Michigan.
21. Anon.: NADA Official Used Car Guide. National Automobile. Dealers Used
Car Guide Company, Eastern Edition, November 1971.
22. Brehab, W. M.: Mechanisms of Pollutant Formation and Control from Automotive
Sources. Presented at Nineteenth Annual Lecture Series, Milwaukee Section,
Society of Automotive Engineers, March 5S 1971.
23. Hoffman, G. A.: Hybrid Power Systems for Vehicles. Symposium on Power
Systems for Electric Vehicles, U.S. Department of Health, Education, and
Welfare. National Center for Air Pollution Control, 1967.
21+2
-------
REFERENCES (cont'd)
2U. Best, G. C., and E. E. Planigan: Allison GT-^(A-The VIP Engine-Versatile
Industrial Power. ASME T2-6T-93, San Fransisco, March 26-30, 1972.
25. Oil & Gas Journal, Editorial Vol. 70, Wo. 3, January 17, 1972, p. V?.
26. Anon.: Semi-Annual Report by the Committee on Motor Vehicle Emissions of
the National Academy of Sciences to the Environmental Protection Agency
Washington, D. C., January 1, 1972.
27. Winfrey, R.: Economic Analysis for Highways. International Textbook
Company, 1969•
28. Anon.: Automobile Insurance and Tax Data Extracted from a lU-State
Study by National Automobile Association. United Services Automobile
Association, December 1971.
29. Anon.: 1968 Owner's Manual, Ford Motor Company, 1967.
30. Anon.: Flat Rate Manual 1971. The Irving-Cloud Publishing Company, 1971-
31. Anon.: Automotive Parts and Accessories Catalogue Wo. 295- J. C. Whitney
and Company, 1971.
32. Claffey, Paul J.: Running Costs of Motor Vehicles as Affected by Road
Design and Traffic. Paul J. Claffey and Associates, National Cooperative
Highway Research Program. Beport No. Ill, Highway Research Board, 1971.
33. Guedet, R. H. and J. E. Louis: Dual Mode Hydromechanical Transmission
as Applied to Gas Turbines. ASME 69-6T-13 Cleveland, March 9-13, 1969.
-------
L-9712U9-T
-------
L-9712H9-7
APPENDIX I
VEHICLE DYNAMICS WITH THE GE HYDROMECHAHICAL TRANSMISSION
Two computer programs vere written to estimate vehicle performance using the
GE hydromechanical transmission (HMT). One program performs full-power acceleration
calculations based on the OAP vehicle performance specifications given in
Kef. 17> and the other program computes fuel economy and emissions for the
Federal Driving Cycle (FDC) as well as for simplified country and suburban
driving.
The basic assumptions used in the development of these programs are
as follows :
• Level road
» No wind
• Air density is related to ambient temperature by the perfect gas law
• Steady-state engine torque curves apply during engine and vehicle
acceleration
• OAP resistance equations apply
The physical constants used in this study are given in Table IX.
The apparent increase in rotational inertia of a rotating mass due to
speed reduction is fundamental to the engine dynamics given in Appendices I and
II, and a brief digression to discuss its derivation is appropriate here.
Consider the rotating mass, connected to a torque source by means of an ideal
reduction gear system (no mass, no losses-), shown in the sketch below.
-------
L-9712U9-7
Let I = rotational inertia
T = applied torque
R = 0)2/02
uij = speed of shaft through which torque is applied
w = speed of rotating mass
in some consistent set of units.
From dynamics, the applied power is equal to the time rate of change of
kinetic energy, or
Tu, = TT
= I
Now
d)2 = BU! , (1-2)
and putting Eq. (1-2) into Eq. (l-l) and rearranging, one has,
&1 = T/R2I , (1-3)
which is the rotational form of Newton's Second Law, referred to the torqued
shaft. The inertia of the rotating mass as seen from the torqued shaft appears
to be increased by the factor H .
Viewed another way, a change in speed of the torqued shaft from ui . to
M! , where the subscripts i and f denote initial and final stages, respectively
f
-------
L-9712^9-7
corresponds to a change in kinetic energy of the rotating mass of % l(iD2 ~ °°2 )
or % R l(ui - W2 2). The factor R appears again when the energy chanfe is *
expressed in terms of a speed different from the speed of the rotating mass.
As described in the main text, the GE-HMT responds to an increase in throttle
setting by "unloading" the engine to allow its speed to come up to the controlled
speed as determined by the throttle setting and the engine operating line. The
rate of engine acceleration following an increase in throttle setting is
determined by the transmission response time. Because the response time of the
HMT is relatively short (fractions of a second), it would be possible, say, to
be driving at some steady speed, depress the accelerator pedal, and have the
vehicle slow down as the transmission "downshifts" to allow the kinetic energy
of the vehicle to help bring the engine to its controlled speed. After the
engine attained controlled speed, the vehicle would then accelerate at constant
engine speed. The deceleration, of course, is undesirable, and in practice the
transmission would be controlled in such a way as to permit a smooth flow of
power to the wheels while at the same time bringing the engine to its controlled
speed. The effect of this HMT response to a change in throttle setting has been
included in the acceleration dynamics by means of a hypothetical "torque-split"
which is described further on.
For this analysis, the engine is assumed to be a. torque source, with no
rotational inertia, connected in series with a rotational inertia equal in
resistance to the sum of the engine and the transmission inertias. The GE-
supplied curve of transmission inertia versus ratio (output speed * input speed),
shown in Fig. 73, was used for this analysis. It should be noted that, in order
for the engine and transmission inertias to be added, the inertia values must
be referenced to the same rotational speed. Thus, if transmission input speed
is used as a reference, the engine inertia referred to the gas generator shaft
must be multiplied by the square of the total engine reduction gear ratio
(gas generator speed f output shaft speed) before being added to the transmission
inertia. '
The sketch below will be used to derive the equations of motion for an'
HMT vehicle.
-------
L-9712H9-T
Equations of Motion - Full-Power Acceleration
•••I"!
X
T2U2
(no rotational inertia) engine and
transmission
inertia
A free-body diagram of the lumped engine and transmission inertias yields
the equation of motion for the engine:
3£
!7
(I-U)
2
where I = combined engine and transmission inertia (slug-ft ), referred to
the engine output speed
T-j_ = engine output torque, minus accessory load (generator, etc.) (H>-ft)
Tg = resisting torque of transmission as seen by the engine (lb-ft)
i. = engine acceleration (rpm/sec)
The engine output shaft speed (o^) is equal to the transmission input shaft
speed, by definition. The transmission input torque (IQ) is related to
-------
L-9712H9-7
•where <>> = engine idle speed (rpm)
u = maximum engine speed (rpm)
max
n = dimensionless torque-split exponent
The purpose of the function described by Eq. (1-5) is to simulate the effect
of the transmission response to full-throttle acceleration. For a fixed value
of n, the resisting torque of the transmission (Tg) will be an increasing
function of engine speed such that, initially, when the engine is at idle,
all of the engine torque (l, ) is available to accelerate the engine. As engine
speed increases, the fraction of the total engine torque available to accelerate
the engine decreases (although the difference T^-Tp, which causes the acceleration
of the engine, may increase) until the engine reaches maximum speed and all of
its torque goes into accelerating the vehicle. Increasing the value of TI
will increase the initial engine acceleration at the expense of vehicle acceleration.
