- Email: [email protected]
Systems design of electric automobiles
Transpn Res. Vol. I, pp. 3-19. Pergamon Press 1967. Printedin Great Britain
SYSTEMS
DESIGN
OF ELECTRIC
AUTOMOBILES
GEORGE A. HOFFMAN University of California, Los Angeles (Received 3 November 1966) 1. INTRODUCTION
: WHY
ELECTRIC
THE AUTOMOBILEwith
AUTOMOBILES?
its internal combustion engine accounts today for most of the passenger transportation in the United States and a rapidly growing portion of European travel. A systems design of electric passenger cars, with the piston engine and gasoline tank replaced by batteries and electric motors, appears therefore to be appropriate for this commencement of an international journal on transportation research. The internal combustion engine was not always a favorite energy-conversion device for propelling passenger cars. At the turn of the century there were more battery-operated electric motor-cars in use in the United States than either steam or gasoline powered. But the severe range and speed limitations of storage batteries in those days soon doomed the electric car to oblivion. Quantity demand and production of electric automobiles ended one-third of a century ago in the United States and never reached sign&ant proportions elsewhere. But in the last decade or so some automotive trends specifically favorable to the reconsideration of electrically driven passenger cars have developed. For example : (1) Electric motor design has progressed very rapidly in recent years. Improvements in electromechanical conversion efficiency and in weight reduction are now at the point where the electric motor merits reinvestigation for automobile traction. (2) In the past decade, the weight of batteries and regenerative fuel cells per unit of stored energy has dropped to a small fraction of their pre-war value. (3) The large increase in city air pollution from the exhausts of the internal combustion engine has become a serious international problem. The socioeconomic losses due to degrading the quality of the urban air is forcing the installation of elaborate exhaust control devices on cars. Mamtfacturing and upkeep cost of future controldevices could become a sizable fraction of today’s engine and fuel costs. Battery-operated electric cars do not contribute signiftcantly to air contamination. (4) The demand for cars is increasing globally with a related proliferation in diversity of automobile models and their personalized use. The rate of increase is greatest for acquiring the first car in the European family, while the rate is steepest for the second car in the United States family. In most cases the automobile is used either for commuting or for household-type trips and is characterized by frequent missions of short range. These suburban cars appear the most readily adaptable to electrical conversion. (5) In the long run, consumer prices for gas and oil are rising proportionately faster than the price of electricity, and will do so for the foreseeable future. Electric propulsion of ground vehicles is therefore steadily becoming more attractive economically. These trends have prompted automotive manufacturers in Great Britain and the United States to reopen recently research and development of battery-operated automobiles after a hiatus of almost half a century, and for these reasons the design of electric cars is reviewed here. 3
D
F
&
KILOGRAMS
KILOGRAMS
ENGINE
TRANSMISSION
WEIGHT
WEIGHT
(POUNDS)
(POUNDS)
KILOGRAMS
TRIM
BODY
WEIGHT
WEIGHT
(POUNDS)
(POUNDS)
Systems design of electric automobiles
5
It is therefore the starting premise of this system design that electric cars should be engineered to resemble or excel present-day cars in most of the following respects: (1) General appearance and diversity of models; (2) Convenience, comfort, passenger capacity and protection, interior design; (3) Performance, top speeds; (4) Handling, agility, ride; (5) Range between refueling; (6) Costs, initial and operating. After a century of development, the weight composition of automobiles has been dictated by the consumer to reflect the above six points and others. For the great variety of cars on the road today, ranging in curb weight? from 1500 lb to over 5000 lb (from 700 kg to 2200 kg), the weight composition is remarkably uniform, both as to proportions of weight and as to linearity with curb weight. Figure 1 is an illustration of the consistency of the weight ratios of major component categories of modem passenger cars designed for the above set of criteria. TABLE 1. WEIGHTAPPOR~ONMENT OF 1966-MODELPASSENGER CARS Component weight divided by curb weight Component Body Trim Glass Engine Transmission Suspensiont Front Rear Wheels Tires Brakes Steering unit Rear axle, driveline? Exhaust system Battery Radiator1 Fuel tank
Mean value
Least value
0.330 0.142 0.032 0.151 0.047
0.243 O-106 o-022 O-118 o-039
0.033 0,027 o-025 0.032 O-038 O-016 o-043 o-01 3 0.013 0.014 0.044
0.025 0.016 0.016 0.026 0.031 0.012 O-036 0.008 0.010 0.007 0.035
t Front-engine, rear-drive models only. $ Water-cooled engines only.
These ratios are listed in Table 1 for sixteen component subgroupings and are the starting point for a successful design of electric cars. Basing the design on the ratios shown in Table 1, assures that the driver at the wheel and his passengers perceive little difference in driving, or riding in, an electric auto versus a conventional car of comparable dimensions. This undifferentiability either from the interior or the exterior is figuratively shown for a typical car in Fig. 2. The f&t step in synthesizing electric cars is the elimination of those components that are not required in electric motor propulsion and retaining all others. The component weight ratio of these retained essential parts of the electric car are shown in the right-hand column t The curb weight is defined as the weight of a fully fueled vehicle, awaiting at the curb and ready to go, but yet lacking its driver and passengers.
6
GEORCJE A. HOFFMAN
of Table 2 to be almost identical to those in the right-hand column in Table 1. The leastvalue ratio was assumed representative of the better-engineered and least-weight products of today. Using least-value ratios, rather than mean values, also reflects the design improvements that are possible when rearranging an automobile for battery operation and electric motors. These advantageous side-effects derived from electric propu@ion are listed in the middle column of Table 2.
CONVENTIONAL
LLLCTNIC
FIG. 2. Diagrammatic comparison between a conventional and an electric automobile. TABLE 2. Warc%rrAPPORTIONMENTINELECIRICAUTOMOBILES
Component Body structure Trim Glass Front suspension Rear suspension Wheels Tires Steering Electric motors Batteries
Alterations possible from conventional counterpart Gage reductions in frame and chassii allowed by redistribution of concentrated weights. Elimination of mid-floor transmission hump and driveline tunnel Reduction of acoustical and thermal insulation. Simplification of dashboard furnishings and instruments. Elimination of air intake grillwork No alterations Equal front and rear weight distribution. Low c-g. battery pack clustered near spring-body junction points Same as above No change No change Lighter steering mechanism from low c.g., equal front and aft weight, 4-wheel traction Elimination of piston-engine, transmission, brakes, axle and driveline, exhaust system, battery, radiator, fuel tank and future exhaust-controls
Weight ratio 024 0.11 0.022 0.022 0022 0.016 0.026 t-ho12 x Y
Sum: 1+30= X+ Y+@47.
Systems design of electric automobiles
7
Thus one can predict approximately the balanced composition of weights in future electric cars as portions of curb weight as in Table 2. The weight of electric motors divided by curb weight is indicated by X, an unknown to be later determined from performance and speed requirements. The ratio (battery weight)/(curb weight) is denoted by Y, which simply equals 053-X. In sum, it appears that slightly more than half of the gross weight of electric cars could be made up of electric motors and energy storage and delivery devices without compromising any other design features of the vehicle. 3. MOTORS,
PERFORMANCE
AND
SPEED
As mentioned earlier, the performance (or acceleration capability) of an acceptable electric car should match that of a comparable weight modem automobile. The performance of passenger cars is often measured by the time required to accelerate from standstill to a desired speed. Since most automobiles are and will be used in or near cities, the performance that electric cars should match might be the acceleration capability of present-day cars at some speed representative of metropolitan travel, say 30 m.p.h. (48 km@). This implies a requirement that the electric-motor power available at half of this speed, 15 m.p.h. be the same as in gasoline-powered cars. Matching horsepowers at 15 m.p.h. actually provides electric cars with better-than-conventional performance up to 15 m.p.h., comparable perperformance for formance for O-30 m.p.h., and somewhat less-than-conventional O-60 m.p.h. acceleration. The horsepower available at the wheels of latest-model cars while accelerating through 15 m.p.h. is shown in Fig. 3 and generally amounts to one-third to one-half of maximum published horsepower. This horsepower is calculated from the manufacturers’ published data about the maximum engine torque, the automatic transmission’s torque-converting ratio at stall (or for manual shift models, the gearbox input/output speed ratio in first), the differential gear ratio at the axle and the number of tire revolutions per mile. The formula for estimating the horsepower at 15 m.p.h., (h.p.,a) is h.p., = 4.76 x 10d x (engine maximum torque) x (converter stall ratio) x (axle ratio) x tire revolutions/mile) where 4.76~10~=~~2~~
,
33,000 = (ft-lb/min)/h.p. 4 = (tire rev/mile)/(tire rev/min at 15 m.p.h.) 2?r = radjrev. This estimate of the available power assumes none of the various torque degradations commonly encountered in engines run under actual, rather than laboratory-controlled conditions, and amounting at times to a one-quarter power loss. These were the torque deteriorations to be expected in electric motors when in general use for automotive propulsion. It may be concluded from the plots in Fig. 3 that electric motors should be selected so as to impart at 15 m.p.h. h-p.,, = (0.02 + OXlO4)W (1) the lower figure being close to the mean of standard-performance passenger cars with curb weight W (lb), and the higher figure for medium- to high-performance cars. The weight and power of some electric motors are plotted in Fig. 4, based on the manufacturers’ specifications. Motors were assumed to be down-geared so as to turn the car’s wheels at a top speed of 1200 rev/min (about 90 or 100 m.p.h.). The horsepower developed was at one-sixth of this maximum rev/mm, namely around 15 m.p.h. The
GEORGE A. HOFFMAN
8
I
Oo Curb
2Coo weight,
FIG. 3. Weight and horsepower
I I
IO
W
( lb
4000 1
(
available at 15 m.p.h. in late-model
100
cars.
