-------
The passenger car with a recuperative transmission and an onboard means of storing
vehicle kinetic energy in a flywheel offers many improvements in emission character-
istics. A basic configuration considered for such a hybrid vehicle drive is shown in
the block diagram of Fig. 1-1.
ENGINE
(40% NORMAL
HP RATING1
TRANSMISSION
TO DIFFERENTIAL
AND WHEELS
FLYWHEEL
ASSEMBLY
Fig. 1-1 Flywheel/Hybrid Drive System
The first obvious improvement is the conservation of energy resulting in less fuel
burned. The availability of a bilateral kinetic-energy storage system in the vehicle
with the high power recuperative characteristics (or power density) needed to recover
vehicle kinetic energy during braking provides several additional means of reducing
emissions. The more important of these improvement potentials brought about by the
presence of the flywheel are as follows:
The flywheel can provide the power required for vehicle acceleration, thereby
reducing the required installed engine horsepower to 40 percent of that now
used for the desired performance. This leads to the possibilities of either
using a smaller engine or modifying the present-size engine for much lower
emissions.
The energy-storage flywheel permits the engine to be operated primarily at
a single point on the torque/speed curve (or over a single line of limited
range on the torque/speed curve) such that emissions may be reduced.
1-5
-------
c:.
No rapid engine accelerations (which presently contribute substantially to
vehicle emissions) are necessary with the flywheel available to provide for
vehicle acceleration requirements.
In the flywheel/hybrid configuration described, no reduction in vehicle size is required
and, similarly, no compromise in vehicle performance from present levels is envi-
sioned. In addition, improvements being made to present types of automotive engines
are all applicable to the hybrid vehicle and further improve emission control potentials.
Finally, when developments in low-emission Rankine and Brayton cycle engines or
modified Diesel engines become available, these may possibly be applied to the
flywheel/hybrid for additional air pollution reduction.
PURPOSE OF THE FLYWHEEL FEASIBILITY STUDY
The purpose of the present study is to determine the feasibility of the flywheel as a
means of attaining low-emission propulsion systems for urban vehicles. A further
objective is the demonstration and performance evaluation of full-sized kinetic-energy
flywheels.
This program is a part of the overall plan of the Advanced Automotive Power Systems
(AAPS) Program. The goal of AAPS is the demonstration within 5 years of a very low
pollution automobile in accordance with the February 10, 1970, Message on Environ-
ment by the President. The planning philosophy of the AAPS Program is to investigate
various technical approaches to very low pollution automobiles (including gas turbines,
steam engines, Diesels, stratified-charge engines, electric hybrids, flywheel hybrids
and flywheel-only systems) and to later select from these the most promising candi-
dates, which will then be developed into prototype systems.
The work done on this Flywheel Feasibility Study and Demonstration Program as
described in this report along with other flywheel activities being conducted by EPA/
APCO will assist EPA/APCO in a preliminary screening to determine the best hybrid
approach electric battery or flywheel.
1-6
-------
KINETIC-ENERGY STORAGE VEHICLE BACKGROUND
The use of the flywheel for the operation of urban vehicles is not new. The most
recent and commercially significant application to date of flywheel energy storage
to land transportation is the Oerlikon Electrogyro (Ref. 1-1).
Twenty years ago, the Oerlikon Engineering Company of Zurich, Switzerland, began
operating the Electrogyro, a passenger bus powered solely by a 5. 25-ft diam flywheel.
Initial success led to the production of these buses and their application in Switzerland,
Germany, Belgium, and the Congo. The last Electrogyro went out of revenue service
in 1969.
A pictorial representation of the Oerlikon Electrogyro is shown in Fig. 1-2. In oper-
ation, three-phase ac power was collected, as shown, from an overhead pole by three
collector arms. This power was used to energize a two-pole squirrel cage induction
I
I I
CAPACITOR
BANK
- AC PROPULSION
MOTOR
MOTOR POLE
CHANGER
MOTOR/GENERATOR
KINETIC-ENERGY
WHEEL ASSEMBLY
Fig. 1-2 Oerlikon Electrogyro Bus
1-7
-------
C
motor which spun the flywheel (operated in 0.7 atmosphere of hydrogen) up to u speed
of about 3,000 rpm (synchronous speed for 50-Hz operation). The bus then operated
using a pole-changing induction drive motor up to an average distance of three quarters
of a mile on power obtained from the kinetic energy stored in the spinning flywheel.
The flywheel was then recharged during a 40-sec dwell at the next charging pole. In
operating the bus, a considerable degree of skill was required by the operator in thai
he had to switch drive motor poles at just the right moment, perhaps as many :is 12
times for each stop/start sequence.
The background of the Electrogyro and other applications form the basis for the updating
of flywheel technology in which advantage can be taken of new materials, electronic-
controls, improved testing techniques, quality control concepts, and recent develop-
ments in seal and bearing technology.
1-8
-------
Section 2
APPROACH
The Flywheel Feasibility Study and Demonstration conducted for the Air Pollution
Control Office of the Environmental Protection Agency was intended to be an objective
assessment of the suitability of flywheel energy storage, either alone or in a heat-
engine hybrid configuration, to propulsion systems for low-pollution urban vehicles.
The program was structured to evaluate the flywheel as a drive system component
for vehicles capable of meeting "real world" operating specifications. These require-
ments were based on typical existing vehicles with full accessory loads - including air
conditioning - and presently available performance in acceleration, maximum speed,
and vehicle range. The final results of the program were to show whether the flywheel,
using presently available materials and technology, is worthy of continued considera-
tion as an alternative means of propulsion for low-pollution urban vehicles. Demonstra-
tions were based on the design,manufacture, and evaluation of full-scale flywheels.
The approach used in conducting the initial study portions of the program was first
to perform applicability studies to find the optimal vehicle/flywheel combinations.
These studies established performance requirements for various vehicles and resulted
in comparative evaluation of various flywheel system configurations. Next, tradeoff
studies were made to evaluate complete drive systems - consisting of flywheels, trans-
missions, and engines for the two selected vehicles from the standpoint of compara-
tive emissions, cost, size, weight, and availability- Finally, a series of design studies
and tests were performed to evaluate flywheel geometries, materials, failure control,
and gyrodynamics.
The applicability study was directed toward four common vehicle types - the city bus,
the delivery/postal van, the family car, and the commuter car in various configura-
tions as shown in Table 2-1. For each of the four vehicle types, the flywheel was
2-1
-------
Table 2-1
VKflTCLl-: CONFIGURATION SKLKCTION
System Configurations
Flywheel Only
Flywheel/Diesel
Flywheel/Rankine Cycle
Flywheel/Gas Turbine
Flywheel/Spark Ignition
Vehicles
City
Bus
A
B
C
D
E
Delivery/
Postal Van
F
G
H
I
J
Family
Car
K
L
M
N
O
Commute i'
Car
P
Q
R
S
T
Select Select
One One
cd both ;is the sole means of propulsion and as ;i hybrid sysk-ni r
e.g. , the Flywheel in conjunction with the Diesel online, tho Uankinr cycle engine.
the 11 ray ton cycle (gas-turbine) engine, and the Otto cycle (spark-ignition) online.
in selecting the two optimum configurations for the study, special consideration w;is
given to the city bus. Therefore, one of the five configurations shown in Table ^- 1
was selected as an optimum flywheel system applicable to a city bus. In addition, one
of the IS flywheel-only or flywheel/heat-engine hybrid configurations was selected
for one of the three smaller urban vehicles.
After selecting the two best applications of the flywheel, a detailed study using a
computer-aided methodology was carried on to review the tradeoffs relating to var-
ious drive configurations - flywheels, engines, and transmissions - with respect to
vehicle performance requirements and emissions resulting from operation over
typical operating profiles. In the drive configuration tradeoff studies, the vehicle
operating profiles used for both the family car and the commuter car were the DTTftW
Urban Dynamometer Driving Schedule, as published in the Federal Register of
November 10, 1970. This 7.5-mile driving cycle consisting of 18 stops and starts -
2-2
-------
represents a portion of Los Angeles streets and freeways. It represents a realistic
city driving profile and was used to quantitatively compare the suitability of various
drives used for passenger cars. The city bus drive configurations were evaluated
using a cycle based on city bus operations data collected for LMSC by the American
Transit Association. The postal van was evaluated for operations over a predicted
cycle representative of typical postal pick-up cycles within the city. These drive
configuration tradeoff studies resulted in the recommendation of optimum flywheel
drives for the two previously selected vehicles on the basis of emission reduction
potentials in combination with cost and availability considerations.
In the same timeframe, a series of flywheel design studies were conducted. In these
studies, attention was directed toward the various available flywheel geometries,
materials, and fabrication techniques. Tests were conducted to evaluate prospective
flywheel materials, and consideration was given to system components such as bear-
ings, seals, vacuum systems, housings, and flywheel assembly suspension systems.
In addition, the gyrodynamic or precession effects experienced with flywheels were
investigated and quantified. The design studies also considered the problems and
potential methods of flywheel failure control.
The results of the flywheel design studies were used in the computer-aided design of
two flywheel types, one for the city bus and one for the selected smaller vehicle. A
large number of plausible flywheel designs provided the opportunity for an objective
choice of the two best candidates. The two designs selected then were detailed and
used to fabricate two flywheels of each type in full-size configurations representative
of typical flywheels which might be applied to future urban vehicles.
A test program was conducted on the four flywheels to verify their energy storage capa-
bilities and power density characteristics during charge and discharge. Spindown
losses of the flywheels at various vacuum levels also were measured and acoustical
noise levels over operating speed ranges were assessed. Final tests on the two
2-3
-------
s
flywheel types included overspeed testing until flywheel failure occurred. The results
of these tests then were evaluated and used to determine the suitability of the flywheel
as a component for urban vehicle propulsion systems.
On the basis of the results of the applicability studies, drive configuration studies.
design studies, and flywheel tests, a series of recommended future program plans
have been prepared and are presented. These program plans present a logical se-
quence of activities for applying the flywheel energy storage principle to future low-
emission urban vehicles having the maximum potentials for near-term air pollution
control.
2-4
-------
Section 3
APPLICABILITY STUDIES
Applicability studies were conducted to determine the most promising configuration
(flywheel only or flywheel/hybrid) for the city bus, and to determine the most prom-
ising combination of vehicle and configuration for the other three smaller vehicles.
First, specifications were developed for each of the vehicles; then, based on con-
current flywheel studies, the flywheel size and volume were determined for each
flywheel only and flywheel/hybrid vehicle configuration. Finally, a comparison of
flywheel size and weight requirements with allowances, in conjunction with other
considerations, led to the selection of vehicles and configurations.
DEVELOPMENT OF VEHICLE SPECIFICATIONS
Specifications for each of the four vehicles under consideration were developed in
conjunction with the Air Pollution Control Office. These specifications are sum-
marized in Table 3-1.
FAMILY CAR SPECIFICATIONS
The family car is a difficult candidate in which to effect emission improvements
because of the severe cost constraints. Despite this, the family car is the prime
candidate for emission reduction because its large population causes the bulk of
automotive air pollution. For this reason, the particular type of family car chosen
is that which has had the highest market percentage - the full-sized six-passenger
family sedan. It is assumed that emission reduction techniques suitable for the
family car could readily be applied to the rest of the automotive population.
The performance specifications in this section for the family car are as outlined in the
Advanced Automotive Power System Program, "Vehicle Design Goals Six-Passenger
3-1
-------
/
?x&,:)00 max
1 .«00 mnx
:ir,
45
rj
>A C = 13
}
(a)
Commuter Car
0. 02 to 5
4 to G2
70
33 6 12
4
5
60 to 300
60
2,500
-
4 , 500
7,000
1 .700
42
160
42
O.BS
4 max
Remarks
Number of stops that vehicle makes per
mile necessitated by traffic or commercial
constraints
length of trip/route divided by time vehicle
Is on trip/route » 'range/'Atr!mKe
Maximum sustained cruise velocity
Maximum sustained velocity on the specified grrulu
Maximum acceleration achievable I proportional to
V (0 to cruise)/At (to reach cruise)]
Length of time vehicle Is stopped (V - 0) while all
accessories are operating. Stop necessitated by
traffic or commercial considerations (such as
loading/unloading passenger from bus) only.
Length of a trip or duty cycle. Maximum value of
one trip if It Is given as n variable is the maximum
length of a trip/duty cycle possible without supple-
menting vehicle energy system. Note: For the
flywheel-only city bus, I range may be achiever! by
one duty cycle with "recharge'1 between duly cycles.
The ratio of recharge limn to duty cycle time shall
not exceed 7 percent.
Payload weight for commercial vehicles
Occupant capacity (passengers anil driver) for
noncommercial vehicles
Curb welghl
Fully loaded total vehicle wfiiuhl
Weight [taelKnublu to nil propulsion HyHliim COJIIJMIII-
onts Including energy HtornKe, conlrulR. rlr..
Maximum vulumu usslimubln Id projmlHiun HyHlnu
cnmponeiila Including energy Hlorugu c'onin>ln. «j(c.
Time from Binrl-up U, usublu |K»wcr outpul
Frontal crosx-scctlonal nrc:i suitable for tho <-:il-
culBtlon of aerodynamic drag
Drag coefficient
-
(n) Not specified.
(b) The range (a to b) means a continuous variable bounded by a and b. Any calculations mode
should be dense enough over the range (a to b) to show the effect of the variable.
(c) Limit for flywheel-only vehicles.
NOTES:
1. Any calculations of rolling resistance due to tires
should be made on the basis of currently available
tires and include the effect of tire width. Decreasing
rolling resistance due to tires by assuming a type of
tire that has unsafe traction characteristics by virtue
of low rolling resistance is not allowed.
2. With respect lo the city bus, the average accelera-
tion (a max) must not be achieved by any Instantaneous
accelerations or rate of acceleration that would cause
passenger discomfort.
3. Any resulting weight distribution that results In
possible handling characteristics significantly dif-
ferent from the normal expected by drivers of the
vehicles should be noted specifically.
4. Any decrease In gross vehicle weight achieved.
for example In the lighter family car and the commuter
car, must not compromise safety considerations.
5. Operation of the vehicle should not be compromised
by ambient weather considerations. Am b I em weather
Is defined as -25' to 110'F.
6. Noise aspects of the various components should
be considered.
3-2
-------
Automobile," (Revision A, November 30, 1970), contained in Appendix A.* These
specifications are in general agreement with the performance capability of present-
day automobiles with the exception that the maximum cruise velocity for the specifica-
tion vehicle is in the neighborhood of 80 mph. (Although not specified explicitly, this
maximum cruise velocity results from the specified velocity requirement of 65 mph on
a 5-percent grade.) Present-day family cars have velocity capabilities considerably
in excess of 80 mph; however, this is felt to be not so much a requirement per sc,
but rather a result of concurrent requirements for good acceleration performance and
economy up to about 80 mph.
Accessory power requirements for each of the four vehicles were estimated by LMSC.
Accessory power requirements for the family car are given in Table 3-2. Air con-
ditioning and power steering loads are somewhat less than for present-day family
cars because the flywheel drive provides a more narrow speed range power takeoff.
In accordance with the requirements stated in "Vehicle Design Goals," emissions for
the family car are calculated on the basis of the revised "DHEW Urban Dynamometer
Driving Schedule" published in the Federal Register of November 10, 1970. This
driving cycle, based on an actual driving experience on Los Angeles roadways, is
7. 5 miles long and includes 18 starts and stops. Velocity and acceleration plots of
this DHEW cycle are shown in Fig. 3-1. In order to accommodate limitations of the
chassis dynamometers used for emission measurement, the measured accelerations
were limited to ±3. 3 mph/sec. This clipping is evident in the acceleration plot. An
acceleration histogram of the DHEW cycle is shown in Fig. 3-2. Except for the
acceleration clipping (and an apparent preference for certain least significant digits
in data taking), the histogram follows the expected distribution.
The sensitivity of road power required for the family car to velocity, grade, and
vehicle weight is shown in the plots of Fig. 3-3. These same data are also replotted
for convenience in Figs. 3-4, 3-5, and 3-6.
* Since the completion of this work there has been a further revision (Revision B,
February 11, 1971) in which vehicle rolling resistance and air drag are slightly re-
duced and vehicle speed and weight are increased. The net result is that the applica-
bility of the flywheel is increased somewhat by this latest revision, because of the
lower ratio of nonrecoverable losses (rolling resistance and air drag) to recoverable
loss (acceleration).
3-3
-------
Table 3-2
ACCESSORY TOWER REQUIREMENTS FOR FAMILY CAR
Accessories
Service Lights - High Beam
Service Lights - Rear
License Plate Lights
Windshield Wiper Motor
Defroster Fan Motor
Heater/Air Conditioning Fan Motor
Clock
Radio
Dash Lights
Alternator Status
Energy Storage Status
Engine Temperature Status
Total Electrical Load
Alternator Losses (Eff. = 0. 85)
Alternator Input
Power Steering Pump
Air Conditioning Compressor
Total Accessory Load
Power Requirements
Watts
46.8
24.0
18.0
12.0
24.0
24. 0
0. (i
60.0
18. 0
1.2
1.2
1. 2
231.0
40.8
271.8
hp
0.4
0. G
4.0
5.0
3-4
-------
6O
145
0.
S^
ft 3O
16
n
100 200 300 400 BOO
600 700 800
TIME (SECONDS)
eoo 1000 noo 1200 1300 vtoo
co
i
(A
9
1OO
2OO 3OO 4OO
500 6OO 700 800
TIME (SECONDS)
9OO 1OOO tlOO 12OO 13OO 14OO
Fig. 3-1 DHEW Urban Dynamometer Driving Schedule - Velocity vs. Time and
Acceleration vs.Time (Revised November 10. 1970 Federal Register)
-------
11O
1OO
90
80
UJ
I
o:
K
D
0 60
0
o
it-
CD
UJ 50
O
Z
01
D
0 4°
z
3O
20
10
O
-I
_
-
..
1 1 ' 1 I I ' r i ~~r ~
3 -7 -6 -5 -4
a
1
I
:
;
:
\l
\l
H : :
|Is:
it; iiy
i)iifihm*
ll llllll'l I'- : \
-3-2-1 6 1 2 3
ACCELERATION INCREMENT (MPH/SEC)
Fig. 3-2 DHEW Urban Dynamometer Driving Schedule Acceleration
Histogram (Revised November 10, 1970 Federal Register)
3-6
-------
co
-q
D
4
H
a
N
H
tr
u
z
H
°-
U
LJ1
IT
H
I
S
"
SDDD Lfc
f7DD L&
H-D tn BD
l/ELBCITY
XDD
1ZD
Fig. 3-3 Family Car Road Power for Various Loaded Weights and Grades
-------
/
c
a
o
H
/\
I
v°
Z -0
h
H
J
u
a
(M
ion HP
no HP
. 120 HP
ID
ERRUE
2D
30
Fig. 3-4 Grade Velocity Variance With hp to Road for 4, 700-lb
Loaded Weight Family Car
3-8
-------
ID
GRRUE
2Q
3D
3-5 Grade Velocity Variation With hp to Road for 5, 000-lb
Loaded Weight Family Car
3-9
-------
10
2D
3D
<%>
Fig. 3-6 Grade Velocity Variation With hp to Road for 5, 300-lb
Loaded Weight Family Car
3-10
-------
COMMUTER CAR SPECIFICATIONS
The commuter car, as defined in Table 3-1, Is of the same general class as the Auwlin
Mini and Fiat 850. The commuter car, however, seats only two, has a top speed of
70 mph, and a range of only 50 miles. The specification obviously bends to accommo-
date a low-emission energy storage means, but is intended to represent the minimum
that the general public would accept as a commuting automobile.
The availability of nonpolluting commuter cars would probably have little effect on total
automotive air pollution during the next decade. Nevertheless, the commuter car could
be a suitable first application for a nonpolluting energy storage device which, with
further development, might gradually be applied to larger automobiles.
The driving cycle used for emission determination is the same DHEW cycle used for
the family car.
Accessory loads for the commuter car are given in Table 3-3, and a plot of velocity
vs. grade is shown in Fig. 3-7.
3-11
-------
Table 3-3
ACCESSORY POWER REQUIREMENTS FOR COMMUTER CAR
Accessories
Power Requirements
Watts
hp
Service Lights - High Beam
Service Lights - Rear
License Plate Lights
Windshield Wiper Motor
Defroster Fan Motor
Heater/Air Conditioning Fan Motor
Clock
Radio
Dash Lights
Alternator Status
Energy Storage Status
Engine Temperature Status
Total Electrical Load
Alternator Losses (Eff. = 0. 85)
Alternator Input
Air Conditioning Compressor
Total Accessory Load
46. 8
24.0
18.0
12.0
18.0
18.0
O.G
60.0
18.0
1.2
1.2
1. 2
219.0
38.6
257. G
0.3
3.0
3.3
3-12
-------
a
a
H
a
a
H
H
^
J
U
ID 20
ERRJ3E <5K>
3D
Fig. 3-7 Velocity vs.Grade for the Commuter Car
3-13
-------
CITY BUS SPECIFICATIONS
The city bus is an attractive candidate for low-pollution propulsion systems because
of the relatively soft constraints in the size, weight, and cost. In addition, the
Federal Government has existing means to subsidize low-pollution city buses through
capital-improvement grants. However, the need for low operating cost is still crit-
ical because this cost is borne by the operator of the bus system. Since the highest
single operating cost is for drivers1 wages, the performance capability of the bus is
very important. (Improved vehicle performance can result in reducer! route times
which in (.urn permit a reduction in the number of buses and drivers required to handle
peak load conditions.) In addition, the cycle life of any energy storage means must be
great enough so that operating cost is not seriously affected.
The city bus performance requirements shown in Table 3-1 are representative of
current production vehicles. The driving cycle used as the basis for. emission cal-
culations is given in Table 3-4. Velocity and acceleration plots for this schedule
are given in Figs. 3-8 and 3-9, respectively. These performance requirements
and driving cycle are based in part on data compiled for LMSC by the American
Transit Association {Table 3-5).
Accessory loads are shown in Table 3-6. Again a limited speed range power takeoff
was assumed for the air conditioner compressor and power steering pump.
A plot of velocity vs.grade for the city bus is shown in Fig. 3-10.
3-14
-------
Table 3-4
CITY BUS CYCLE
Velocity Mode
(mph)
0 to 19. 2
19.2
19.2 to 25.2
25.2
25. 2 to 0
0
Time
(sec)
6.0
3.0
3.0
2.0
9.0
13.0
36
Distance
(ft)
99
112
71
112
132
___0
529
Acceleration Rate(a)
(mph/ sec)
3.2
0
2
0
-2.8
Average Velocity = 10 mph
Stops = 10/mile
(a) Accelerations at constant road hp except maximum
jerk = 2 mph/sec^
Table 3-5
AMERICAN TRANSIT ASSOCIATION STATISTICS FOR CITY BUSES
City
Los Angeles
New York
San Francisco
Maximum
Speed
(mph)
24
Average
Speed
(mph)
12.8
9.3
15.99
Maximum
Route
Length
(miles)
17.3
12.07
12.18
Median
Route
Length
(miles)
7
5
5.79
Maximum
Grade
(percent)
7 to 10
13.0
21.0
Stops
per
Mile
11
10
9-15
3-15
-------
ia 20 30
TIME
Fig. 3-8 Velocity vs. Time for City Bus Driving Schedule
in
\
I
D.
