NASA TMX-64623 Rover Mobility Systems

8/3/2019 NASA TMX-64623 Rover Mobility Systems

1/41

N A S A T EC H NIC AL

M E M O R A N D U M

NASA TMX-64623

M O BI LI TY SYSTEMS AC TIV ITY FOR LUNAR ROVERS AT MS

Clyde S. Jon es, Jr . and Frank J. NolaA s rionics Laboratory

September 9 , 1971

NASA

M S F C - F o r m 3190 ( R e v June 1971)


2/41

1 . R E P O RT N O .

NASA TM X-64623

2. G O V E R N M E N T A C C E S S I O N N O. 3. R E C I P I E N T ' S C ATA L O G N O .

4 . T I T L E A N 0 S U B T I T L E

Mobility Systems Activity for Lunar R overs at MSFC

7 . A U T H O R I S )

Clyde S. Jones, J r , and Frank J . Nola9 . P E R F O R M I N G O R G A N I Z AT I O N N A M E AN D A D D R E SS

September 9, 1971

8. P E R F O R M I N G O R G A N I Z AT I O N R E P O RT U

10. W O R K U N I T N O .

19. S E C U R I T Y C L A S S I F . (of thls report) 20 . S E C U R I T Y C L A S S I F. (of thls pmge) 21 . NO. O F PA G E S

George C. Marshall Space Flight CenterMarshall Space Flight Cen ter , Alabama 35812

12 . S P O N S O R I N G A G E N CY N A M E A ND A D D R E S S

National Aeronautics and Space AdministrationWashington, D. C. 20546

22 . P R I C E

1 1 . C O N T R A C T OH GRANT NO.

Technical Memorandum

Unclassified

I 14. S P O N S O R I N G A G E N C Y C O D E

Unclassified 42 3. 00

II S . S U P P L E M E N TA R Y N O T E S

Prepared by Astrionics Laboratory, Science and Engineering

16. A B S T R A C T

The Apollo Lunar Roving Vehicle (LRV) mobility system is descri bed. Special emphasia isgiven to the redundancy asp ect s and to the selection of the dr ive mot ors. A summary ch art ofthe performance on the l u n a r surf ace during the Apollo 15 flight is included. An appendix givesdet ail s on some MSFC development work on high efficiency drive sys te ms and compares thesesystems to the selected system.

17. K E Y W O R D S

Lunar Koving VehicleBrushless motorsRedundancyMobility s y s t e m


3/41

TABLEOF CONTENTS

--

INTRODUCTION . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .MOBILITYSYSTEMS . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .MLRV MOBILITY SYSTEM . . . . . . . . . . . . . . . . . . . . . . . . . . . . .REDUNDANCY . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .POSSIBLE IMPROVEMENTS IN REDUNDANCY . . . . . . . . . . . . . . . .

MLRV MOBILITY SYSTEM OPERATION . . . . . . . . . . . . . . . . . . . .STEERING . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .PERFORMANCE . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .APPENDIX . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .R E F E R E N C E S . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

Page

1

2

2

3

4

4

5

6

8

36

ii i


4/41

LIST OF ILLUSTRATI ONS

Figure

I .2 .

3 .

4 .

5 .

6.

7

8 .

9 .

10

I 1

12 .

13 .

14 .15 .

16

Title Page

Breadbo ard vehicle . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 20Breadboard vehicle with re mote control operations . . . . . . . . 21

Mobility system components . . . . . . . . . . . . . . . . . . . . . . 22Power distribution . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 23Control and display console . . . . . . . . . . . . . . . . . . . . . . . . 24

Block diag ram of t raction dr ive . . . . . . . . . . . . . . . . . . . . . 25Block d i a g r a m of s t e e r i n g c o n tr o l . . . . . . . . . . . . . . . . . . . . 26

LRVmoto r s . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 27Stall torqu e vs curr ent for br ushl ess and series-woundbrush motor . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 28Torque vs speed maximum performance curv es . . . . . . . . . . . 29Efficiency vs to rque . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 30Efficiency vs torque . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 31Perc ent incr eas e in driving range of brushless motorover series motor as a function of torque . . . . . . . . . . . . . . . 32