1000
ELECTRIC MOTOR WEIGHT, M, (LB)
FIG. 4. Weight and power of electric motors. 0, Automotive starters and motors; 0, 24,000 rev/min polyphase squirrel-cage induction motor, driven frequency converter, pulsed operation, oil cooled; + , train and auto a.c. traction prietary; A, industrial a.c. motors, uncooled, continuous use; V, industrial uncooled, continuous use.
d.c. traction by variable motors, prodc. motors,
Systems design of electtic automobiles
9
duration and frequency of the power pulses were assumed to be patterned after suburban and light-trafhc driving conditions to be discussed later in this study. The automotive class of motors shown in Fig. 4 are polyphase squirrel-cage induction types, oil-cooled and driven with a continuously variable frequency converter. One is an “in&the-wheel” motor presently used in the integral motor-wheel drive of an experimental army truck and off-road heavy vehicles (Slabiak, 1966). Another is the drive motor of the demonstration electric car of an automotive manufacturer (Marks, 1966). These automotive electric-traction motors weigh about 1.2 lb/h.p. and compare favorably with aircraft electric power systems that can weigh as little as O-8lb/h.p. Lightweight homopolar d.c. motors are also reported under development in England, ranging down to 1 lb/h.p. (Huntington, 1966). With the motor weight to power ratio of 1.2 lb/h.p., the power delivery of these future automotive-traction motors then is h.p. = 0.833M
(2)
where M (lb) is the total weight of the electric motors in each car. Combining expression (1) with (2) (that is, matching the initial O-30 m.p.h. performance of modem gasoline engines) gives for the electric motor’s weight M= (0.024 f 0~005) W
In other words, about 3 per cent of the weight of electric cars should be assigned to traction motors to achieve intermediate levels of accelerative performance. Assigning another 1 per cent of the vehicle’s weight to motor controls and static converters gives a value of 0.04 for X in Table 2. This leaves Y to be O-49, that is: almost half of the car’s weight could be assigned to batteries if so desired. It is interesting to note here that the torque performance of the electric motors (when driven at constant power beyond 15-25 m.p.h.) parallels quite satisfactorily the wheel-axle torque vs. speed curves of conventional internal combustion engine automobiles up to moderate highway speeds (see Fig. 5). This eliminates the problem of driver readaptation to electric propulsion in city and suburban traffic. The characteristics of the lighter motors in Fig. 4 seem to assure that the electric car responds and performs as a piston-enginepowered car would to the driver’s acceleration-pedal demands under the most frequent stop-and-go traffic conditions. In Fig. 5 the automatic-transmission curve (dashed) is a composite of the manufacturer’s data on the latest model cars with standard option engines, whereas the low-slip electric induction-motor curves are based on assumed operation at constant wattage. The lower electric-motor curve represents a standard performance electric vehicle, while the upper curve stands for an intermediate-to-high performance option. The polyphase squirrel-cage induction motor regulated for constant power by a cycle converter offers several automotive advantages such as being brushless and requiring no commutation, i.e. it promises to be almost maintenance-free during an automobile’s normal life. Furthermore, its propulsive power capability is linear with its weight, eliminating many restraints on the automotive engineer’s choice of number of motors per vehicle. Thus he need not contend with weight differences when considering whether to install only one electric motor (to drive the front wheels), two electric motors (one in front and one in the rear) or four electric motors (each driving a wheel). If the four-motor scheme is selected, the suspension engineer faces two alternatives: attaching each motor to the frame and driving the wheel through an axle with two universal joints, or having the motor integral within the rim of the wheel and driving it through planetary or other type reduction gears. This last arrangement is the most felicitious on first observation, but integrating the motor
GEOROE A. HOFFMAN
10
with the wheel may result in a harsh ride and unpleasant jounce. These might be attenuated by radically redesigning the suspension to account for the doubling of the present unsprung weight when adding 1 per cent of W (each motor) to each wheel-tire assembly. Unusual problems of heat dissipation into the tires from the motor would also have to be faced. It is our opinion that the space savings, simplicity and uncluttered design afforded by the motor-in-the-wheel will eventually outweigh its drawbacks. motor
It-
ol 0
20 vehicle
40 speed,
drivrn
v
at
60
, (mph)
80
100
120
,
0 FIG. 5.
“9
km h
k’:o/h
Wheel torque vs. speed characteristics of conventional automobiles and constant-power electric-motor-driven cars.
The maximum speed of electric automobiles on a level road can now be calculated by equating the constant propulsive power delivered by the motors with the power required to overcome aerodynamic drag, tire rolling resistance and any other frictional dissipative forces (the last being negligibly small compared to aerodynamic and tire resistance). The tire rolling drag force of passenger-car tires is between 0.01 and 0.02 of the curb weight and increases with speed. For good-quality properly inflated tires (the lower bound of the band in Fig. 6), the rolling tractive force is about (O-01+ 5 x 1O”v) x w
(3)
The aerodynamic drag force at 60 m.p.h. of various 1960-model vehicles is shown in Fig. 7, indicating that this resistance for recently styled automobiles may be assumed to be 30+0~015(w+
150)
(4)
where 30 lb is the drag force at 60 m.p.h. of a driver’s body reclining, as in the car seat, and 150 lb is the weight of the driver. Aerodynamic drag of cars is quadratic with speed, so that expression (4) may be multiplied by (~/60)~ to yield the drag-speed relation of automobiles to be [0.0083 + 4.2 x 103 W+ 1SO)]V* (5) The tota force (lb) resisting automobile motion- i.e. the sum of expressions (3) and (5)-when multiplied by the speed, u, and the appropriate dimensional conversion factors,
Systems design of electric automobiles
11
is close to the total automotive drag power required from the electric motors. This tractive power requirement is then 0.00267 x ((0.01 + 5 x 10” v) x W+ [09083 f4.2 x 10d( W-t 1So)] u”) x D where o.m267
=
(6)
lai7 Wec)/m.p.h. 550 [(ft-lb)/sec]/h.p.
01
I
0 FIG.
1
1
20 rprrd,
v,
40 (mph)
80
60
100
6. Passenger car tire characteristics at speed.
200 r
01
1
1
0
I
.
curb FIG.
wright
plur
onr
I
4ooo
2000 drivrr
61
(lb)
7. Weight and aerodynamic drag of vehicles at 60 m.p.h.
At the top speed of the car, t), the required horsepower in (6) (with V substituted for u) matches and balances the motor deliverable power, presented in expression (l), to average about 0.02 W (h.p.) (7) for standard to intermediate performance vehicles.