E a
V H
z
D
H
h
tE a
tr «
u
j
u
u
xo 20 aa
TIME
4-0
Fig. 3-9 Acceleration vs. Time for City Bus Driving Schedule
3-16
-------
Table 3-6
ACCESSORY POWER REQUIREMENTS FOR CITY BUS
Accessories
Power Requirements
Watts
hp
Service Lights - Regular Beam
Service Lights - Rear
License Plate Lights
Destination Sign Light
Interior Lights (5 at 40 w)
Identification Lights (2)
Entrance/Exit Lights (2)
Clearance Lights (4)
Windshield Wiper Motor
Defroster Fan Motor
Clock
Dash Lights
Alternator Status
Energy Storage Status
Other Gages
Total Electrical Load
Alternator Losses (Eff. = 0.9)
Alternator Inputs
Power Steering Pump
Air Conditioning Compressor and Fan
Total Accessory Load
50
40
36
36
200
20
Intermittent
40
24
48
0.6
36.0
1.2
1.2
18.0
551.0
61.2
612.2
0.8
1.5
18.0
20.3
3-17
-------
/
T
a
a
H
a
n
xs
I
H
U
J
LJ
a
IM
20
ID
GRRUE
Fig. 3-1.0 Velocity vs. Grade for City Bus
3D
3-18
-------
DELIVERY/POSTAL VAN SPECIFICATIONS
The delivery/postal van, although of small population, could be an attractive first
application of a low-pollution propulsion system. As in the case of the city bus,
size, weight, and cost constraints for the delivery/postal van are soft, and the
purchase of postal vans by the Federal Government could be directed toward low-
pollution vehicles. The delivery/postal van specifications are given in Table 3-1;
and the driving schedule in Table 3-7. These are representative of current vehicles
involved in urban delivery and mail collection. Velocity and acceleration plots for
this schedule are given in Figs. 3-11 and 3-12, respectively.
As shown in Table 3-8, the accessory loads are almost negligible.
Figure 3-13 shows a plot of velocity vs. grade for the delivery/postal van.
Table 3-7
DELIVERY/POSTAL VAN CYCLE
Velocity Mode
(mph)
0 to 16
16
16 to 32
32
32 to 0
0
Time
(sec)
4
11
4
9
8
60
96
Distance
(ft)
73
396
92
330
165
.0
1,056
Acceleration Rate
(mph/sec)
4.0
4.0
-4.0
Average Velocity = 7. 5 mph
Stops = 5/mile
3-19
-------
/\
I
D- *
£ N
V
h
H
Q *
J
U
ID 20 30
TIME <5ECHNU5>
Fig. 3-11 Velocity vs. Time for Deliver/Postal Van Driving Schedule
/N
in
\
i
Q_
z
B
H
h
tr °
o: ? 1
u
j
u
u
u _
4-
10 5O 3D
TIME <5ECaNH]5>
Fig. 3-12 Acceleration vs. Time for Delivery/Postal Van Driving Schedule
3-20
-------
Table 3-8
ACCESSORY POWER REQUIREMENTS FOR DELI VERY/POSTAL VAN
Accessories
Service Lights High Beam
Service Lights - Rear
License Plate Lights
Windshield Wiper Motor
Defroster Heater Fan Motor
Clock
Dash Lights
Alternator Status
Energy Storage Status
Engine Temperature Status
Total Electrical Load
Alternator Losses (Eff. = 0. 85)
Alternator Input
Total Accessory Load
Power Requirements
Watts
46.8
24.0
18.0
12.0
24.0
0.6
18.0
1.2
1.2
1.2
147.0
25.9
172.9
hp
0.2
0.2
3-21
-------
a
D
rl
a
a
XX
I
H
J
U
a
P4
o xo 20
GRRJ1E <5K>
Fig. 3-13 Velocity vs.Grade for Delivery/Postal Van
3Q
3-22
-------
VEHICLE/CONFIGURATION SELECTION
Preliminary flywheel size and weight estimates were made for both flywheel only and
flywheel/hybrid configurations of each of the four vehicles. This was done to elim-
inate from further consideration those combinations of vehicle and configuration that
are not feasible because of excessive flywheel size or weight.
Flywheel design studies (described in Section 5) showed two types of flywheels that
appeared attractive - one, a uniform biaxial-stress steel disc; the other, a uniaxial-
stress bar (or log) type flywheel of glass filaments. The bar-type flywheel has the
higher theoretical energy density (w-hr/lb) but lower energy per unit volume (w-hr/
cu ft), so size and weight estimates were made for both types of flywheels.
For flywheel-only vehicles, the required flywheel capacities to meet various parts of
the performance specifications are given in Table 3-9. The resulting flywheel sixes
and weights are as shown in Table 3-10. For the family car, commuter car, and
delivery/postal van, the required flywheel weight exceeds the maximum allowable
weight of the entire propulsion system. The bar flywheel for the city bus has a
volume greater than that allowed for the total propulsion system. The only remain-
ing possibility, then, is the city bus with a steel disc flywheel. The flywheel weight,
however, is over 2/3 the total propulsion system weight. This flywheel weight is
excessive, but if the range (between recharging) of the bus were reduced to approxi-
mately 10 miles, then the flywheel-only bus would appear to be feasible. Because of
the excessive size and weight of flywheels for the flywheel-only vehicles, further
consideration was limited to flywheel/hybrid vehicles.
The flywheels for the hybrid vehicles are sized on the basis of vehicle kinetic energy
at maximum speed. The rationale for sizing the flywheel in this manner is the "Total
Kinetic Energy Control" technique. By means of a linear control loop, heat-engine
output is controlled so as to maintain the sum of the vehicle kinetic energy and the
flywheel kinetic energy at a fixed value approximately equal to the vehicle kinetic
energy at maximum speed. Thus, the flywheel supplies the power required for
3-23
-------
/
«
Table 3-9
KD FI.YWIIKKI, CAPACITIES FOR FLYWHEEL-ONLY VEHICLES
Vehicle
Family Car
Commuter Car
City Bus
Delivery/Postal Van
Energy Required
for Maximum
Grade Operation
(kw-hr)
2.91
10.0
1.88
Energy Required
for Vmax Over
Range
(kw-hr)
136(a)
17.2
52.0
42.5
Energy Required
for Trip Profile
Over Range
(kw-hr)
114(a)
14.1
97.2
51.5
(a) 200 miles per NAPCA "Vehicle Design Goals Six-Passenger Automobile.
(see Appendix A).
;:i;ccleration, and the heat engine supplies the power necessary lo overcome aero-
dynamic drag, rolling resistance, grade, and associated drive-train losses. This
"Total Kinetic Energy Control" method is discussed more fully in Section <\.
The flywheel energy capacities required for each of the four hybrid vehicles arc
shown in Table 3-11. These capacities are such that the resulting weights and
volumes are only a small fraction of the allowable values for the total propulsion
system. Therefore all four hybrid vehicles - using either flywheel type are
feasible with respect to flywheel weight and volume.
Although each of the smaller hybrid vehicles were shown to be potentially feasible.
the family car was chosen for further study because it causes the bulk of today's
automotive air pollution.
The applicability studies thus resulted in the following choice of vehicles for further
study:
Flywheel/hybrid family car
Flywheel/hybrid city bus
3-24
-------
Table 3-10
FLYWHEEL SIZES AND WEIGHTS FOR FLYWHEEL-ONLY DESIGN
Vehicle
Type
Family
Car
Commuter
Car
City Bus
Delivery/
Postal
Van
Required
Flywheel
Capacity
(kw-hr)
136
17.2
97.2
51.5
Maximum
Allowable
Propulsion
System Weight(a)
(Ib)
1,500
600
6,000
1,700
Maximum
Allowable
Propulsion
System Volume (a)
(cu ft)
28
16
175
42
Constant Stress
Maraging Steel
Flywheel
(23. 6 w-hr/lb,
6. 8 w-hr/cu in. )
Weight
(Ib)
5,763
729
4,119
2,182
Volume
(cu ft)
11.57
1.46
8.27
4.38
Bar Flywheel
(S-1014)
(25. 1 w-hr/lb,
0. 226 w-hr/cu in. )
Weight
(Ib)
5,418
685
3,873
2,052
Volume
(cuft)
348
44.0
249
132
CO
Ol
(a) Propulsion system includes flywheel, transmission, clutches, driveshaft, differential, axle, and controls.
-------
Table 3-11
REQUIRED FLYWHEEL CAPACITIES FOR
FLYWHEEL/HYBRID VEHICLES
Vehicle
Required Flywheel
Capacity
(w-hr)
Family Car
Commuter Car
City Bus
Delivery/Postal Van
372
115
GG5
155
3-26
-------
VJXVK
Section 4
CONFIGURATION TRADEOFFS
As a result of the applicability studies (Section 3), the flywheel/hybrid family car and
the flywheel/hybrid city bus were selected for further study. Configuration tradeoffs
therefore were directed toward selecting the best transmission type and the best heat-
engine type for each of these two vehicles.
TRANSMISSION SELECTION
The function of a transmission for a flywheel/hybrid vehicle is to couple together the
heat engine, the flywheel, and the vehicle such that emissions from the heat engine
are reduced with minimum sacrifice in overall vehicle performance. To achieve
this, acceleration loads are handled by a reversible exchange of kinetic energy be-
tween the vehicle and the flywheel, while cruise power is supplied by the heat engine.
TRANSMISSION REQUIREMENTS
Ratios. Torque provided by the flywheel is proportional to the rate of change of fly-
wheel speed; thus, infinitely variable (stepless) ratios are required in the flywheel-
vehicle path in order to produce reasonably smooth torque and preclude excessive
shifting losses. Similarly, the transmission speed ratio must not wander or oscillate
because this would cause objectional variations in vehicle acceleration and possible
operational instability.
Ratio Range. To provide reasonably good efficiency, the vehicle must operate over a
fairly wide speed range without intentional dissipation (clutch slipping). Thus, a
vehicle speed range (without dissipation) of 10:1 was assumed. Since the flywheel
slows as the vehicle accelerates, the overall ratio range required is approximately
10 times the flywheel speed range.
The selection of flywheel speed range is a tradeoff between required flywheel size and
required transmission size. The portion of the flywheel energy that is available is
given by
4-1
-------
where
W./W = ratio of available flywheel energy to total flywheel energy*
R = flywheel speed range
Increasing the flywheel speed range increases the transmission speed range that is
required, and therefore may increase transmission size; but it also reduces the
flywheel size that is required.
Recuperation. The exchange of kinetic energy between the flywheel and the vehicle
must be reversible. When the vehicle is moving forward, the flywheel must deliver
power to the vehicle during acceleration and recuperate power from the vehicle during
braking. When the vehicle is reversing, flywheel power may be desirable but is not
essential, and recuperation during braking in reverse is of little value.
Efficiency. In order to effect maximum heat-engine emission reduction, the trans-
mission must be efficient over a wide range of vehicle speeds. Efficiency is especially
critical in the flywheel-vehicle path because in urban driving high rates of acceleration
and repetitive recuperation can cause the average power to and from the road to be
several times the average engine power. Thus, a given percentage loss in flywheel-
vehicle path could be reflected as a much larger percentage increase in engine brake
specific fuel consumption and resulting emissions.
Compatibility With Current Practice. The flywheel transmission must be compatible
with current automotive practice if it is to go into production in the next 5 or 10 years.
The usual design constraints of size, weight, and component availability are largely
inflexible in this time span. Transmission noise should not be worse than with current
vehicles and driver feel should not be adversely affected. Maintenance and repair must
be capable of being accomplished using presently available facilities with a minimum
of personnel training.
* Flywheel weight does not vary inversely with this expression since allowable working
stress is affected by stress range for the same cycle life.
4-2
-------
TRANSMISSION CONTROLS
One of the first Items for consideration in the transmission studies was the determina-
tion of flywheel speed as a function of vehicle speed. The simplest approach might be
to control the heat engine in an on-off (full throttle on - declutched off) mode so as to
drive the flywheel up to approximately its half-energy speed, reserving the remainder
of the flywheel capacity for recuperative braking. This control mode, however, has
two drawbacks: first, the flywheel size is double that required for vehicle accelera-
tion from 0 to maximum speed; second, the flywheel energy, and therefore the per-
formance capability of the vehicle is history dependent. For example, a vehicle
cruising for some time just below maximum sustained cruise speed would have a fly-
wheel charged to one half total capacity. This would permit a rapid pass maneuver at
speeds well above the maximum sustained cruise speed. Immediately following such
a maneuver, however, the flywheel energy might be completely exhausted. Then if ah
identical maneuver is attempted, the vehicle will "bog down" with its speed limited to
maximum sustained cruise speed. In fact, the loss of available power could occur in
mid-maneuver. The dangers inherent in such vehicle behavior are readily apparent. *
After considering .several other control techniques, a control method was devised
which can be used to minimize history dependence of vehicle performance and which
obviates flywheel oversizing for recuperation capacity. This approach is the Total
Kinetic Energy (TKE) control concept.
TOTAL KINETIC ENERGY CONTROL
The basic idea of the TKE approach is to control the heat-engine output so as to hold at
a constant value the sum of the kinetic energies of the flywheel and the vehicle. The
speed of the vehicle is squared and scaled to give the vehicle kinetic energy. Likewise,
the flywheel speed is squared and scaled to yield the flywheel kinetic energy. These
two kinetic energy values are added to give the total kinetic energy of the flywheel-
vehicle combination. From this TKE value is subtracted a predetermined quantity,
essentially equal to the kinetic energy of the fully loaded vehicle at maximum cruise
*An analysis of flywheel/hybrid family car performance in the DOT High-Speed
Pass Maneuver is given in Appendix E.
4-3
-------
velocity. The resulting error signal is then used as a feedback signal to control (ho
heat-engine output so as to minimize the error signal and thereby hold the TKE fixed.
The net result of this TKE control approach is that the flywheel provides power for
vehicle acceleration (and stores power during vehicle deceleration) while the heat
engine supplies only cruise power and accessory power. The cruise power includes
losses due to rolling resistance, aerodynamic drag, grade, and associated transmis-
sion and flywheel losses. It should be noted that neglecting transmission losses, a
sudden and sustained acceleration of the vehicle does not affect the TKE if the heat
engine continuously supplies cruise power.
In order to minimize emissions as fully as possible, this basic TKE approach may be
modified somewhat. For instance, it may be desirable to allow an excess of TKE
during regenerative braking on downhill runs in order to convert the potential energy
of the vehicle into flywheel kinetic energy. In no case, of course, would it be per-
missible to allow the flywheel to overspeed.
Because of the necessary lag in heat-engine response to a control error signal, the
driver hears a slow build-up in engine noise sometime after making a sudden acceler-
ation. This inherently disconcerting affect can be eliminated by means of the com-
pensated feedback control system as shown in Fig. 4-1. The Power Control separates
the accelerator pedal power command signal into an acceleration power signal and a
cruise power signal. The cruise power is estimated as a direct function of vehicle
speed, and this cruise power signal is subtracted from the commanded power to give
the acceleration power signal. The cruise power signal adjusts the speed of the heat
engine. The heat engine throttle is preprogrammed on the basis of engine speed for
minimum emissions. The Gear Ratio Control divides the acceleration power by vehi-
cle speed to give a torque command which in turn is integrated to give a gear ratio
command to the vehicle transmission. The TKE (vehicle plus flywheel) tends to drift
due to cruise power estimation error and grade power, so a separate loop is used to
calculate the TKE and compare it to a preset reference. The resultant error signal
is used to linearly override the power signal to the heat engine so that the TKE is
maintained constant. This control scheme results in a vehicle feel for the driver
which is very close to that of a conventional family car except that engine noise is
only slightly affected by acceleration.
4-4
-------
ACCELERATOR
BRAKE
PRNDL
TO WHEEL
DIFFERENTIAL
+
ENERGY j
r>
-
REFERENCE
Fig. 4-1 Control System Block Diagram
The final form of the transmission controls is dependent upon required future testing
both of emissions and vehicle "feel" in various control modes. Since testing of proto-
type vehicles in various control modes is necessary, the use of all-electronic controls
for prototype vehicle propulsion systems is strongly recommended. Production
control systems may or may not be all-electronic because of cost.
TRANSMISSION CONFIGURATIONS
The two basic transmission configurations that were considered are shown in Fig. 4-2.
The double transmission, as the name implies, is really two transmissions: a wide-
range transmission between the flywheel and the vehicle, and a narrow-range trans-
mission between the flywheel and the engine. The single transmission is a wide-range
4-5
-------
transmission linking the vehicle with the flywheel and/or engine, where the flywheel
and heat engine may be clutched in and out independently.
DOUBLE TRANSMISSION
ENGINE
TRANSMISSION
(1.5:1 RANGE)
FLYWHEEL
TRANSMISSION
(15:1 RANGE)
DIFFERENTIAL
VEHICLE
SINGLE TRANSMISSION
VEHICLE
Fig. 4-2 Basic Transmission Configurations
The double transmission offers complete control flexibility since the heat-engine
speed may be controlled independent of either flywheel speed or vehicle speed.
Cruise efficiency, however, tends to be degraded by having two transmissions in the
engine-vehicle path.
The single transmission has better cruise efficiency by virtue of having only one
transmission in the engine-vehicle path, but heat-engine speed is tied directly to fly-
wheel speed. Since the flywheel must slow down as the vehicle speeds up, so must
the heat engine slow down as the vehicle speeds up. This means that maximum engine
speed occurs at vehicle standstill, while at maximum vehicle speed the heat engine is
at two thirds of its maximum speed (still assuming a 1.5:1 flywheel speed range).
These speed relationships mean that the heat engine is being used in an awkward fashion.
4-6
-------
In addition, driver "feel" is upset because of engine sound. The single transmission,
however, is lighter, smaller, and lees expensive than the double transmission.
For prototype vehicles it is recommended that a double transmission be employed so
that all control modes of interest may be investigated. The single transmission would
be simulated by "locking-up" the engine/flywheel transmission.
Both single and double configurations were given continued consideration in the con-
figuration tradeoffs.
TRANSMISSION TYPES
Initially, four types of transmissions were to be given consideration - mechanical,
hydrostatic, hydrokinetic, and electric. Each type was evaluated in terms of its appli-
cability to the flywheel/vehicle transmission. This transmission constitutes the major
technical challenge since it involves providing a bilateral power flow between two large
inertias. The engine/flywheel transmission represents no significant technical chal-
lenge. Purely mechanical gear transmissions are not suitable because they cannot
provide the stepless changes in gear ratio required for reasonably smooth and efficient
operation. Hydrokinetic transmissions are not applicable for the overall transmission
because their output torque is controlled by the input/output speed ratio.
Hydrostatic Transmissions. The hydrostatic transmission consists of a hydraulic
pump connected hydrauiically to a hydraulic motor. Variable control can be obtained
either by means of bypass valving or by a variable stroke in one or both units. Because
of the need for high efficiency, only the variable stroke types were considered.
Block diagrams of double and single configurations of hydrostatic transmissions are
shown in Fig. 4-3. For both configurations, the hydrostatic units would be integrated
into a single transmission assembly. The single transmission would be rather con-
ventional, but the double transmission would be special because of the integration of
an additional pump.
4-7
-------
DOUBLE TRANSMISSION
VEHICLE
SINGLE TRANSMISSION
VEHICLE
Fig. 4-3 Block Diagrams of Hydrostatic Transmissions
Conventional industrial-type hydrostatic transmissions are excessively heavy (and thcre-
forealso too costly) for mass-produced vehicles. Sizing therefore was based on aircraft-
type components. A curve of weight versus rated power for aircraft-type hydrostatic
pumps and motors is shown in Fig. 4-4. Volume was estimated by LMSC at 24 cu
in. /lb, and cost at $1. 00/lb for passenger cars and $2. 00/lb for commercial vehicles.
Efficiency versus speed for a hydrostatic transmission is given in Fig. 4-5.
Electric Transmissions. Electric transmissions can take many forms, employing ac
machines with thyristor cycloconverters or dc link inverters or dc machines with
thyristor choppers. The dc machine plus chopper was ruled out because in high-volume
production it is larger, heavier, and more expensive than the ac machine with its power
conditioning. Cycloconverters were ruled out because of the 2:1 oversizing of the
4-8
-------
3,000-PSI
CONTINUOUS DUTY
AltOtAFT WEIGHTS
I ODD
IATEO FOWH (Hf)
Fig. 4-4 Weight vs. Rated Speed for Hydrostatic Motors and Pumps
U
z
u
H
U
H
U-
L
U
CD
5PEEID <5K>
TK
1DD
Fig. 4-5 Efficiency vs. Rated Speed for a Hydrostatic Transmission
4-9
-------
generator (Ref. 4-1) and the speed relationships which they require. Consideration,
then, was limited to ac machines with inverters operating in dc link systems. The
single and double configurations of these systems are shown in Fig. 4-6. In both con-
figurations, during regenerative braking, the motors and inverters function in the
negative slip region as generators and rectifiers.
DOUftlf TRANSMISSION
HEAT
ENGINE
^
GENERATOR
RECTIFIER
MOTOR
AND
INVERTER
MOTOR
AND
INVERTER
GEARBOX AND
DIFFERENTIAL
VEHICLE
SINGLE TRANSMISSION
HEAT
ENGINE
CLUTCH
VEHICLE
Fig. 4-6 Electric Transmissions
Variable speed experimental ac motor drive systems using inverters and cyclocon-
verters to energize the motor have been described extensively in literature. Motor
speeds of about 12, 000 rpm have been generally reported, and in some cases motor
speeds as high as 20, 000 rpm are used. By proper control of the frequency, the ac
motor can be made to have a speed torque characteristic (as shown in Fig. 4-7) similar
to the dc series motor used for traction application. The ac motor thus can provide
high starting and acceleration torque several times the rated torque of the motors.
4-10
-------
TORQUE
CONTROLLED SLIP
SPEED
Fig. 4-7 Torque-Speed Characteristics for Squirrel Cage Induction Motor
Weight, volume, and cost versus rated torque for catalog squirrel cage induction
motors are given in Figs. 4-8, 4-9, and 4-10, respectively.
Cost of the modified high-speed motors in high-volume production was estimated by
LMSC to be the same as the single quantity cost of a simpler standard catalog machine.
The cost is very close to $1. 00/lb.
Efficiency versus power rating for a line of squirrel cage induction motors is shown in
Fig. 4-11. The general trend is one reason why electric transmissions are used in
Diesel locomotives and large mine trucks, but not in smaller vehicles. Below about
1, 000 hp the machine losses become too large. Although experimental electric drives
can show substantially higher efficiencies, the cost of such systems is presently
prohibitive for application to family cars.
4-11
-------
a
a
a
a
H
S B
j2
l-
I
m
LJ
a
rl
-:
M
ID
1DD
1DDD
TQRQUE
Fig. 4-8 Weight vs. Bated Torque for Squirrel Cage Induction Motors
4-12
-------
H
L_
r^
L_
1 Q
1 -j
1 pi
~1
1
x/
z
i . .
, ! H
i
1
Nk
H .
/
j
/
\/
/
L
HB^M|^MMi
J^^
J
' M
X
9
,
-1
^<
^
1
T
fe
W
^
,
*i
,-
<<
11
*
*
ID 1DD
TERQUE
1DDD
Fig. 4-9 Volume vs. Rated Torque for Squirrel Cage Induction Motors
4-13
-------
a
a
a
a
H
a
a
-in-
s/
H
Ul
u 2
a
H
*-
1DD
1DDD
THRQLJE
Fig. 4-10 Cost vs. Rnled Torque for Squirrel Cage Induction Motors
4-14
-------
Ol
u
z
u
H
U
H
U_
b.
U
Q
Q
H
a
U1
1DD
1DDD
HC3R5EPC3NER
Fig. 4-11 Efficiency vs. Rated Power for Squirrel Cage Induction Motors
-------
Kl'ficicncy as a function of speed for combined motor and inverter is shown in Fig. 4-12.
This curve is based on LMSC motor testing. Again, better efficiencies have been at-
tained with high-cost experimental drives.
Weight and volume curves for a line of aircraft generators is shown in Figs. 4-13 and
4-14, respectively. These curves were used without modification. Cost was esti-
mated by LMSC at $1. 00/lb for passenger cars and $2. 00/lb for commercial vehicles.
Full load efficiency versus rated power for this line of aircraft generators is shown in
Fig. 4-lF>. As mentioned before in discussing motors, the low efficiency at the- lowc-r
power ratings has limited the application of electric transmissions to high-power
(greater than 1, 000 hp) vehicles.
Power Splitting Transmissions. In addition to the four types of transmissions
(mechanical, hydrostatic, hydrokinetic, and electric) initially given consideration, a
fifth type - the power splitting transmission was also studied. The power splitting
transmission, as shown in Fig. 4-16,is a combination of a mechanical-differential and
a hydrostatic transmission. The mechanical-differential gearing is normally plane-
tary and the hydrostatic transmission can tie between the second input of the differen-
tial and either the output or the other input. Input coupling was assumed because the
relatively narrow speed range of the flywheel provides a better tie point than the
infinite speed range of the vehicle.