Brushless motor electronics . . . . . . . . . . . . . . . . . . . . . . . 33Efficiency v s torque for maximum torque speed . . . . . . . . . . . 34Homopolar inductor rotor . . . . . . . . . . . . . . . . . . . . . . . . . 35

iv


5/41

TECHNICAL MEMORANDUM X-64623

M O B I L I T Y SY ST EM S A C T I V I T Y

FOR LUNAR ROVERS AT MSFC

INTRODUCTION

The Mar shal l Space Flight Center (MSFC) has been ac tive in lun ar vehicletechnology f o r a number of ye ar s. Ear ly work was done in study pr og ra ms ofvehi cles such a s the Mobile Labo rato ry (MOLAB) and lat er the I h a l Mode LunarRoving Vehicle (DLRV) . A t least thr ee diffe rent Mobility T es t Articles (MTA)were built by different contractors involved in the program to test various

mobility concepts, especially in the wheel area. These programs wereca rr ie d only through the NASA Phas e B contrac tual stag e; parti ally b ecau se offunding but ultimately beca use of the ac cel era ted need fo r a Manned LunarRoving Vehicle (MLRV) fo r the Apollo 15 mis sio n. The MOLAB and DLRVpro gra ms looked toward la rg e vehicles with a broad range of operationalmodes and capabil ities and surviv al for long period s of t ime in a very hostileenvironmen t. The advent of the MLRV and its accelerated program allowedth es e capa bilities and modes of operation to be reduced in the sev erity of th eirrequirements to compress the delivery schedule. For example, the tempera-ture environment w a s reduced from the *120" C required for a complete lunarcycle t o only 0" to 100C for a partial lun ar day, and the operational time w a s

reduced from as much a s two ye ar s to only a few hou rs. One as pe ct of theproblem that somewhat incre ased in complexity w a s a weight res tr ic ti on ofapproximately I S 1 kg. The ac tual final weight of the MLRV was somewhathigher ( 2 2 7 kg) , but considering that the payload requ ireme nt is about 450 kg,th is weight is still reasonably ambitious.

During the MLRV contractual cycle, some para llel development ofhardware ha s been pursued at MSFC to su ppor t the MLRV design and toinvestigate the po ssible fu tu re o r extended u s e of the r ove r for teleoperator-type programs.used by the Astrioni cs Laboratory to study mobility sys tem s in gener al and

the s teerin g system of the MLRV in parti cula r. Figure 2 shows anothersimpl e and economical vehicle fabricated in-house and instrumented f orremote control operations. This vehicle ha s been successfully operated froma mountain about 4. 8 km distant with a communications delay time in the com-mand link to simulate lunar operation. This program has been quite succ ess-ful and lead s to the belief that teleoper ation of vehicles of t his type can be suc-cess fully accomplish ed within the pre sen t technology.

Figure I hows an aust er e breadboard vehicle which w a s


6/41

MOBILITY SYSTEMS

In all the sy ste ms that have been considered at MSFC, the electricmoto r has been proposed as the b as ic device f or locomotion.block diag ram of the fundamental components which normal ly make up themobility system.

Figure 3 is a

A battery is the customary energy source, although in the DLRV andMOLAB type vehicles supplementary s our ces such as radio-isotope thermo-electric generator (RTG) o r solar a rra y were considered.used for thro ttle and steeri ng control is a hand c on tro lle r which usually con-sists of a potentiometer o r other device that produces an elec tric al signalproportional to hand-controller movement. The hand-controller voltage isproc essed by con trol elec troni cs that vary the amount of bat tery powerapplied t o the motor. Both wheel and stee rin g moto rs a r e usually matched t o

the load via a transmission of appropriate ratio. The selection of the motorand transm ission is probably one of the most difficult trade-offs and is dis-cussed in some depth in the appendix. The par ts of the mobility s yst em notconsidered here ar e the suspension and brake system.

The device

MLRV MOBILITY SYSTEM

The MLRV mobility system w a s designed and built by the DelcoElectronics Division, General Motors Corp. , Goleta, California, under asubc ontr act t o The Boeing Co. Bas ic specifications furn ished by MSFC whichdictated the design constr aints a r e a s follows:

i . Performance capabilities.

a. Climbing a 25-deg slop e fully loaded.

b. Speed of 16 km/hr.

c. Negotiating a 30-cm high obstacle with both wheels in contact

at zer o velocity and a 70-cm wide cr ev as se for both wheels at ze ro velocity.