12
GEORGE
A. HO-
Equating expressions (6) and (7), rearranging terms and rounding off numbers gives the equation of the maximum speed, V (m.p.h.), to be
The real solution of equation (8) is a function of W whose intricacies are beyond the scope of this paper. It is suflkient to state that the solution of (8) varies from V = 89 m.p.h. for W= 2000 lb to V = 100 m.p.h. for W= 5000 lb, and that 2000~ Wc5000 is a reasonable expectation for the future mass-accepted electric cars. It appears then that matching weight composition and initial performance of gasolinepowered cars can result in the design of future electric automobiles capable of top speeds of about 90-100 m.p.h. (145-160 km/hr) when allotting 4 per cent of the curb weight to the most appropriate electric-traction motors. A maximum speed of 100 m.p.h. is quite respectable and useful for most driving conditions presently envisionable, particularly when it is available in conjunction with accelerative performance quite close to that of the competitor vehicle. 4. CHOICES
OF BATTERIES
AND
RANGES
As seen in the last section, Y, the weight portion allocable for electrical energy storage and delivery, is left to be about half of the curb weight-specifically Y = 0.49. Three types of electrochemical energy-storage and delivery systems are considered here as taking up this half of the curb weight: 1. Conventional rechargeable batteries, currently available for commercial and aerospace applications. 2. Regenerative fuel cells, with metal fuels and ambient air as the oxidizer source (hereafter called metal-air batteries for brevity). 3. Future high-energy batteries, presently only in the predevelopmental stage. Some use molten-salt electrolytes, while others are theoretically operable at room temperatures. Table 3 lists the salient characteristics of industrial batteries, on the assumption that an intensive upgrading effort aimed at a life of a thousand or two charge-discharge cycles would be successful in adapting these batteries for powering automobiles in the near future. Next, Table 3 presents (for illustrative purposes only) the estimated properties of two hypothetical metal-air batteries. The first, the Zn-air battery is hypothetical in the same sense that it requires much further development and refinement before it can qualify for mass-production and automotive use, even though it is operative today and a prototype TABLE 3. PROPERTIES OF CANDIDATE BATTERIES FOR ELECTRIC AUTOMOBILES
Type
Symbol
(a) Industrial batteries Lead-acid Pb-acid Nickel-cadmium Ni-Cd Silver-zinc Ag-Zn (b) Hypothetical metal-air batteries Zinc-air Zn-air Metal-air (c) Hot batteries Sodium-sulphur Na-S Lithium+hlorine Li-Cl
Energy density d (W-hr/lb)
Discharge time (hr)
Source for value of energy density, d
10 20 30
0.5 1 1.5
Manufacturers’ data (Shair, 1966a) (Shair, 1966a)
50 70
3 4
(Shipps, 1966a) (Charkey and Dalin, 1966)
100 100
6 6
(Ford Motor Company, 1966) (General Motors Company, 1966)
Systems design of electric automobiles
13
be running a vehicle this year or next. In the second entry of the metal-air couple systems, the metal may be magnesium, calcium,. iron or lithium. Rechargeable batteries of this kind are hypothetical since they are yet to be designed or successfully cycled, and the value shown for their energy density is only our guess as to its eventual level, after development in the near future. A flow diagram of hypothetical meta-air batteries is given in Fig. 8, with a description of its modus operandi. might
BATTERY
TERMINALS
--n
d.“.o .o.o
““:iNG >
EXCESS
n+
AIR
OXIDE FILTER TANK
. .
0. . .o.o.
COMPRESSOR
LlOUlD ELECTROLYTE TANK
FIG. 8. Diagram of a metal-air battery. The metal-air battery derives energy from the conversion of a metal to its oxide on exposure to pressurized air. The flow during a discharge cycle is depicted schematically, with the oxide particles represented as dots. For recharging, a voltage is applied across the terminals; the flow of particles is reversed so that they are carried by the electrolyte into the cell and electrolyzed into metal and oxygen. The metal is redeposited on the anode and the oxygen bubbles off at the cathode and out of the excess-air vent.
The last class of batteries in Table 3 are labelled “hot batteries” since they require elevated temperatures to melt the electrolyte and to initiate their functioning, (about 300°C operating temperature for Na-S batteries and 600°C for L&Cl cells). This class is representative of the higher-energy cells under investigation in the automotive industry. Their energy density as displayed in Table 3 is lower than the most optimistic industrial estimates and again reflects our guess as to their actual capability after a few years development. Aside from the obvious drawback of high-temperature operation only, these batteries utilize liquid metals that react explosively with air, and require highly toxic substances in large quantities. Hydrocarbon and several other fuel cells are omitted from Table 3, either because of their non-regenerative nature or because of their excessively low-energy densities, even though much of the research activity in direct energy conversion is in this area (Barak, 1966). These types of fuel cells are quite feasible as electric power generating units for cars, in the intermediate run, preceding the ultimate advent of lightweight low-cost batteries.
14
GEORCSB A. HOFFMAN
Also omitted from Table 3 are the newer lithium batteries claimed to have achieved or surpassed 100 W&r/lb (1.6 x 10s J/kg) in the laboratory (l&e&erg, 1966; Shair, 1966b). Commerical availability of lithium batteries in quantity seems to be several years away, and capability for thousands of discharge-charge cycles is yet to be demonstrated. Thus came our temporary omission of lithium batteries from this report, but not necessarily from considerations for more distant future possibilities. In all entries of Table 3 the energy density, d, depends on the frequency and length of deep-discharge time. The time for discharge shown is assumed roughly to be the anticipated range divided by a representative mean speed. The distance a battery-operated car can travel before requiring either a recharge or refueling is here defined as the range of the vehicle. The range is determined by the maximum energy deliverable by the motor, which is
0.92 YWd
or
0.45 Wd W-hr
(9)
where 0.92 is our assumption for the average motor efficiency in converting electrical energy to mechanical work. Homopolar motors so far exhibit lower efficiency than this value, while induction motors are often more efficient than assumed here. The energy in (9) is expended in the work done by the motors to overcome tire and aerodynamic drag. This work is the sum of the drag forces-expressions (3) and (5)-times the range, R, when the car is traveling at a steady speed, V, on a level road on a windless day. This work @lb) equals ((0.01 + 5 x 10d u) W+ [0*0083+ 4.2 x lO-@(W+ 150)] us}x (R, miles) x (5280 ft/mile)
(10)
Equating energy and work, i.e. expressions (9) and (lo), gives the range of electric cars driven at steady speed to be d R* mile = oG442+2*21 x 10-%+[(0~395/FV)+ 1*86x lO-s]uz The range under this purely fictitious driving condition is shown in Fig. 9. The width of the band is representative of the dependence of R on W, the lower bound being for W = 2000 lb and the upper for W = 3000 lb. More realistic travel simulations than steady-speed driving are needed, particularly to account for accelerations and decelerations. Two such driving conditions are assumed as shown in Fig. 10. Condition I represents urban travel in moderately congested tra& with considerable stop-and-go driving. Condition II defines the pattern for suburban driving in moderately congested trafhc with considerable stop-and-go driving. Condition II defines light tra5*not on expressways. These assumptions were modelled after the actual record of engine utilixation in a variety of urban settings. With the mission profiles assumed in Fig. 10, the energy expended per trip, in steadyspeed cruising, can be shown to be rather small (a twenty-fifth and less) compared to the energy required per trip for acceleration or grade-climbing. The steady-speed energy requirements may then be neglected without introducing much error and can be compensated by the stipulation that the work done in decelerating is dissipated in heat-t This leaves only the acceleration and hill-climbing phases to be considered, conservatively t Dynamic braking by the electric motors was assumed here to be available down to speeds of 5-10 m.p.h. A friction brake would take over below these speeds for full stops and parking. Regenerative braking (i.e. battery-charging when decelerating) is another scheme that merits attention for energyrecovery in electric automobiles. It is not incorporated in thll design study because it extends the range by only a quarter or less, but at some penal@ ln control intricacies and costs, particularly for circulating electrolyte batterles. Regenerative braking is certainly technologically feasible at this time for electric cars with conventional batteries.
Systems design of electric automobiles
15
to be done at close to the full power level. (By contrast, in Conditions I and II traffic, the median engine utilization by late-model cars is usuahy only about half of the available engine power.) The full power assumption should also compensate for the omission of the cruising power from the following calculations and should account for slightly lessthan-average performance of electric automobiles at 30-60 m.p.h. assumed
0
50 ’
km/h
100 ’
I50 km/h YtF
500 km
20 sprod FIG.