Thus, a high value of n will allow the engine to reach its maximum speed
sooner than a relatively lower value. Viewed another way, the engine will be
at its maximum speed at a lower vehicle speed for the greater value of n•
It is emphasized that the torque split of Eq. (l-5) is merely a device to
simulate the effect of transmission response. However, the technique was
believed to be satisfactory, by GE personnel, for estimating vehicle performance
(Ref. 3).
At any instant the transmission output shaft speed may be computed from
Eq. (1-6) assuming no wheel slippage:
30 Ra ,, ,--.
_ v (1-6)
3 IT r '
OJ
where r = driving wheel radius (ft)
RQ = rear-axle ratio
cL
v = vehicle speed (fps)
w. = transmission output shaft speed (rpm)
-------
L-9712U9-7
Application of the definition of transmission efficiency, Eq. (1-7) , yields an
expression for the transmission output torque, (Eq. (1-8).
•power out _ T3"3 (1-7)
"t power in T2W£
Thus
2> , (1-8)
where T = transmission output torque (lb-ft)
n, = transmission efficiency.
"C
The transmission efficiency is found as a function of input power and output
speed from Fig. 23. The torque applied to the wheels can be found by an application
of the definition of rear axle efficiency:
] (1-9)
v
where T, = rear-axle torque (ib-ft)
n = rear-axle efficiency
a,
Uij = rear-axle speed (rpm.)
The aerodynamic and rolling resistance forces are given by Eqs . (1-10)
and (I-ll), respectively, and the vehicle acceleration is given by Eq. (1-12).
Faero = *
Froll
= W(
T
a = ;r - Faero - Froll
250
-------
L-9T12U9-7
where a
A
~
aero
roll
= vehicle acceleration (ft/sec )
= vehicle frontal area (ft )
= aerodynamic drag coefficient
= aero dynamic resistance force (ib)
= ro-LlinS resistance force (ib)
= acceleration of gravity (ft/sec )
r^,r r = rolling resistance coefficients (units of Ib and fps)
rw = radius of driving wheel (ft)
W = vehicle weight (ib)
W = apparent vehicle inertial weight (ib)
p = ambient air density (slug/ft )
The -apparent inertial weight W is equal to the gravitational weight plus the
equivalent weight of rotating masses such as the wheels, axles, etc.
Equations (I-U) and (1-12) may be integrated to derive the full-power
acceleration history of the vehicle.
Equations of Motion - Driving Cycle Computations
The problem of computing vehicle performance for a driving cycle is somewhat
different than for a full -power acceleration. For a given driving cycle, the
transmission output, rather than input, conditions are known at discrete
and specified intervals of time, and there are in general no differential
equations of motion to integrate. Rather, quantities such as fuel consumed
and emissions produced are calculated for each point, summed over all points
of the driving cycle, and then averaged as needed, in the computation of
fuel economy in miles per gallon.
251
-------
L-9712^9-7
Let a and v be the acceleration and speed, respectively, of a vehicle
at some point in a driving cycle. Then a rearrangement of Eq. (1-12) allows
solution for the required torque (T^) at the driving wheels.
The transmission output speed (.w^) can be obtained from Eq. (.1-6) and the
transmission output torque (T^) can be solved for from Eq, (1-9) :
Now, knowing the transmission output torque and speed, it is necessary to find
the input torque and speed in order to determine the engine operating point
(torque and speed).
Since the transmission efficiency map (Fig. 23) is expressed in terms of
input power and output speed, it is necessary to implement a trial-and-error
procedure to find the transmission input conditions. A separate subroutine was
devised to perform this function for the HMT. In essence, the computational
procedure of the subroutine finds, by a trial-and-error algorithm, a transmission
input power which satisfies both the efficiency map of Fig. 23 and the definition
of overall transmission efficiency, Eq. (1-7). Now the engine operating speed
must be found, since at this point only the transmission input power, but not
the particular combination of torque and speed (T2 and u)2, respectively),
is known. This is accomplished by referring to the operating line of the engine,
expressed as a function of power and speed. Then, with the transmission input
power and speed known, the input torque is obtained from the definition of
power .
Finally, the engine output torque is calculated as the sum of the transmission
input torque, the accessory-load torque, and the torque required to accelerate
the engine to its next operating point in the driving Cycle. The engine
acceleration torque can be estimated by performing the above computations for
the next interval or point on the driving cycle to determine the engine operating
speed. Then the average engine acceleration torque during the present interval-
is assumed to be the torque required to accelerate the engine from its present
speed to the speed at the next point at a uniform rate. This is expressed
252
-------
mathematically by Eq., (1-15)
T
accel
where T , = average engine acceleration torque (ib-ft)
accel
At = time interval between the present and next driving cycle intervals (sec)
to . = engine speed for the next point on the driving cycle (rpm)
nexo
unov = Presetvt engine speed (rpm)
The term of Eq. (1-15) in parentheses is the assumed average engine acceleration
(rpm/sec) during the present interval.
253
-------
L-9712U9-T
25**
-------
L-9712U9-7
APPENDIX II
VEHICLE DYNAMICS WITH THE BORG-WABNER
AUTOMATIC TRANSMISSION
The basic assumptions and physical constants used in the Borg-Warner
transmission vehicle derivations are given in Appendix I and Table IX.
Additional assumptions applying to the multispeed transmission analysis are
as follows:
• Maximum acceleration only
• Engine and clutch inertias are lumped together
• Gearbox efficiency is constant
« Inertias and losses in the gearbox (including internal clutches and
bands) are neglected as such, but are assumed to be included with the
lumped, constant gearbox efficiency.
• Upshifts occur only at maximum engine speed
* Clutch has inertia valve mechanism to limit engine deceleration
on upshifts
» Coefficient of friction between clutch surfaces is uniform and
constant
• Clutch output shaft speed drops instantaneously at each upshift.
For the purposes of this analysis, the vehicle acceleration dynamics have
been divided into three sections which correspond to three different regimes
of clutch operation (for maximum acceleration). The clutch pressure plate (or
engine) speed is referred to as the clutch input speed and the clutch disc
(or gearbox input) speed is referred to as the clutch output speed. The clutch
performance map of Fig. Ill will be useful for understanding the ensuing
discussion. The three modes of clutch operation are as follows:
Regime 1 - Engine acceleration. Clutch input speed is greater than clutch
output speed and the engine is accelerating under the torque difference between
the maximum available engine torque and the torque transmitted by the clutch
for a given engine speed (see Fig. 71*) • The clutch input and output speeds
255
-------
L-971249-7
FIG. Ill
CLUTCH PERFORMANCE MAP
Ul
a.
0.
I-
U
NOT IN VEHICLE ACCELERATION REGIMES
REGIME H
'"IDLE
CLUTCH INPUT SPEED -
"LOCK
256
-------
L-9712U9-7
(UK and IDO, respectively) are defined for Regime 1 by the relations below.
The subscripts "idle" and "lock" refer to the engine idle and clutch lockup
speeds, respectively.
"idle * <°i * "lock (II-1)
0 £ UQ < w. (H-2)
Regime 2 - Engine acceleration. The clutch is locked and clutch input and
output speeds are the same. The clutch speeds for this regime are defined by
Eq. (II-3). The subscript "max" refers to maximum engine speed.