2. Four wheels of appropriate si ze and design to produce the aboveperformance.

3. A range of 120 km.

2


7/41

4. A life of 7 8 hr during the luna r morning.

5 . A design such th at no-single-point fa il ur e wi l l abort the missionand no-second fai lur e w i l l endanger the crew.

The specifications resulted in a desi gn with the following ha rdw are andperformance parameters:

I.

improve traction.An 81-cm-diameter wi re mesh wheel with chevr ons attached t o

2. A series-wound brush-type mot or mated to an 80:1 harmonic drivetransmission forms the wheel dr ive unit. This unit is hermetically sealed andthe two run in a nitrogen atmosphere at approximately 5 . 1 7 N/cm2. The tr an s-mission is lu br ic at ed with Kry tox 143AZ oil.

3. To meet the no-single-point-failure spec ifica tion , a syst em ofredundancy w a s conceived and is disc usse d in mo re detail in the followingparagraphs.

REDUNDANCY

Almost eve ry element of the MLRV ha s som e meas ur e of redundancy.A s shown in Figure 4, the energy source w a s made redundant by using twosilver-zinc batteries of nominal 36-V, 120-A-hr capacity t o excite the fourpower buse s ( A through D) . A l l loads can select one bus from each batteryvia s witches on the control and display console (F ig. 5 ) . The low-level(*15-V) power supplies fo r th e drive control electronics (DCE) box are fedin parallel from both batter ies such that the loss of ei ther battery duringdynamic operation would not affect its output. This is done because onlyone supply is operating (the other is in standby), and the loss of this supplyw i l l eliminate all vehicle control.

The wheel dri ve sys tem redundancy is a result of the four-wheel driveaspect. Ther e ar e four motor-tran smission units and, within the DCE box, fourpower amplifiers, each of which can be tu rned on or off f ro m the console via theDRIVE POWER switches (Fig, 5) . The hand contr oller h as only two throttlepotentiometers and two pulse width modulators (PWM) to d rive the four poweramplifiers. These two PWM modules are fed through f ou r DRIVE ENABLEswitches on the console allowing the astronaut to select ei ther PWM to driveany of the four wheel drives.

3


8/41

Steering redundancy is obtained by having identical front and r e a rst ee ri ng sys tem s. This double Ackerman ste er in g not only gives redundancybut re sul ts in a ver y short turning radius. Either ste eri ng unit can be bothele ctr ica lly and mechanically decoupled. The rear unit can be mechanically

recoupled, lout the front system cannot.

In addit ion, the low-level power supply is standby redundant asmentioned earlier.

POSSIBLE IMPROVEMENTSIN REDUNDANCY

The redundancy sche me desc rib ed , although meeting the specifiedc r it e r ia of no-single-point -f ai lu re , could be improved upon at no cos t in

weight and poss ibly provide a cost savings. A more natural , s impler schemewould be to make the driv e sy ste ms completely independent, i . e . , the throttlepotentiometer, PWM, low-level power supp ly, and power amp lifi er. Th is quad-rup le redundant sche me elim ina tes much of the switching on the console, there byreducing as tron au t workload. The DRIVE ENABLE, P W M SELECT, and h15-Vswitches could be elim ina ted completely. Although it would appear t o incr eas eelectronic par ts count, this is not the case.sup pli es mus t each be capable of supplying power fo r all elect ronic s and onlyone is active. An in-house redes ign consi sting of six active sm al l supp lies fo rindividual dr ive sy ste ms w a s found to r equ ire less pa rt s and to weigh signifi-cantly less, Although four PWM modules are needed instead of two, theseunits require so few par t s as to be insignificant. Two component failures in thepresent system can cause los s of al l wheel drives (e. g. , a potentiometer feed-ing PW M No. I nd a component in P W M No. 2) , but this is not the case withth e modified sch eme . The advantage i n the existing system is that a singlefai lur e in circu itry p rio r t o the DRIVE ENABLE switches can be compensatedf o r by the switching schem e. However, th e number of components involved isonly about 15 as compared to approximately 90 in the power amplifier circuitry;the ref ore , the probability of a failure in this area is remote.