9. Rans
40
, v,
60
80
IOZ
(mph)
of electric cars driven at steady speed.
FIG. 10. Assumed engine utilization in city-driving situations.
The energy expended between recharges when traveling in the metropolis is then (O-02IV) x (0.352”) x 3600 x 550/2656
where 0.02 IV = O-35 = T=
3600 = 550 = 2656 =
or
522 WT W-hr maximum available motor power, from expression (1) portion of Conditions I and II trip time spent in acceleration travel time available between recharges or refuelings (hr) sec/hr (ft-lb/sec)/h.p. (ft-lb)/W-hr
(11)
GEORGEA. HOFFMAN
16
Equating (9) with (11) gives both the urban and suburban driving time between recharges to be T = 0.086d hr The variability in city-driving characteristics is greatest in the block speed (total trip length+ total trip time), achieved in a variety of settings and locations, tragic conditions and driver behavior. It is only too well known that block speeds of 20 m.p.h. and lower are experienced at peak traffic hours in densely populated regions or under inclement weather, while block speeds of 40 m.p.h. or higher can occasionally be achieved in some suburban driving at off-peak times and in good weather. The range of city-driven electric cars is therefore presented in Fig. 11 for a variety of block speeds, this range being simply T x (block speed).
IS0
range
200
, nim )
km
100
battery
energy
density, d,
(W-h/lb)
FIG. 11. Range of city-driven electric automobiles.
Range was only slightly affected by Win the steady-speed cruise mode, and now-under metropolitan driving conditions-the range of electric automobiles seems independent of the vehicle overall weight. One may gather from this observation that the present spectrum of car weights (from 2000 to 5000 lb) and cotigurations (from small to large-luxurious) offered to the consumer for his choice need not change with the substitution of electric energy for combustion. The prospective electric car owner of the future need not consider vehicle size and weight when selecting the range between recharges that he can tolerate. This is analogous with today’s designs where the smallest to the largest passenger cars have more or less comparable ranges between refuelings. Figures 10 and 11 imply that conventional batteries are presently unsuitable for many consumer range demands. Much better suited are the metal-air and other yet to be developed high-energy density batteries since these promise for the future quite respectable ranges between refueling and recharging. It should be, noted that the mission profiles in Fig. 10 are more arduous than one can normally expect. A driver’s more moderate use of the accelerator pedal than the full power assumed in (1 l), and travel along routes with few stop signs or with well-coordinated lights, could result in ranges up to twice those exhibited in Fig. 11. They approach and could surpass the ranges of gasoline-powered cars with a single tankful of fuel. In conclusion : electric cars of adequate performance and top speed could be designed in our lifetime for ranges of 200-300 km if metal-air batteries are successfully adaptable to automotive use. In the more distant future higher energy batteries could provide twice such ranges.
Systemsdesignof electricautomobiles 5. SOCIO-ECONOMIC
EFFECTS OF THE FUTURE AUTOMOBILES
17 USE OF ELECTRIC
A limited sample of the societal implications of all-electric automotive propulsion in the future is investigated in this section. The first item is the forecast of a probable schedule for the introduction of electric cars into the automotive market with u priori assumption that the vehicle will prove comparable in design and lifetime costs with the non-electric cars extant at the time. High-energy batteries
>R8D)
Motors and >Y> controls Ntimbrr of prototyprs producrd
I
Full occrptancr
FIG. 12. Possiblescheduleof introductionfor electricautomobiles.
One such schedule was worked out in Fig. 12 for either the United States or the panEuropean market. It is based on factoring mass-production and mass-acceptance of radically innovated cars by two orders of magnitude every 4 years. In other words, stipulating ten prototypes operating at the end of this decade with conventional speeds and performances (but not yet range), we escalated the yearly production one hundredfold every 4 years. With such a sustained enormous program of expansion only by the mid-80’s could the industry produce electric cars at a yearly rate comparable to that of gasoline-engine cars. Even after the electric automobile’s complete acceptance by that time, it will take another dozen or so years for the extant combustion-engine vehicles to wear out and become extinct. 2
18
GEORW A. HOFFMAN
Under these &cumstances, it appears from Fig. 12 that only toward the turn of thy twenty-tist century could one expect extensive use of electric automobiles in our dail travel activities. The palpable economic effects on the consumer’s pocketbook will be important Consider first the initial (and, later, the operating) costs to the car owner in terms of the difference between electric and non-electric cars, both produced at rates of, say, 106 to 10 units per year. The production unit cost--and therefore price per unit .weight-for half o. the car weight (structure, trim, suspension, wheels, tires) should be essentially the same fol both types of cars. Then the lighter-bodied electric car shows some price advantage hen (see Fig. 13). In the remaining half of the vehicle weight, a major cost jump will lo encountered, due to the cost of batteries being intrinsically higher per pound-perhaps 1.5 times?-than the cost of conventional car components. Thus, as a whole, electric cars might tend to be more expensive at least by a quarter than their conventional counterparu of similar size and weight. ELECTRIC
* 5 2 ENGINE
AND
AUTOMATIC
B 5
TRANSMISSION
8 WHEELS,
TIRES.
STEERING
5 u ELECTRIC
4
WHEELS,
MOTORS TIRES,
STEERING
Y 2 (13 E
STRUCTURE
Fro. 13. Comparative manufacturing costs of electric and gasoline automobiles.
The mass marketing of electric automobiles will require a special (but not too novel) task: to convince the prospective automobile buyers to pay considerably more at the time of the purchase of the new vehicle than they would pay for gasoline-engine cars; and to trade off this sign&ant increment (one-quarter of the initial cost) against the equally significant decrement in propective operating costs over the ensuing years of car use. The unchangeable operating costs will still be : the interest on the capital, tire replacement, road and highway taxes, maintenance, insurance and licensing fees. But a sizeable lowering in operating expenditures will be derived by switching from gasoline and oil to electric energy. A reduction of about one-halfin fuel costs might be expected. Integrated over the years of the car’s life, this would more than offset the initial purchase price increase. t Lead-acid batteries are one and a half times as expensive per pound as the average price/lb of the whole automobile. Zinc and nickel, candidate metals for fuel or electrodes in metel-air batteries, are cheaper per pound than lead, while lithium is several times as expensive.
Systems
design of electric automobiles
19
In essence : one prediction is that electric cars will be costlier to purchase and somewhat cheaper to operate, with the operational cost savings over the years adding up to a net benefit. Automobile manufacturers should look favorably to the mass-production and marketing of all-electric passenger cars that are comparable to present cars. The added options of highly intricate and sophisticated motors and advanced batteries being substituted for piston-engines and automatic transmissions might spell some handsome profits. While the oil and gas industry might be dismayed at the coming of the all-electric vehicle, electric utilities will welcome the advent of electric cars: the demand for electric power consumption would about double. The price of electricity might be reduced by one-tenth or one-fifth, because of this doubled capacity and in view of the heavy night-time power demands as batteries are being recharged at home. The mounting problems of urban air pollution now plaguing many cities and mainly due to the emissions from the internal combustion engine should be greatly alleviated by battery-operated cars. One of the most significant benefits from electric cars in the future might prove to be the abatement of automotive exhaust. These emissions are extracting from urban man an enormous toll in health and in the degradation of the quality of the cities’ air. REFERENCES BARAK M. (1966). Fuel cells-present position and outstanding problems. Adu. Energy Conversion 6, 29-55. CHARKEY A. and DALIN G. A. (1966). Metal-air battery systems. Proc. 20th Annual Power Sources Conf., pp. 19-83. EI~ENBER~M. (1966). Electrochimica Corporation Announcement, November. Ford Motor Company. (1966). News release of 3 October. General Motors Company. (1966). News release of 28 October. HOFFMANG. A. (1966). The electric automobile. Sci. Am. 215, 34-40. HUNTINGTONR. (1966). . . . Electric car project. Aatocar 125, 1006. MAR= C. (1966). General Motors news release of 28 October. SHAIIXR. C. (1966a). Sealed secondary cells for space power stations. J. Spacecr. & Rockets 3, 68-70. SHAIR R. C. (1966b). Gulton Industries news release of 5 October. SHIPPSP. R. (1966). Secondary metal-air system. Proc. 20th Annual Power Sources Co&, pp. 86-88. SLABUK W. (1966). An A.C. individual wheel drive system for land vehicles. Society of Automotive Engineers Paper 660134.