(oi = w0 £ umax (II-3)
Regime 3 - Engine, deceleration. This regime occurs immediately after an upshift
when the clutch output, shaft speed drops following the gear change. 'Clutch
input speed is greater than clutch, output speed and the engine is decelerating
at a constant rate under the bypass action of the inertia valve. The clutch
speeds for Regime 3 are defined by the equations below -
UT ,<'(!>• g a) fll-lt)
, lock i max VJ.J.-H;
The equations of motion are different for each clutch performance regime,
and are derived separately. Before proceeding with the derivations It should
be noted that because of the action-reaction principle, the clutch output1 torque
is equal to the input torque, regardless of the input and output speeds. , -Because
of this, and the definition of clutch efficiency, the clutch efficiency is
numerically equal to the, clutch slip ratio (output speed * input speed) .
The engine is assumed to' be a torque source with no rotational inertia, connected
in series with a:rotational inertia equal in resistance to the sum of the engine
and clutch inertias. As in the case of the GE-HMT, the engine inertia is
referenced to the engine output (clutch input) speed.
257
-------
L-9T12H9-7
The sketch below will be used to derive the equations of motion. For
generality, the number of speeds in the transmission was considered variable
rather than fixed (8-speed).
Engine
J\_
In
Transmission
Clutch
Engine and Clutch
Inertia
Differential
2'2
Eegime 1
A free-body diagram of the combined engine and clutch inertias yields the
following equation for the engine acceleration
a
e ir
(II-6)
where I = combined engine and clutch inertias (slug-ft )
Te = engine output torque (ib-ft)
T^ = torque transmitted by clutch (ib-ft) - see Fig. 7^
(o = engine acceleration (rpm/sec)
Because of the action-reaction principle,TQ (clutch output torque) is equal to T-
258
-------
L-9T121+9-7
(clutch input torque), and a).j_ (clutch input speed) is equal to toe (engine speed)
by definition.
The transmission output speed and torque are given by Eqs. (II-7) and (II-8),
respectively. Equation (I1-8) was obtained from the definition of transmission
efficiency.
Tl = VoRn ' to
where Rn = UO/(JD^ = transmission gear reduction ratio for the nth gear
T, = transmission output torque (ib-ft)
n^ = transmission gearbox efficiency
dij = transmission output speed (rpm)
In a similar fashion, the rear- axle speed and torque are given by Eqs. (II-9)
and (11-10), respectively.
-
where R = i&i/ia^ = rear-axle gear reduction ratio
Tp = rear-axle torque (ib-ft)
n = rear-axle effiency
a
u2 = rear-axle speed (rpm)
For compactness, let P (ib) be the sum of the aerodynamic and rolling resistance
which are given in Eqs?S(l-10) and (l-ll) in Appendix I. Then the vehicle acceleration
is given by Eq. ( 11-11 ).
V T
\r-y,
259
-------
L-9T12H9-T
where a = vehicle acceleration (ft/sec2)
g = acceleration of gravity (ft/sec2)
V = equivalent inertial weight of vehicle (lb)
Equations (II-6) and (ll-ll) may be integrated to yield the full-power acceleration
history of the vehicle during Regime 1 clutch operation. This mode of acceleration
ends when the clutch output speed becomes equal to the clutch input (engine)
speed. The clutch output speed (.u ) can be determined as a function of the vehicle
speed by applying the kinematical expression for the rotational speed of a wheel
which is rolling without slipping, and then combining with Eqs. (II-T) and (II-9)•
The result of these manipulations is given by Eq. (11-12).
Regime 2
The equations of motion derived for Regime 1 also apply to Regime 2, but
since the clutch is locked (,ii)0= w^), the engine and vehicle equations of motion
can be combined. One may begin by taking Eq. (ll-ll) for the vehicle acceleration
and using Eqs. (II-8) and (II-10) to express T~ in terms of the clutch torque,
!.;_ (which is equal to T). The result is Eq. f 11-13) .
In Eq. (II-6) the engine acceleration may be expressed in terms of the vehicle
acceleration by noting that to equals ui (because the clutch is locked) ,
differentiating Eq. (11-12), and substituting the result into Eq. (II-6).
The result is Eq. (ll-lh) .
260
-------
L-9T12H9-7
Finally, eliminating T.j_ between Eqs. (11-13) and (11-lU), and rearranging, yields
Eq. (11-15) which expresses vehicle acceleration directly in terms of engine
torque.
n n.R R T
a t a n e ^
*res
a =
Equation (11-15) is identical to Eq.. (Il-ll) except for extra terms in the
denominator which may be considered as augmenting the inertial mass of the vehicle.
The additional term is due to the effect of the kinetic energy that must be
supplied to the engine inertia as the vehicle accelerates. This term can add
50? or more to the effective mass of a gas turbine vehicle in first gear because
of the gear reductions involved.
Equation (11-15) may be integrated to yield the acceleration history of
the vehicle while the clutch is locked. The engine speed is related to the
vehicle speed in this regime by Eq. (11-12). Equation (11-15) is valid until
the engine attains maximum speed at which point an upshift is assumed to occur.
Begime 3
The equations of motion for the vehicle during Regime 3 acceleration are
the same as those for Regime 1 except that the engine is now decelerating (after
the upshift) at a constant rate due to the effect of the inertia valve.
Assuming a constant rate of engine deceleration, Eq. (II-6) may be rewritten
to solve for the clutch input torque.
Ti = Te + L_ I a , (11-16)
where a = engine deceleration (rpm/sec).
The effect of the engine deceleration is to cause an apparent increase in the
amount of torque available from the engine. The additional torque, of course,
arises from the kinetic energy surrendered by the rotating engine as it decelerates.
Equations (II-7) through (Il-ll) apply to Regime 3, and Eq. (Il-ll) may be
integrated to yield the acceleration history during Regime 3 clutch operation.
261
-------
L-9712^9-7
Sequence of Clutch Operation
For the dynamic model of the controlled slipping clutch used in this report,
the following describes the sequences of clutch operation which occur during
a full-power acceleration. It is assumed the vehicle is initially at rest
with the transmission in first gear, and the engine at idle. Time begins when-
the acceleration pedal is (instantaneously-) floored. The clutch performance
map of Fig. Ill will be used to illustrate the discussion.
When the acceleration is begun, the state of the clutch is represented
by point A on Fig. 111. Here, the engine is at idle and the output shaft of
the clutch is stalled because the vehicle has not yet begun to move. The subsequent
motion of the vehicle can now describe various types of trajectories on the
clutch performance map depending on the choice of vehicle parameters, particularly
the effective engine inertia. For a vehicle with a relatively small engine
inertia, the engine will quickly accelerate to clutch lockup speed (lo^oc^)
and then remain constant while the vehicle speed increases to bring the clutch
output speed to W]_ock- This mode of acceleration is described by trajectory
ABHD on Fig. 111. During the portion of the curve BHD, the engine speed is
constant and the vehicle accelerates under constant torque. When the clutch
output shaft reaches to-, ^ (point D) , the clutch locks up (because clutch
torque becomes equal to engine torque at this point) and the remainder of the
first gear acceleration is in Regime 2 along line DGE.