The pre sen t low-level power

M L R V M O B I L I T Y S YS TE M O PE R AT IO N

Figure 6 is a block diagra m of the DCE showing one complete dri veFor throttle operation a signal is fed from the hand-controllercheme.

potentiometers into each P W M module. This is a dc voltage input proportional

4


9/41

to hand-controller position and the output is a constant-amplitude pul se whosewidth is proportional to the voltage input. The outputs fro m the two PWM'sare fed to the control console where one is selected by the DRIVE ENABLEswitch t o excite the wheel-drive power amplifie r. The signal is applied tothe power-inhibit gate (GI) here it is passed on to the drive-power amplifierprovided none of the three inhibit signals is present . These inhibit signalsare derived as follows:

1. When the current exceeds a prescribed level (about 25 A), thecurrent- level detector t r ig gers a one-shot multivib rator which produces aninhibit signal fo r a pres et t ime.

2. When the forw ard or reverse switch on the hand control is changed,the reversing -relay control gate (Gz ) is enabled provided the wheel speed isbelow 1 km/h r det ermi ned by the wheel-speed ci rcui try. The output of th i sgate t r i g g e r s a one-shot multivibrator which drive s the rev ersi ng relay andat the sa me time inhibits GI fo r 60 m s so that no curre nt will be present inthe mo tor when the relay switches.

3. When the brake handle on the hand c ont roll er is pulled back mor ethan 13 deg, GI w i l l be inhibited so that no power can be applied. Notice thatthis inhibit is set such that both br ak es and power can be pres ent at the samet i m e to allow the astronaut to start off on hi lls without rol ling backward.

The signal from G I is then impresse d on the power-amplifier circ uitr ywhich dr ive s the motor. The motor speed is a function of the dc voltage input

from the hand controller.

STEERING

The st eer in g syste m on the MLRV con sist s of Ackerman-type linkageon th e front and rear, giving a turning radius of 3 m. Steering is powered bya position-control ser vo drive with a split-field-series motor driving througha geared tran smis sion for actuation. The split field allows for directionreversal without the need for relays. The stee ring will go from stop to stop

in approximately 6 s for an average rate of about 15 deg/s.

Figure 7 is a block diagram of the s tee ring system used on the MLRV.The steering-command potentiometer on the hand controller feeds a voltageinto the control electronics proportional to the desired wheel position. This

5


10/41

voltage is comp ared with a voltage from the feedback potentiometer, and theoutput is an er r o r voltage which is fed to a volta ge-li miter circuit. The out-put voltage of the limiter is comp ared with a voltage which is derived fromre sis to rs that sample the motor curre nt in the two fields of the motor. The

difference in the two voltages is the error, signal which is then amplifiedthrough one of two channels, depending on pol ari ty, and thi s driv es the moto rin such a direction as to reduce the e r r o r in the command and feedbackpotentiometers to zero.

PERFORMANCE

The ov eral l per forman ce of the MLRV on the Apollo 1 5 mission w a sver y good. The ast ron aut s w e r e pleased with the performance and felt thatthe ba sic design did not need improvement. A t least two minor anomaliesw e r e assoc iated with the mobility sys tem , but the se caused no degradation inperformance.

When power was first applied to MLRV on the lunar surface, it w a snoted that no voltage indication was pre sen t on bat ter y number two. A quickcheck of equipment being fed fro m th is batt ery indicated that the problem w a sa faulty mete r. This caused no performance problems but caused a lo ss ofdata feedback for perfo rman ce evaluation.

When the front ste eri ng w a s tested p rio r to embarking on the firstlunar t ravers e , it w a s found to be inoperative. A quick recyc ling of sw itchesand circuit brea ker s and some simple tests produced no solution. The frontsteering w as subsequently powered down and the vehicle w a s operated usingrear ste eri ng only. The astronau ts found thi s ste eri ng very respon sive andwere reluctant to make any fur th er tests on the front ste eri ng during the firstt rav ers e to determine the status. A t the beginning of the second traverse, thefront stee ring switches and circui t bre ake rs were cycled a fe w t imes, andwhen power w a s applied, the front steering w a s found t o be operational. Infor-mation a s to what caused thi s anomaly is unavailable, and the conclusion w a sthat a number of things could have caused the obs erved res ul ts. When the

second tra ve rs e began, the astronauts felt that the steer ing w a s too sensitiveand turne d off the re a r steerin g. However, because of a tendency fo r thewheels t o drif t , it ,was turne d back on, and after a few minutes of experienceno fur the r steerin g difficulties were reported.