SYSTEMS
DESIGN
OF ELECTRIC
AUTOMOBILES
GEORGE A. HOFFMAN University of California, Los Angeles (Received 3 November 1966) 1. INTRODUCTION
: WHY
ELECTRIC
THE AUTOMOBILEwith
AUTOMOBILES?
its internal combustion engine accounts today for most of the passenger transportation in the United States and a rapidly growing portion of European travel. A systems design of electric passenger cars, with the piston engine and gasoline tank replaced by batteries and electric motors, appears therefore to be appropriate for this commencement of an international journal on transportation research. The internal combustion engine was not always a favorite energy-conversion device for propelling passenger cars. At the turn of the century there were more battery-operated electric motor-cars in use in the United States than either steam or gasoline powered. But the severe range and speed limitations of storage batteries in those days soon doomed the electric car to oblivion. Quantity demand and production of electric automobiles ended one-third of a century ago in the United States and never reached sign&ant proportions elsewhere. But in the last decade or so some automotive trends specifically favorable to the reconsideration of electrically driven passenger cars have developed. For example : (1) Electric motor design has progressed very rapidly in recent years. Improvements in electromechanical conversion efficiency and in weight reduction are now at the point where the electric motor merits reinvestigation for automobile traction. (2) In the past decade, the weight of batteries and regenerative fuel cells per unit of stored energy has dropped to a small fraction of their pre-war value. (3) The large increase in city air pollution from the exhausts of the internal combustion engine has become a serious international problem. The socioeconomic losses due to degrading the quality of the urban air is forcing the installation of elaborate exhaust control devices on cars. Mamtfacturing and upkeep cost of future controldevices could become a sizable fraction of today’s engine and fuel costs. Battery-operated electric cars do not contribute signiftcantly to air contamination. (4) The demand for cars is increasing globally with a related proliferation in diversity of automobile models and their personalized use. The rate of increase is greatest for acquiring the first car in the European family, while the rate is steepest for the second car in the United States family. In most cases the automobile is used either for commuting or for household-type trips and is characterized by frequent missions of short range. These suburban cars appear the most readily adaptable to electrical conversion. (5) In the long run, consumer prices for gas and oil are rising proportionately faster than the price of electricity, and will do so for the foreseeable future. Electric propulsion of ground vehicles is therefore steadily becoming more attractive economically. These trends have prompted automotive manufacturers in Great Britain and the United States to reopen recently research and development of battery-operated automobiles after a hiatus of almost half a century, and for these reasons the design of electric cars is reviewed here. 3
D
F
&
KILOGRAMS
KILOGRAMS
ENGINE
TRANSMISSION
WEIGHT
WEIGHT
(POUNDS)
(POUNDS)
KILOGRAMS
TRIM
BODY
WEIGHT
WEIGHT
(POUNDS)
(POUNDS)
Systems design of electric automobiles
5
It is therefore the starting premise of this system design that electric cars should be engineered to resemble or excel present-day cars in most of the following respects: (1) General appearance and diversity of models; (2) Convenience, comfort, passenger capacity and protection, interior design; (3) Performance, top speeds; (4) Handling, agility, ride; (5) Range between refueling; (6) Costs, initial and operating. After a century of development, the weight composition of automobiles has been dictated by the consumer to reflect the above six points and others. For the great variety of cars on the road today, ranging in curb weight? from 1500 lb to over 5000 lb (from 700 kg to 2200 kg), the weight composition is remarkably uniform, both as to proportions of weight and as to linearity with curb weight. Figure 1 is an illustration of the consistency of the weight ratios of major component categories of modem passenger cars designed for the above set of criteria. TABLE 1. WEIGHTAPPOR~ONMENT OF 1966-MODELPASSENGER CARS Component weight divided by curb weight Component Body Trim Glass Engine Transmission Suspensiont Front Rear Wheels Tires Brakes Steering unit Rear axle, driveline? Exhaust system Battery Radiator1 Fuel tank
Mean value
Least value
0.330 0.142 0.032 0.151 0.047
0.243 O-106 o-022 O-118 o-039
0.033 0,027 o-025 0.032 O-038 O-016 o-043 o-01 3 0.013 0.014 0.044
0.025 0.016 0.016 0.026 0.031 0.012 O-036 0.008 0.010 0.007 0.035
t Front-engine, rear-drive models only. $ Water-cooled engines only.
These ratios are listed in Table 1 for sixteen component subgroupings and are the starting point for a successful design of electric cars. Basing the design on the ratios shown in Table 1, assures that the driver at the wheel and his passengers perceive little difference in driving, or riding in, an electric auto versus a conventional car of comparable dimensions. This undifferentiability either from the interior or the exterior is figuratively shown for a typical car in Fig. 2. The f&t step in synthesizing electric cars is the elimination of those components that are not required in electric motor propulsion and retaining all others. The component weight ratio of these retained essential parts of the electric car are shown in the right-hand column t The curb weight is defined as the weight of a fully fueled vehicle, awaiting at the curb and ready to go, but yet lacking its driver and passengers.
6
GEORCJE A. HOFFMAN
of Table 2 to be almost identical to those in the right-hand column in Table 1. The leastvalue ratio was assumed representative of the better-engineered and least-weight products of today. Using least-value ratios, rather than mean values, also reflects the design improvements that are possible when rearranging an automobile for battery operation and electric motors. These advantageous side-effects derived from electric propu@ion are listed in the middle column of Table 2.
CONVENTIONAL
LLLCTNIC
FIG. 2. Diagrammatic comparison between a conventional and an electric automobile. TABLE 2. Warc%rrAPPORTIONMENTINELECIRICAUTOMOBILES
Component Body structure Trim Glass Front suspension Rear suspension Wheels Tires Steering Electric motors Batteries
Alterations possible from conventional counterpart Gage reductions in frame and chassii allowed by redistribution of concentrated weights. Elimination of mid-floor transmission hump and driveline tunnel Reduction of acoustical and thermal insulation. Simplification of dashboard furnishings and instruments. Elimination of air intake grillwork No alterations Equal front and rear weight distribution. Low c-g. battery pack clustered near spring-body junction points Same as above No change No change Lighter steering mechanism from low c.g., equal front and aft weight, 4-wheel traction Elimination of piston-engine, transmission, brakes, axle and driveline, exhaust system, battery, radiator, fuel tank and future exhaust-controls
Weight ratio 024 0.11 0.022 0.022 0022 0.016 0.026 t-ho12 x Y
Sum: 1+30= X+ Y+@47.
Systems design of electric automobiles
7
Thus one can predict approximately the balanced composition of weights in future electric cars as portions of curb weight as in Table 2. The weight of electric motors divided by curb weight is indicated by X, an unknown to be later determined from performance and speed requirements. The ratio (battery weight)/(curb weight) is denoted by Y, which simply equals 053-X. In sum, it appears that slightly more than half of the gross weight of electric cars could be made up of electric motors and energy storage and delivery devices without compromising any other design features of the vehicle. 3. MOTORS,
PERFORMANCE
AND
SPEED
As mentioned earlier, the performance (or acceleration capability) of an acceptable electric car should match that of a comparable weight modem automobile. The performance of passenger cars is often measured by the time required to accelerate from standstill to a desired speed. Since most automobiles are and will be used in or near cities, the performance that electric cars should match might be the acceleration capability of present-day cars at some speed representative of metropolitan travel, say 30 m.p.h. (48 km@). This implies a requirement that the electric-motor power available at half of this speed, 15 m.p.h. be the same as in gasoline-powered cars. Matching horsepowers at 15 m.p.h. actually provides electric cars with better-than-conventional performance up to 15 m.p.h., comparable perperformance for formance for O-30 m.p.h., and somewhat less-than-conventional O-60 m.p.h. acceleration. The horsepower available at the wheels of latest-model cars while accelerating through 15 m.p.h. is shown in Fig. 3 and generally amounts to one-third to one-half of maximum published horsepower. This horsepower is calculated from the manufacturers’ published data about the maximum engine torque, the automatic transmission’s torque-converting ratio at stall (or for manual shift models, the gearbox input/output speed ratio in first), the differential gear ratio at the axle and the number of tire revolutions per mile. The formula for estimating the horsepower at 15 m.p.h., (h.p.,a) is h.p., = 4.76 x 10d x (engine maximum torque) x (converter stall ratio) x (axle ratio) x tire revolutions/mile) where 4.76~10~=~~2~~
,
33,000 = (ft-lb/min)/h.p. 4 = (tire rev/mile)/(tire rev/min at 15 m.p.h.) 2?r = radjrev. This estimate of the available power assumes none of the various torque degradations commonly encountered in engines run under actual, rather than laboratory-controlled conditions, and amounting at times to a one-quarter power loss. These were the torque deteriorations to be expected in electric motors when in general use for automotive propulsion. It may be concluded from the plots in Fig. 3 that electric motors should be selected so as to impart at 15 m.p.h. h-p.,, = (0.02 + OXlO4)W (1) the lower figure being close to the mean of standard-performance passenger cars with curb weight W (lb), and the higher figure for medium- to high-performance cars. The weight and power of some electric motors are plotted in Fig. 4, based on the manufacturers’ specifications. Motors were assumed to be down-geared so as to turn the car’s wheels at a top speed of 1200 rev/min (about 90 or 100 m.p.h.). The horsepower developed was at one-sixth of this maximum rev/mm, namely around 15 m.p.h. The
GEORGE A. HOFFMAN
8
I
Oo Curb
2Coo weight,
FIG. 3. Weight and horsepower
I I
IO
W
( lb
4000 1
(
available at 15 m.p.h. in late-model
100
cars.