For large engine inertias, the vehicle (and hence clutch output shaft)
may accelerate more rapidly in comparison with the engine, and the output
shaft speed of the clutch may increase to match the input speed before i
(curve AC), and the clutch will be locked for the duration of first gear
(line CDGE).
At point E an upshift occurs and the output shaft speed drops to point
F. This commences Regime 3 operation and the engine begins to decelerate
at a constant rate while the vehicle continues to accelerate. If the engine
deceleration is rapid enough, the engine will reach
-------
L-9712H9-7
If the engine deceleration rate is low, the output shaft speed of the
engine will increase to meet the engine speed, and lockup will occur (path FGE).
At point E,the next upshift occurs and the process repeats. Successive upshifts
do not in general follow the same trajectories because vehicle acceleration
varies with speed.
It should be noted that the preceding description of the clutch and gearbox
operation was derived from the dynamics of the idealized mathematical models
used in this study. The dynamics of an actual vehicle may be somewhat different
due to the various internal transmission losses and inertias which were neglected
in this analysis. These factors would tend to soften the sharply defined rates
of change indicated by the model; for example, instantaneous clutch lockup.
Despite these qualifications, the net acceleration predictions are believed to
be accurate.
263
-------
L-9712U9-7
26k
-------
APPENDIX III
COMPUTER PROGRAM DESCRIPTION
This appendix describes the three vehicle performance computer programs
which were written for this study. Program users' instructions, including
listings of all programs and subroutines , are included in a separate volume
(L-9712^9-8). The vehicle dynamics on which the three programs are based are
described in Appendices I and II.
Vehicle Acceleration Program - Hydromechanical Transmission
This program was written to predict the acceleration performance and top
speed of a vehicle equipped with the GE-HMT transmission. The program output
consists of a time history of the vehicle's acceleration, including parameters
such as speed, elapsed distance, engine speed and power, and transmission
efficiency. For each case (a particular vehicle), the program calculates the
acceleration history of the vehicle from a standstill to top speed. While this
is occurring, the program monitors the vehicle's performance and performs
interpolations to derive the acceleration data for the OAP performance goals,
such as 0-60 mph acceleration time, etc. At the end of each case, the program
prints a summary of the OAP acceleration performance. A print option permits
suppression of the acceleration history printout so that only the heading,
which consists of a summary of the engine and vehicle data, and the OAP
acceleration results are printed. A sample printout, with the acceleration
history suppressed, is shown in Table XXIa.
The program utilizes a variable step size (in time) for integrating the
vehicle's motion. The calculations are set up so that below 15 mph the step
size is 0.1 sec, because of high engine acceleration rates. Above 15 mph,
the step size is calculated such that the vehicle speed change between intervals
is 1.0 mob.. In addition, the intervals are calculated so that the vehicle
speeds appear as integer values on the printout. Above 80 mph, a step size
of 2.0 sec is used. The program is fairly rapid, with an average execution time
(on a UWIVAC 1108) of about h sec for each case.
The OAP performance goals state that acceleration times are measured from
the instant when the accelerator is depressed. Thus the acceleration history,
with the vehicle initially at a standstill, is initiated with the engine at idle,
and the program computes the subsequent build-up of engine and vehicle speeds.
265
-------
TABLEXXIa
SAMPLE COMPUTER PRINTOUT
* * * UARL VFHICLE PERFORM." NCK PRO = 2.30
OUTPUT SHAFT INERTTA (SLUG-FT**?) = .7f,
EfJGIfJt NUMBER = 3
PtRFOKMANCF JLGRADATION FACTOR = .8^
INITIAL DJGINE SPEED FRACTION CAT 25 MPH) = .62'*
tllGII-jE KUMUP TIMEr SEC IAT 2b MPH) = l.llb
INITIAL ENGINE SPEED FRACTION tAT 50 MPH) ^ .73f.
LNGINE RUNUP TIME' SEC CAT 5C MPH) = .SUO
VEHICLE AND PERFORMANCE
VEHICLE WEIGHT =
ACCESSORY HOUFR (HP) =
CDA PRODUCT IFT**2) =
ROLLING RESISTANCE COEFFICIENTS
FO (LB/LB) =
Fl (LB/LB-WPH)*E-05 =
F2 (LB/LB-VPH**2)*F-07 =
REAR AXLE RATIO =
WEIGHT FRACTION On DPIVING WHFEL^
TRANSMISSION TORQUE SPLIT EXPONENT
AMBIENT TEMPERATURE (F) =
ROAD-TIRE ADHESION COEFFICIENT =
»*.«J
5?.
.U15
?.nn/l
.50
1.00
1U5.
1.0
* * SUMMARY OF ACCELERATION PERFORMANCE * *
(EPA VEHICLE DESIGN GOALS - SIX PASSENGER AUTOMOBILF)
* fl-60 MPH ACCELERATION TIVF (SEC) = 16.40
* DISTANCE COVERED IN TO SFC FROM STANDSTILL ISEO = 332.4
* 23-70 WH ACCELERATION TIFF, (SEC) = 1P.48
* HOT HIGH-SPEED PASS MANEUVER TIME CSFC) = l<4.i+5
* DOT HIGH-SPEED PASS MANEUVER 01STANCF (FT) = 1332.4
* TOP PPEEO IMPH) = 93.«
-------
L-9712lt9-7
However, mathematical difficulties with the transmission torque-split model
preclude starting acceleration runs from nonzero vehicle speeds. For these
cases, which begin at 25 and 50 mph for the GAP maneuvers, the engine is assumed
to be initially at rated (maximum) speed. The resultant performance is then
corrected by adding on a calculated engine run-up time from the 25- and 50-mph
steady-state engine speeds.
Vehicle Acceleration Program - Borg-Warner Mechanical Transmission
This program predicts the acceleration performance of a vehicle equipped
with the Borg-Warner mechanical transmission and a controlled slipping clutch.
The program output, like the HMT acceleration program, consists of a time
history of the vehicle's acceleration, including vehicle speed, elapsed
distance, engine speed and power, power lost in the clutch, and the transmission
gear number. The acceleration is initiated from a standstill with the engine
at idle, and interpolations are performed during the acceleration to compute
the GAP performance data. The program terminates when the vehicle has reached
top speed or when the DOT high-speed pass maneuver has been completed, whichever
comes first. At the end of each case, the program prints a summary of the GAP
acceleration performance data. A print option permits suppression of the acceleration
history printout so that only the heading and summary are printed, as in the
HMT program.
A variable computation step size is used which depends on the vehicle
acceleration and the clutch performance regime. The step sizes are "based on
a. vehicle speed change of 0.5 mph or a change of engine speed of 5$ of the
difference between the idle and lock, or lock and maximum speeds (depending
on the clutch regime), whichever is smaller. The total computation time for
each case is short, averaging about 3 sec on a UHIVAC 1108. A sample printout
is shown in Table XXIb.
Because of the close ratio spacing of the transmission (small rpni changes
at each shift), Borg-Warner feels that the transmission shifts should be controlled
only by engine speed, and not influenced by throttle position as in conventional
automatic transmissions. This means that, for a given vehicle speed, the engine
speed is the same regardless of throttle position. Thus, the 25-70 mph acceleration
time and DOT high-speed pass maneuver interpolations are not corrected for engine
run-up time as in the case of the HMT.