6


11/41

A summ ary of mobility system p erformance obtained from d ata sentback durin g vehicle operatio n on the moon is shown in Table i . Although thedistance traveled w a s about 8 km less than planned for this mission, the speedw a s 25 percent higher than expected. Energy rate s were als o lower than hadbeen expected beca use of much lower frictional f or ce s from the lun ar so il than

had been predicted. The astronau ts stated that the vehicle handled quite w e l land negotiated slo pes with ease. Probably the steep est slope encountered w a sbetween 10 and 15 deg, but they were of the opinion tha t tract ion and veh icleperformance were such that much ste epe r slo pes could have been covered.

7


12/41

APPENDIX

The propulsion system for the Lunar Roving Vehicle (LRV) is anapplication for electric mot ors which must meet a la rg e combination of

stringent requirements. Among thes e requirements are variable torque,var iab le spee d, light weight, high efficiency, high tor que for obsta cle climbing,and high speed fo r maximum scie ntific exploration time . Dynamic brak ing,though not a f ir m requirement, is a desi rabl e feature. In addition, the syst emmust operate in the thermal vacuum of the lunar environment.

Despite the advantages of the s imp ler brush-type motor system f or theshor t t h r eed ay life of the LRV, it w a s felt that the short delivery schedulewould not allow suffic ient tim e fo r qualification of th is type of m oto r t o oper at ein the lunar environment. Br us hl es s mo to rs had the advantage that no newtechnology and no extensive test program would be requ ired t o qualify th ei r us ein a the rma l vacuum, Experience gained fro m in-house and contr acto r-supported research on candidate motor sys tems led to the conclusion that thepermanent magnet (PM) brus hles s dc motor had the best potential fo r satisfy-ing these requirements [ i ].

In the preliminary specifications for the mobility subsystem of the LRV,the P M brushless motor w a s indicated as being the choice candidate. However,the contractor w a s allowed to select the motor of his choice ( ter med base line) ,but w a s contractually obligated to c a rr y along the development of an al terna tesyst em until it could be shown that the baseline system could perform satis-

fact orily under simulated lunar conditions.

The base line propulsion mo to r, developed by the Ga rr et t Corp. for theLRV, is a four-pole brush-type series motor coupled to a trans miss ion withan 8 O : i speed reduction.Machinery Co. , is known a s a harm onic drive. The moto r and the wave gen-erator of the harmonic drive are hermetically sealed and operate in5.17 N/cm2 of dr y nitrogen. The al te rn at e mot or, developed by GeneralElec tric Co. , is a six-pole P M brushless motor coupled to a planetary-spurgear combination with an 80: i speed reduction. The ent ire syst em is open t ovacuum. The two mot ors a re shown in Fig ure 8.

This transmission, developed by the United Shoe

The LRV is a four-wheel vehicle weighing 227 kg (e ar th ). The fullyloaded weight with two astr ona uts pl us scientific equipment and payload is7 2 6 kg (ear th) . A motor at each wheel is driven by an electronic controllerwhich r espo nds t o commands fro m potenti omet ers mounted on the pivot of a

8


13/41

hand controller. A l l driving functions including forward and reverse driving,speed control, braking, and steerin g are initiated by the hand controller.The power source f or the vehicle is two silver-zinc bat teri es nominally r ate dat 36 V with a total weight of 64 kg. The ba tte rie s can be operated individuallyo r in paralle l and have a combined capacity of 240 A - h r.

The maximum specified speed when op eratin g on a smooth m ar e is16 lun/hr corresponding t o a wheel speed of 116 rpm. The predicted smoothma re torque required at the wheel is 5 .76 N-m. The torque requir ed fo robstacle climbing and for spanning cr ev as se s is 106 N-m at 3 rpm. Two othertorque-speed points specified are for climbing a 6- and a 25-deg slope. Th es eare, respect ively, 20 N-m at 60 rpm and 52 N-m at 25 rpm. With re fer en ceto the motor shaft and allowances for g ea r efficiency, the specified torquespeed requirements of the motor are 0 .1 N-m at 9280 rpm, 0 . 2 7 N-m at4800 rpm, 0. 68 N-m at 2000 rpm, and 1. 56 N-m at 240 rpm . Since th ereq uir eme nts of the propulsion syste m are established, a comparison of thebaseline series motor and the alternate P M brushless motor is given. Thecomparison is divided into the following topics: torqu e ve rsu s speed capab ility,weight and efficiency, braki ng, ele ctro nic co nt ro lle r, and growth capability.Also included is a brief discussion of ge ar reduc er tran smi ssio ns and othermotor sys tem s which have been proposed for l unar vehicles.