1000
ELECTRIC MOTOR WEIGHT, M, (LB)
FIG. 4. Weight and power of electric motors. 0, Automotive starters and motors; 0, 24,000 rev/min polyphase squirrel-cage induction motor, driven frequency converter, pulsed operation, oil cooled; + , train and auto a.c. traction prietary; A, industrial a.c. motors, uncooled, continuous use; V, industrial uncooled, continuous use.
d.c. traction by variable motors, prodc. motors,
Systems design of electtic automobiles
9
duration and frequency of the power pulses were assumed to be patterned after suburban and light-trafhc driving conditions to be discussed later in this study. The automotive class of motors shown in Fig. 4 are polyphase squirrel-cage induction types, oil-cooled and driven with a continuously variable frequency converter. One is an “in&the-wheel” motor presently used in the integral motor-wheel drive of an experimental army truck and off-road heavy vehicles (Slabiak, 1966). Another is the drive motor of the demonstration electric car of an automotive manufacturer (Marks, 1966). These automotive electric-traction motors weigh about 1.2 lb/h.p. and compare favorably with aircraft electric power systems that can weigh as little as O-8lb/h.p. Lightweight homopolar d.c. motors are also reported under development in England, ranging down to 1 lb/h.p. (Huntington, 1966). With the motor weight to power ratio of 1.2 lb/h.p., the power delivery of these future automotive-traction motors then is h.p. = 0.833M
(2)
where M (lb) is the total weight of the electric motors in each car. Combining expression (1) with (2) (that is, matching the initial O-30 m.p.h. performance of modem gasoline engines) gives for the electric motor’s weight M= (0.024 f 0~005) W
In other words, about 3 per cent of the weight of electric cars should be assigned to traction motors to achieve intermediate levels of accelerative performance. Assigning another 1 per cent of the vehicle’s weight to motor controls and static converters gives a value of 0.04 for X in Table 2. This leaves Y to be O-49, that is: almost half of the car’s weight could be assigned to batteries if so desired. It is interesting to note here that the torque performance of the electric motors (when driven at constant power beyond 15-25 m.p.h.) parallels quite satisfactorily the wheel-axle torque vs. speed curves of conventional internal combustion engine automobiles up to moderate highway speeds (see Fig. 5). This eliminates the problem of driver readaptation to electric propulsion in city and suburban traffic. The characteristics of the lighter motors in Fig. 4 seem to assure that the electric car responds and performs as a piston-enginepowered car would to the driver’s acceleration-pedal demands under the most frequent stop-and-go traffic conditions. In Fig. 5 the automatic-transmission curve (dashed) is a composite of the manufacturer’s data on the latest model cars with standard option engines, whereas the low-slip electric induction-motor curves are based on assumed operation at constant wattage. The lower electric-motor curve represents a standard performance electric vehicle, while the upper curve stands for an intermediate-to-high performance option. The polyphase squirrel-cage induction motor regulated for constant power by a cycle converter offers several automotive advantages such as being brushless and requiring no commutation, i.e. it promises to be almost maintenance-free during an automobile’s normal life. Furthermore, its propulsive power capability is linear with its weight, eliminating many restraints on the automotive engineer’s choice of number of motors per vehicle. Thus he need not contend with weight differences when considering whether to install only one electric motor (to drive the front wheels), two electric motors (one in front and one in the rear) or four electric motors (each driving a wheel). If the four-motor scheme is selected, the suspension engineer faces two alternatives: attaching each motor to the frame and driving the wheel through an axle with two universal joints, or having the motor integral within the rim of the wheel and driving it through planetary or other type reduction gears. This last arrangement is the most felicitious on first observation, but integrating the motor
GEOROE A. HOFFMAN
10
with the wheel may result in a harsh ride and unpleasant jounce. These might be attenuated by radically redesigning the suspension to account for the doubling of the present unsprung weight when adding 1 per cent of W (each motor) to each wheel-tire assembly. Unusual problems of heat dissipation into the tires from the motor would also have to be faced. It is our opinion that the space savings, simplicity and uncluttered design afforded by the motor-in-the-wheel will eventually outweigh its drawbacks. motor
It-
ol 0
20 vehicle
40 speed,
drivrn
v
at
60
, (mph)
80
100
120
,
0 FIG. 5.
“9
km h
k’:o/h
Wheel torque vs. speed characteristics of conventional automobiles and constant-power electric-motor-driven cars.
The maximum speed of electric automobiles on a level road can now be calculated by equating the constant propulsive power delivered by the motors with the power required to overcome aerodynamic drag, tire rolling resistance and any other frictional dissipative forces (the last being negligibly small compared to aerodynamic and tire resistance). The tire rolling drag force of passenger-car tires is between 0.01 and 0.02 of the curb weight and increases with speed. For good-quality properly inflated tires (the lower bound of the band in Fig. 6), the rolling tractive force is about (O-01+ 5 x 1O”v) x w
(3)
The aerodynamic drag force at 60 m.p.h. of various 1960-model vehicles is shown in Fig. 7, indicating that this resistance for recently styled automobiles may be assumed to be 30+0~015(w+
150)
(4)
where 30 lb is the drag force at 60 m.p.h. of a driver’s body reclining, as in the car seat, and 150 lb is the weight of the driver. Aerodynamic drag of cars is quadratic with speed, so that expression (4) may be multiplied by (~/60)~ to yield the drag-speed relation of automobiles to be [0.0083 + 4.2 x 103 W+ 1SO)]V* (5) The tota force (lb) resisting automobile motion- i.e. the sum of expressions (3) and (5)-when multiplied by the speed, u, and the appropriate dimensional conversion factors,
Systems design of electric automobiles
11
is close to the total automotive drag power required from the electric motors. This tractive power requirement is then 0.00267 x ((0.01 + 5 x 10” v) x W+ [09083 f4.2 x 10d( W-t 1So)] u”) x D where o.m267
=
(6)
lai7 Wec)/m.p.h. 550 [(ft-lb)/sec]/h.p.
01
I
0 FIG.
1
1
20 rprrd,
v,
40 (mph)
80
60
100
6. Passenger car tire characteristics at speed.
200 r
01
1
1
0
I
.
curb FIG.
wright
plur
onr
I
4ooo
2000 drivrr
61
(lb)
7. Weight and aerodynamic drag of vehicles at 60 m.p.h.
At the top speed of the car, t), the required horsepower in (6) (with V substituted for u) matches and balances the motor deliverable power, presented in expression (l), to average about 0.02 W (h.p.) (7) for standard to intermediate performance vehicles.