The power splitting transmission configurations are shown in Fig. 4-17. Operation
is similar to that of the hydrostatic transmission, but the power splitting transmis-
sion has better efficiency since much of the power is handled mechanically. The power
splitting transmission, despite its greater size and weight, has lower cost than the
hydrostatic.
A curve of efficiency versus speed for a power splitting transmission is shown in
Fig. 4-18. This particular transmission is the Sundstrand DMT-250 which is slated
for high-volume production for on-the-road truck application by the end of 1972. Size,
weight, and cost estimates, by LMSC, also were based on extrapolations of data for
this transmission.
4-16
-------
SPEED (%)
Fig. 4-12 Efficiency vs. Speed for Combined Motor and Inverter
4-17
-------
s*-f{
a
a
a
rl
cD
J
I-
I
ID
H
U
a
1DD
1DDD
H0R5EPBWER
Fig. 4-13 Weight vs. Rated Power for Aircraft Generators
4-18
-------
I-
L_
I
U
u
V
u
z
n
J
H
a
rl
ID
1DD
1DDD
HQR5EPEIWER
Fig. 4-14 Volume vs. Rated Power for Aircraft Generators
4-19
-------
Q
1DD
1DDD
HQR5EPQNER
Fig. 4-15 Efficiency vs. Rated Power for Aircraft Generators
4-20
-------
INPUT
(FROM FLYWHEEL)
HYDROSTATIC
TRANSMISSION
MECHANICAL
DIFFERENTIAL
OUTPUT
(TO VEHICLE)
Fig. 4-16 Power Splitting Transmission
DOUBLE TRANSMISSION
ENGINE
TRANSMISSION
(3:\ RANGE)
pi YWHFFI
>OWER SPLITTING
TRANSMISSION
GEARBOX AND
DIFFERENTIAL
VEHICLE^
SINGLE TRANSMISSION
ENGINE
CLUTCH
POWER SPLITTING
TRANSMISSION
GEARBOX AND
DIFFERENTIAL
VEHICLE
FLYWHEEL
Fig. 4-17 Power Splitting Transmission Configurations
4-21
-------
a
a
H
u
Z a
u w
H
U
H
L. ,
L N
U
iaa
5PEEI]
Fig. 4-18 Efficiency vs. Speed for a Power Splitting Transmission
Family Car Transmission Comparison. A comparison of electric, hydrostatic, and
power splitting transmissions for both single and double transmissions suitable for
application to the family car is shown in Table 4-1. The efficiency is based on opera-
tion of the specified family car over the DHEW Urban Dynamometer Driving Schedule
with the assumption that all braking is recuperative. The efficiencies of each trans-
mission type are normalized on the basis of 1.00 efficiency for the best configuration
(power splitting single).
The size, weight, and cost figures for the various transmissions represent estimates
by LMSC based on equivalent production quantities. The electric transmissions are
worst on all accounts; the costs are probably prohibitive. The best transmission on
the basis of highest efficiency (which minimizes emissions) and lowest cost is the
power splitting transmission. The weight and volume of the power splitting trans-
mission, although greater than those of the hydrostatic, are not excessive.
4-22
-------
Table 4-1
FAMILY CAR TRANSMISSION COMPARISON
Item
DREW Schedule Efficiency
(Normalized)
Volume (cu ft)
Weight (Ib)
Cost ($)
Electric
Single
0.588
3.4
488
641
Double
0.369
4.0
536
689
Hydrostatic
Single
0.745
3.3
280
403
Double
0.626
3.7
289
469
Power Splitting
Single
1.000
3.7
311
261
Double
0. 835
5. 1
391
341
On these bases, therefore, the power splitting transmission was chosen as the best
one for the flywheel /hybrid family car.
Bus Transmission Comparison. A comparison of electric, hydrostatic, and power
splitting transmissions for both single and double transmissions suitable for applica-
tion to the city bus is shown in Table 4-2. The efficiency is based on operation of the
specified fully loaded bus over the city bus cycle with the assumption that all braking is
recuperative. The efficiencies of each transmission type are normalized on the basis
of 1. 00 efficiency for the best configuration (power splitting single).
Table 4-2
BUS TRANSMISSION COMPARISON
Item
Profile Efficiency
(Normalized)
Volume (cu ft)
Weight (Ib)
Cost ($)
Electric
Single
0.6254
16
2,023
6,088
Double
0.4267
17
2,077
6,291
Hydrostatic
Single
0.7681
20.6
1,937
3,335
Double
0.6615
23.8
2,085
3,900
Power Splitting
Single
1.000
25.5
2,255
2,205
Double
0.8495
35.5
2,922
2,872
4-23
-------
The si/o, weight, and cost, fibres for the various transmissions represent estimates
l>y LMSC based on equivalent production quantities. As with the case of the family car,
the electric transmission types arc seen to be prohibitively expensive. Again, on the
bases of best efficiency and lowest cost, the power splitting transmission was chosen
as being the best for the city bus.
HEAT ENGINE SELECTION
The heat engines to be considered are the spark-ignition, Diesel, gas turbine, and
Rankine cycle engines. Primary attention was given to cost, size, weight, efficiency,
and emission rates of carbon monoxide (CO), unburned hydrocarbons (HC), and oxides
of nitrogen (NOr), but the engine study also considered engine sound level and smoke
A.
emission.
HEAT ENGINE SURVEY
To support the configuration tradeoffs, realistic heat-engine data in the. range of 2<>
to 200 hp were required, representing the most desirable engine characteristics that
could reasonably be expected to be commercially available in the near future. Data for
this purpose were acquired by means of an exploratory survey of applicable heat engines
available, and letters of inquiry to the manufacturers of the engines of interest.
The specific objectives of the engine data survey were to derive the relationships
between engine continuous power rating and cost, specific fuel consumption, weight,
envelope volume, and exhaust emissions. For each engine type studied, these re-
lationships were established in a form suitable for input to the computerized para-
metric study. It was recognized that the usefulness and validity of the study depends
on the availability, in the 1971 1975 timeframe, of commercial versions of the engines
considered, in appropriate sizes, substantially in accordance with the performance
attributed to them and without the necessity for implemental development funding. Thus.
the stratified-charge engine, which may offer improvements over the common spark-
ignition engine if it becomes available over a range of sizes in the near future, was
not included. Similarly, rotating combustion engine data were not used because this
4-24
-------
engine's 1975 commercial availability in a range of sixes is not assured at this time
even though such engines are now used in automobiles (the NSU RO-80 for example).
The Stirling engine concept was considered to be in an inadequate state of vehicular
development for 1975 availability.
Engine preliminary data sources are listed in Refs. 4-2 through 4-7, Final data for
the engines of interest were obtained directly from the manufacturers. Exhaust
emission data were obtained primarily from Ref. 4-8 and were in part confirmed by
a small amount of data obtained from engine manufacturers. Information on engine
developmental status was derived in part from Ref. 4-9.
In order to simplify the task of sorting the most promising Diesel and spark-ignition
engines for use in the study, a selective elimination of engines from Refs. 4-2, 4-3,
and 4-4 was performed. As a weight, volume, and cost consideration, all low- and
medium-speed engines were eliminated from consideration on the basis that these
engines would be heavy and costly. In general, only engines of 1,800 rpm or more
were considered. As a "noise pollution" consideration, turbosupercharged engines
were eliminated. (Ref. 4-10. Also see the NAPCA Vehicle Design Goals, Appendix A.
This process of elimination resulted in a quantity of 160 eligible Diesel engines within
the range of 15 to 150 continuous horsepower, and 100 eligible spark-ignition engines.
All turbine engines determined to be within the range of 15 to 150 hp were considered,
and all "small" Rankine engines were considered.
Engine weights and piston displacements versus continuous horsepower were plotted
for the eligible Diesel and spark-ignition engines, and then straight lines were drawn
through the weight and displacement data in such a way that an adequate range of
engines lay below the lines. These Diesel and spark-ignition engines were regarded
as the preferred engines for purposes of the study, and was the subject of letters of
inquiry to the manufacturers to obtain confirming and additional data. Letters of
inquiry were also sent to all known or possible manufacturers of small turbine and
Rankine engines for whom addresses could be determined. A quantitative summary of
engines investigated, inquiries made, and responses obtained is given in Table 4-3.
4-25
-------
Table 4-3
QUANTITY OF ENGINES INVESTIGATED
Type
SI 'ARK-IGNITION
Eligible Engines
Manufacturer's Data Requested
Useful Response Received
DIESEL
Eligible Engines
Manufacturer's Data Requested
Useful Response Received
GAS TURBINE
Eligible Engines
Manufacturer's Data Requested
Useful Response Received
RANKINE
Eligible Engines
Manufacturer's Data Requested
Useful Response Received
Domestic
Engines
97
37
23
34
22
21
4
7
r>
14
3
Foreign
Engines
3
3
3
]26
37
12
2
2
0
Indeterminate
1
1
Total
TOO
40
26
1.60
M
:.{;{
<;
f)
n
15
4
ENGINE TREND DATA
The parameters of cost, specific fuel consumption, weight, and volume all vs. con-
tinuous horsepower were evaluated by a computerized "least squares curve fit" to
determine the most appropriate equation relating each parameter to the continuous
power rating. For each parameter, curves of the following type were compared with
the actual data:
Y = A + (BX)
Y = A(e)BX
4-26
-------
Y = A(X)B
***(!)
i
Y =
A + (BX)
Y = A + (BX)
where
Y = value of the parameter (cost, specific fuel consumption, weight, and
volume) .
X = corresponding continuous power rating
e = base of natural logarithms
A = Y intercept
B = factor
The equations of these curves were then used to determine the parameters at particu-
lar values of continuous rated power. A compilation of the data used to establish these
trend curves is given in Appendix C.
Emission curves for the various engines were supplied by EPA/APCO. These curves
are based on rather rudimentary data and do not include any cold start factors. Much
more data is needed and many more calculations must be made to properly calculate
the emissions. The emission figures quoted in this report should be considered only
approximate and preliminary results. These curves of HC, CO, and NOX as a function
of continuous horsepower are shown in Figs. 4-19, 4-20, and 4-21, respectively. These
curves were used as the basis for all emission calculations.
Spark-Ignition Engines. The conventional automobile spark-ignition engine is con-
strained by its cycle to a range of air/fuel ratios (between extremes of approximately 8
and 17), and in order to avoid detonation is limited to relatively low compression ratios
(less than approximately 12) compared with the Diesel. Its exhaust emissions therefore
4-27
-------
a
a
D
SPRRK IGNITION
DIESEL
RRNKINE
IZE5EL <6E5T>
RRNKZNE
TURBINE
SPRRK IGNITIHN <&E5T>
4-O BO 12D
HHR5EP0WER
TUR6INE <6EST>
16D
Fig. 4-19 HC Emissions vs. Continuous Horsepower
4-28
-------
a
a
a
a
a
a
e
a
a
a
e
a
D.S
I
I
I
tfl
o
H
i
i
a
a
i
a
a
a
DIESEL
-TURB.INE
RHNKINE
:5PRRK IGNITI0N
-RRNKINE <6EST>
-5PRRK IGNITION
-TURblNE
-IICSEL <6E5T>
BO 12D
HQR5EPCZIWER
16Q
Fig. 4-20 CO Emissions vs. Continuous Horsepower
4-29
-------
a
a
a
IXE5EU
TURfelNE
TUR&INE <6E5T>
3JIE5EU <6E5T>
5PHRK IBNZTION
RFWKINE
SPHRK IENITZQN <6E5T>
RRNKINC <&EST>
4-D BD 12Q
HHR5EPQWER
Fig. 4-21 NO Emissions vs. Continuous Horsepower
X
4-30
-------
£#££
tend to follow a characteristic pattern, and largely must be controlled external to the
cylinder or by avoiding power transients and idling conditions. The high state of develop-
ment of the basic engine and its high rate of production over a range of power values
results in a relatively compact, light, low-cost, readily available engine.
From Table 4-3 it is seen that letters of inquiry were sent to manufacturers repre-
senting 40 percent of the total quantity of eligible engine models. These represented
the most promising models for the application, derived by selective and preferential
elimination processes. Useful responses were received covering 65 percent of the
engines so investigated. However, emissions data were so meager as to be useless
for establishment of trends; hence, Ref. 4-8 was concluded to be unmodified by manu-
facturer's information, and was regarded as the effective emissions data source for
this study.
For the spark-ignition engine, the following relations were used to express the values
of the parameters:
Function Ordinates Equation
Cost vs. hp Yvs. X Y = A(X)B
Specific Fuel Consumption vs. hp Yvs. X Y = A + (rr
Weight vs. hp Yvs. X Y = A(X)B
Volume vs. hp Y vs. X Y = A(X)B
The computer plots of the four parameters showing all data points used for the curve
fit are shown in Figs. 4-22 through 4-25.
Specific fuel consumption, volume, cost and weight charts do not include the effect of
advanced emission control techniques and therefore are only approximate.
4-31
-------
a:
j
j
H
H
I-
U1
ID
U
SO 1DO ISO
HORSEPOWER
CO
Fig. 4-22 Spark-Ignition Cost vs.
Continuous Horsepower
/s
cfl
i
-I
H
LJ
SO 100 ISO
HHR5EPI3UJER
Fig. 4-24 Spark-Ignition Weight vs.
Continuous Horsepower
I
i'
I
\
cfl
d ".
u
U.
Ul
SO 1DD ISO
HORSEPOWER
Fig. 4-23 Spark-Ignition Specific Fuel Con-
sumption vs. Continuous Horsepower
XN
I-
L.
u
u
51
D
J
Q
SO 1DO ISO
HORSEPOWER
Fig. 4-25 Spark-Ignition Volume vs.
Continuous Horsepower
-------
Diesel Engines. The Diesel cycle inherently permits operation over a wide range of
air/fuel ratios, and permits higher compression ratios than the spark-ignition engine,
thereby providing lower specific fuel consumption. Due to its greater size and weight,
costly fuel-metering and injection system, and lower production rate, the Diesel is
more expensive than the spark-Ignition. It also is noisier because of the rapid rate
of pressure rise in the cylinder and injection system noise. Its higher air/fuel ratio
tends to result in much lower CO content in the exhaust, and HC content can be lower
with appropriate attention given to design of injection equipment. Despite the fact that
its higher air/fuel ratios tend to result In lower NO , its high combustion temperature
A
and pressure cause NO to be high.
Letters of inquiry were sent regarding 37 percent of the total quantity of eligible
Diesel models. Useful responses covered 56 percent of the models so investigated.
As with the spark-ignition engines, these represented the most promising Diesels
for the application, derived by two processes of elimination.
Extensive emissions data were received from one manufacturer, which provided a
general confirmation of the data of Ref. 4-8 for Diesels. Only meager emissions
data were received from the remainder of the manufacturers.
The parameters were again evaluated by computer curve fit to determine the best
equations relating the parameters to the continuous power rating. The following
relations were used:
Function Ordinates Equation
Cost vs. hp Yvs. X Y = A(X)B
1
Specific Fuel Consumption vs. hp Yvs. X Y =
A + (BX)
B
Weight vs. hp Yvs. X Y = A(X)
Volume vs. hp Y vs. X Y = A(X)B
4-33
-------
The computer plots of the four parameters showing all data points are shown in
Figs. 4-26 through 4-29.
Any reductions in cost, weight, or NO emissions would make the Diesel more
A
attractive.
Gas Turbine Engines. In comparison with the reciprocating internal combustion
engine, the nonregenerative gas turbine is much lighter, smaller, and simpler me-
chanically. However, continuously sustaining the high gas temperatures necessary
to obtain acceptable thermal efficiency and maintaining the production tolerances nec-
essary require expensive metallurgical, design, and production processes. For this
reason, turbine costs are generally appreciably higher. The turbine operates with
overall air/fuel ratios much higher than the spark-ignition engine, and higher than the
Diesel, in order to limit turbine inlet temperatures. Because its combustion process
is continuous - not intermittent as with the reciprocating engines - and occurs in a
continuously hot combustor, it tends to be a very low producer of CO and HC. Its
high air flow rates require special attention to intake and exhaust design to avoid
bothersome noise levels.
Turbines may be attractive for flywheel applications because of the good speed
compatibility.
The references used to collect engine data indicate that very few turboshaft gas tur-
bines exist in the range of 20 to 200 continuous shaft horsepower. For this engine
category, Refs. 4-2 and 4-4 indicate half a dozen such engines available. Letters
of inquiry were sent regarding these engines, and to several additional manufacturers
from Ref. 4-6. A total of five useful responses were obtained, an insufficient quan-
tity to support the drawing of trend curves. No emissions data were received. Data
related to fuel consumption, weight, and envelope volume were so diverse as to pro-
hibit any conclusion regarding a curve shape. The probable reason for this is the
strong growth of turbine power with time. Three of the turbines are growth versions
4-34
-------
u
en
HDR5EPQWER
Fig. 4-26 Diesel Cost vs. Continuous
Horsepower
H
I
log
H *
Id
3
a en 100 ica
HQR5EPQWER
Fig. 4-28 Diesel Weight vs. Continuous
Horsepower
a:
i
I a
a.*.
I
\
u
L.
CO 1DO 1KO
HI3R5EPCJWER
Fig. 4-27 Diesel Specific Fuel Consumption
vs. Continuous Horsepower
so 1.00 xso
HDR5EPDWER
Fig. 4-29 Diesel Volume vs. Continuous
Horsepower
-------
of approximately 1960 technology, which would tend to show lower values of specific:
fuel consumption, weight, and volume. Another of the turbines is a new development
and has the capability of doubling its present power in a few years, but it would re-
quire funding to accomplish this. The fifth turbine appears to be just in the planning
stage. Only three of the responses included cost figures.
In view of the small amount of useful small-turbine data, supplemental data were used
to derive curve shapes.
a. The three cost values, when plotted versus power, lay very nearly on a
straight line. As a check, four other points were added for known engine1
costs within the power range of 300 to 800 hp. These supplemental points
verified the applicability of the small-engine cost data, and permitted the
establishment of a cost curve of the form Y = A -i (BX). This plot with
the points used is shown in Fig. 4-30.
b. The five specific fuel consumption data points would not support the estab-
lishment of a curve in which confidence could be placed. Therefore, data
from an additional 30 turbine engines were taken from Ref. 4-4, covering
continuous power values up to 1, 150 hp. These data, representing a more
or less average degree of engine growth, were combined on a plot with the
five small engines, and the computer curve fit program employed to deter-
mine the relationship of specific fuel consumption to continuous power.
Investigating the same curve forms as for the spark-ignition and Diesel
engines, the relation used was
Y = A (X)B
This computer plot with the points is shown in Fig. 4-31.
4-36
-------
s«
oooac
v a
-S-
in
n
u
14
H
£
I
\ 0
Q. H
I
\
cfl
J M
V
u
L.
Ul
\
120O
HHRBEPQWER
300
HDR5EPC1WER
i
CO
-a
Fig. 4-30 Gas Turbine Cost vs. Continuous
Horsepower
cfl
J
\/
h
I
ID
H
U
Z
* <*
300
HC3R5EPDWER
12OO
Fig. 4-32 Gas Turbine Weight vs. Con-
tinuous Horsepower
Fig. 4-31 Gas Turbine Specific Fuel Con-
sumption vs. Continuous Horsepower
/N
H
L.
U
v
y
E
I]
J
E
300
HC3R5EPC3WER
1ZOO
Fig. 4-33 Gas Turbine Volume vs. Con-
tinuous Horsepower
-------
irtr
Weight data for the five small turbines were not considered reliable for
197r> engine weight prediction. Two engine weights were only prediction
estimates, and the other two engine weights were abnormally low because
the engines are at an advanced state of power growth. Again, Ref. 4-4
data for 30 additional engines were used to supplement the small-engine
data. The resultant relationship between weight and continuous rated power
used was
Y = A (X )B
This plot with points is shown in Fig. 4-32.
d. Precisely the same comments noted for weight also apply to the small-
engine volume data. Reference 4-4 data again were used as a supplement,
and the resultant relationship used was
Y - A (X)B
This computer plot with points is shown in Fig. 4-33.
To regard these equations as being applicable to a specific range of 1975 production
engines is somewhat unrealistic. Any turbine engine that is expected to reach pro-
duction status by 1975 must now be in the planning stage, and will need to go into
design promptly. Therefore, the only small turbines whose availability can really
be depended upon are those few for which the data were obtained; moreover, the
parametric relations derived do not necessarily "fit" these specific engines. However,
the relations are considered to be fair representations of small-turbine-engine capa-
bility, and as such are appropriate for use in the study.
Rankine Cycle Engines. The Rankine cycle engine and its associated equipment con-
stitute, a larger and heavier system than the other engine types considered, and its
4-38
-------
efficiency is considerably below that of reciprocating engines. For purposes <>!' mm-
l)ustion efficiency and smoke elimination it is operated over a narrow ran^o of :iir/
fuel ratios with approximately 10 to 20 percent excess air over that required for opti-
mum combustion with the fuel used. It has the potential of being a very low producer
of CO, HC, and NO because combustion is continuous and complete, occurring at
J*t
moderately high temperatures in a hot chamber, and because the air/fuel ratio is
higher than the stoichiometric mixture. System cooling size and weight requirements
are high because the working fluid must be condensed and recirculated. In contrast
to the production status of reciprocating engines, and the emerging production status
of automotive turbines, there is no evidence of the availability of large-scale produc-
tion plans or facilities for the Rankine cycle engine.
The status of steam or vapor engines applicable to this study was indeterminate;
therefore, letters of inquiry were sent to all of the potential data sources indicated by
Refs. 4-5, 4-6, and 4-7. The relatively small number of useful responses - 4 out of
15 might be due to the small size of most companies dealing with small vapor
engines, or to a reluctance, in the face of a potential future vapor engine market, to
share useful technical information which has been acquired by individually sponsored
research and efforts. The small response made necessary a considerable degree of
extrapolation to derive parametric relations for vapor engines.
Since both water and organic working fluid Rankine engine systems are presently under
investigation, with several proponents for each type fluid, a decision was required
regarding the fluid on which to base the study input data. It was decided to base the
inputs on a water (steam) cycle, primarily on the basis that a higher efficiency ap-
pears attainable with steam than with the organic fluids reviewed. Other considerations
were the strong historical familiarity with steam and the better availability of steam
engine data. Data are based on the use of a reciprocating expander, rather than a
steam turbine, because of its better efficiency at partial loads.
a. The relationship of cost to continuous power rating of the steam power
system involved some degree of approximation, with very little data
4-39
-------
c
available. One item of data was an auto manufacturer's estimate that a
100-hp steam power system would cost approximately 40 to 86 percent
more than the 220-hp spark-ignition engine system which it would hypo-
thetically replace. (The comparison included an automatic transmission,
drive shaft, and rear axle for each system.) Correcting the spark-ignition
engine maximum output to continuous power by an assumed factor of
75 percent, the comparison then applies between a 100-hp steam system
and a 165-hp spark-ignition system.
A steam engine manufacturer estimated that a steam engine would cost
about as much as a spark-ignition engine including semi-automatic trans-
mission. Using these values, the following estimates were made:
Spark-
Ignition
Power
' (hp)
90
165
Equivalent
Steam
Power
(hp)
54.5
100
Engine
Cost
($)
510
700
Automatic
Transmission
Cost
($)
70
165
1 fix
A \J "
Spark-
Ignition
Cost
($)
816
1120
Spark-
Ignition
Engine +
Transmission
Cost
($)
580
865
Based on the above, it was decided to estimate the steam engine cost at
approximately twice the spark-ignition engine cost at the same horsepower.
This computer plot is shown in Fig. 4-34.
b. The specific fuel consumption relation was based on assumed brake thermal
efficiency of 15 percent at the 10-hp engine size, 20 percent at the 160-hp
engine size, and a fuel heating value of 18,400 Btu/lb, with linear interpo-
lation between these points. This plot with the points used is shown in
Fig. 4-35.
c. The engine weight curve was based upon the response from one source,
which indicated a weight of 500 Ib for a 36-hp system and 950 Ib for a 95-hp
system. A straight line interpolation was assumed between these points,
and was extrapolated from them. This plot with point used is shown in
Fig. 4-36. It is noteworthy that the weight at 95 hp is approximately veri-
fied by the value of 957 Ib cited for a 100-hp system in Ref. 4-11.