Torque Versus Speed Capability

A plot of the above specified performance points is typical of the

torque vers us speed characte ristics of a series motor . Since the speed of adc motor is limited mainly by t he counter electr omoti ve for ce (emf) gener ated bythe armature, it can be see n that by c ontrolling a single current , the se rie smotor can cover a broad torque speed range.lar ge arm atu re and field current cr ea te s two s trong magnetic fields which

output torque , the ar ma tu re and field curre nt is reduced. The weakenedfield now generates less counter emf in the armature and allows the motorto develop high speed.

F o r low-speed operation a

couple to produce a high-output torque. For high-speed operation at low-

Ideally, the series motor torque va ries as the squ ar e of the curre nt.

The torque v e r s u s curren t curve for the L R V motor shown in Figure 9reveal s a near linear relationship, The linearity is the result of bearingand brush friction at the low-torque end and of co re s atu rat ion at the highend. The specified maximum torque of 1 . 5 6 N-m is reached at a currentlevel of 22 A and is limited to this value by the electronic controller.

9


14/41

The maximum torqu e ve rs us speed capability of the series motor whenpowered from the specified 36-V source is shown in Figure 10.has a significantly higher than required speed capability at high-torque levels.At low-torque levels, however, whe re most of the driving is done, the speedcapability is minimal. The moto r has a top speed of 8400 rpm at the initiallypredicted smooth ma re torque of 0. I -m. Information gained fro m rece ntlunar explorations indicates that the smooth mar e torque may be more thantwice the initially specified value.moto r capability is approximately 6000 rpm correspondin g to a vehicle speedof 10 lan/hr.

The motor

Under thes e conditions , the maximum

A disadv antage of the P M brushless motor in the LRV application is th ehigh a rma tur e curren t req uired to c over the specified torque speed range.Unlike the variable field series moto r the field streng th of the P M brushlessmotor is constant and, in general, is high to minimize weight.

If the armature is wound with a la rg e number of tu rn s to produce hightorque at a reasonable current level, th e generated counter emf wil l be highand wi llJ imi t the moto r to low speed. Reducing the number of tu rn s for high-speed operation correspondingly red uces the to rque constant and must be com-pensated by proportionately increasing the ar mat ure curre nt to establish theampe re-tu rn product require d for high torque.ran ge, the exact maximum curren t requireme nt for a P M brushless motorcan ea sily be c alcu lated by equating th e mechanica l output power of t he mo torto its electrical output power.

Fo r a given torque-speed

This is given by the following equation.

where T is the maximum output torque requ ired in N-my S is the maximumspeed in ra d/s , V is the counter emf in volts developed at maximum speed,and I is the maximum a rmatu re curr ent in amperes. The voltage availablefor counter emf is the battery voltage less the voltage dropped in the elect roniccontr oller and in the ar ma tu re resistanc e. Assuming about a 2-volt dro p,V in the above equation becom es 34 volts for the LRV application. Substitutingthe other known requirem ents and applying prope r conversion fact or s givesthe following maximum c urr ent :

In addition to making the co ntr oll er design difficult, thi s amount of c ur re ntwould produce high losses in the c ont rol ler and in the cablin g and would resultin a moto r with a power output capability s ev er al ti me s gr ea te r than wouldev er be re quire d by the vehicle.