267
-------
* UAKL VLnlCLL
TABLEXXIb
SAMPLE COMPUTER PRINTOUT
CKHAllCu PK.V-Krt. - "AXI^U'i ACCF.LF RAT !0f. CAPAflLITY * * *
ro
*•
to
IiiUCiLt.T&il»\EI. siAS
i.NGI.-Jt: , N-bPtlEU SLIPPING CLUTCH Tf>ANS''ISSrON)
UJGII.E uATA
-lilEJ._= __________ .....
MAXIMUM t-.G. iPEEJ (RP.'i) =
MAXIi-iUM OUTPUT SHAFT 'j
Li^GIuL ijEAK RLuUcriOI-l =
IULE Sh'LtU 1-KHCnON =
I \J\-\L >*>LiLu IKHM) ^
OUTPUT bH«F i i,4E_HTIA
EhGIUt I^UMbt-K =
PtRFUKMANCE. UtGKADATIOh FACTuK
CLUTcrt LOCKUP SHtED FKACTiOiJ =
RATIU( lj= 4.50
13.Q.G
7000.
11. 7/1
.11
1
.Bh
l( P)- ____
HATIuC 3)=
RATIO { l*) =
HATIOl b)z
HATIu( b)=
7)=
2,56
2.06
1.69
1.40
1.18
i Tno
VEHICLE WO "EKFOR^AMCE DATA
VEHICLE WEIGHT (L") = 3970.
ACCESSORY POWFR (HP) = l+.r)
CLA PRODUCT (FT**2) = 12'
.
-------
L-9T12U9-7
Fuel Economy and Emissions Program - Hydromechanical Transmission
The purpose of this program is to estimate the fuel economy and emissions
of a vehicle which uses the GE HMT. The program output consists of a summary
of the fuel economy (mpg) and emissions produced over the Federal Driving
Cycle (FDC), and for steady-state speeds of 20, 30, ho, 50, 60, and 70 mph.
The fuel economy in Btu/mi and the average power for each of the above cases
are also given. Finally, the average mpg for the simplified suburban and
country routes, and for the composite of the latter two routes and the FDC,
are given. A sample printout is shown in Table XXII. A print option permits
a listing of various vehicle parameters such as engine power, engine speed, and
transmission efficiency for each of the 1370 intervals of the FDC. The program
requires about 25 sec of Univac 1108 computing time for each case.
At each point (or interval) in the cycle, the fuel consumed and the emissions
produced are calculated in the following manner. Knowing the vehicle speed and
acceleration, the road-load torque is computed. Then, as described in the
section on vehicle dynamics, the transmission input torque and speed are computed
and added to the engine accessory load torque plus an additional torque which is
calculated to account for engine acceleration effects. Then, knowing the engine
output speed and torque (engine operating point), the fuel flow is computed by
numerical interpolation from the engine fuel flow map which is stored in
tabular form.
The pollutant emissions produced during the interval are obtained by
numerical interpolation for the emissions indices (ib of pollutant produced
per 1000 Ib of fuel consumed) for each species and then multiplying the index
by the fuel consumed. Finally, the fuel consumed and pollutants produced are
summed over all the intervals of the cycle and then averaged as required.
-------
TABLE
SAMPLE COMPUTER PRINTOUT
« * * MARL VEHICLE" PERFOHt ANCE t'POOKn'1 - FU£L FCM'10i-iY * » *
(SINGLE-SHAFT GAS TURBINE FKGlNEt GE HYOPOMLCHAi.'ICAL TKAMSMSriON)
ENGINE DATA
MAXIMUM POWER (HP) =
MAXIMUM G.G. SPEED IRPMJ =
MAXIMUM OUTPUT SHAFT SPEED (KPM> =
ENGINL GEAR REDUCTION =
luLE SPEED FRACTION =
IDLE SPEED (RPM) -
GAS GtNERATOR INERTIA (LBN'-IN»*2) =
OUTPUT SHAFT INERTIA (SLUG-FT»*2) :
TRANSMISSION NO-LOAD POWER (HP)
ENGINfc. NUMBEK =
OPTIMUM FUEL FLO* LINL NUMBER =
EMISSIONS IUUEX FLOW NUMBER =
130.0
136000.
2700.
39.3/1
.500
1350.
2.30
.76
.00
3
2
3
VEHICLE rtNu PERFORMANCE DATA
VFHlCLt HiEIGilT (L.B) = ^850.
ACCESSORY POlvER (HP) = 1.3
CDA PRODUCT (FT**?) = 12,
POLLING RESISTANCE COEFFICTENT-;
FO (Lu/LB) = - .015
Fl CLB/LB-wPH)*E-05 = 2.15
F2
-------
L-9712^9-7
APPENDIX IV
HISTORICAL ENGINE-RELATED OWNERSHIP COSTS
For analysis purposes, present engine-related auto ownership costs have been
separated into four major categories: fixed costs, out-of-pocket running costs,
maintenance costs , and repair costs. Each of these categories is divided into
its components which will be discussed in detail later.
At the outset it is assumed that the car will be driven by three owners
over a seven-year lifetime for a total distance of 105,000 miles at a constant
rate of 15,000 miles per year. The initial purchaser of the car is assumed to
own and operate it for the first four years, the second owner for two years,
and the third for one year. Data from Ref. 16 indicate that this would be a
typical ownership history. An ownership cost history for the engine-transmission.
complex is shown in Table XXIII, details of which are discussed below.
Fixed Costs
Fixed costs are taken as those costs which are mainly independent of
odometer mileage. These include the capital costs (depreciation, interest, and/or
finance charges) plus insurance and taxes.
The engine-transmission share of the original selling price of a. 1971
medium-priced U-door sedan has been separately established as $5l8 for the engine
and $160 for the transmission for a total of $678. Data from Ref. 21 indicate
that the wholesale value of a car (and, presumably, all its components) can
be closely approximated by a declining-balance depreciation rate of 27.7$
per year. That is, at the end of any year the car's wholesale value is
72.3/5 of what is was one year previously. Extrapolation of the ratio of wholesale-
to-retail prices to the time of initial sale indicates an initial wholesale
value of 8*t.5/S of the original retail sales price. Hence, an initial wholesale
value of $573 is found for the engine-transmission complex.
Statistical data on automobile financing (Ref. l6) indicate that about
one-third of all new car buyers pay cash and that the remaining two-thirds
finance. It is assumed that those who finance pay one-third cash and borrow the
balance for three years at 12% with uniform annual payments. From this it follows
271
-------
L-971249-7
TABLE XXIII
Engine/Transmission-Related Ownership Costs
1971 Automobile
(Costs in Dollars)
0 1
Miles - 15000
EMiles - 15000
Fixed Costs :
Cash Cost 678 -
Loan 298 210
Equity 275 20 4
Value 573 4l4
Deprec-
iation - 264
Interest - 36
Vestcharge - l6
Insurance - 25
Taxes - 77
Total Fixed Costs 4l8
Fuel & Oil Costs 387
Maintenance :
Interval - 6 mos 28
Interval - 12 mos 43
Interval - 24 mos -
Total Maintenance 71
Repairs :
Distributor (l4)* 12
Exhaust' (39)
Water Pump (52)
Carburetor (45)
Fan Belt (51)
Fuel Pump (52)
Generator (52)
Transmission (66)
Engine Block (70)
Radiator (76)
Oil Pump (109)
Total Repairs 12
Grand Total 888
2
15000
30000
TIT
J — L J_
188
299
115
25
12
25
10
187
387
42
43
4o
125
12
-
-
-
-
-
-
-
-
-
-
12
711
Year
3
15000
45000
217
217
82
13
11
25
10
141
387
28
43
-
~71
12
26
24
29
-
-
-
-
-
-
-
91
690
4
15000
60000
91 ft
CJ.U
7D
l U
87
157
60
_ _ _
13
6
10
89
387
42
86
40
168
12
-
-
-
4
15
32
-
-
-
-
63
707
5
15000
75000
76
113
105
8
5
6
17
141
387
28
43
-
~
12
-
-
-
-
-
-
178
93
-
-
283
882
6
15000
90000
82
82
31
4
5
6
10
56
387
42
43
40
125
12
26
24
29
_
-
_
_
-
16
_
107
675
7
15000
105000
__
59
59
23
_ _ .