4-40
-------
Ui
cr
IE
J
J
i-
ui
D
U
CO IDE)
HORSEPOWER
Fig. 4-34 Steam Engine Cost vs. Con-
tinuous Horsepower
Q.
I
\
-J*
U
L.
Jl
CO 100 XCO
HQRSEPCIWER
Fig. 4-35 Steam Engine Specific Fuel Con-
sumption vs. Continuous Horsepower
-Jg
H *
LJ
Z
SO 100 ISO
HE3R5EPC3WER
200
Fig. 4-36 Steam Engine Weight vs. Con-
tinuous Horsepower
t-
U_
is
U
V
U
£ s
D
J
E
SO 1OQ ISO
HBRSEPOWER
Fig. 4-37 Steam Engine Volume vs. Con-
tinuous Horsepower
-------
d. The displacement volume curve was based on a value of 15. l> cu t't at oti tip.
from the data of one of the responses, and a value of 8. 0 cu ft at f> hp from
another data input. A straight line was drawn through these points, and
then was generally verified by an estimated value of 34. 5 cu ft at 100 hp,
derived from another data input by proportioning. This plot with points
used is shown in Fig. 4-37.
In summary, the following relations are determined to be most applicable
to Rankine engines:
Function Ordinates Equation
Cost vs. hp Yvs. X Y = A(X)B
Specific Fuel Consumption vs. hp Yvs. X Y = A + (BX)
Weight vs. hp Y vs. X Y = A + (BX)
Volume vs. hp Y vs. X Y = A + (BX)
ENGINE COMPARISON
Plots of cost, specific fuel consumption, weight, and volume as a function of contin-
uous horsepower for all four engine types are replotted in Figs. 4-38, 4-39, 4-40,
and 4-41, respectively. The gas turbine is seen to be excessively expensive, and
the Rankine engine to have excessive volume. The Diesel and spark-ignition engines
are acceptable on the basis of all four parameters. The spark-ignition engine has
the lowest cost, and the Diesel the lowest specific fuel consumption.
The heat-engine power rating required for the hybrid configuration of each of the
four vehicles is shown in Table 4-4. These road power requirements are based on
vehicle performance requirements: for the family car, a sustained 65 mph on a
5-percent grade, and for the other vehicles, sustained maximum speed on the level.
As shown in Table 4-4, the heat-engine power required is the sum of the road power,
drivetrain losses, and accessory loads. The drivetrain losses are based on the
efficiency characteristics of the power-splitting transmission.
4-42
-------
Table 4-4
HEAT-ENGINE SIZING FOR HYBRID VEHICLES
Item
Family Car
Commuter Car
City Bus
Delivery/Postal
Van
Road Power
(hp)
87.9
22.8
101.6
32.4
Drivetrain
Losses
(hp)
9.8
5.7
25.4
8.1
Accessory Load
(hp)
5.0
3.3
32.3
0.2
Engine Power
(hp)
102.7
31.8
159.3
40.7
The resultant heat-engine power ratings for the family car and city bus were used in
conjunction with the engine trend curves to determine heat-engine weight, volume,
cost, and specific fuel consumption for these two vehicles.
Emissions for these engines were calculated by computer-running each of the engines
in both the city bus and the family car. Emission characteristics were those supplied
by EPA/APCO. For the family car, the operating profile was the DHEW Urban
Dynamometer Driving Schedule; for the city bus the profile used was the City Bus
Cycle. For both vehicles, a model of the previously selected power splitting trans-
mission was used.
At each discrete second of cycle time, the engine power was calculated as the sum of
rolling resistance, aerodynamic drag, accelerator power, transmission loss, and
accessory loads. This power level was used in conjunction with the emission curves to
determine HC, CO, and NO emissions for the 1-sec interval. The emissions for
Jv
each second were then summed to yield the total emissions which in turn was divided
by the total cycle distance to give the emissions per mile.
The results of these computer runs for the family car are shown in Table 4-5. Emis-
sion and parameter values for an average 1970 family car are included for comparison.
On the basis of minimum cost and emissions, the spark-ignition engine was chosen as
4-43
-------
CO 1OO ICO
HBRSEPEWER
zoo
Fig. 4-38 Comparison of Cost vs. Con-
tinuous Horsepower
CO 100 1KO
HDRSEPBUIETR
zoa
Fig. 4-39 Comparison of Specific Fuel Con-
sumption vs. Continuous Horsepower
i
gs
H «
U
3
G CO 1OO
HQR5EPDWER
Fig. 4-40 Comparison of Weight vs. Con-
tinuous Horsepower
t-
Lu
U
NX
LI
Z
D
J
SO 1DO ICO
HC3R5EPBNER
Fig. 4-41 Comparison of Volume vs. Con-
tinuous Horsepower
-------
/
*z
best for the family car. A comparison of these emissions with various standards is
presented in Table 4-6. Again, the results are estimates and more refined analysis
with better data including cold start effects is needed.
Table 4-5
FAMILY CAR ENGINE SELECTION
Item
Volume (cu ft)
Weight (lb)
Cost ($)
Cruise Specific Fuel
Consumption (Ib/hp-hr)
Emissions
(gin/mile)
j\
Emissions (gm/mile)
1975
Standard
0.45
4.7
0.5
Average
New
1970 Car
3.25
36.9
3.22
Flywheel/Hybrid Car
With Worst
Gasoline Engine
0.69
2.57
1.11
With Best
Gasoline Engine
0.08
1.42
0.59
(a) Effective Date: Model Year 1976.
4-45
-------
The results of computer runs to determine city bus emissions are presented in Table
4-7. Parameter values for a typical 1970 city bus arc included for comparison.
Table 4-8 shows a comparison of these emissions for the spark-ignition engine and
Diesel engine with the 1973 California standard. On the basis of minimum cost and
emissions, the engine chosen as best for the city bus is the spark-ignition engine.
Table 4-7
BUS ENGINE SELECTION
Item
Volume (cu ft)
Weight (11))
Cost ($)
Cruise Specific Fuel
Consumption (Ib/hp-hr)
Emissions
(Km/mile) (a)
nc:
CO
N0x
1970
Bus
76.5
4,688
8,138
0.38
Power Splitting Transmission
Diesel
55.8
3,743
4,318
0.404
Best
1.81
1 . 85
9.04
Worst
2.65
67.8
29.4
Turbine
31.8
2,594
9,648
0.859
Best
0.063
3. 17
12.2
Worst
0.520
34.8
21.3
Rankine
70.9
3,834
3,748
0.703
Best
1.49
12.2
2.83
Worst
2.26
18.5
4.23
Spark-Ignition
47.6
3, 181
3,238
0.480
Best
0. 540
9.27
3.84
Worst
4.f>2
1G.7
7.23
(a) Includes all accessories.
Table 4-8
CITY BUS EMISSIONS
Item
HC
CO
NOX
Emission (gm/mile)
1973
California
Standard
11
17
25
Flywheel/Hybrid Bus
Worst
Gasoline
Engine
4.52
16.7
7.23
Best
Gasoline
Engine
0.54
9.27
3.84
Worst
Diesel
Engine
2.65
67.8
29.4
Best
Diesel
Engine
1.81
1.85
9.04
4-46
-------
Section 5
FLYWHEEL DESIGN STUDIES
In the same time frame as the applicability study and configuration tradeoff study por-
tions of the program, design studies relating to flywheel technology were conducted.
These design studies were later specifically directed toward the flywheel sizes and
capacities which were established on the basis of the completed configuration tradeoffs.
FLYWHEEL GEOMETRIES
The large variety of flywheel geometries may be categorized into a few basic types.
For the purpose of the design studies, consideration was given to six basic types
a pierced uniform disc (i.e. , having constant thickness normal to the plane of rotation
and a center hole), an unpierced uniform disc, a constant-stress disc, a truncated
conical disc (having the manufacturing advantages of the uniform disc while approach-
ing the energy storage efficiency of the ideal constant-stress disc), a rim-type fly-
wheel, and a "bar" type of configuration. In the last, the material is arranged in u
prismatic or cylindrical bar rotated about an axis perpendicular to the diameter at its
midpoint.
If constraints other than those due to stress considerations are removed from periph-
eral velocity, flywheel specific energy (in in. -Ib/lb) is equal to the ratio of working
stress (in Ib/in. ) to density (in Ib/in. j times a dimensionless shape factor for the
particular flywheel geometry. Flywheel shape factors for various geometries are given
in Table 5-1.
Characteristics of the six flywheel geometries are presented in the succeeding
paragraphs.
5-1
-------
Table 5-1
FLYWHEEL SHAPE FACTORS FOR VARIOUS GEOMETRIES
Geometry
Pierced Disc (ID/OD = 0. 1)
Bar (Single Filament)
Thick Rim (ID/OD = 0.8)
Thin Rim (ID/OD = 1)
Disc (No Hole)
Truncated Conical Disc
Exponential Disc (/ = 5- 48)
Shape Factor
0. 305
0.333
0.438
0.500
0.606
0.806
0.807
PIERCED DISC FLYWHEELS
Figure 5-1 shows a computer-generated plot of the shape (half-thickness) of the
pierced disc along with the tangential and radial stresses as a function of radius.
Since the stresses at the hole edge approach twice that at the center of a solid disc,
the pierced disc makes poor use of material unless the material is built up at the
proximity of the center hole to overcome the effect of the tangential stress buildup.
RIM FLYWHEELS
The rim-type flywheel is probably the most familiar flywheel geometry. In the rim-
type flywheel, material is concentrated near the periphery. The rim material is
stressed primarily as a hoop in the tangential direction with a minimal web attach-
ing the hoop material to the center hub or shaft of the flywheel. The shape of the
rim-type flywheel is shown in the computer plot of Fig. 5-2 along with tangential and
radial stress plots. The rim flywheel appears to be an excellent candidate geometry
for filamentary composite materials such as fiberglass bonded with epoxy. These
5-2
-------
en
CO
RBHZU5
Fig. 5-1 Stress Plots for a Pierced Disc Flywheel
-------
Cn
I
a
a
H
in
a.
v a
a
in H
in
u
K
in
TRNBENTIRL
RR3XRL
H Q
v
Ul N
in
u
x a
u
H H
RR2XU5
6 B
xa
X2
Fig. 5-2 Stress Plots for a Rim Flywheel
-------
materials characteristically have high strength in the direction of the filaments ;m
-------
z
H
v
1
H
J
C
f 6 H
RRDXLJS
Fig. 5-3 Stress Plots for a Flat Disc Flywheel
-------
en
i
R R U I U 5
Fig. 5-4 Stress Plots for an Exponential Disc Flywheel
-------
CONICAL DISC FLYWHEELS
A conical disc represents an intermediate case between the flat unpierced disc and the
constant-stress disc flywheels. A computer plot of a typical conical flywheel is shown
in Fig. 5-5 along with tangential and radial stress plots. The conical disc has almost
the same energy density as the constant stress geometry, and the essential simplicity
from the standpoint of manufacture of the conical disc often makes it a more desirable
shape than the constant stress geometry.
BAR-TYPE FLYWHEELS
The bar flywheel geometry makes use of a log or bar rotated about a transverse axis
and was included in the flywheel design studies because of the suitability of this shape
for the utilization of filamentary composite materials having high-strength uniaxial
characteristics. The bar flywheel configuration is shown in Fig. 5-6.
A simple means of considering the energy density capabilities of the bar flywheel may
be obtained by comparing a single filament rotated about its midpoint with a single
filament in a theoretical rim. In the rim, the filament would be stressed uniformly
throughout its entire length. However, in the single filament bar the stress decreases
with the distance from the center of rotation of the filament, and, as seen in Table 5-1
the single filament bar has an energy density only two-thirds that of a single filament
thin-rim flywheel. Improvement in the energy density characteristics of the bar fly-
wheel can be brought about by shaping the cross-section of the bar exponentially, again
making use of the filamentary material only in its uniaxial direction. In the ultimate
case, the shaped bar-type flywheel can approach the energy density of the theoretical
rim. This shaping, however, imposes sh'ear property requirements on the bonding
matrix which may be difficult or impossible to meet.
5-8
-------
O1
I
R H I I U S
12
Fig. 5-5 Stress Plots for a Conical Disc Flywheel
-------
c
Fig. 5-6 Bar Flywheel Configuration
FLYWHEEL MATERIALS
A survey of materials that could be obtained in mill-run quantities was conducted as a
part of the flywheel design studies. Only materials with known characteristics that
could be applied to the design, fabrication, and testing of flywheels within the time-
frame of the present contract were considered. A total of 11 materials were chosen
on the basis of high strength and/or low cost, and are shown in Table 5-2. Recom-
mended working stress was derived from design studies in which the design cycle life
of the flywheel was assumed to be 10 million cycles. Material cost was based on mill-
run quantities.
In order to facilitate a quantitative comparison, a normalized cost was calculated as
follows: material cost was divided by the working-stress-to-density ratio and then
divided by the resulting value for the least expensive material. Thus, the normalized
cost represents the cost for each material to provide an equivalent energy storage
capability for a given flywheel configuration.
The first material shown is 18NI-400 maraging steel, which is the highest strength
steel currently available in mill-run quantities with an ultimate tensile stress level of
409 ksi and a yield stress of 400 ksi. The material is specified by the supplier as
having a cycle life of 10 million when used at a working stress level of 260 ksi. The
$2.25/lb cost of the maraging steel in mill-run quantities is expected to remain at
approximately this level since maraging steel is composed of relatively large portions
5-10
-------
Oi
I
Table 5-2
FLYWHEEL MATERIALS
Material
1SNT-400
(Maraging Steel)
18NI-300
(Maraging Steel)
4340 Steel
1040 Steel
1020 Steel
Cast Iron
2021-T81
(Aluminum)
2024-T851
(Aluminum)
6A1-4V
(Titanium)
E -Glass
S-Glass
Density
(P) 3
0.289
0.289
0.283
0.283
0.283
0.280
0.103
0.100
0. 160
0.075
0.072
Poisson's
Ratio
0.26
0.30
0.32
0.30
0.30
0.30
0.33
0.33
0.32
0.29
0.293
Ultimate
Tensile
(Ftu) ksi
409
307
260
87
68
55
62
66
150
200
260
Yield
Tensile
400
300
217
58
43
37
52
58
140
-
-
Working
Stress
(a)ksi
260
200
130
36
25
20
26
35
82
67
87
CT
(x io6)
0.900
0.692
0.459
0.127
0.088
0.071
0.252
0.350
0.512
0.890
1.210
Material
Cost
(S/lb)
2.25
2.25
0.60
0.30
0.30
0.30
0.53
0.50
4.00
0.42
0. 75
Normalized
Cost
(S/Ib)
5.30
6.89
2.78
5.00
7.23
8.94
4.45
3.03
16. 55
1.00
1.31
-------
of cobalt and nickel. The 18NI-300 maraging steel is quite similar to the 400 grade
except that the ultimate and yield tensile stress levels are somewhat lower. This mate-
rial is expected to have a greater fracture toughness than the 400-grade material ami
thus ultimately may be more desirable because of its greater resistance to the develop-
ment and propagation of cracks. It is conceivable that a test program would reveal that
the working stress level capability of the 300 grade would approach that of the 400 grade
for equivalent cycle life and fracture toughness.
AISI 4340-grade steel is a relatively common high-strength hot-rolled steel that has
an attractive combination of characteristics in particular, low cost and a high recom-
mended working stress level. The $0. 60/lb cost shown in Table 5-2 is for vacuum arc
remelted 4340-grade steel, which has excellent stress uniformity and a very high level
of homogeneity. AISI 1040 and 1020 steels are common steels used in a wide variety of
industrial applications. These materials have much lower working stress levels, but
at the same time much lower material cost. Cast iron also was considered as a candi-
date although its stress level is the lowest of any of the ferrous metals considered.
Two grades of relatively high-strength aluminum are shown in Table 5-2. These materi-
als have merit factors better than those of the low-grade steels and cast irons. Alumi-
num, however, by virtue of its lower density, has poorer volumetric efficiency when
applied to kinetic-energy wheels than the heavier ferrous materials. The remaining
isotropic or metallic material considered was titanium, which has a relatively high
working stress level along with a low density, but is prohibitively high in cost.
Two nonmetallic materials were considered E-glass and S-glass filamentary com-
posites. These materials are similar in chemical composition but have substantial dif-
ferences by virtue of the finishing operations used. S-glass is a premium quality glass
fiber that is well suited to filament-winding applications. Its high tensile strength makes
it ideal in applications where major considerations are high strength and low weight. It
also is used in many parallel lay-down lamination constructions. The characteristics
of E-glass are very similar to S-glass, although, as shown in Table 5-2, the recom-
mended working stress level and the material cost of the E-glass are somewhat lower.
5-12
-------
The normalized costs shown in the right-hand column of Table 5-2 indicate that the1
most cost-effective material is E-glass, followed by S-glass and then 4340-gradi' stc.ol.
The filamentary composites, however, seem to be readily applicable to only tho l>;ir-
type flywheel geometry. These filamentary composites might be used in a rim-type
flywheel, but at the present time the web attachment and balancing requirements appear
to be significant problems. The 4340-grade steel appears to be an excellent candidate;
material for application to low-cost flywheels in a disc configuration. The 400-grade
maraging steel also appears to be an attractive candidate for many applications by vir-
tue of its high working-stress-to-density ratio; and if flywheel weight is a major con-
sideration, and cost is not critical, maraging steel might be the best material. All 1]
materials were considered in the flywheel selection that followed.
COMPUTER-AIDED FLYWHEEL SCREENING
When the results of applicability studies were completed, a computer-aided screening
of various flywheels for the two selected vehicles was performed. The study results
showed that the flywheel size requirements were relatively small since in hybrid con-
figurations the flywheel is required to provide energy only for vehicle acceleration. In
the case of the family car, the required kinetic energy from the flywheel was found to
be 395 w-hr. In order to provide a reasonable engineering contingency, it was decided
to size the family car flywheel at a level of 500 w-hr. In a similar manner, the con-
figuration tradeoffs showed that the city bus flywheel capacity requirement was 604 w-hr.
Again, an engineering contingency was added, raising the total city bus flywheel capacity
to 1 kw-hr.
The screening of the two hybrid vehicle flywheel types was accomplished with the use
of a digital computer, combining all the input information previously derived for fly-
wheel shapes and materials. The six previously described flywheel geometries, in con-
junction with the 11 flywheel materials, gave a possible combination of 66 flywheel con-
figurations. However, this number of configurations was reduced by eliminating certain
combinations that would appear to be unreasonable. For example the use of filamentary
composites in disc-type flywheels was eliminated because the filamentary materials arc'
5-13
-------
not well suited for winding into discs. Similarly, the use of the more expensive iso-
tropic materials such as the maraging steels and titanium in pierced discs and rim-
type wheels also was eliminated because of the low energy density associated with
these flywheel geometries. On this basis, a total of 33 plausible combinations of the
six flywheel geometries and 11 materials were considered. In order to have the fly-
wheel design as accomplished by the computer cover a broad range of dimensional
ratios, five variations ranging from maximum reasonable radius to minimum reason-
able radius (highest speed) and associated thickness dimensions were utilized. Finally,
the two flywheel capacities, 0. 5 and 1. 0 kw-hr, were considered for the computer-
aided flywheel screening. Thus, the combination of 33 plausible geometry/materials,
along with five dimensional ratios and two capacities, gave 330 possible flywheel con-
figurations to be screened by the computer.
The computer calculated flywheel weight and dimensions along with maximum flywheel
speed. In addition, the computer calculated the flywheel assembly weight, volume,
and cost. Here, the assembly was assumed to incorporate a housing suitable for main-
taining a vacuum, bearings, and seals. Flywheel assembly cost was estimated as
three times flywheel material cost plus $3. 00/lb for the housing.
To reduce the amount of data to be reviewed, several constraints were applied to the
computer to eliminate flywheel types which for one reason or another would be unrea-
sonable for final consideration. The first of these constraints was flywheel assembly
weight. It was assumed that the flywheel assembly weight should not exceed half the
weight specified for the entire vehicle propulsion system, whether family car or bus.
The second constraint was on flywheel assembly volume, where again it was assumed
that the assembly volume should not exceed 50 percent of the total volume specified for
the propulsion system for either the car or the bus. The third set of constraints was
applied to the dimensions of the flywheel assemblies. In the case of the family car, it
was assumed that the flywheel assembly radius should not exceed 2 ft and that the
assembly height should not exceed 1 ft. These constraints were assumed to represent
maximum realistic values for fitting the flywheel assembly into a family car.
5-14
-------
^"^ '&ST/1
Similar dimensional constraints were applied for the city bus; flywheel assembly radius
would not exceed 3 ft, nor the assembly height 2 ft. A speed constraint of 24, 000 rpm
also was applied. The selection of this speed was based on the availability of relatively
inexpensive high-quality seals and bearings as a function of speed. The high availability
of seals and bearings up to about 24, 000 rpm is probably due to the common usi> of
400 H/ electrical power for the operation of two-pole electric motors as i.s common
practice in commercial and military aircraft systems. Bearings and sun Is for );irj;c
rotating masses at speeds higher than 24,000 rpm generally must be specifically
designed, are relatively high in cost, and are not readily available.
A final constraint placed upon the flywheel assembly was that of cost. The assumption
used was that for family cars or city buses a cost greater than $0. 50/w-hr of energy
stored would be unreasonably high. The constraints were then applied by the digital
computer which selected 24 acceptable candidate flywheels suitable for application to
the flywheel/heat-engine hybrid vehicles. The computer-calculated values of size,
weight, and cost for these flywheel assemblies are shown in Table 5-3.
SELECTION OF HYBRID VEHICLE FLYWHEEL DESIGNS
Of the 24 acceptable hybrid flywheels, 11 are suitable for application to the family car
with the energy storage capacity of 500 w-hr. These flywheels include pierced disc,
flat disc, conical disc, and exponential geometries, but exclude all bar and rim geome-
tries by virtue of the imposed constraints. The optimum flywheel for the family car
from the standpoint of minimum assembly cost, size, and weight is shown in Table 5-3
to be the constant-stress disc of 4340-grade steel. The cost shown represents the pro-
jected assembly cost in automotive quantities after amortization of necessary tooling.
Thus, the decision was made to fabricate two identical 46-lb, 24, 000-rpm flywheels
from 4340-grade steel in the exponential geometry, as described in Section 6.
The acceptable hybrid flywheel configurations suitable for the city bus are seen in
Table 5-3 to cover essentially the same range as those for the family car with the excep-
tion that two of the filamentary composite flywheels in the bar configuration are shown
5-15
-------
01
1
Table 5-3
HYBRID FLYWHEELS
Capacity
(kw-hr)
0.5
0.5
0.5
0.5
0.5
0.5
0.5
0.5
0.5
0.5
0.5
1.0
1.0
1.0
1.0
1.0
1.0
1.0
1.0
1.0
1.0
1.0
1.0
1.0
Material
4340
4340
1040
2021-T81
2024-T851
4340
2021-T81
2024-T851
4340
2021-T81
2024-T851
4340
4340
1040
2021-T81
2024-T851
4340
2021-T81
2024-T851
4340
2021-T81
2024-T851
E-Glass
S -Glass
Geometry
Pierced Disc
Solid Disc
Solid Disc
Solid Disc
Solid Disc
Conical
Conical
Conical
Constant -Stress
C onstant -Stress
Constant -Stress
Pierced Disc
Solid Disc
Solid Disc
Solid Disc
Solid Disc
Conical
Conical
Conical
Constant -Stress
Constant -Stress
Constant-Stress
Bar
Bar
Flywheel
Speed
(rpm)
21,749
22,821
13,453
13, 394
21,940
23,410
18,698
22,970
24, 000
24,000
24,000
17, 308
21,746
10,678
13,394
17,714
22,873
18,097
20,857
24,000
24,000
24,000
15, 132
19r 199
Flywheel
Weight
(lb)
111.7
57.6
206.6
105.0
75.8
52.0
92.1
66.2
42.4
75.2
56.62
225.9
115.1
413.1
210.1
151.5
99.5
180.2
130.8
83.8
150.3
110.0
87.6
63. 3
Assembly
Weight
(lb)
124.8
74.9
222.0
127.9
91.5
69. 1
112.3
82.5
59.5
90.9
73.52
242.0
133.7
433.1
233.0
171.7
119.3
199.5
149.8
104.8
166.0
127.9
136.5
108.9
Assembly
Volume
0.75
0.39
1.39
1.95
1.45
0.20
1.0
0.75
0.12
0.56
0.43
1.52
0.78
2.78
3.89
2.89
0.4
2.0
1.49
0.22
1.10
0.83
9.90
7.63
Assembly
Cost
(S)'a>
240.44
155.65
232.40
235.65
160.70
145. 18
201.08
148. 17
127.51
166. 50
135.63
454.85
262.86
431.52
402.60
287.88
238.26
344.30
293.22
213.66
286.00
218.56
260. 59
274. 54
(a) Assembly cost elements are described on p. 5-14.