12
GEORGE
A. HO-
Equating expressions (6) and (7), rearranging terms and rounding off numbers gives the equation of the maximum speed, V (m.p.h.), to be
The real solution of equation (8) is a function of W whose intricacies are beyond the scope of this paper. It is suflkient to state that the solution of (8) varies from V = 89 m.p.h. for W= 2000 lb to V = 100 m.p.h. for W= 5000 lb, and that 2000~ Wc5000 is a reasonable expectation for the future mass-accepted electric cars. It appears then that matching weight composition and initial performance of gasolinepowered cars can result in the design of future electric automobiles capable of top speeds of about 90-100 m.p.h. (145-160 km/hr) when allotting 4 per cent of the curb weight to the most appropriate electric-traction motors. A maximum speed of 100 m.p.h. is quite respectable and useful for most driving conditions presently envisionable, particularly when it is available in conjunction with accelerative performance quite close to that of the competitor vehicle. 4. CHOICES
OF BATTERIES
AND
RANGES
As seen in the last section, Y, the weight portion allocable for electrical energy storage and delivery, is left to be about half of the curb weight-specifically Y = 0.49. Three types of electrochemical energy-storage and delivery systems are considered here as taking up this half of the curb weight: 1. Conventional rechargeable batteries, currently available for commercial and aerospace applications. 2. Regenerative fuel cells, with metal fuels and ambient air as the oxidizer source (hereafter called metal-air batteries for brevity). 3. Future high-energy batteries, presently only in the predevelopmental stage. Some use molten-salt electrolytes, while others are theoretically operable at room temperatures. Table 3 lists the salient characteristics of industrial batteries, on the assumption that an intensive upgrading effort aimed at a life of a thousand or two charge-discharge cycles would be successful in adapting these batteries for powering automobiles in the near future. Next, Table 3 presents (for illustrative purposes only) the estimated properties of two hypothetical metal-air batteries. The first, the Zn-air battery is hypothetical in the same sense that it requires much further development and refinement before it can qualify for mass-production and automotive use, even though it is operative today and a prototype TABLE 3. PROPERTIES OF CANDIDATE BATTERIES FOR ELECTRIC AUTOMOBILES
Type
Symbol
(a) Industrial batteries Lead-acid Pb-acid Nickel-cadmium Ni-Cd Silver-zinc Ag-Zn (b) Hypothetical metal-air batteries Zinc-air Zn-air Metal-air (c) Hot batteries Sodium-sulphur Na-S Lithium+hlorine Li-Cl
Energy density d (W-hr/lb)
Discharge time (hr)
Source for value of energy density, d
10 20 30
0.5 1 1.5
Manufacturers’ data (Shair, 1966a) (Shair, 1966a)
50 70
3 4
(Shipps, 1966a) (Charkey and Dalin, 1966)
100 100
6 6
(Ford Motor Company, 1966) (General Motors Company, 1966)
Systems design of electric automobiles
13
be running a vehicle this year or next. In the second entry of the metal-air couple systems, the metal may be magnesium, calcium,. iron or lithium. Rechargeable batteries of this kind are hypothetical since they are yet to be designed or successfully cycled, and the value shown for their energy density is only our guess as to its eventual level, after development in the near future. A flow diagram of hypothetical meta-air batteries is given in Fig. 8, with a description of its modus operandi. might
BATTERY
TERMINALS
--n
d.“.o .o.o
““:iNG >
EXCESS
n+
AIR
OXIDE FILTER TANK
. .
0. . .o.o.
COMPRESSOR
LlOUlD ELECTROLYTE TANK
FIG. 8. Diagram of a metal-air battery. The metal-air battery derives energy from the conversion of a metal to its oxide on exposure to pressurized air. The flow during a discharge cycle is depicted schematically, with the oxide particles represented as dots. For recharging, a voltage is applied across the terminals; the flow of particles is reversed so that they are carried by the electrolyte into the cell and electrolyzed into metal and oxygen. The metal is redeposited on the anode and the oxygen bubbles off at the cathode and out of the excess-air vent.
The last class of batteries in Table 3 are labelled “hot batteries” since they require elevated temperatures to melt the electrolyte and to initiate their functioning, (about 300°C operating temperature for Na-S batteries and 600°C for L&Cl cells). This class is representative of the higher-energy cells under investigation in the automotive industry. Their energy density as displayed in Table 3 is lower than the most optimistic industrial estimates and again reflects our guess as to their actual capability after a few years development. Aside from the obvious drawback of high-temperature operation only, these batteries utilize liquid metals that react explosively with air, and require highly toxic substances in large quantities. Hydrocarbon and several other fuel cells are omitted from Table 3, either because of their non-regenerative nature or because of their excessively low-energy densities, even though much of the research activity in direct energy conversion is in this area (Barak, 1966). These types of fuel cells are quite feasible as electric power generating units for cars, in the intermediate run, preceding the ultimate advent of lightweight low-cost batteries.
14
GEORCSB A. HOFFMAN
Also omitted from Table 3 are the newer lithium batteries claimed to have achieved or surpassed 100 W&r/lb (1.6 x 10s J/kg) in the laboratory (l&e&erg, 1966; Shair, 1966b). Commerical availability of lithium batteries in quantity seems to be several years away, and capability for thousands of discharge-charge cycles is yet to be demonstrated. Thus came our temporary omission of lithium batteries from this report, but not necessarily from considerations for more distant future possibilities. In all entries of Table 3 the energy density, d, depends on the frequency and length of deep-discharge time. The time for discharge shown is assumed roughly to be the anticipated range divided by a representative mean speed. The distance a battery-operated car can travel before requiring either a recharge or refueling is here defined as the range of the vehicle. The range is determined by the maximum energy deliverable by the motor, which is
0.92 YWd
or
0.45 Wd W-hr
(9)
where 0.92 is our assumption for the average motor efficiency in converting electrical energy to mechanical work. Homopolar motors so far exhibit lower efficiency than this value, while induction motors are often more efficient than assumed here. The energy in (9) is expended in the work done by the motors to overcome tire and aerodynamic drag. This work is the sum of the drag forces-expressions (3) and (5)-times the range, R, when the car is traveling at a steady speed, V, on a level road on a windless day. This work @lb) equals ((0.01 + 5 x 10d u) W+ [0*0083+ 4.2 x lO-@(W+ 150)] us}x (R, miles) x (5280 ft/mile)
(10)
Equating energy and work, i.e. expressions (9) and (lo), gives the range of electric cars driven at steady speed to be d R* mile = oG442+2*21 x 10-%+[(0~395/FV)+ 1*86x lO-s]uz The range under this purely fictitious driving condition is shown in Fig. 9. The width of the band is representative of the dependence of R on W, the lower bound being for W = 2000 lb and the upper for W = 3000 lb. More realistic travel simulations than steady-speed driving are needed, particularly to account for accelerations and decelerations. Two such driving conditions are assumed as shown in Fig. 10. Condition I represents urban travel in moderately congested tra& with considerable stop-and-go driving. Condition II defines the pattern for suburban driving in moderately congested trafhc with considerable stop-and-go driving. Condition II defines light tra5*not on expressways. These assumptions were modelled after the actual record of engine utilixation in a variety of urban settings. With the mission profiles assumed in Fig. 10, the energy expended per trip, in steadyspeed cruising, can be shown to be rather small (a twenty-fifth and less) compared to the energy required per trip for acceleration or grade-climbing. The steady-speed energy requirements may then be neglected without introducing much error and can be compensated by the stipulation that the work done in decelerating is dissipated in heat-t This leaves only the acceleration and hill-climbing phases to be considered, conservatively t Dynamic braking by the electric motors was assumed here to be available down to speeds of 5-10 m.p.h. A friction brake would take over below these speeds for full stops and parking. Regenerative braking (i.e. battery-charging when decelerating) is another scheme that merits attention for energyrecovery in electric automobiles. It is not incorporated in thll design study because it extends the range by only a quarter or less, but at some penal@ ln control intricacies and costs, particularly for circulating electrolyte batterles. Regenerative braking is certainly technologically feasible at this time for electric cars with conventional batteries.
Systems design of electric automobiles
15
to be done at close to the full power level. (By contrast, in Conditions I and II traffic, the median engine utilization by late-model cars is usuahy only about half of the available engine power.) The full power assumption should also compensate for the omission of the cruising power from the following calculations and should account for slightly lessthan-average performance of electric automobiles at 30-60 m.p.h. assumed
0
50 ’
km/h
100 ’
I50 km/h YtF
500 km
20 sprod FIG.
9. Rans
40
, v,
60
80
IOZ
(mph)
of electric cars driven at steady speed.