5
6
10
94
387
28
43
_
IT
12
_
_
_
4
15
32
_
_
_
18
81
633
EYears
1126
2709
702
649
5186
^Interval, months
272
-------
L-9712^9-7
that, in aggregate, new car purchases are financed $6% cash and kb% by loan. Thus
a hypothetical purchaser pays $380 cash for the engine-transmission and borrows
$298. His initial equity, however, is the difference between the initial wholesale
value and the loan, $275, and not the $380 which he actually paid.
In addition to the 12$ typical loan interest, a vestcharge of 6% on the owner's
equity is also levied. This is in accordance with principles discussed in Ref. 2k
and represents the income which might be expected from an appropriate alternative
investment.
Depreciation is charged at the aforementioned rate of 27.7% of the previous
year-end value except that the first year's depreciation figure includes the
difference between initial cost and initial wholesale value. This same exception
applies to the fifth year, which is the first year of ownership by the car's second
owner.
Financing the second ownership follows the same principles as the first but
with somewhat different proportions. Automobile financing statistics (Ref. l6)
indicate that 52$ of second-hand car buyers pay all cash and that k8% finance.
Here again, those who finance are assumed to pay one-third down and two-thirds on
account at 12$, l»ut this second loan is assumed to mature in two years instead of
three. This process results in a hypothetical second owner paying 68$ cash and
borrowing 32%. Meanwhile, the ratio of wholesale value to retail price has
declined from an original 8U.5$ to 72.0$. Thus, the second buyer's cost is 1.39
times wholesale value, whereas the comparable first buyer's cost factor was 1.183.
Reference 28 was used as the source for insurance and tax costs. These data
(averages of 5 U. S. cities) can be regarded as at least typical though not neces-
sarily equal to national averages. The data of Ref. 28 were first expressed as
percentages of initial cost and applied in that manner to the present application.
The philosophy regarding insurance is that liability, property damage, and
medical payments are properly chargeable to the driver, not the vehicle, and should
not be shared as an engine-transmission cost. Comprehensive (fire and theft) and
collision ($50 deductible) were found to be 0.9$ and 2.8$, respectively, of initial
cost. It was assumed that comprehensive would be carried for the life of the car
but that collision would be carried only for the first three years.
Sales tax averaged 3$ for the 5 cities examined and is applied in this instance
to both retail sales of the vehicle. A federal excise tax of 7$ is applied to the
initial sale of the vehicle. The annually recurring taxes, registrations, licenses,
inspections, etc., averaged l.Ul$ and are applied here throughout the car's life.
273
-------
L-9712^9-7
TABLE XXIV
ENGINE MAINTENANCE COSTS
Interval
(mos. or kmi.) Item Cost.
6 Change oil 5
Change oil filter 5
Clean breather cap k
Check and adjust belts k
12 Replace plugs 12'
Replace points 5
Replace PCV 3
Replace fuel filter 3
Clean cooling system 5
Tune-up 15 '
2k Replace belts 10
Replace hoses 15
Replace air filter 5
Change coolant 10
-------
L-9712^9-7
It will be noted that the total capital (fixed) costs of $1126 amount to
of the original price.
Out-of-Pocket Running Costs
Engine-related out-of-pocket costs are simply fuel and oil consumption. Here,
the fuel consumption rate of Ref. 15 was used, 13.8 miles per gallon*, and an oil
consumption rate of 3A qt/1000 mi, not including oil changes, was assumed. Fuel
(gasoline for conventional cars) was assumed to cost 3l|.90/gal, including state and
federal tax, and oil,
Maintenance Costs
Routine maintenance operations and mileage intervals are shown in Table XXXIV.
These are typical of owner's manual recommendations (Ref.29 ). Also shown are
estimated costs "by item. A summary of these costs appears in the maintenance cost
section of Table XXIII. The service station flat-rate manual (Ref.30 ) and parts
catalogues (Ref. 31) were used as guides in estimating maintenance costs.
Repair Costs
In spite of maintenance, component failures do occur, necessitating occasional
repairs. The engine-transmission components, the average mileages at which they
require repairs, and the average costs of repairs are listed in the Repairs Section
of Table XXIII. These data were taken from Ref.32 and represent repair experience
on fleets of cars owned and operated by the Bureau of Public Roads and by the
highway departments of each of the 50 states. Although the data of Ref. 32 do not
so specifically state, it has been established that these fleets were subjected to a
schedule of maintenance comparable to Table XXIV in addition to the repair work
presented.
Total costs -for each year of a car's life are shown at the bottom of Table XXIII
and, in addition, are graphically illustrated in Fig. 112. The "present worth" of
each of these total annual costs is established in Table XXV. The sum of the
"present worths" provides a final single figure for purposes of ownership cost
comparison, the "present worth" of total ownership costs of an engine-transmission
combination over a seven-year useful life.
*This estimate is inconsistent with Ref. 15, which indicates a 13.96 mpg national
average for cars of all weights, including subcompacts and compacts. UARL simula-
tions consistently indicate that the mileage for the UOOO-lb car is 11 to 11.5 mpg,
depending on driving cycle, which is more consistent with the Ref. 15 data.
275
-------
L-971249-7
FIG. 112
ENGINE-TRANSMISSION RELATED OWNERSHIP COSTS, 1971 AUTOMOBILE
1000
800
600
O
o
400
200
-FIRST OWNER-
-SECOND OWNER-
YEAR END
2T6
-------
L-9712^9-7
TABLE XXV
PRESENT WORTH OF OWNERSHIP COSTS
1971 Automobile
Year
1
2
3
It
5
6
7
Cost
8Ul
711
690
707
882
675
583
PWF (0.06)
0.9k3k
0.8900
0.8396
0.7921
0.7VT3
0.7050
0.6651
PW
793
633
579
560
659
)*76
388
IPW
793
li*26
2005
2565
322U
322^
1+088
277
-------
L-9T12U9-7
278
-------
L-9712U9-7
APPENDIX V
DESIGN OPTIONS FOR IMPROVED FUEL ECONOMY
The study ground rules and the conservative design assumptions result in
base-line engines which result in vehicle fuel economies far short of the ultimate
potential of these engines. In this appendix a number of design options are con-
sidered which serve to increase the fuel economy. With these options, the SSS-10
could achieve'the OAP design goal of 10 mpg in a 197^ demonstration engine, and
exceed it in subsequent production engines.