-------
to fall within the constraints. The much greater volume that is available for thi1 fly-
wheel assembly in the city bus permits the glass bar flywheels to be acceptable Candi-
dates although the volumetric efficiencies of these flywheel configurations arc relatively
low as previously noted. Review of the 1.0 kw-hr flywheel designs of Table 0-3 again
reveals that the 4340-grade steel flywheel in an exponential configuration has the highest
cost effectiveness of any of the designs. This design then appears to be the logical
choice for fabrication and testing of flywheels suitable for hybrid city busses. However,
it was decided that little additional information would result from fabrication and test of
the 1. 0 kw-hr flywheel in essentially the same configuration and of the same material
as chosen for the 0. 5 kw-hr flywheel. Thus, the decision was made to fabricate a fly-
wheel in a very different configuration the bar flywheel of S-glass filamentary com-
posite bonded with epoxy. Although the cost effectiveness of the S-glass flywheel is
shown in Table 5-3 to be somewhat less than that for the E-glass design, the potential
of continued cost reduction in the basic S-glass material is highly probable. Thus, in
the time period of ultimate production of flywheels in this configuration, the S-glass
bar would likely be the type having the greatest cost effectiveness.
FLYWHEEL GYRODYNAMIC CONSIDERATIONS
The gyrodynamic torques which result from the rotation of a flywheel perpendicular to
its axis of rotation are simply the product of flywheel momentum and the applied angu-
lar rate. Since the flywheel kinetic energy is proportional to the product of flywheel
moment of inertia and the square of flywheel rotational speed, while momentum is
related to the moment of inertial times the rotational speed, the gyrodynamic forces
may be minimized by decreasing the flywheel moment and increasing design speed.
An analysis of gyrodynamic forces was made based on the chosen family car flywheel
design. The 0.5-kw-hr family car flywheel with a maximum operating speed of 24. 000
2
rpm has a moment of inertia of 0.42 Ib-ft/sec . In normal vehicle driving operations,
the maximum vertical acceleration that is assumed to be applied to the sprung mass
of the car is 0. 3 g. The flywheel system is assumed to be in the sprung mass. This
level is based on vehicle ride studies indicating that 0. 3g is the vertical acceleration
tolerance limit for most drivers (Ref. 5-1). Similarly, the assumption can be made
that the maximum turning velocity at which the car would be operated is one radian per
5-17
-------
second or a 0. 9 g turn at 20 mph. Table 5-4 shows the equivalent pitch, roll, or yaw
forces that would be applied to the front, side, or rear of the family car resulting from
flywheel effects during maximum assumed maneuvers. These maximum force values
arc shown in Table 5-4 for three orthogonal flywheel spin axis orientations. With the
flywheel spin axis vertical, the maximum pitch force that would be experienced due to
vertical acceleration would be equivalent to a 25-lb upward or downward force on the
hood of the car directly above the front axle. Similarly, the maximum roll force
(applied as an upward or downward force at the car door handle) resulting from vertical
acceleration is \ 43 Ib. These force levels are seen to be relatively small in contrast
to force variations on the suspension which can occur due to vehicle loading. Although
these forces are small, they may cause a different driver "feel" than conventional sys-
tems, but these effects can only be assessed properly in a prototype vehicle.
In the case of the longitudinal flywheel spin axis orientation (such as the engine flywheel
in most existing cars), Table 5-4 shows that the maximum pitch force resulting from
vehicle turning is ± 100 Ib. The equivalent yaw force (applied to the side of the front or
rear fender above the wheel) which results from the vertical acceleration is ±25 Ib.
The final case shown in Table 5-4 the transverse flywheel orientation results in a
maximum roll force due to vehicle turning of ±167 Ib, which is roughly equivalent to
the effect on the suspension of the addition of a 250-lb passenger near the car door.
Table 5-4
FLYWHEEL GYRODYNAMIC FORCES
Flywheel Spin Axis
Vertical
Longitudinal
Transverse
Maximum
Pitch Force
(Ib)
± 25
±100
Maximum
Roll Force
(Ib)
± 43
±167
Maximum
Yaw Force
(Ib)
±25
±25
5-18
-------
As an additional comparison, it can be shown that the summation of family car wheel
and tire momentum values can be approximately half that of the 0. 5 kw-hr flywheel.
Similarly, the present engine flywheel momentum is nearly one-third that of the pro-
posed hybrid flywheel. Since these momentum values have little effect on the hand! inn
characteristics of the present car, it is concluded that the flywheel gyrodynamic forces
in the family car are only a minor consideration. Further, it may be concluded that
the choice of flywheel spin axis orientation can be based primarily on vehicle packaging
considerations. Thus, gyrodynamic considerations do not appear to constrain the
vehicle designer. The effect of these forces on a vehicle operating in an environment
with a low coefficient of friction (e.g. , ice) needs to be studied further.
Precession forces also tend to reduce flywheel bearing life; in fact, the largest bearing
loads are those due to precession. Assuming a bearing spacing of 6 in. , the peak
precession forces on the bearings are approximately 2, 000 Ib for maneuvers previously
described.
Similar conclusions regarding gyrodynamic forces can be drawn for the hybrid city bus
flywheel. The larger vehicle dimensions and heavier suspension of the bus serve to
minimize the effect of the increased forces.
FLYWHEEL SAFETY
The hazard of an uncontained flywheel failure is so great that it must be positively pre-
vented. Among the possible causes of flywheel failure are the following:
Material defect Vacuum failure
Accidental overspeed Vehicle accidents
Bearing failure
Generally, there are three modes of flywheel failure: (1) elastic/plastic deformation,
which is the usual result of overspeed or excessive heating of a sound flywheel; (2) dis-
integration, which can result from fatigue stress or material defects; and (3) disloca-
tion, which can result from the failure of both bearings as might occur with a vehicle
accident. There are two basic approaches to flywheel failure control: (1) avoid failure
by conservative design and inspection, and (2) shield so that a flywheel failure can do
no damage.
5-19
-------
c
CONSERVATIVE DESIGN AND INSPECTION
The most common current practice for minimizing the hazards of flywheel and rotor
failures is to provide adequate design safety factors in combination with quality control
of materials, processes, and finished parts. This approach is commonly used in auto-
motive, aircraft, and industrial systems. In the case of the automotive engine flywheel
failure, the bell housing is incapable of containing flywheel disintegration; however,
material design margins assure a reasonable degree of safety against failure. The con-
siderable technical effort that has been directed toward turbine rotor safety is applica-
ble to flywheels because of the similarity in design considerations and geometries.
In addition to being designed so as to avoid failure, flywheels may also be designed so
as to fail in a nonhazardous manner. For instance, turbine rotors commonly employ
a necked section just inside the buckets so that bucket failure will not cause failure of
the rest of the rotor. Circumferential grooves may be cut in homogeneous disc fly-
wheels so that crack propagation is inhibited and overspeed will result in bands of
material being thrown off in succession from the outside inward. Filamentary com-
posite thick-rim flywheels may be expected to exhibit this type failure. Computer
programs are available for designing flywheel shapes so that stresses can be arranged
for graceful failure, which can be contained by a relatively light housing.
Inspection requirements for flywheels may be classified into production inspection,
periodic inspection, and continuous in-situ inspection. Applicable nondestructive pro-
duction test techniques include x-ray, ultrasonic, magnaflux techniques, and overspeed
testing. Unshielded flywheels should receive overspeed testing and ultrasonic (or
x-ray) testing on a 100-percent sample basis. Periodic testing is best accomplished
by overspeed testing, and could be conducted during periodic vehicle inspections. In-
situ inspection can be accomplished with noncontact probes that sense elastic or plastic
deformation, and possibly by detection of the acoustic signature that has been found to
accompany incipient crack propagation (Ref. 5-2).
5-20
-------
/T
SHIELDING
The second basic approach to flywheel failure control is shielding. This approach is
necessary if accidental overspeed cannot be positively prevented. Annular shields can
be used to completely contain the flywheel in case of a failure. For the type of fly-
wheels involved in vehicle propulsion, however, the weight of the annular shield may
be several times the weight of the flywheel itself. This technique may also incorpo-
rate inertia bands - annular rings that absorb much of the flywheel energy in coming
up to rotational speed as the impacting flywheel or flywheel pieces slow down. Another
possibility is the use of lightly stressed peripheral bands integral with the flywheel.
Other techniques that are currently used for rotor burst protection, but which do not
appear attractive for this application, are: target shields, where only personnel and
critical equipment are shielded; deflectors, which deflect flywheel pieces away from
personnel and critical equipment; and plane orientation, where personnel and critical
equipment are kept out of the plane of flywheel rotation.
The decision was made to design the flywheels for this program on the basis that no
containment would be provided. This approach is in keeping with current automotive
practice. Because of the greater energy stored, flywheel containment by lightweight,
low cost methods is certainly an area that deserves future study.
FLYWHEEL ACCESSORIES
The environmental requirements of a flywheel applied to an urban vehicle propulsion
system were reviewed as a part of the design studies. The two flywheels considered
for the hybrid car and city bus have supersonic peripheral operating speeds. This
leads to the requirement for flywheel operation in a low-pressure environment with
the conccmittant need for a housing, an evacuation system, and rotating seals. In
addition, the flywheel must be supported on a bearing system suitable for the speed
range and capable of sustaining the shock, vibration, and gyrodynamic loads resulting
from vehicle operation.
5-21
-------
FLYWHEEL HOUSING
High-energy-density flywheels must be operated in vacuum so that the pumping losses
will not be excessive. An empirical relationship, which shows the magnitude of
windage power losses, was used to make gross determinations. This equation (sup-
plied by EPA/APCO) is based on the analyses of Daily and Nece (Ref. 5-3) as well as
Mann and Marston (Ref. 5-4). The steady-state windage loss calculation is for an
unshrouded disc rotating in a noncompressible gas in the subsonic region. Limited
test data acquired in the present program have shown the equation to be reasonably
accurate (perhaps *10 percent) for nonshrouded flywheel operation in the supersonic
region. The losses with an optimized shroud may be reduced substantially.
The equation is as follows:
k 0. 8 / XT \ 2. 8 , > A -02
10
where
hp = steady-state windage hp
t = tip thickness in in.
R = outer radius in in.
P = air pressure in psia
T = temperature in 8R
N = rotational speed in rpm
H = air viscosity in Ib/hr-ft
A plot of windage loss as a function of pressure for the 0. 5-kw-hr family car flywheel
was made based on the following flywheel parameters:
N = 24,000 rpm
R = 10.22 in.
5-22
-------
t = 0.576 in.
H = 0. 0431 Ib/hr-ft (at 60° F)
T = r>19.4°R
This plot, shown in Fig. 5-7, indicates that a vacuum level of 0.01 atmosphere (7.G
mm hg) or better is desirable if pumping losses are to be held to 2. 5 hp. Even with
the use of an optimized shroud, this vacuum level is required if windage losses above
1 hp are unacceptable.
The housing to hold the flywheel bearings and seals must then be capable of withstand-
ing atmospheric pressure. Such a housing is most simply made from a steel or alumi-
num casting in two halves that can be line-bored for the location of bearings and seals.
The ultimate housing design should give full consideration to the tradeoffs involved
with material cost, fabrication cost, size, weight, internal (shrouding) contour, bear-
ing and seal replacement, and mounting arrangement with respect to the vehicle.
FLYWHEEL BEARINGS AND SEALS
The constraint placed on flywheel maximum operating speed (24,000 rpm) in the
computer-aided flywheel design was imposed so that the selection of bearings and seals
could be made from a relatively wide range of commercially available items.
A review of bearing requirements has shown that angular-contact ball bearings are
well suited to the hybrid flywheel application for either the bus or car flywheels. The
flywheel as mounted on the bearing system is expected to be always rotating in the
subcritical mode. Thus, the effective center of mass of the flywheel will rotate in a
small circle as permitted by the radial clearance in the bearings. The diameter of
this circle of rotation must be held to a minimum in order to prevent excessive radial
forces on the bearings. The recommended bearing systems for the flywheels should
make provision for preloading the bearings to maintain the proper radial clearance.
The angular-contact bearings may be used in either single or tandem configurations
5-23
-------
Ol
I
CO
D ID 20
PRESSURE
Fig. 5-7 Windage Loss vs. Pressure for 0. 5 kw-hr Family Car Flywheel
3D
-------
depending on permissible flywheel unbalance, bearing quality, and life considerations.
A lubrication system supplying clean oil or oil/air mist in preference to grease should
be provided for the bearings, because of high rotating speeds.
Several rotating seal types suitable for the hybrid flywheel application are commer-
cially available. Since there is no need for a "hard" vacuum, the more exotic aero-
space types were not considered. An inexpensive rubbing seal type which appears to
be well-suited to the application is the spring-loaded lip seal using low friction carbon-
impregnated Teflon. Similarly, the spring-loaded carbon face seal types commonly
used in automotive air conditioning compressors appear to be acceptable.
EVACUATION SYSTEM
A relatively simple vacuum system can provide the 0.01 atmosphere evacuation of the
flywheel housing. The volume of air to be removed during evacuation is quite small
since the hybrid vehicle flywheel housing will be closely contoured to the flywheel.
One obvious means of doing most of the evacuation of the housing is by the use of the
intake manifold of a reciprocating engine connected through a check valve to the housing.
Tests have shown vacuum levels of about 6 in. of Hg or 0.2 atmosphere to be available
by this means. The additional evacuation required to reach the 0. 01-atmosphere
level can be accomplished by several types of simple, inexpensive vacuum pumps. A
small flywheel-driven or engine-driven pump could be used. Methods of providing the
low-pressure environment is an area that requires further study.
5-25
-------
6rX(f
Section 6
FLYWHEEL DETAIL DESIGN, FABRICATION, AND TESTING
The Flywheel Design Studies (Section 5) resulted in selecting an AISI 4340 steel flywheel
in the exponential geometry for the family car and an S-glass bar-type flywheel for the
city bus. Detailed designs were made for the two types of flywheels; two flywheels of
each type were fabricated and tested.
DETAILED DESIGN OF FAMILY CAR FLYWHEEL
The design study recommended a flywheel fabricated of AISI 4340 steel and indicated
that a 46-lb weight could be obtained. This weight was based on the configuration
shown in Fig. 5-4.
Fatigue data show that AISI 4340 heat treated to 260 ksi can tolerate a static load of
150 ksi, with an imposed sinusoidal load of 27-ksi amplitude for 10-million cycles, with
a stress concentration factor of 2.0 applied. This stress concentration factor is more
than adequate to account for the flywheel-to-hub transition.
Using such data, it was decided to use a heat treat of 270 ksi and a working stress of
130 ksi.
The following constraints were put upon the design:
(a) Rotational speed (a;) - 24, 000 rpm (max.)
(b) Working stress - 130 ksi (max.)
(c) Diameter 24 in. (max.)
(d) Thickness of the disk > 0.25 in. at any point
The configuration in Fig. 5-4 was used as the initial design. The early dropoff of
both the radial and tangential stress made it apparent that mere mass could be added
6-1
-------
to the periphery of the wheel, increasing the stresses throughout most of the flywheel.
By increasing the stresses toward the rim, the overall efficiency could be improved.
A small rim was added, approximately 0.5 in. deep at the rim and tapering back into
the flywheel. Coordinates of the wheel geometry then were put into the data log of the
computer, and the stresses were calculated along with a shape factor, which was a
measure of the overall efficiency of the configuration. Coordinates then were changed
slightly, and a second stress analysis and shape factor were computed. An optimum
design was obtained by iterating the design until the shape factor reached a maximum.
When the optimum design was plotted and examined, it was noted that the configuration
was nearly conical from the start of the hub and out to a radius of 7.25 in. For ease
in manufacturing, it was decided to let this part of the configuration become conical.
The new configuration is thus a hub section, a conical section extending out to 7.25-in.
radius, and a constant radius flare. The major diameter of the flywheel is 20.446 in.
with a maximum disk thickness of 0. 957 in. at the center and a rim thickness of
0.576 in. Figure 6-1 shows this configuration and a plot of its tangential and radial
stresses. The falloff in stress near the hub was deemed desirable in order to offset
the stress concentration caused by the shaft. The shape factor was increased to 0.832.
Appendix B presents the computer program used for the stress analysis.
FABRICATION OF FAMILY CAR FLYWHEEL
The steel flywheels were fabricated from forged billets of 300M VAR steel - a modified
AISI E4340 chromium nickel molybdenum alloy. It is modified by adding vanadium and
a higher silicon content to produce high strength levels without the necessity of increas-
ing the carbon content. A vacuum remelt steel was used; this eliminates the major
occlusions in the material. Mill run cost of 300M VAR is about $0. 60/lb.
After rough forging to produce a radial flow of the disk material, the billets were rough
machined and heat treated. This steel will fully harden to a depth in excess of 4 in. ;
thus, no problems were encountered in achieving full hardening. Samples of the material
were heat treated at the same time for material testing. Heat treatment specified was
6-2
-------
os
co
H
in
a.
in
in
u
tr
i-
in
a
a
a
a
H
TRNBENTIRL
RFOIRL
RRJ1IUS
-------
260-ksi minimum tensile. All samples were in excess of 290-ksi ultimate tensile with
a uniform yield stress of 245 ksi. After heat treat, the billets were tested ultrasonical-
ly (Specification MIL-I-8950, Class A) to verify that no occlusions remained in the
material.
Final machining was done with a pattern and tracer. While being machined, the fly-
wheel was supported in a chuck by its major diameter; then, the flywheel was turned
over, and the other side was machined. This method provided the needed support to
the thin outer section just inside the rim.
After final machining, the flywheels were magnaflux inspected to verify that no cracks
had occurred during fabrication. Next, the flywheels were dimensionally checked to
see that they were within tolerance. The finished flywheels then were X-rayed, and
the films were examined for any visible flaws.
Finally, the flywheels were balanced by grinding material from the rim. The dynamic
balance achieved brought the spin axis and center of mass to within 0. 0002 in. of each
other. Figure 6-2 shows the finished flywheel.
Fig. 6-2 Steel Flywheel - 0.5 kw-hr
6-4
-------
TESTING OF FAMILY CAR FLYWHEEL
NONDESTRUCTIVE TESTS
Nondestructive tests were conducted on the steel flywheels prior to spin testing.
These tests are discussed in the section on fabrication.
TENSILE SPECIMEN TESTS
Tensile tests were conducted on specimens fabricated from the same billets that were
used to make the steel flywheels. These material tests showed that the steel had un
ultimate tensile strength of 290 ksi minimum.
FLYWHEEL NO. 1 - SPIN TESTING
All spin testing was conducted in the LMSC high-speed spin chamber; this is a contain-
ment structure of heavy steel plates. The chamber is in a pit below ground level and
is sealed to support a vacuum. An air turbine provides spin power, and a disk brake
provides high torque levels for extracting power. Torque is measured with a hollow-
ring-type torque transducer, and speed is measured with a magnetic pickup and a
60-tooth gear. An automatic digital data reading system was used.
The first steel flywheel was tested to verify pumping losses and energy density (dis-
integration speed).
Pumping Losses Test. Tests were conducted at various levels of pit vacuum to deter -
mine the pumping losses of the flywheel. No actual torque measurements were made;
driving torque was estimated from the power output curves of the air turbine at various
speeds and inlet line air pressure. Table 6-1 compares the power losses calculated
from driving torque and speed with the losses calculated by using the EPA/APCO-
supplied equation (see page 5-22).
6-5
-------
c:
Table 6-1
COMPARISON OF POWER LOSS CALCULATIONS
Flywheel Speed
(rpm)
14,110
12,530
Vacuum.
(psia)
8.35
14.7
Loss (From
Turbine
Characteristics)
14
13
Loss (Calculated
From EPA/APCO
Equation) (hp)
15.3
17.3
Disintegration Test. This test was to determine the speed which would result in fly-
wheel disintegration and thereby would verify the flywheel stress safety factor.
For this test, the flywheel was suspended below the turbine. Two USS Tl steel inertia
rings with inside diameters of 24 in. and outside diameters of 35.5 in. were placed
between the flywheel and the inside wall of the pit. The rings were 6 and 4 in. in depth,
respectively. The inertia rings were to absorb rotational energy of the flywheel at dis-
integration and protect the inside wall of the pit.
The flywheel was slowly accelerated until disintegration occurred at 35,590 rpm. This
represents a peripheral velocity of 3,169.72 ft/sec, an energy density of 26.091 w-hr/lb,
and a total energy stored of 1. 096 kw-hr.
Failure occurred at a calculated stress of 285,875 psi. This stress compares favorably
with an average ultimate tensile stress of 293 ksi measured on four test specimens
made from the flywheel material. This verifies that the design analysis is within the
accuracy of the measured tensile values of the material.
The flywheel failure may have been initiated at an occlusion on the surface shown in
Figs. 6-3 and 6-4.
6-6
-------
.' Fig. 6-3 Close-up of Occlusion in Failed Steel Flywheel
-
Fig. 6-4 Failed Steel Flywheel
6-7
-------
c
As shown in Figs. 6-4 and 6-5, a large section of the flywheel remained in one piece.
The rim and the thin outside section of the flywheel were fragmented by secondary
failure.
The effect on the pit was that the inertia rings did not rotate and had been gouged to a
depth of approximately 3/4 in. in the worst section.
Fig. 6-5 Pit After Disintegration of No. 1 Steel Flywheel
FLYWHEEL NO. 2 - SPIN TESTING
Power density, spindown, and accoustic noise tests were conducted on the No. 2 steel
flywheel. During the course of testing the flywheel was cycled over its full speed range
36 times.
Power Density Tests. Power density measurements were made to verify that at least
5,000 w/lb of usable power could be extracted from the flywheel, at various speeds
(Table 6-2).
6-8
-------
Table 6-2
RESULTS OF POWER DENSITY TEST, NO. 2 STEEL PLYWHEK1,
For: 40-lb flywheel I = 0.4203 lb-ft-sec2
Average kw = 0.142(10) T N
^ ' ' avir. ave.
Average kw =
Date
3/4
3/4
3/5
3/5
3/5
3/5
3/5
3/5
Run
No.