FIG. 10. Assumed engine utilization in city-driving situations.
The energy expended between recharges when traveling in the metropolis is then (O-02IV) x (0.352”) x 3600 x 550/2656
where 0.02 IV = O-35 = T=
3600 = 550 = 2656 =
or
522 WT W-hr maximum available motor power, from expression (1) portion of Conditions I and II trip time spent in acceleration travel time available between recharges or refuelings (hr) sec/hr (ft-lb/sec)/h.p. (ft-lb)/W-hr
(11)
GEORGEA. HOFFMAN
16
Equating (9) with (11) gives both the urban and suburban driving time between recharges to be T = 0.086d hr The variability in city-driving characteristics is greatest in the block speed (total trip length+ total trip time), achieved in a variety of settings and locations, tragic conditions and driver behavior. It is only too well known that block speeds of 20 m.p.h. and lower are experienced at peak traffic hours in densely populated regions or under inclement weather, while block speeds of 40 m.p.h. or higher can occasionally be achieved in some suburban driving at off-peak times and in good weather. The range of city-driven electric cars is therefore presented in Fig. 11 for a variety of block speeds, this range being simply T x (block speed).
IS0
range
200
, nim )
km
100
battery
energy
density, d,
(W-h/lb)
FIG. 11. Range of city-driven electric automobiles.
Range was only slightly affected by Win the steady-speed cruise mode, and now-under metropolitan driving conditions-the range of electric automobiles seems independent of the vehicle overall weight. One may gather from this observation that the present spectrum of car weights (from 2000 to 5000 lb) and cotigurations (from small to large-luxurious) offered to the consumer for his choice need not change with the substitution of electric energy for combustion. The prospective electric car owner of the future need not consider vehicle size and weight when selecting the range between recharges that he can tolerate. This is analogous with today’s designs where the smallest to the largest passenger cars have more or less comparable ranges between refuelings. Figures 10 and 11 imply that conventional batteries are presently unsuitable for many consumer range demands. Much better suited are the metal-air and other yet to be developed high-energy density batteries since these promise for the future quite respectable ranges between refueling and recharging. It should be, noted that the mission profiles in Fig. 10 are more arduous than one can normally expect. A driver’s more moderate use of the accelerator pedal than the full power assumed in (1 l), and travel along routes with few stop signs or with well-coordinated lights, could result in ranges up to twice those exhibited in Fig. 11. They approach and could surpass the ranges of gasoline-powered cars with a single tankful of fuel. In conclusion : electric cars of adequate performance and top speed could be designed in our lifetime for ranges of 200-300 km if metal-air batteries are successfully adaptable to automotive use. In the more distant future higher energy batteries could provide twice such ranges.
Systemsdesignof electricautomobiles 5. SOCIO-ECONOMIC
EFFECTS OF THE FUTURE AUTOMOBILES
17 USE OF ELECTRIC
A limited sample of the societal implications of all-electric automotive propulsion in the future is investigated in this section. The first item is the forecast of a probable schedule for the introduction of electric cars into the automotive market with u priori assumption that the vehicle will prove comparable in design and lifetime costs with the non-electric cars extant at the time. High-energy batteries
>R8D)
Motors and >Y> controls Ntimbrr of prototyprs producrd
I
Full occrptancr
FIG. 12. Possiblescheduleof introductionfor electricautomobiles.
One such schedule was worked out in Fig. 12 for either the United States or the panEuropean market. It is based on factoring mass-production and mass-acceptance of radically innovated cars by two orders of magnitude every 4 years. In other words, stipulating ten prototypes operating at the end of this decade with conventional speeds and performances (but not yet range), we escalated the yearly production one hundredfold every 4 years. With such a sustained enormous program of expansion only by the mid-80’s could the industry produce electric cars at a yearly rate comparable to that of gasoline-engine cars. Even after the electric automobile’s complete acceptance by that time, it will take another dozen or so years for the extant combustion-engine vehicles to wear out and become extinct. 2
18
GEORW A. HOFFMAN
Under these &cumstances, it appears from Fig. 12 that only toward the turn of thy twenty-tist century could one expect extensive use of electric automobiles in our dail travel activities. The palpable economic effects on the consumer’s pocketbook will be important Consider first the initial (and, later, the operating) costs to the car owner in terms of the difference between electric and non-electric cars, both produced at rates of, say, 106 to 10 units per year. The production unit cost--and therefore price per unit .weight-for half o. the car weight (structure, trim, suspension, wheels, tires) should be essentially the same fol both types of cars. Then the lighter-bodied electric car shows some price advantage hen (see Fig. 13). In the remaining half of the vehicle weight, a major cost jump will lo encountered, due to the cost of batteries being intrinsically higher per pound-perhaps 1.5 times?-than the cost of conventional car components. Thus, as a whole, electric cars might tend to be more expensive at least by a quarter than their conventional counterparu of similar size and weight. ELECTRIC
* 5 2 ENGINE
AND
AUTOMATIC
B 5
TRANSMISSION
8 WHEELS,
TIRES.
STEERING
5 u ELECTRIC
4
WHEELS,
MOTORS TIRES,
STEERING
Y 2 (13 E
STRUCTURE
Fro. 13. Comparative manufacturing costs of electric and gasoline automobiles.
The mass marketing of electric automobiles will require a special (but not too novel) task: to convince the prospective automobile buyers to pay considerably more at the time of the purchase of the new vehicle than they would pay for gasoline-engine cars; and to trade off this sign&ant increment (one-quarter of the initial cost) against the equally significant decrement in propective operating costs over the ensuing years of car use. The unchangeable operating costs will still be : the interest on the capital, tire replacement, road and highway taxes, maintenance, insurance and licensing fees. But a sizeable lowering in operating expenditures will be derived by switching from gasoline and oil to electric energy. A reduction of about one-halfin fuel costs might be expected. Integrated over the years of the car’s life, this would more than offset the initial purchase price increase. t Lead-acid batteries are one and a half times as expensive per pound as the average price/lb of the whole automobile. Zinc and nickel, candidate metals for fuel or electrodes in metel-air batteries, are cheaper per pound than lead, while lithium is several times as expensive.
Systems
design of electric automobiles
19
In essence : one prediction is that electric cars will be costlier to purchase and somewhat cheaper to operate, with the operational cost savings over the years adding up to a net benefit. Automobile manufacturers should look favorably to the mass-production and marketing of all-electric passenger cars that are comparable to present cars. The added options of highly intricate and sophisticated motors and advanced batteries being substituted for piston-engines and automatic transmissions might spell some handsome profits. While the oil and gas industry might be dismayed at the coming of the all-electric vehicle, electric utilities will welcome the advent of electric cars: the demand for electric power consumption would about double. The price of electricity might be reduced by one-tenth or one-fifth, because of this doubled capacity and in view of the heavy night-time power demands as batteries are being recharged at home. The mounting problems of urban air pollution now plaguing many cities and mainly due to the emissions from the internal combustion engine should be greatly alleviated by battery-operated cars. One of the most significant benefits from electric cars in the future might prove to be the abatement of automotive exhaust. These emissions are extracting from urban man an enormous toll in health and in the degradation of the quality of the cities’ air. REFERENCES BARAK M. (1966). Fuel cells-present position and outstanding problems. Adu. Energy Conversion 6, 29-55. CHARKEY A. and DALIN G. A. (1966). Metal-air battery systems. Proc. 20th Annual Power Sources Conf., pp. 19-83. EI~ENBER~M. (1966). Electrochimica Corporation Announcement, November. Ford Motor Company. (1966). News release of 3 October. General Motors Company. (1966). News release of 28 October. HOFFMANG. A. (1966). The electric automobile. Sci. Am. 215, 34-40. HUNTINGTONR. (1966). . . . Electric car project. Aatocar 125, 1006. MAR= C. (1966). General Motors news release of 28 October. SHAIIXR. C. (1966a). Sealed secondary cells for space power stations. J. Spacecr. & Rockets 3, 68-70. SHAIR R. C. (1966b). Gulton Industries news release of 5 October. SHIPPSP. R. (1966). Secondary metal-air system. Proc. 20th Annual Power Sources Co&, pp. 86-88. SLABUK W. (1966). An A.C. individual wheel drive system for land vehicles. Society of Automotive Engineers Paper 660134.