Transmission Design Options
An extended optimization study of engine and transmission combinations is
required and has been recommended to ,OAP. Aside from the obvious result that
transmission efficiency improvements/lead to corresponding fuel economy improvements,
additional benefits of transmission optimization can be attained by:
/
1. permitting the engine to be sized smaller in order to achieve
identical performance, thus providing better part-load fuel
economy, and
2. providing better response and thus permitting lower idle speeds
and better idle.fuel consumption.
/
Reasonable expectations of the/result of such a study are that fuel economy could
be increased by from 10 to 20$' over the FDC as a result of using currently pro-
posed, but more efficient, transmission concepts, and optimizing the engine and
transmission together. Furthermore, potential improvements exceeding 50% can be
postulated. Several transmission concepts appear worthy of more detailed in-
vestigation, and are discussed in the following:
I
General Electric Hydromechanieal Transmission (GE HMT)
The GE HMT was specified for the base-line vehicle in this study since it has
been demonstrated in actual hardware, provides excellent drivability, and has
prospects for reasonable Manufacturing cost with a relatively low technological
risk. A major drawback, viowever, is that its average efficiency over the FDC is
only 50$. This efficiency could undoubtedly be improved by readily conceived
modifications which might include an overdrive, for example, to provide maximum
efficiency for both the FDC and for freeway operation. This effect can be seen
by referring to Figs. 9/1 and 92. For example, specification of an overdrive which
allows a 2:1 drive axle ratio at speeds greater than 35 nqrh, and 3:1 at speeds less
279
-------
L-9712k9-7
TABLE XXVI
SSS-10 Fuel'Economy Optimization
Driving Cycle
Line
1
2
3
h
5
6
1
8
9
10
11
12
13
Ik
15
Modification
Base-line
GE HMT With Overdrive
Sundstrand DMT
Borg Warner n-speed
Tracer
Transmission n = 100%
3700# Vehicle Test Weight
3k80# Vehicle Test Weight
Aero Drag of 80$
Radial Belted Tires
Combined Vehicle
Modifications 5,8,9 & 10
Engine Reoptimization
Combined System
Modifications: (1971*)
(1977-78 Performance)
(Post-80 Performance)
FDC
7.51
7.90
8.05
8.26
7.76
9.0!4
7.61
7.77
7.6l
7.92
8.38
9.00
10.05
10.55
11.08
OAP
Suburban
11.01
11.3k
11.56
12.12
12.55
13. 6U
11.11
11.26
11. 2k
12.05
13.95
12.12
15.3.5
16.12
16.89
OAP
Country
12.02
12.02
12.65
13.21
12.82
13.82
12.12
12.26
12.93
13.10
Ik. 9k
11.50
lit. 23
1U.95
15.66
OAP
Composite
9.77
10.07
10.^7
10.71*
10. it?
11.71
9.88
10. Ok
10.07
10.50
11.63
10.8k
12.77
13. kl
Ik. 05
280
-------
L-9712U9-7
than 35 mph, would improve fuel economy by permitting maximum efficiency for both
operating modes.- This modification plus a 3% efficiency gain at low-power operation
(below 10$ power) would yield the results shown on line 2 in Table XXVI (7.90 mpg
on the FDC, with corresponding improvements on the other driving cycles examined).
Sundstrand Dual-Mode Transmission (DMT)
The Sundstrand DMT has been described in Ref, 33 as it pertains to commercial
applications. Like the GE HMT, it is a split-path combination hydrostatic and
mechanical device which provides infinitely variable output characteristics. It
is more efficient, particularly in the lower speed ranges, at the expense, of
greater complexity and a probable higher manufacturing cost. Although exact per-
formance data of a suitably sized DMT were not evaluated, an average increase of
efficiency of 15$ in the low-speed range and 5% in the high-speed range might be
expected. Efficiency improvements of this magnitude would lead to the results
shown in line 3 of Table XXVI •
Borg-Warner n-Speed Transmission
The Borg-Warner-suggested concept (Ref. 1 ) of a 6, 8 or 10-speed geared
transmission coupled to the SSS-10 through a controlled-speed slipping clutch would
provide relatively high transmission efficiency at the expense of some off-optimum
engine operation and the possible requirement for a somewhat larger engine to
provide equivalent performance. The overall effect expected might be to provide
a fuel economy gain of approximately 10$, as shown on line k of Table XXVI (8.26 mpg
on the FDC).
Tracor Transmission
Data on the performance of the Tracor traction (friction) drive transmission
was furnished by EPA for consideration. These data are shown in Fig. 113, and were
adapted for use in the simulation program as shown in Fig. 11U. The results of the
fuel economy evaluation using this data is shown on line 5 of Table XXVI (7-76 mpg on
the FDC).
Other Transmission Possiblities
Other likely transmission possibilities include modified torque converter
automatics, geared multispeed manuals, variable-sheave pulley and chain traction
devices, other toroidal devices, etc. The last three in particular show high
promise for higher overall transmission efficiency. While detailed study is
required to establish the feasibility of any of these devices, line 6 shows the
fuel economy which would be obtained if transmission efficiency (n) was 100?.
In other words, this line represents the theoretical limit for a base line vehicle
with no transmission losses other than reduction gearing and rear axle losses.
281
-------
TRACOR TRANSMISSION EFFICIENCY FOR VARIOUS OUTPUT/INPUT SPEED RATIO
100
90
z
UJ
u
Of.
UJ
Q.
U
UJ
u
u.
u.
UJ
80
55 70
—
3
60
50
SCALE CHANGE
10
20
30 40 50 60
INPUT POWER, HP
70
80
90
100
110
120 130
-------
ASSUMED TRACOR TOROIDAL TRANSMISSION EFFICIENCY FOR VARIOUS VEHICLE SPEEDS
•o
I
CD
U)
1.0
0.8
>- 0.6
u
0.4
0.2
ASSUMED EFFICIENCY
" OF SLIPPING CLUTCH
J_
_L
TOROIDAL TRANSMISSION
0 0.2 0.4 0.6 0.8 1.0 1.2
SPEED RATIO - OUTPUT/INPUT
VEHICLE SPEED PARAMETER _ V/V*
1.4 1.6 1.8 2.0
Tl
O
-------
L-97121+9-7
Probable Vehicle Test Weight
The section of this report concerning RECOMMENDED ENGINE SELECTION (page 221)
presented estimates of likely vehicle weight, which were shown in Table XIX. It is
strongly believed that the SSS-10 has a significant advantage in that its low weight
can further reduce the overall weight of the vehicle designed for its use, with a
corresponding beneficial effect on fuel economy. From line 13 of Table XIX, an
estimated adjusted gross weight of 1+1+00 Ib is shown. Deduction of TOO Ib as per
GAP specifications would then lead to a vehicle test weight of 3700 Ib. The effect
on fuel economy of a reduction from 3850 Ib for the baseline vehicle to 3700 Ib
can be derived from Fig. 93 and is shown on line 7 of Table XXVI (7.6l mpg on the
FDC).