1
2
1
2
3
4
5
6
Speed
Increment
(rpm)
4, 570 to 1,850
3, 470 to 400
10, 130 to 9,980
9, 980 to 9,370
9, 370 to 8,590
8. 590 to 7,750
7 ,750 to 7.420
11,780 to 11.090
11, 090 to 9,620
9, 620 to 7,890
7, 890 to 7,700
16,530 to 15,520
15,520 to 13,460
13,960 to 13,350
13,350 to 13.260
23,730 to 22,850
22, 850 to 21,860
21,860 to 21,520
21,520 to 21,400
12,010 to 11,200
11, 200 to 9,680
9, 680 to 8,070
8, 070 to 6,720
6, 720 to 6,560
12,840 to 11,400
11, 400 to 10,440
Average
Speed
(rpm)
3,210
1,935
10,055
9,675
8,980
8.170
7,585
11,435
10,355
8,755
7,795
16,025
14,740
13,655
13,305
23,290
22,355
21,685
21,460
11,605
10,440
8,875
7,375
6,640
12,120
10,920
AKE (3600)
interval
Average
Torque
(Ib-ft)
69.1
89.2
6.9
34.9
52.9
61.2
50.3
32.0
88.8
119.1
67.0
28.4
106.9
108.8
16.4
23.3
70.7
20.9
8.7
42.5
114.5
121.8
122.2
23.4
88.3
76.0
Average
kw Based
on Speed
and Torque
31.5
24.5
9.85
47.96
67.47
71.02
54.19
51.97
130.61
148.11
74.18
64.64
223.82
211.02
30.99
73.77
224.50
64.38
26.52
70.56
169.79
153.54
128.36
22.07
152.01
117.88
w/lb
Based on
Speed and
Torque
685
533
215
1,043
L.466
1,544
1,178
1.130
2,839
3,219
1,612
1,405
4,866
4,587
674
1,603
4,880
1,400
576
1,523
3,691
3,338
2,790
479
3,304
2,563
Initial
KE
(w-hr)
18.13
10.45
89.1
86.45
76.21
64.05
52.13
120.45
106.75
80.33
54.03
237.17
209.08
167.71
154.70
488.78
453 . 20
414.78
401.98
127.08
108.88
81.33
56.53
39.20
143.10
112.81
Final
KE
(w-hr)
2.98
0.14
86.45
76.21
64.05
52.13
47.79
10G.75
80.33
54.03
51.46
209.08
167.71
154.70
152.62
453.20
414.78
401.98
397.51
108.88
81.33
56.53
39.20
37.35
112.81
94.61
Average
kw Based
on KE
Change
54.54
37.12
15.17
62.48
74.20
72.73
2G.48
72.56
139.93
139.29
13.61
171.40
252.43
79.38
12.69
217.10
234.43
78.10
27.27
111.05
168.10
151.32
.105.74
11.29
184.82
111.09
w/lb
Based
on KE
Change
1, 186
807
330
1 , 359
1,613
1,581
r,7Ci
1 , r,77
:MH::
II.IKJH
2%
M,72G
5,487
1,720
27(5
4,717
5,099
1,698
593
2,414
H,G54
:i,290
2,299
240
4,017
2,414
6-9
-------
The test objective of 5, 000 w/lb was exceeded at average speeds of 14,740 rpm and
23,355 rpm on runs 3 and 4, respectively.
Spindown Tests. Spindown tests were conducted by taking the flywheel to 24,000 rpm
and uncoupling it from the turbine. The pit was evacuated to 0.2-mm Hg for this test.
Rotational speed was measured periodically. After 20 min. of spindown, the pit was
vented to the atmosphere. The bearing and the brake disk were the major sources of
drag - approximately 3 hp between 24,000 and 20,000 rpm. Bearing drag was high
because of oil-flooding of the bearings. The 6-in.-diam. brake disk was running at
atmospheric pressure. Figure 6-6 provides a time history of the No. 2 steel flywheel
spindown.
Acoustic Test. Prior to the start of the spindown test, a background noise recording
was made. All equipment that would not be operating during spindown was turned off.
The background noise level was recorded on a General Radio 1525A Data Recorder.
The type 1560 P40K preamplifier and microphone was placed 10 ft from the spin pit.
System calibration was accomplished with a General Radio Sound Level Calibrator,
type 1562A. The noise level was recorded during the entire spindown test. At the
conclusion of the test, a post-test calibration was made. Reference sound pressure
level (SPL) is 20 micronewtons per square meter.
The data were reduced by playing the tapes back through a Spectral Dynamics Model
SD 301 Real Time Analyzer and recording the data on a Mosley X-Y plotter. Figures
6-7 through 6-13 show samples of the reduced data, including sound level and frequency
calibration; the only significant noise, above background level, is at the rotational
frequency. It should be noted that the maximum sound pressure level (SPL) measured
throughout the test was 84.2 dB at 19,500 rpm, at 19,000 rpm the maximum level was
70.2 dB, and at 18,600 rpm the SPL had again increased to 82 dB. From 18,600 rpm,
the SPL continually decreased until at 9,800 rpm it was down to 71 dB.
Throughout the test, there was a high SPL recorded at approximately 700 cycles.
Because there was no frequency shift of this noise, it was attributed to background
machinery, such as a pump, and should not be considered in the noise level of the
flywheel.
6-10
-------
c
Obviously, the noise to be expected from a flywheel/hybrid vehicle cannot be extrapolated
from these results since the flywheel housing plays an important role in the flywheel
assembly acoustical signature.
20
2 15
x
I
Sf,
a
< 10
2
22MMHG
PIT VENTED
760 MM HG
01 2 3 4 5 6 7 & 9 10 11 12 13 14 15 16 17
TIME (SEC x 1C"2)
Fig. 6-6 Spindown Test, EPA/APCO No. 2 Steel Flywheel
6-11
-------
c
f CALIBRATION
Z-1U dBat 1,000 Hz I
100
FREQUENCY (Hi)
1,000
Fig. 6-7 No. 2 Steel Flywheel Acoustic Noise Test Calibration
100
1,000
FREQUENCY (Hz)
Fig. 6-8 No. 2 Steel Flywheel Acoustic Noise Test Background Level
6-12
-------
c
100
1,000
FREQUENCY (Hx)
Fig. 6-9 No. 2 Steel Flywheel Acoustic Noise Test- 22, 800 rpm
f
_j
>
a
50
40
1,000
FREQUENCY (Hi)
Fig. 6-10 No. 2 Steel Flywheel Acoustic Noise Test-19,500 rpm
6-13
-------
c
100
1,000
FREQUENCY (Hi)
Fig. 6-11 No. 2 Steel Flywheel Acoustic Noise Test-19,000 rpm
1,000
FREQUENCY (Hz)
Fig. 6-12 No. 2 Steel Flywheel Acoustic Noise Test-18,600 rpm
6-14
-------
FREQUENCY (Hz)
1,000
Fig. 6-13 No. 2 Steel Flywheel Acoustic Noise Test - 9,800 rpm
DETAILED DESIGN OF CITY BUS FLYWHEEL
BODY TAPER DESIGN
In a bar-type flywheel, the major stress is longitudinal, i. e., in a direction along the
filaments. This longitudinal stress must be carried by the filaments themselves. How-
ever, there also is a transverse stress normal to the direction of the filaments. This
transverse stress is much lower than the longitudinal stress if the length of the bar is
much greater than the width. This transverse stress, as low as it is, can be very
important because it must be handled by the epoxy matrix which has a much lower
tensile strength than the glass fibers. At the center of the bar the transverse contrac-
tion strain caused by longitudinal stress is greater than the lateral elongation caused
by lateral stress. On the basis of a strain failure theory, there should be no problem
here. However, as the end of the bar is approached, the longitudinal stress goes to
zero; thus, so does the lateral contraction resulting from longitudinal stress. The
lateral expansion, however, remains undiminished; consequently, there is a danger of
6-15
-------
splitting or fraying the ends. This splitting would be progressive and would result in a
longitudinal delamination of the flywheel. For this reason, the ends of the flywheel were
provided with special tapers designed to maintain the lateral strain compressive through-
out the length of the flywheel. Although the taper (Fig. 6-14) is required only in the
lateral direction, the compound lateral-axial taper was used for ease of fabrication.
Fig. 6-14 Bar-Type Flywheel
The second bar-type flywheel was identical to the first except that a cross-ply construc-
tion was employed to provide additional protection against delamination resulting from
lateral stresses. Table 6-3 shows the ratio of longitudinal to transverse interleaves.
Table 6-3
RATIO, LONGITUDINAL TO TRANSVERSE INTERLEAVES
Axial Distance From Centerline
(in.)
Longitudinal/Transverse
Interleave Ratio
0 to 1.6
1.6 to 2.2
2.2 to 2.6
2.6 to 3.385
3:1
5:1
10:1
1:0
6-16
-------
HUB DESIGN
The attachment of the flywheel body to the hub is a major design problem of the bar-
type flywheel. At speed, there is a strain incompatibility between the body and the hub;
the longitudinal strain in the bar is much greater than that in the hub. Therefore, the
attachment must permit relative longitudinal displacement between the body and the
hub. However, to maintain balance, there must be very little displacement between
the center of gravity of the body and the intended rotating axis of the hub.
Figure 6-16 shows the two pieces of the hub prior to assembly. The slotted sleeve was
welded into the hub. This hub assembly then was slipped over the body of the flywheel.
The interior surfaces of the hub were coated with mold release, and the flywheel body
was prepared for bonding. A potting compound was drawn up the slots and cured. The
resulting potted fingers on the body lock with the hub slots to prevent relative rotation
or translation, but because the hub surfaces were coated with mold release, relative
radial growth caused by centrifugal stress is uninhibited. Translation normal to the
longitudinal axis also is prevented because the walls of the hub slots are parallel.
Fig. 6-15 Bar Flywheel Hub Prior to Assembly
6-17
-------
c
FABRICATION OF CITY BUS FLYWHEEL
Two glass bar flywheels were fabricated by using the same configuration but different
layup procedures. Both bar flywheels were constructed of Ferro Corporation S-1014
glass fiber with 828/1031/NMA resin. Resin content was 22 percent, by volume.
Prepreg material was laminated to form ten sections approximately 1-in. thick, 36-in.
long, and 8-in. wide. These sections were placed in an autoclave for curing, laminated
together to form billets 8 in. x 8 in. and 36-in. long, and the entire billet was cured.
The final configuration was ground from the billet (Fig. 6-16).
Fig. 6-16 Grinding the Glass Bar Flywheel
When the flywheel grinding was completed, the hub was bonded to the bar. The fly-
wheels then were balanced dynamically to 0.3 in-oz, by grinding material off the
tapered ends.
6-18
-------
TESTING OF CITY BUS FLYWHEEL
FATIGUE PROPERTIES
Fatigue of uniaxially oriented glass fibers in an epoxy matrix is difficult to test; there-
fore, little data are available. Because of the very high strength of the material (ap-
proximately 280 ksi) and the need to distribute the load evenly into the test specimen.
it is difficult to fabricate specimens that can tolerate the dynamic loads required for
fatigue testing.
Samples were fabricated and tested in a Sontag Fatigue test machine, but in each case
the sample failed because of inadequate bonding of the holding tabs. This test series
was discontinued after failure of several specimens.
HITCO, a Division of Armco Steel Corporation, provided information which showed
that the ratio of fatigue strength (Ff) to the tensile ultimate (F, ) for fiberglass
materials is approximately 0. 33 at 10-million cycles. With an F. = 260 ksi, a
working stress of 87 ksi can be used.
TENSILE SPECIMEN TESTS
Tensile tests were conducted on specimens fabricated from the same material batching
used for the fiberglass wheels. The fiberglass showed an average tensile ultimate of
265 ksi, which was above the allowable required to meet the fatigue requirements.
CREEP TESTS
Creep tests were conducted on fiberglass coupons. Loadings for the creep tests were
125 and 200 ksi. The tests were conducted in an ambient environment. The approximate
temperature during the test was 82° F with a relative humidity of 37 percent. Tests
were conducted for 30 days. An average creep of 0.003 in. was recorded on all coupons.
No creep was experienced after 15 days under load. Some of the measured creep could
have been in the attachment of the fiberglass coupons to the metal loading tabs. The fact
that no creep was measured during the last 15 days of loading indicates that creep
should not cause problems.
6-19
-------
x
c
NO. 1 GLASS FLYWHEEL TESTS
No. 1 glass flywheel, which was of unidirectional construction, was tested to demon-
strate energy density, power density, spindown, and disintegration speed.
The flywheel design speed was 20,000 rpm, but that speed was not reached prior to
disintegration.
Table 6-4 provides data from the energy and power density tests.
Spindown. A spindown test was conducted on the flywheel. The spin pit could not be
evacuated below 5-mm Hg, so significant drag was experienced (Fig. 6-17).
14
13
2 12
5 11
oe
Q 10
UJ
UJ
O-
-PIT PRESSURE 5-MM HG
TIME (SEC a 10"2)
Fig. 6-17 Spindown Tests, EPA/APCO No. 1 Glass Flywheel
Disintegration Speed Test. The No. 1 glass flywheel disintegrated at 15,070 rpm,
while an attempt was being made to bring the flywheel to the design speed of 20, 000 rpm.
As shown in Figs. 6-18 and 6-19, the flywheel and hub were completely destroyed.
6-20
-------
Table 6-4
NO. 1 S-GLASS FLYWHEEL TEST DATA
o>
to
Energy Density
Test rpra
Run 1 13,710
Run 2 Prior to Burst 15,070
w-hr w-hr/lb
469.91 6.10
567.76 7.37
Po wer Density
Speed
Increment
(rpm)
5,030 to 4,990
4, 990 to 4, 930
4,930 to 4,850
4,850 to 4,750
4, 750 to 4, 630
4,630 to 4,540
4,440 to 4,310
4, 310 to 4,120
4,120 to 3,880
3,880 to 3,750
3,660 to 3,530
3,530 to 3,280
3,280 to 2,910
2,910 to 2,520
2,520 to 2,120
2,120 to 2,000
Average
Speed
(rpm)
5,010
4,960
4,890
4,800
4,690
4,595
4,375
4,215
4,000
3,815
3,595
3,405
3,095
2,715
2,320
2,060
Average
Torque
(Ib-ft)
4.6
12.5
15.3
19.1
24.0
22.7
3.6
31.1
43.7
48.7
12.4
39.7
66.7
78.1
89.6
78.5
Average
kw Baaed
on rN
5.88
15.82
19.096
23.4
28.7
26.6
4.02
33.5
44.6
47.4
11.4
34.5
52.7
54.1
53.1
41.3
w/lb
Based
on rN
76.4
205.5
248.0
304.0
373.0
346.0
52.2
435.0
579.0
616.0
148,0
448.0
684.0
703.0
689.0
536.0
Initial
KE
63.25
62.25
60.76
58.81
56.41
53.59
49.28
46.44
42.43
37.63
33.50
31.15
26.90
21.17
15.88
11.24
Final
KE
62.25
60.76
58.80
56.41
53.59
51.53
46.44
42.43
37.63
35.16
31.15
26.90
21.17
15.88
11.24
10.00
Average kw
Based on
KE Change
6.1
9.1
11.9
14.64
17.2
12.6
17.3
24.4
29.3
15.1
14.25
25.9
34.9
32.3
28.3
7.94
w/lb
Based on
KE Change
79.24
118.0
155.0
190.0
223.0
163.0
225.0
317.0
380.0
196.0
185.0
337.0
454.0
419.0
368.0
98.0
-------
Fig. 6-18 No. 1 Glass Flywheel After Disintegration
Fig. 6-19 No. i Glass Flywheel Hub and Bearing Support After Flywheel Disintegration
6-22
-------
The massive disintegration that occurred after failure makes it difficult to determine
the failure mode. The break in the hub has the appearance of a tensile failure, thus
indicating that a high transverse force was involved. Such a force might be attributed
to transverse delamination or a secondary failure of some sort. At burst, the energy
density was 7. 37 w-hr/lb, and the total energy was 567. 8 w-hr.
NO. 2 GLASS FLYWHEEL SPIN TEST
During the energy density test of this flywheel, disintegration occured at 14,690 rpm.
As shown in Figs. 6-20 and 6-21, the flywheel and hub were again completely destroyed
although the added transverse fibers did serve to hold most of the longitudinal layers
together. Again, the hub has the appearance of having been torn apart in tension.
Fig. 6-20 No. 2 Glass Flywheel After Disintegration
6-23
-------
Fig. 6-21 Major Pieces of No. 2 Glass Flywheel After Disintegration
6-24
-------
The longitudinal layers showed heavy damage, such as might have resulted from a
transverse blow, on one end and very little damage on the other end.
On the basis of the above evidence, there is an indication that failure may have oc-
curred by the fiberglass moving longitudinally out of the hub. The severe damage to
one end of the fiberglass might be attributed to that end hitting the pit wall. (The
running clearance between the fiberglass and the pit wall was approximately 1 in.)
The tensile failure of the hub may be attributed to the large transverse force applied
to the hub when the fiberglass impacted the pit wall.
The possibility of the fiberglass moving longitudinally out of the hub may be attributed
to insufficient lubrication provided in the assembly procedure. As explained on p.
6-17, radial displacement between hub and fiberglass was to have been made possible
by the lubricating properties of the mold release. There is, however, a large initial
breakaway shear stress involved in mold release lubrication. Thus, the lubrication
between the potted fingers on the fiberglass and the hub may have been inhibited by a
very large value of initial (one-time-only) stiction. At the moderately high stresses
at failure speed, one side of the hub may have overcome this stiction and the resulting
sudden longitudinal unbalance may have pulled the fiberglass out of the hub.
If further work with flywheels of this type is desired, any hub/fiberglass interface
problems could probably be solved with better lubrication, possibly by teflon-coating
the hub prior to assembly.
6-25
-------
Section 7
HYBRID DESIGN CONFIGURATION
Results of the applicability studies and configuration tradeoffs have shown that the best
flywheel applications are flywheel/heat-engine configurations for city busses and
family cars. The configuration tradeoffs also have shown that for both vehicles the
optimum transmission is the power-splitting hydrostatic transmission, and the opti-
mum heat engine is the spark-ignition engine. In addition, these studies have shown
that no major changes would be required in either the city bus or the family car to
accommodate these flywheel/hybrid drives. Several drivetrain layouts were made in
order to show a number of ways in which a flywheel/hybrid drive system could be
incorporated into current production vehicles.
FAMILY CAR
Figure 7-1 shows a phantom view of a conventional drivetrain arrangement in a fuLl-
sized family car. Replacing the conventional hydrokinetic transmission by a flywheel
drive transmission, the resulting arrangement would be as shown in Fig. 7-2. The
engine has been moved forward to accommodate the greater length of the flywheel
drive transmission. Although the same engine size is shown in both cases, a smaller
engine could be used with the flywheel/hybrid drive.
Figure 7-3 shows a conceptual layout for an engine-mounted flywheel/hybrid drive
transmission. No attempt was made to minimize the length of the transmission,
although this could certainly be done.
Another possible arrangement for the flywheel drive is shown in Fig. 7-4. Here the
conventional hydrokinetic transmission has been replaced by a simple torque damper.
The flywheel drive transmission has been incorporated into an independent rear
7-1
-------
c
suspension transaxle package. Conceptual layout of this flywheel transaxle is shown
in Fig. 7-5, and a configuration block diagram is shown in Fig. 7-6. Drivetrain pack-
aging advantages are immediately apparent. The smaller engine and removal of the
front transmission allow more crumple space to meet head-on collision safety require-
ments. The transmission hump is no longer required, and the driveshaft tunnel can be
substantially reduced or possibly eliminated due largely to the body mounting of the
transaxle package. Weight distribution is improved slightly, and more flexibility in
choosing weight distribution is provided. The independent rear suspension (IRS) is
more expensive than the conventional Detroit solid axle, but offers advantages in
handling and ride quality which in themselves have justified IRS for many foreign
manufacturers.
CITY BUS
A conventional drivetrain arrangement for a 53-passenger city bus is shown in Fig.
7-7. Figure 7-8 shows a flywheel/hybrid drivetrain arrangement for the city bus,
and a conceptual layout of the flywheel/hybrid transmission is shown in Fig. 7-9.
Although the flywheel drive transmission is somewhat larger than a conventional bus
transmission, its use makes possible a smaller engine approximately one half the
size of the conventional engine - such that the overall package of engine plus trans-
mission is approximately the same or slightly less.
Although these studies did not involve consideration of various types of fuels, the use
of fuels such as compressed natural gas (CNG) for the city bus is attractive on the
basis of both cost and emissions. When existing engines are converted to CNG, a
serious loss in power capability results. The replacement of the existing transmissions
with flywheel drives would more than restore the resulting loss in acceleration
capability and provide additional reductions in emissions and operating cost.
7-2
-------
Fig. 7-1 Drivetrain Arrangement for Conventional Engine -Mounted
Hydrokinetic Transmission
Fig. 7-2 Drivetrain Arrangement for Engine-Mounted Flywheel/Hybrid
Transmission
7-3
-------
ENGINE
HYDROSTATIC
UNITS
PLANETARY SPEED
REDUCER , DRIVE CLUTCH
BAND CLUTCH
VER-RUNNING\\ REVERSE .-VEHICLE PLANETARY
CLUTCH \\ /DIFFERENTIAL
ENG. PLANETARY DIFF.
VEHICLE
HYDROSTATIC
UNITS
Fig. 7-3 Engine-Mounted Flywheel/Hybrid Transmission Layout
-------
FUEL
I
01
FLYWHEEL
TRANSAXLE
Fig. 1-4 Drivetrain Arrangement for Transaxle Flywheel/Hybrid Transmission
-------
FLYWHEEL
PLANETARY SPEED
REDUCER
ENGINE PLANETARY
DIFFERENTIAL
REAR AXLE
CLUTCH
VEHICLE
PLANETARY
DIFFERENTIAL
ENGINE
HYDROSTATIC
UNITS
VEHICLE
HYDROSTATIC
UNITS
Fig. 7-5 Transaxle Flywheel/Hybrid Transmission Layout
-------
-1
I
-1
HEAT
ENGINE
40% NORMAL
HP RANGE
2:1 RPM RANGE
TORSIONAL
DAMPEN ER
FLYWHEEL
TRANSMISSION
(ENG/FW)
0.5 KW-HR
16,000 TO 24,000 RPM
TRANSMISSION
(FWAEH)
CRUISE POWER
3:1 RATIO RANGE
PEAK ROAD POWER
15:1 RATIO RANGE
FLYWHEEL TRANS AXLE PACKAGE
DIFFERENTIAL
ROAD
PLANETARY/HYDROSTATIC POWER
SPLITTING TRANSMISSION
Fig. 7-6 Transaxle Flywheel/Hybrid Transmission Configuration
-------
Fig. 7-7 Drivetrain Arrangement for Conventional City Bus
7-8
-------
Fig. 7-8 Drivetrain Arrangement for Flywheel/Hybrid City Bus
7-9
-------
.0
I
I-1
o
f_- "A
\
Fig. 7-9 Layout of Flywheel/Hybrid Transmission for City Bus
-------
Section 8
CONCLUSIONS AND RECOMMENDATIONS
The overall conclusion of the Flywheel Feasibility Study and Demonstration program
is that flywheel energy storage is a feasible means of emission reduction for full per-
formance vehicles. Although the specific energy of present-day flywheels is not suf-
ficient to make flywheel-only vehicles feasible, flywheel/heat-engine hybrid vehicles
are entirely feasible. The flywheel/heat-energy hybrid drive may permit the full size
family car to achieve low emission levels while enhancing performance and gas
mileage in city driving.
Specific conclusions and recommendations are given in the following paragraphs:
1. Pure flywheel drives are not suitable for any of the four vehicles because
the specific energy of the flywheel-only drive system is too low. Using a
figure of 25 w-hr/lb as the highest possible flywheel specific energy avail-
able today, the family car would require about a 10:1 improvement in spe-
cific energy, and the commuter car or delivery/postal van would need roughly
a 2:1 improvement. It is not likely that these improvements could be attained
within the next 5 years. The city bus is a borderline case; a pure flywheel
drive would be suitable If recharging could be accomplished at distances of
approximately 10 miles or less. Of course there may be little relative
merit in pure energy storage drives if the source of recharge power itself
generates more pollution per unit of energy than a flywheel/heat-engine
hybrid vehicle (Ref. 8-1).
2. Flywheel/heat engine hybrid drives are suitable for all four vehicles in
particular and for most vehicles in general. For hybrid drives, specific
energy is not critical; the weight of the steel flywheel is on the order of
1 percent of the gross vehicle weight. Specific power requirements are
8-1
-------
c.
easily met by the flywheel per se, but dictate transmission requirements.
Although the technical feasibilities of hybrid drives for the three smaller
vehicles were indicated, the family car was chosen for further study
because of its high population.
3. The proper system control technique is that of total kinetic energy (TKE)
control. The TKE approach can give the vehicle a nearly non-history-
dependent performance capability and obviates flywheel oversizing for re-
cuperation. For prototype vehicles, an all-electronic control system is
recommended so that various control modes can be investigated.
4. The preferred transmission configuration is the double-transmission type
because it permits independent speed relationships between flywheel and
vehicle as well as between engine and flywheel. The double configuration
is essential for prototype vehicles so that various control modes can be
evaluated. Evaluation of a single-transmission mode would be made with
the proper section of the double-trans miss ion locked up.
5. Of the four types of transmissions studied (mechanical, hydrostatic,
hydrokinetic, and electric), the best type is a combination of mechanical
and hydrostatic - the power splitting transmission. This transmission is
comprised of a planetary differential with a variable hydrostatic bypass.
Hydrokinetic transmissions are unsuitable to the flywheel/hybrid by their
very nature of operation. Electric transmissions are prohibitively
costly.
6. The spark-ignition engine is the best presently available heat engine for
hybrid vehicles, both on the basis of minimum cost and on the basis of mini-
mum emissions. The basic flywheel/hybrid approach, however, is applicable
to all four engine types investigated (Otto, Brayton, Diesel, and Rankine)
and eliminates certain drawbacks with certain types. For instance, in a
flywheel/hybrid system, the lag in the gas turbine response would be of no
consequence; and for the steam engine, the expander, and more importantly
the condenser, would be sized on the basis of cruise plus accessory power
rather than peak power.
8-2
-------
7. Based on emissions data supplied by EPA/APCO, a flywheel/hybrid family
car using a presently available spark-ignition engine can achieve low
emission levels.
8. Based on emissions data supplied by EPA/APCO, a flywheel/hybrid family
car using one of several engines of projected 1975 availability would have
emissions substantially below a flywheel/hybrid with a contemporary engine.
9. Based on emissions data supplied by EPA/APCO, a flywheel/hybrid city
bus using a presently available spark-ignition engine would have substan-
tially reduced emissions. In addition, operating cost would be lower than
that of a conventional bus.
10. The optimum flywheel geometry is a modified exponential or conical disc
with a built-up rim.
11. The optimum flywheel material is common high-strength steel such as
AISI 4340.
12. The design of steel disc flywheels is simple and straight forward, and
performance is highly predictable. Manufacturing, inspection, and test
techniques are well established for similar hardware such as turbine rotors.
For these reasons, future development work in steel disc flywheels per sc
is considered to be unnecessary.
13. The practicality of filamentary composite flywheels remains unproven.
The energy density potentials of filamentary composite flywheels is insuf-
ficient for pure flywheel vehicles and unnecessary for flywheel /he at- engine
hybrid vehicles. For these reasons, little merit is seen in continued
development of filamentary composite flywheels.
14. Gyrodynamic effects are not expected to present serious problems, but
vehicular testing is recommended. Steering effects should be investigated
with various road surface conditions, tire types, vehicle weight distribu-
tions, and flywheel attitudes.
8-3
-------
15. Flywheel burst containment techniques should be investigated analytically, and
empirical tests should bo conducted to verify design approaches.
10. Work in the area of aerodynamic shrouding for flywheels at supersonic tip speeds
is recommended. Empirical techniques devised for turbine rotor research are
directly applicable.
17. Detailed design studies are recommended for flywheel ancillaries such as bear-
ings, seals, and evacuation systems.
18. The development and test of prototype family cars and city buses is recommended.
The recommended system components are: spark ignition engine, power splitting
transmission (planetary differential plus hydrostatic bypass) of the double con-
figuration, all-electronic controls, and a steel conical disc flywheel. This sys-
tem would be retrofitted into a full sized six-passenger family sedan of high
production. An identical propulsion system would be installed on a dynamometer
where emissions could be measured. Various control modes thus could be tested
simultaneously for emissions as well as vehicle "feel" and "driveability." The
double transmission is required in order to give complete flexibility to the system
for testing all types of control modes. For instance, a single transmission would
be simulated by "locking-up" the engine-flywheel transmission. The all-electronic
controls are also essential for test flexibility.
Program plans for the development and test of prototype vehicles are given in Section 9,
Future Program Plans.
8-4
-------
Section 9
FUTURE PROGRAM PLANS
The results of the Flywheel Feasibility Study and Demonstration as described in this
report show the high potentials for automotive emission reduction offered by kinetic-
energy-storage vehicle drives. In particular, the applicability studies conducted for
EPA/APCO have shown that the most desirable vehicle drive configurations are
flywheel/heat-engine hybrids utilizing spark-ignition engines in a family car and city
bus. The results of the computer-aided calculation of emissions from a car meeting
the "Vehicle Design Goals Six-Passenger Automobile" (Revision A, November 30, 1970)
when operated over the DHEW Urban Dynamometer Driving Schedule (Federal Register,
November 10, 1970) show that low emissions may be possible with a flywheel/spark-
ignition engine hybrid drive. Similar reductions in emissions from a city bus result
from the flywheel/spark-ignition engine hybrid drive when operated over the stipulated
City Bus Cycle. Flywheel/hybrid drives with other engine types may offer the potential
of even lower emission levels when and if these engines become available. In addition,
anticipated improvements in flywheel materials provide the possibility of even more
efficacious flywheels from the standpoint of energy density for future vehicle applica-
tions. However, the flywheels that were designed, built, and tested on the present
contract have shown that completely acceptable kinetic-energy-storage hybrid propulsion
systems can now be demonstrated without further engine or flywheel material improve-
ments. The succeeding paragraphs present plans for a logical sequence of activities
that will lead to operational, full-scale, low-pollution cars and busses configured with
flywheel/heat-engine hybrid propulsion systems.
A review of various approaches toward the development of practical hybrid vehicle
drives has been conducted. As a result, conducting the following three development
programs in combination is felt to represent the most expeditious means of achieving
the integration of flywheel/spark-ignition engine drives into fully operational proto-
type passenger cars and buses.
9-1
-------
' /
.w
PROGRAM 1 - FLYWHEEL/HYBRID DRIVE SYSTEM DEVELOPMENT
The objective of this program (of approximately 8 months duration) is the full-scale
feasibility demonstration and performance verification of a flywheel/spark-ignition
engine drive system using presently available components. The system power capabil-
ity would be in the general range suitable for a bus or car. The "breadboard" drive
system would be comprised of the following components:
a. A 0. 5 kw-hr or 1. 0 kw-hr steel flywheel similar to flywheels built and
tested by LMSC on the present contract
b. Flywheel housing consisting of a structure supporting a seal and bearing
system based on the findings from the present contract tests
c. Flywheel speed-reducing gearbox
d. Spark-ignition engine of suitable capacity
e. Transmission comprised of hydrostatic components to simulate control
interfaces and operation of a power-splitting transmission. (This trans-
mission configuration will be used to simulate the ultimate power-splitting
transmission such that tests can be conducted without delay.)
f. Clutch system
g. Ancillary cooling, vacuum, and control systems
The "breadboard" drive would be assembled and extensively tested using a program-
mable laboratory dynamometer to simulate vehicle rolling, drag, grade, and inertia
loads in all four torque-speed quadrants of operation. The specific goals of the planned
program are as follows:
a. To verify the operational characteristics of a full-scale flywheel/hybrid
drive suitable for ultimate application to a city bus and/or passenger
automobile
b. To define the control requirements and control interfaces of a flywheel/
hybrid drive
c. To determine the transmission requirements of a flywheel/hybrid drive
d. To select speed ranges for heat-engine and flywheel
9-2
-------
sfYr
PROGRAM 2 - INTEGRATION OF A FLYWHEEL/SPARK-IGNITION HYBUID DUIVK
SYSTEM INTO PROTOTYPE PASSENGER CARS
This is a 27-month program culminating in the delivery to EPA/APCO of five flywheel/
spark-ignition engine hybrid family cars. The program is divided into logical, serial
phases to provide an efficient sequence of work tasks with minimized development risk
and several significant program review points. The major objectives and activities of
the planned program phases are the following:
Phase 1 - Design, fabricate and test two identical and complete full-sized preproto-
type drive systems (flywheel, engine, transmission, controls and accessories). Design
would be based on dynamometer testing described above of a "breadboard" flywheel/
engine/transmission system used to determine general operating characteristics,
control interfaces, component requirements, and pollution reduction potentials. The
design program would include a computer-aided transmission analysis, flywheel
fatigue and failure control studies and tests, and system optimization studies. The
air pollution reduction characteristics of one drive would be established by exhaust
gas sampling while operating over a simulated DHEW Urban Dynamometer Driving
Schedule. The vehicle performance and driver "feel" characteristics would be
determined by operating a testbed vehicle with the second drive system in comparison
with a conventional control vehicle.
Phase 2 Fabricate five complete prototype drive systems with spares and assemble
into full-sized six-passenger automobiles. Shakedown and performance testing would
be conducted prior to delivery of vehicles to EPA/APCO in a ready-to-test condition.
PROGRAM 3 - INTEGRATION OF A FLYWHEEL/SPARK-IGNITION HYBRID DRIVE
SYSTEM INTO PROTOTYPE CITY BUSES
The planned 28-month development program, leading to delivery of three full-scale
prototype hybrid city buses suitable for operational demonstration in revenue service,
is divided into three logical phases. The first phase is directed toward the acquisition
9-3
-------
and analysis of operational bus data to establish realistic city bus operational cycles
and performance requirements.
The second and third phases closely parallel the two phases of the passenger car pro-
gram described above. The second phase (which could be carried on almost concur-
rently with the Phase 1 determination of city bus profiles and performance require-
ments) consists of the design, fabrication, and test of a full-scale preprototype city
bus drive suitable for extensive laboratory dynamometer testing. The design would be
based on the results of the "breadboard" test program of the flywheel/hybrid drive
development described above. Final sizing would be based on the findings of the
Phase ] operational bus data study. The end results of the Phase 2 program would
establish the performance and air pollution reduction characteristics of the hybrid
bus drive in simulated operations over typical city bus driving cycles. Phase 2 would
also include a computer-aided transmission analysis as well as investigation and testing
of potential problem areas of flywheel safety, fatigue cycles, and gyrodynamics.
The Phase 3 program plan consists of the fabrication of three full-sized bus drives
with spares followed by assembly into three buses. The buses would be extensively
tested for performance and safety prior to delivery of vehicles to EPA/APCO in a
ready-to-test condition.
9-4
-------
Section 10
REFERENCES
1-1 "The Oerlikon Electrogyro, Its Development and Application for Omnibus Ser-
vice," Automobile Engineer. Dec 1965
2-1 "DREW Urban Dynanometer Driving Schedule," Federal Register. 10 Nov 1970
4-1 David L. Plett and Harold G. Carlson, "Performance of a Variable Speed Constant
Frequency Electrical System," IEEE Transactions on Aerospace. Vol. 2, No. 2,
Apr 1964
4-2 Worldwide Diesel and Gas Turbine Catalog. Vol. 35, 1970
4-3 Automotive Industries 52nd Annual Engineering Specifications. 1970
4-4 Listing of U.S. Reciprocating (Gasoline) Engines, U.S. Gas Turbine Engines,
and Leading International Gas Turbines; Aviation Week and Space Technology;
Mar 1970
4-5 S. William Gouse, Jr., "The Search for Low Emission Vehicle Propulsion,"
American Society of Mechanical Engineers, No. 69-WA/Enr-4.
4-6 List of Rankine Cycle Engine Companies and Small Gas Turbine Engine Companies
transmitted by Dr. Karl Hellman (EPA/APCO) letter dated 16 Jul 1970.
4-7 Report of Hearing Proceedings, Subcommittee on Air and Water Pollution of
the Committee on Public Works, Committee on Commerce, U.S. Senate, 90th
Congress, 1968
4-8 Brake Specific Emission Curves transmitted by Dr. Karl Hellman (EPA/APCO) for
" spark ignition, diesel, gas turbine, and Rankine engines as modified by mutual
agreement in telephone conversation, 18 Aug 1970
4-9 Arthur F. Underwood, "Requiem for the Piston Engine?", Machine Design,
6 Aug 1970
10-1
-------
4-10 R. E. Wolfgang Hempel, "Does Turbocharging Increase Diesel Engine Noise? -
Observations on the Generation, Emission, and Reduction of Diesel Engine
Noise," SAE Paper No. 680406, May 1968
4-11 Thermo Electron Corporation, Conceptual Design. Rankine-Cycie Power System
With Organic Working Fluid and' Reciprocating Engines for Passenger Vehicles.
Report No. TE4121 -133-70, Jun 1970
5-1 D. E. Goldman and H. E. von Gierke, "The Effects of Shock and Vibration on
Man, " Naval Medical Res. Inst. , No. 60-3
5-2 Francis J. Lavoie, "Nondestructive Testing. " Machine Design. Sep 4, 1969
5-3 J. W. Daily and R. E. Nece, "Chamber Dimension Effects on Induced Flow and
Frictional Resistance of Enclosed Rotating Disks, " Journal of Basic Engineering,
Mar 1960
Robert W. Mann and Charles H. Marston, "Friction Drag on Bladed Disks in
Housings, " Journal of Basic Engineering, Dec 1961
6-1 Lockheed Aircraft Corporation, Stress Memorandum No. 114, S-N Curves, p. 15
8-1 Paul D. Agarwal, "Electricity Not Such a Clean Fuel," Automotive Engineering,
Feb 1971
10-2
-------
Appendix A
NAPCA VEHICLE DESIGN GOALS - SIX-PASSENGER AUTOMOBILE
(REVISION A, NOVEMBER 30, 1970}
The design goals presented below are intended to provide:
A common objective for prospective contractors.
Criteria for evaluating proposals and selecting
a contractor.
Criteria for evaluating competitive power systems
for entering first generation system hardware.
The derived criteria are based on typical characteristics of the class of
passenger automobiles with the largest market volume produced in the U.S.
during the model years 1969 and 1970. It is noted that most weight and volume
characteristics presented are maximum values and the performance character-
istics are intended as minimum values. Prospective contractors who take
exceptions must Justify these exceptions in their proposals and relate
these exceptions to the technical goals presented herein.
1. Vehicle weight without propulsion system - Wo.
W0 is the weight of the vehicle without the propulsion system
and includes, but is not limited to: body, frame, glass and
trim, suspension, service brakes, seats, upholstery, sound
absorbing materials, insulation, wheels (rims and tires),
accessory ducting, dashboard Instruments and accessory wiring,
and all other components not included in the propulsion system.
It also includes the air conditioner compressor, the power
steering pump, and the power brakes actuating device.
W0 is 2700 Ibs.
2. Propulsion system weight - Wp.
W_ includes the energy storage unit (including containment),
power converter (including both functional components and
controls) and power transmission to the driven wheels. It also
includes the exhaust system, pumps, motors and fans necessary
lor operation of the propulsion system, and any propulsion
system heating or cooling devices.
The maximum allowable propulsion system weight, Wpm, is 1600 Ibs.
The propulsion system may be lighter than the maximum allowable.
A-l
-------
3. Vehicle curb weight - Wc.
Wc - W0 + Wp.
The maximum allowable vehicle curb weight, Wcm, is 4300 Ibs.
(2700 + 1600 max. - 4300).
4. Vehicle test weight - Wt.
Wt = W0 + Wp + 300 Ibs. Wt is the vehicle weight at which all
accelerative maneuvers, fuel economy and emissions are to be
calculated. (Items 8b,c,d).
The maximum allowable test weight, Wtm, is 4600 Ibs. (2700 +
1600 max. + 300 - 4600).
5. Gross vehicle weight - W .
O
Wg = W0 + Wp + 1000 Ibs. Wg is the gross vehicle weight at
which sustained cruise grade velocity capability is to be
calculated. (Item 8 e). The 1000 Ibs. load simulates a full
load of passengers and baggage.
The maximum allowable gross vehicle weight, Wgm, is 5300 Ibs.
(2700 + 1600 max. + 1000 - 5300).
6. Propulsion system volume - V_.
Vp shall be packagable in such a way that the volume encroach-
ment on either the passenger or luggage compartment is not
significantly different than today's (1970) standard full
size family car. Necessary external appearance (styling)
changes will be minor in nature. Vp shall also be packagable
in such a way that the handling characteristics of the
vehicle do not depart significantly from a 1970 full size family
car.
The maximum allowable volume assignable to the propulsion
system, Vpm, Is 35 ft3.
7. Emission Goals.
The vehicle when tested for emissions in accordance with the procedure
outlined in the Nov. 10, 1970 Federal Register shall have a
weight of Wt. The emission goals for the vehicle are:
Hydrocarbons* 0.14 grams/mile maximum
Carbon monoxide 6.16 grams/mile maximum
Oxides of nitrogen** 0.4 grams/mile maximum
Particulates 0.03 grams/mile maximum
A-2
-------
*Total hydrocarbons (using 1972 measurement procedures)
plus total aldehydes. Aldehydes will not be more than
10 percent by weight of the hydrocarbons or 0.014
grams/mile, whichever is greater.
**Measured or computed as N02«
The powerplant should not produce objectionable levels of smoke,
odors, or other pollutants which are not included above.
8. Start up, Acceleration, and Grade Velocity Performance.
a. Start up: The time to reach 65 percent full power is
A3 sec. Ambient conditions are 14.7 psia pressure,
60°F temperature.
Cold (-40°F) start-up techniques shall be equivalent to
the typical automobile spark-ignition engine.
b. Acceleration from a standing start: The minimum distance
to be covered in 10.0 sec. is 440 ft. The maximum time to
reach a velocity of 60 mph is 13.5 sec. Ambient conditions are
14.7 psia, 85°F. Vehicle weight is Wc. Acceleration is on a
level grade and initiated with the engine at the normal idle
condition.
c. 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. Ambient conditions are
14.7 psia, 85°F. Vehicle weight is Wc, and acceleration
is on level grade.
d. 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 vehicle 100 feet behind
the back of a 55 foot truck traveling at a constant 50 mph
to a position where the back of the vehicle is 100 feet
in front of the front of the 55 foot truck is, 15 sec. and
1400 ft. The entire maneuver takes place in a traffic
lane adjacent to the lane in which the truck is operated.
The vehicle speed during the maneuver cannot exceed 80 mph.
Ambient conditions are 14.7 psia, 85°F. Vehicle weight is
Wt, and acceleration is on level grade.
e. Grade velocity: The minimum sustained cruise velocity on
a 5 percent grade with the air conditioner operating
is 65 mph. Ambient conditions are 14.7 psia 85°F.
Vehicle weight is Wg.
A-3
-------
-4-
The vehicle must be capable of providing performance (Paragraphs
8b, 8c, 8d, 8e) within 5% of the stated 85°F values, when operated
at ambient temperatures from-20°F to 105°F.
9. Minimum vehicle range without supplementing the energy storage
will be 200 miles. The minimum range shall be calculated for,
and applied to each of the two following modes: 1) A city-
suburban mode, and 2) a cruise mode. Mode 1 is the driving cycle
which appears in the Nov. 10, 1970 Federal Register. For vehicles
whose performance does not depend on the state of energy storage,
the range may be calculated for one cycle and ratioed to 200 miles.
For vehicles whose performance does depend on the state of energy
storage the Federal driving cycle must be repeated until 200 miles
have been completed. Mode 2 is a constant 70 mph cruise on a
level road for 200 miles. The vehicle weight for both modes shall
be, initially, Wt. The ambient conditions shall be a pressure of
14.7 psia, and temperatures of 60°F, 85°F and 105°F. The vehicle
minimum range shall not decrease by more than 5% at an ambient
temperature of -20°F.
For hybrid vehicles, the energy level in the power augmenting
device at the completion of operation will be equivalent to the
energy level at the beginning of operation.
10. System thermal efficiency will be calculated by two methods:
1. A "fuel economy" figure based on the number of Btu per
mile required to drive the vehicle over the 1972 Federal
driving cycle which appears in the July 15, 1970 Federal
Register. Vehicle weight is W£.
2. A "fuel economy" figure based on the number of Btu per
mile required to drive the vehicle at constant speed, In
still air, on level road, at speeds of 20, 30, 40, 50, 60,
70, and 80 mph. Vehicle weight is Wt.
In both cases, the system thermal efficiency shall be calculated with
sufficient electrical, power steering and power brake loads in service
to permit safe operation of the automobile, air conditioning and
heater loads being excluded.'
Ambient conditions are 14.7 psia and temperatures of 60°F, 85°F
and 1058F.
11. Air Drag Calculation; The product of the drag coefficient, Cd»
and the frontal area, Af is to be used in air drag calculations.
The product CdAf has a fixed value of 13 ft2. The air density used
in computations shall correspond to the applicable ambient air
temperature.
A-4
-------
12. Rolling Resistance, R, is expressed in the equation R W/50
[1 + (l.A x 10~3V) + (1.2 10~5V2)] Ibs. V is the vehicle
velocity in ft/sec. W is the vehicle weight in Ibs.
13. Accessory power requirements: The accessory power demands are
defined to include: the air conditioning compressor, the
power steering pump, the alternator, and the power brakes
actuating device. The accessories also include a device for
heating the passenger compartment if the heating demand is not
supplied by waste heat.
The maximum intermittent accessory load, Paim, is 10 HP (plus the
heating load, if applicable). The maximum continuous accessory
load, Pacm, is 7.5 HP (plus the heating load if applicable).
If accessories are driven at variable speeds, the above values
apply. If the accessories are driven at constant speed, the
values will be reduced by 3 HP.
14. Passenger comfort requirements: Heating and air conditioning
of the passenger compartment shall be at a rate equivalent to
that provided in the present (1970) standard full size family
car.
Present automobile heating rate practice is approximately 30,000
Btu/hr. For an air conditioning system at 110°F ambient, 30°F
and 40% relative humidity air to the evaporator, Che rate is
approximately 13,000 Btu/hr.
15. The propulsion system shall be operable within an expected ambient
temperature range of -40° to 125°F.
16. Propulsion system design (normal operation) lifetime will be
3000 hours. Normal maintenance may include replacement of
accossable minor parts of the propulsion system via a usual
maintenance procedure, but the major parts of the system shall
be designed for a 3000 hour minimum life.
17. Noise:
a. Maximum Noise Test: The maximum noise generated by the vehicle
shall not exceed 77 dbA when measured in accordance with SAE
J968a. Note that the noise level is 77 dbA whereas in the SAE
J968a the level is 86 dbA.
b. Low Speed Noise Test: The maximum noise generated by the
vehicle shall not exceed 63 dbA when measuri-ii in accordance
with SAE J968a except that a constant vehicle velocity of
30 mph is used on the pass-by, the vehicle hiiing in high gear
or the highest gear in which it can be oper.-ii.od at that speed.
A-5
-------
c. Idle Noise Test: The maximum noise generated by the vehicle
shall not exceed 62 dbA when measured in accordance with SAE
J968a except that the engine is idling (clutch disengaged or
in neutral gear) and the vehicle passes by at a speed of less
than 10 mph. The microphone will be placed at 10 feet from
the centerline of the vehicle pass line.
18. The vehicle shall comply with all current Department of
Transportation Federal Motor Vehicle Safety Standards.
Revision A (Nov. 30, 1970)
A-6
-------
Appendix B
COMPUTER PROGRAM DESCRIPTIONS
Program Name: /Fl.O Flywheel/
Size:
Data Inputs:
File Inputs:
13056 characters
Variable
Weight
Frontal area
Drag coefficient
Engine size
Flywheel size
Maximum jerk
Profile
Unit
Ib
sqft
hp
kw-hr
mph-sec/sec
Contains elapsed time, acceleration, acceleration type (constant
horsepower or constant acceleration), and grade for trip profile
Emissions
Contains log plots of HC, CO, and NO
J\.
Program Functions:
Develops incremental emissions as a function of heat engine horse-
power
Computes constant road horsepower acceleration limited by jerk
rate
Computes transmission efficiencies
Program Outputs:
For each second:
Time (sec)
Acceleration (mph/sec)
B-l
-------
- Velocity (mph)
- Drag (Ib)
- Rolling resistance (Ib)
Horsepower out
- Cumulative horsepower out
- Average horsepower out
Horsepower in
Cumulative horsepower in
- Average horsepower in
- Net horsepower in system
Flywheel energy
- Distance traveled (ft)
Cumulative grams of HC
- Cumulative grams of CO
- Cumulative grams of NO
A.
Summary Outputs:
Maximum horsepower required
- Cruise horsepower
Input horsepower required
- Average horsepower out
Maximum flywheel energy
Emissions in:
Ib
Ib/hr
Ib/mi
gm
gm/hr
gm/mi
Comments: Several versions of this program have been developed for specific
purposes. For example, some versions initiate with the flywheel
energy at zero; others initiate with the flywheel at maximum energy
and then automatically continue the run until the flywheel is again at
maximum energy. Many subprograms have been developed to
simulate different transmission types.
B-2
-------
Program Name: /F51. 1 VELGRD/
2304 Characters
Data Inputs: Variable Unit
Weight Ib
Frontal area sq ft
Drag coefficient
Maximum velocity mph
Program Functions:
Calculates road horsepower required to cruise at maximum
velocity on 0- percent grade
Using calculated horsepower (or an optional input horsepower),
develops a plot of velocity vs. grade percent for grade from 0 to
30 percent (or higher)
Program Outputs:
Cruise horsepower on 0-percent grade
Table of velocity and grade
File of velocity and grade to be used by X-Y plotter
B-3
-------
Program Name: /F10.4 FWGEN/
Size:
Data Inputs:
1536 Characters
Variable
Increment
Radius
Units
in.
in.