Additional weight savings may be possible. For example, the Tracer trans-
mission is estimated to weigh 130 Ib versus the 185 Ib assumed in Table XIX. In
addition, its higher input speed permits weight reduction in the engine gearbox of
approximately 30 Ib. If this transmission proves to be satisfactory, vehicle
weight would decrease by an additional 11+2 Ib [1.67 x (55 + 30)]. Fuel weight
will also decrease for a vehicle delivering 10 mpg over the FDC, from the 187 Ib
assumed in Table XII, to ll+O Ib. An additional decrease of 220 Ib (ll+2 + 1.67 x 1+7)
would account for both of these effects, and the results on fuel mileage are
shown on line 8 of Table XXVI for the resulting 3^80-Ib test weight vehicle. (Note,
this result accounts for the weight reduction, but not the efficiency improvement
of the Tracer transmission. Multiple effects are discussed later.)
Vehicle Resistance
It is obvious that vehicle resistance determines a large part of the power
required to propel the vehicle, and thus has a first-order effect on fuel con-
sumption. As discussed in the Vehicle Evaluation Section, OAP-specified resistance
equations call for a standard vehicular frontal area, air drag coefficient, and
rolling drag coefficient in order to provide a common basis for analytical
evaluation of competing propulsion systems. In the design.of the demonstration
vehicle, there is opportunity to reduce the resistance expected for the actual
vehicle. For example, the small size of the SSS-10 would probably permit a some-
what smaller aerodynamic cross section (A) than the 25 ft2 specified by the OAP,
while the reduction of radiator cross section and some underpanning (which may be
required in any event for the exhaust system) could also reduce the air drag
coefficient (CAD). The present C^A value of 12 ft2 (0.1+8 x 25) could be reduced
20$, to 9.6 ft2, with a C^ of O.Uo and an A of 2k ft2 (or, alternatively, a C^ of
0.1+1+ and a frontal area of 21.8 ft2).
*U.S.Government Printing Office: 1974—747-787/318 Region No. 4
281+
-------
L-9712U9-7
The changes in fuel economy for the base-line vehicle solely due to a
reduction in aerodynamic resistance of this magnitude are shown in line 9 of
Table XXVI.
With regard to reducing rolling resistance, it is well known that the use of
radial belted tines reduces tire resistance considerably. Line 10 of Table
shows the effect on fuel consumption of the base-line vehicle when radial belted
tires are substituted with a 20% reduction in rolling resistance.
Combined Vehicle Modifications
All of the above fuel economy optimization procedures have been considered
independently of each other. A detailed study of the system interactions has not
been made, but a cursory analysis clearly indicates that significant fuel economy
improvements are possible when the above elements are combined in a reasonable
fashion. Line 11 presents the expected results from such a cursory analysis
pertaining to combining certain of the above vehicle modifications. The calculations
were made from the combination of the Tracer transmission (line 5), an expected
vehicle test weight of 3^80 Ib (line 8), aerodynamic design improvements (line 9)>
and radial belted tires (line 10). Fue]_ mileage on the FDC is now shown to be 8.38
mpg, while overall mileage is 11.63 mpg. These figures represent improvements of
11$ and lk%, respectively, in mpg, over those for the base-line vehicle. Note
that potential improvements in the engine itself have not been considered, as yet.
Engine Reoptimization
In the course of an actual engine development program, considerable opportunity
exists for further reoptimization of the engine to improve its fuel consumption.
Improvement goals for three candidate areas for reoptimization might be as follows:
(l) reduce idle speed an additional 5$ of design speed in order to reduce idle fuel
consumption by an additional 8 to "LQ%; (2) improve the intake and exhaust ducting
arrangements to reduce the predicted installation losses at low airflow in order
to improve low part-power fuel economy by an additional 5%; (3) reoptimize all'
engine components and rematch as necessary to further reduce idle and low-power
fuel consumption by 5 to 1.0%, accepting, if necessary, a concurrent 10% reduction
in full-power fuel consumption.
The net result of successful pursuit of these three goals would be to improve
FDC fuel economy of the SSS-10 by 20$, with a concomittant reduction on the OAP
Country cycle of 5$. This result is shown on line 12 of Table XXVI, showing 9.00
mpg on the FDC.
285
-------
L-9712U9-7
Combined System Modifications
The combination of the suggested vehicle modifications (line 11) and engine
efficiency improvement goals (line 12) results in the system estimates shown on
line 13, in which the GAP specification of 10 rnpg over the FDC is indicated. Con-
siderable further study is needed to establish the probability of achieving the
performance shown on line 13. However, it is clear that a reasonable possibility
of making significant improvements in the fuel economy of the SSS-10 demonstrator
automobile exists.
With regard to eventual production automobiles, further development of the
engine and refinement of the optimization processes described above lead to
expectations of future performance 5 to 10% better than shown, for 1977-78 and
post-1980 periods, respectively. The resulting values are shown in lines 1^ and
15. (It should be noted that the estimated post-1980 performance of the SSS-10
exceeds that predicted for the RCSS-8 base-line vehicle.)
A large difference exists between the transmission possibilities of line 6
and the line 5 data used for the values shown in lines 11, 13, I1* and 15. The
assumption of satisfactory development of high-efficiency transmissions combined
with the other modifications described above would yield fuel mileage estimates
of up to 17.mpg for the post-1980 period.
The proper optimization of any.system with complicated interactions, such as
a new type of low-polluting automobile,, involves not only a large analytical effort,
but extensive field and customer testing. Therefore, the above comments on fuel
economy optimization should not be regarded as definitive, but rather indicative
that if fuel economy is important enough, the SSS-10 automobile has the potential
to deliver over 10 mpg on both the FDC and the OAF composite cycles.
Driving Cycle Effects
The relative fuel economy of the several engines considered herein depends
on the driving cycle. The OAF driving cycle contains a very small proportion
of freeway driving; only 1/9 (by time) of the driving is at 70 mph and an additional
1/9 at 60 mph. This preponderance of slow-speed operation penalizes the SSS-10
engine with its somewhat worse part-load fuel consumption. The following table
illustrates this effect for both the base-line vehicle and the post-1980 fuel-
economy-optimized vehicle (line 15) of Table XXVI.
286
-------
L-9T121+9-7
FDC
OAF Suburban
OAP Country
OAP Overall**
60 mph
50 mph
Base-line
SSS-10
Mileage
(line 1)
7.5
11.0
12.0
9.8
12.2
11.5
i/mi.*
Relative
Cost
H.I
2.8
2.6
3.2
2.5
2.7
OC-70
Mileage
9.5
12.6
12.5
11.3
12.9
11.5
Relative
Cost
3.7
2.8
2.8
3.1
2.7
3.0
Post-1980
Fuel Economy
Optimized SSS-10
Mileage Cost
(line 15)
11.1
16.9
15.7
llt.l
15.7
15.0
2.8
2.0
2.0
2.2
2.0
2.1
Thus, the urban driver (FDC) of the base-line SSS-10 auto vill find his fuel cost
higher (O.it#/mi) by about the same amount that the freeway driver (70 mph) finds his
cost lower (0.3<#/mi), while the typical driver of the base-line SSS-10 auto (OAP
Overall) will notice no significant difference (O.l^/mi). The driver of the
, post-1980 SSS-10 vehicle will have a consistent 0.7#/mi to 0.9^/mi cost saving
compared with present experience.
*
**
SSS-10 fuel at 31^/gal; OC-70 fuel at 35^/gal.
Combination of equal portions of FDC, OAP Suburban, and OAP Country, by time.
287
-------
|