Equations
Equations for flywheel
Geometry at varying intervals
From 0 to radius
Program Functions:
Develops incremental half-thickness of flywheel from 0 to R based
on input equations
Program Output:
- A file containing radius increments and half-thickness for use in
Program /F10.2 FWSTRESS/
B-4
-------
Program Name: /F10.2 FWSTRESS/
Size: 5376 Characters
Data Inputs: Variable Unit
Speed Rad/sec
Material density Ib/cu in.
Poissons Ratio
Young's Modulus
Temperature coefficient
Temperature difference deg-F
Outside radius in.
File Inputs: Incremental radius and half-thickness of flywheel from O to R
(outside radius)
Program Functions:
First accepts a trial value of the tangential stress of the outermost
increment
- Calculates total force acting on flywheel
Zeros in by reiteration on the actual outside tangential stress so
that the incremental forces will equal the total force
Generates the following outputs upon finding the correct
stress:
Program Outputs
Incremental
Radius (in.)
Thickness (in.) from input file
Temperature (deg-F)
Tangential stress (psi)
Radial stress (psi)
B-5
-------
r
Summary:
Mean tangential stress (psi)
2
Moment of inertia (Ib-in.-sec )
Kinetic energy (in.-lb)
- Weight (Ib)
- Energy density (in.-lb/lb)
- Shape factor
Comments: This program included an option to create a file so that an X-Y plotter
can generate a plot of the flywheel cross section and the associated
tangential and radial stresses across the cross section
B-6
-------
Appendix C
HEAT-ENGINE TREND DATA
MFG
MODEL TYPE BHP WT COST VOL SFC
FARYMANN
FARYMANN
FARYMANN
LYCOMING
LYCOMING
LYCOMING
LYCOMING
VOLKSWAGON
VOLKSWAGON
PEUGEOT
PEUGEOT
PEUGEOT
PEUGEOT
PEUGEOT
DAIMLER
ONAN
ONAN
ONAN
ONAN
ONAN
KOHLER
KOHLER
KOHLER
DEUT2
DEUT2
PERKINS
PERKINS
PERKINS
PERKINS
PERKINS
PERKINS
FORD
FORD
FORD
FORD
FORD
FORD
FORD
FORD
FORD
R
P
S
W-32
W-42
W-34
U-44
122
126
XDP4/88
XDP4/90
XDP6/90
XM
XM1
OM636
CCKA
CCKB
NY
D060
DJ120
K321
K482
K662
FIL410
F2L410
4*108
3-152
4.154
4.236
NA-120
NA-180
091GF
104GF
B8PE
B8PD
200GF
C5PF
C5PG
C4PD
C4PB
D
D
D
D
D
D
D
S
S
D
D
D
S
S
D
S
S
S
D
D
S
S
S
D
D
D
D
D
D
D
D
S
S
S
S
S
S
S
S
S
16
20
25
18
20
36
40
34
45
52
56
81
50
55
36
12
14
18
12
22
11
14
19
11
21
36
35
55
65
90
144
44
60
36
51
90
113
124
143
158
240
335
370
306
313
433
433
206
220
390
414
561
319
319
425
168
172
125
270
438
119
180
250
264
359
480
700
555
750
1040
1340
254
254
535
535
337
473
473
720
720
375
387
423
697
702
1177
1185
375
412
590
608
1070
401
431
736
239
248
265
541
869
234
401
506
473
593
744
764
892
852
1185
1852
438
458
468
489
520
555
572
583
754
3-9
4.3
4.3
8-2
8.2
12.3
12>3
11 .6
11.7
6-8
6.8
11.0
6.7
6.7
8*4
3.8
3-8
4*4
5.9
9.6
3»2
4.7
8*6
5.6
7.4
5.5
12-7
11*3
9.2
18*1
27-2
5.3
5-3
9.9
9-9
11*8
15.6
15-6
22>2
22*2
460
460
460
480
485
510
485
.480
.490
485
485
485
.530
530
450
750
630
620
400
400
620
690
600
488
461
470
420
430
.370
380
415
C-l
-------
MFC
MODEL TYPE BHP WT COST VOL SFC
FORD
FORD
FORD
FORD
FORD
G.M.C.
G.M.C.
G.M.C.
G.M.C.
G.M.C.
G.M.C.
G.M.C.
G.M.C.
G.M.C.
G.M.C.
SO LOR
WILLIAMS
KLOCKNER
TURBOTRON
MO TO REN
SO LOR
KLOCKNER
AI RESEARCH
WILLIAMS
WILLIAMS
A I RESEARCH
MO TO REN
ALLISON
TURBOMECA
ALLISON
MO TO REN &
TURBOMECA
ALLISON
TURBOMECA
AI RESEARCH
TURBOMECA
TURBOMECA
TURBOMECA
UACL (P&W)
ROLLS-ROYC
COPB
C3PY
242DF
330DF
363DF
911619
911620
991894
6002802
6002818
305C
351C
401M
DH478
DH637
10KW
WR-26
T216
?
6012L
TITAN
T112
GTP30
WR2-6
WR24-7
TSE-36
6022 A2
250C18
TAA230
T63
6022 A3
2C6
T63
OREDEN4
TSE231
2A
3B
14A
ST6
103/503
D
D
D
D
D
S
S
S
S
S
S
S
S
D
D
T
T
T
T
T
T
T
T
T
T
T
T
T
T
T
T
T
T
T
T
T
T
T
T
T
44
89
68
97
106
41
43
82
108
120
112
125
150
117
159
22
80
85
68
94
110
120
130
130
170
210
210
270
300
310
320
340
340
355
400
450
465
510
510
600
585
880
754
965
956
412
450
410
950
980
950
950
1271
45
225
176
85
96
85
79
105
56
85
178
190
136
290
155
320
315
190
250
174
285
400
352
269
655
869
1126
913
1320
1356
403
407
413
336
348
640
710
805
10.
10.
14.
25.
25.
25.
25.
35.
1.
6.
8<
5-
3.
3-
1.
5-
2'
2<
12>
9.
9.
9.
14.
12
10
3.
4'
7
12
9
9
1
.0
.5
>8
8
8
.9
1
.2
.0
.7
.8
.1
.7
.5
.2
.3
.1
.7
0
3
.6
4
.0
.7
8
.3
4
.2
.5
8
580
530
490
490
.500
446
490
491
1-200
590
1*370
580
1.280
1-070
960
1.070
1.030
1.080
630
740
720
.880
.710
670
950
650
580
640
620
670
590
680
860
C-2
-------
MFC
MODEL TYPE BMP WT COST VOL SFC
CZECH
AI RESEARCH
A I RESEARCH
TURBOMECA
LYCOMING
GEN ELECT
GEN ELECT
GEN ELECT
ROLLS-ROYC
M601
TPE331
TPT331
24
T53
T58
T58
T58
H1200
T
T
T
T
T
T
T
T
T
620
680
715
755
935
1060
1060
1130
1150
299
355
343
341
496
315
315
309
314
7.
14<
13.
11.
11.
6-
7.
11-
12.
.8
.6
.2
.9
.5
.9
1
.5
.8
640
650
600
540
700
660
.630
620
.640
C-3
-------
Appendix D
COMPUTER-AIDED STRESS ANALYSIS OF DISC TYPE FLYWHEELS
by R. Gilbert and J. Spill man
The following stress analysis follows the method of Stodola* (pp. 373f), but the differ
ence form of the equations is used rather than the differential form. An isotropic
material is assumed. It also is assumed that there is a negligible stress gradient in
the axial direction. Considering an incremental piece of the flywheel,
the equation for the summation of forces may be derived.
The volume,
The mass,
AV = (xA6) (Ax) (y)
(D.I)
Am =
g
g
(D.2)
* Steam and Gas Turbines, A. Stodola, Vol. I, McGraw-Hill, 1927,
D-l
-------
The radial acceleration,
a = xu>2 (D.:i)
The centrifugal force on an element due to its own acceleration,
F = ma (D.4)
2
F = £^-x2yAxA0 (D.5)
O
The tangential force on each radial face,
T = yAxat (D.6)
The radial force on inner circumferential face,
R = yxA0ar (D.7)
The radial force on outer circumferential face,
R' = (y + Ay) (x + Ax) A0 (OT + Aap (D.8)
The sum of the forces in the radial direction = 0 .
R' - R - TA0 + F = 0 (D.9)
Expanding Eq. (D.8)
R1 = (yx + yAx + xAy + Ay Ax) (a + Aa ) A0 (D.10)
D-2
-------
R' = (xy
-------
from Eq. (D.17),
a = Ee + vo (D.li))
from Eq. (D.18),
at ' Eet
(D.20)
from Eqs. (D.19) and (D.20)
a - Ect
Eer + V°t =
(D.22)
a. - i>2(7. = Eve + Eet (D.23)
at = ~~~2(t + V*T} (D<24)
1 - v
Similarly, by symmetry,
° = ^-o(. + "O (D.25)
1 - v
The circumferential strain at A,
D-4
-------
D-5
(D.27)
(D.28)
The length of A - B after elongation,
Ax + A(Ax) = (x + Ax + £_) - (x +
JD
+ Ax
(D.2!))
(D.30)
(D.32)
A (Ax) =
(D.33)
Ax Ax
D 4)
^ '
But by definition,
-------
From Eq. (D.27)
= Xt (D.37)
Combining above with Eq. (D.18)
= |ar) (D.40)
Ax - xvAar - J^CT Ax)
Combing Eqs. (D.39) and (D.44)
/ Aat
D-6
i / Aat
= 4 (XT* + ^ - xv-r-^ - i/a ) (D.42)
AxElAxt Ax r/ v/
Aa\
- (D.44)
-------
A a
Ax Ax
a - a,
which is the stress compatibility equation.
The summation of forces Eq. (D.16) and the stress compatibility Eq. (D.45) ca.n now
be combined to determine the tangential and radial stresses in a flywheel of given shape,
material, and peripheral velocity.
COMPUTER METHOD
The wheel is divided into a number of circumferential elements (onion rings) and the
stresses are calculated starting from the outside of the outermost element and working
inward.
D-7
-------
r~
From (D.I6),
g ~2yAx (D.46)
Axya = (xya ) - (xya ) (D.47)
r 1 T 2
pu2 2
(xya) = (xya) - yaAx + £- x yAx (D.18)
2 1 g
[2 T /
(xya) - ya Ax + ^-x2yAx /(xy) (D.49)
r 1 l g J/ 2
where variables (x , y, a , a.) without subscripts (1 or 2) are mean values. For thin
slices, sufficient accuracy can be obtained by using arithmetic averages.
For the outermost slice, ar- = 0 , but a^ is not known. A value for a^ must be
assumed and total stress plots obtained. The tangential stress profile when integrated
over the half wheel will deviate from the known mean targential stress as calculated
from the following equation:
x
(a) mean = - =- - (D.50)
1 X
vAx
x=o
A second assumption for o-j. at the outside is then made and successive iteration
eventually brings the integrated stress profile and the mean tangential stress into agree-
ment. The resulting stress profiles are then correct. The computer program employs
D-8
-------
a binary closure technique such that the initial assumption for ov need not be at all
close to the actual value . For near-optimum energy storage flywheels of the kind
considered here, a reasonable initial assumption for o^ at the outside element is one
half the mean tangential stress per Eq. (D.50).
Since Eq. (D.49) requires the mean value of a, , an initial calculation is made using
1
at for a, , and after at is calculated (see Eq. (D.53) below) crt is set to -z (at +
at ) and ) -Ax (D.51)
t r v ' x v '
Aa = (a ) - (a ) (D.52)
1 r 1 l 2
/ \ CTr " at
(a ) = (a ) -via -a I - (1 + v) Ax (D.53)
t 2 t 1 \TI T2i x
Thermal stresses due to a radial temperature gradient may be superimposed by adding
the term Ee (t - t ) to the right hand side of Eq. (D. 53),
where
o n
E = Young's Modulus, Ib/in. (30 x 10 for steel)
i fi
e = thermal coefficient of expansion, °F (7.3 x 10" for steel)
t -- temperature, °F, at the end (inside) of the increment
£
t = temperature, °F, at the beginning (outside) of the increment
D-9
-------
Calculations of total weight, inertia, kinetic energy, etc., are calculated in the normal
fashion by summing the values for each increment. The flywheel shape is provided by
u data file which may be loaded from tabular data or from an equation or sets of
equations.
The computer program and a typical run are attached.
D-10
-------
PROGRAM /F10.2 FW STRESS/
100 AS="PROGRAM /10-2 FWSTRESS/" IUPDATED OCT 7,1970
110 REAL A*A1*B,C*D*D1*E*F1*G*I*M,N*P,Q,R,S*T/T1*T2,V
120 DIM R(3),T(3)*X(300)*YC300)
130 PRINT FOR 1=1 TO 10
140 PRINT "DATE OF RUNl ":DATE
150 PRINT FOR 1=1 TO 3
160 A=1*M=2*E«3>Q=0*D1=2»6*S2=1*S3=0
170 READ W*P,V*E2*E3*T5
180 G=386«089
190 GOSUB 2000
200 S*T2=l»2*P*Wt2/G*Fl/Al/2
210 J=.1*S
220 FOR I=-10 TO 25 STEP+1
230 IF 2t(I+l)
-------
PROGRAM /F10.2 FV SIRESS/
470 PRINT IN FORM "35B QQ.4D KVH'/" IK/
480 PRINT IN FORM "' WEIGHT
- 3Q.DD' LBV"»W2
490 PRINT IN FORM MI ENERGY DENSITY
- ' 6Q.DD1 IN-LBS/LBV'rKl
500 PRINT IN FORM "35B 4Q.D ' WH/LBV":K1/12/2-6552E3
510 PRINT IN FORM SHAPE FACTOR
- ' Q.4D/":P/T4*K1
520 PRINT FOR 1=1 TO 2
530 PRINT" RADIUS THICKNESS TEMPERATURE TANGENTI
AL RADIAL"
540 PRINT " (IN) (IN) (DEG-F) STRES
S STRESS"
550 PRINT SPACE(38):"(PSI) (PSI)"
560 PRINT"
570 PRINT
580 T4*R,T1*W2*M2=0*T«S*T6=0
590 X=X( 1>,Y=Y(1)
600 1=1, Z=0
610 GO SUB 4000
620 FOR 1=2 TO N- 1
630 IF Tl>P*Wt2/G*Fl + 1000 OR TKO THEN 900
640 D=X(I-1)-X(I)
650 X=X(I-l)-D/2
660 Y=(Y(I)+Y(I-l))/2
670 T(M),T(E)=T
680 R(E)=R
690 R=0
700 B=0
710 T6=T6*T5/X( 1)*D
720 IF X(I)=0 THEN 880
730 R(A)=(X(I-l)*Y(I-l)*R(E)-Y*T(M)*D+P/G*Wt2*Xt2*Y*D
740 R(M)=(R(A)+R(E)>/2
750 T(A)=T(E)-V*(R(E)-R(A))-(1+V)*((R(M)-T(M))/X)*D+E
2*E3*T5/X(1 )*D
760 B=B+1
770 T(M)=(T(A)+T(E))/2
780 IF INT(T(A)*1000)=INT(R*1000) THEN 810
790 R=T(A)
800 GO TO 730
810 U2=W2+Y*D*PI*P*(X(I) + X(I-D)
820 M2=M2-KY*D*PI*P*(X(I)+X(I-l)))/G*(Xt2)
830 Tl = Tl-»-T(M)*Y*D
840 T=T(A)>R=R(A)*X=X(I)*Y=Y(I)
850 T4=MAX(T4,R*T)
D-12
-------
PROGRAM /F10-2 FW STRESS/
860 GOSUB 4000
870 GO TO 920
880 I=N-1
890 GO TO 920
900 I=N-1
910 IF TKO THEN Z=-l ELSE Z=l
920 NEXT 1
930 IF X(N-l)<>0 THEN 970
940 T1=T1+T(A>*Y*D
950 W2=W2+Y*D*PI*P*X
960 M2=M2+(Y*D*PI*P*X(N-2))/G*X»2
970 R1=SQRTCM2*G/V2>
980 K=(W2/(2*G))*R1«2*W»2
990 K1=K/W2
1000 IF Q=l THEN 1030
1010 GOSUB 3000
1020 GO TO 250
1030 IF X(N-1X>0 THEN 1050
1040 PRINT IN FORM "32.4D 4B 2Z«4D 6B 3Z-D/"tO*Y(N-1>*
T6
1050 PRINT FOR 1=1 TO 10
1060 STOP
2000 OPEN /D10-2 DATA/*INPUT*1
2010 ON ENDFILE(l) GO TO 2150
2020 N=l
2030 DIM X(300)*Y(300)
2040 INPUT FROM UX(N>*Y(N)
2050 FUA1=0
2060 N=N+1
2070 INPUT FROM ltX(N>*Y(N>
2080 D=X(N-1)-X(N)
2090 X=X(N-D-D/2
2100 Y=(YCN> + Y(N-l»/2
2110 Fl=Fl+Xt2*Y*D
2120 A1=A1+Y*D
2130 N=N+1
2140 GO TO 2070
2150 CLOSE 1
2160 PRINT
2170 PRINT "TOTAL FORCE="lFl*P*Vlt2/G
2180 PRINT
2190 PRINT "CALC FORCE 0»S.STRESS"
2200 PRINT " "
2210 RETURN
3000 IF S2<100 THEN 3030
3010 Q=l
3020 RETURN
3030 S3=S3+1
3040 IF ABS(P*WT2/G*F1-T1X10 THEN 3230
D-13
-------
S4=l ELSE S4=-l
3100 ELSE 3120
"»T2
"»T2
PROGRAM /F10-2 FW STRESS/
3050 IF S3>1 THEN 3080
3060 IF TKP*Wf2/G*Fl THEN
3070 GO TO 3130
3080 IF S2 = -l THEN 3130
3090 IF Tl
-------
DATE OF RUN: 10/08 16»34
PROGRAM / 10.1.2 FWSTRESS/
FLYWHEEL PARAMETERS-*
SPEED 2512 RAD/SEC
MATERIAL DENSITY .285 LBS/CU-IN
POISSONS RATIO .3
YOUNGS MODULUS 30.0E+06
TEMP. EXP. COEFFICIENT 7.3E-06
TOTAL TEMP. DIFFERENCE »0 DEG-F
OUTSIDE RADIUS 10-2230 INCHES
FLYWHEEL DESIGN FACTORS
MEAN TANGENTIAL STRESS 116867 PSI
MOMENT OF INERTIA 5.049 LB-INt2-SEC
KINETIC ENERGY 15931389 IN-LBS
5000 KWH
WEIGHT 46.40 LB
ENERGY DENSITY 343325.93 IN-LBS/LB
10.8 WH/LB
SHAPE FACTOR .7547
RADIUS THICKNESS TEMPERATURE TANGENTIAL RADIAL
(IN) (IN) (DEG-F) STRESS STRESS
0
.0
>0
.0
>0
.0
.0
.0
0
0
.0
.0
.0
.0
0
.0
.0
.0
0
62032
71740
80151
87458
93818
99362
104201
108425
112110
115320
118109
120522
122595
124360
125843
127065
128042
128787
129306
129566
0
19683
36515
50947
63352
74035
83249
91200
98064
103983
109076
113444
117167
120314
122942
125095
126811
128119
129045
129643
D-15
-------
Appendix E
ANALYSIS OF FLYWHEEL/HYBRID FAMILY CAR PERFORMANCE IN
DOT HIGH-SPEED PASS MANEUVER
An analysis of the family car with a flywheel/heat-engine hybrid drive in repetitive
cycles of the DOT High-Speed Pass Maneuver was conducted.
The DOT High-Speed Pass Maneuver is described in the "NAPCA Vehicle Design
Goals Six-Passenger Automobile (Revision A, November 30, 1970)" as follows:
"The maximum time and maximum distance to go from an initial velocity
of 50 mph with the front of the vehicle 100 feet behind the back of a 55 foot
truck traveling at a constant 50 mph to a position where the back of the
vehicle is 100 feet in front of the front of the 55 foot truck is, 15 sec. and
1400 ft. The entire maneuver takes place in a traffic lane adjacent to the
lane in which the truck is operated. The vehicle speed during the maneu-
ver cannot exceed 80 mph. Ambient conditions are 14. 7 psia, 85° F. Vehi-
cle weight is W , and acceleration is on level grade."
E.I INITIAL ASSUMPTIONS
The family car hybrid drive was assumed to have the following characteristics in
accordance with the aforementioned "Vehicle Design Goals":
Family car length = 20 ft
Continuous hp available to road = 65.9 hp
Maximum hp to road during constant hp acceleration = 184 hp
Flywheel energy storage capacity =391 w-hr
Maximum flywheel energy required for acceleration = 355 w-hr
Vehicle weight = 4000 Ib
Maximum transmission hp output = 184 hp
Average transmission efficiency during maneuver = 90%
Minimum deceleration (braking) time = 1/3 acceleration time
E-l
-------
c~
E.2 ANALYSIS
Two boundary cases were analyzed to verify the suitability of the flywheel/heat-engine
drive for repetitive maneuvers. The first case represents the poorest vehicle per-
formance that will meet the maneuver criteria. Here, the family car is accelerated
at constant horsepower such that it travels exactly 375 ft further than the truck in
L5 sec. The car is then decelerated to 50 mph in 5 sec, at which time the maneuver
is repeated.
E. 3 CASE I CALCULATIONS
Based on the Total Kinetic Energy (TKE) concept, the available flywheel kinetic energy
(KE) at the start of the maneuver is:
Initial Flywheel KE =
1 -
(initial velocity)
2
(final velocity)
(Max KE Reqd) + Reserve KE
or
Initial Flywheel KE =
1 - ^1(355) +
(80)2J
36 = 252 w-hr
During acceleration, the rolling, drag, and transmission losses are supplied by the
heat engine. At end of acceleration, velocity is 78.5 mph and remaining flywheel
energy is
KEf = 355
1 -
(78.
(80)2 J
+ 36 = 49 w-hr
The recuperated vehicle kinetic energy during deceleration (without increasing trans-
mission hp) then is
E-2
-------
_ 184 (0. 9) (746) (5) = 12
KEd " 3600 7 W
In addition, the heat engine can simultaneously supply energy to the flywheel as follows
_ 65. 9 (746) (5) _ .
KEeng-fw ' (0.9)3600 " 76 W"hr
Thus, the total energy available to the flywheel at the end of the maneuver is 297 w-hr.
However, the TKE control would hold the flywheel at approximately 252 w-hr. Thus,
the maneuver could be repeated continuously.
E.4 CASE 2 CALCULATIONS
The second case analyzed was for the highest performance available, by using the fly-
wheel and heat engine to supply a constant 184 hp to the road until the car is 375 ft
ahead of the truck. The time of 1/3 the acceleration period is then used for recupe ra-
tive braking. This scenario would be the most probable in actual two-lane road driving
operations since the exposure is minimized.
Again, as in Case 1, the initial flywheel KE is 252 w-hr at the start of the maneuver.
The acceleration time (at constant hp) to reach a point 375 ft ahead of the truck is
11.5 sec and the final velocity reached is 80 mph.
The flywheel KE at the end of acceleration is
KEf = 0 + 36 = 36 w-hr
The energy recuperated during braking is
. 184(0..KM) P. ») . JM w.hr
E-3
-------
The make-up energy simultaneously available during deceleration from the engine is
KE = 65.9(746)0.83) =
eng-fw (0.9) (3600) 5H w hr
Thus, the available energy to the flywheel is 226 w-hr. Since only 216 w-hr of fly-
wheel energy is required to repeat the maneuver immediately, the flywheel status is
sufficient for an identical maneuver on the second pass. On the third and successive
repetitive pass maneuvers without dwell, the time to pass is increased from 11.5 sec
to 13 sec (still well within the 15-sec limit). In order to provide for repetitive high-
performance pass maneuvers up to the 200-mile family .car range, the imposition of a
dwell time of 0. 5 sec after the second and subsequent repetitive pass maneuvers is
required.
E-4
-------