10


15/41

The alternate P M bru shle ss motor developed fo r the L R V employs atechnique which el iminat es the above problem and res ul ts in a low-currentsys tem which cover s the enti re torque-speed range with relatively highefficiency. The st at or winding is tapped at one-fourth of its full number ofturns. A t low speeds, where high torques may be required, the full windingis connected a cr os s the electro nic controller. When the motor speed reac hesapproximately 2300 rpm, a speed se nse winding in the m otor initiates a com-mand to a re la y which disco nnect s the full winding and sw itches t o the one-fourt h tu rn tap. The reduced counter emf now allows the motor to run up toits maximum specified s peed while developing torque up t o one-fourth of maxi -mum value. Thi s technique reduces the maximum cu rr en t requir ement to11.3 A o r to about one-half of that req ui red fo r the series motor. The entireswitching pro ces s is performed automatically by the con troller and is analogoust o automatic gea r shifting in an automobile. The tap changing oc cu rs in lessthan 10 ms. The motor cur ren t is inhibited f or 15 m s du ring the tap changingpr oc es s to elim inate the possibility of any inductive voltages which may appe arac ro ss the windings.

The torque versu s cur rent curve for the P M brushless motor shown inFigure 9 indicates a lin ea r relatio nship to 200 per cen t of rate d torqu e. Sincethe torque is normally limited to 1.56 N-m by the contro ller , the motor has atleast a 100-percent safet y mar gin fo r demagnetization. The ultima te torqu ecapability of the motor ha s not been deter mined becau se of power limi tation sin the electronic controller.

The maximum torque v ers us speed capability of the br us hl es s motor is

comp ared with the series motor in Figure 10. Although the series motor candevelop higher spee ds at the higher torque lev els, the predicted driving condi-tion s indicate the vehicle w i l l operate less than 5 per cen t of the tim e in th isrange. * The P M brushless motor has far sup erio r performance in the antici-pated maximum driving range and can propel the vehicle at 18 km/hr at theinitially predicted smooth mare torque compared to 14 lun/hr f or the seriesmotor. Note al so that the P M bru shl ess motor can maintain the specified16-km/hr speed a+ orque levels almost four time s the minimum requiredvalue. This is equivalent t o climbing a 12-deg slope. The overspeed capa-bility of the P M brushless motor is inherent in the system because thedeveloped counter emf is sinusoidal. Cur ren t can be forc ed into the windings

during tha t portion of the cycle when the counte r emf is less than the saturation

1. Results fro m the flight indicate almost all operations at less than 27 N-mof torque.


16/41

l imits of the electron ic cont rolle r. This cha rac ter ist ic could have been usedto reduce the maximum motor curre nt 10 percent but w a s designed in asreserve capability.

We i g h tan d Eff i c i ency

Weight and efficiency of an electric motor a re directly related. Sincemo st of the inefficiency of a motor is caused by copper lo ss , it is apparentthat in crea sing the diame ter of a motor increases the available slot area andallows the use of lar ge r w i r e . The series motor for the L R V weighs 2 . 7 kgcompared to 1 . 6 kg for the P M bru shl ess motor. Both moto rs have a copperloss (12R) of approximately 10 0 W at peak rated torque of 1 .56 N-m. Thisbri ngs t o light se ve ra l outstanding advantages of the P M brushless motor overa variable field brush-type motor. First, t o approac h the efficiency of th e

P M brushless motor, the series motor mu st be made almost twice as heavy.The prima ry reason for this is that the 12R los s inherent i n generating fieldflux in the series motor is significant; whereas, the P M field is essentiallyfree of lo ss es . Even if efficiency were not a prime consideration, the ser iesmo tor mu st be la rg e beca use of the difficulty assoc iated with rem oving heatfr om the rot or. Since the motor w a s initially designed to operate in a vacuum,the primary sour ce for heat removal is by radiation to the s tat or. With theP M brushless motor, no heat is generated in the rotor. The heat generatingelement, the s tator, is in contact with the vehicle, and the heat is easilyremoved by conduction as w e l l as by radiation.

A family of cur ves re lat ing efficiency and speed as a function of torquefo r t he P M brush less motor and the series motor systems are shown inFigures 11 and 12. These curves include the lo sse s in the electronic control-le r f o r both sy st em s. Although the series mot or efficiency app roac hes that ofthe P M brushless motor at seve ral points on the cur ve, these occur at torquelev els which are seldom encountered and contribute little to the overall missionefficiency. In the anticipated maximum duty torque-spee d ran ge , the P Mbrushless motor system is 20 to 30 percent mo re efficient than the seriesmot or syst em. The significance of th is is bes t shown by the cur ves inFigure 13. F or a given batter y life, these c urves depict the percent incre ase

in driving range real ized with the P M brush less motor system over the seriesmotor system at vario us torque levels. On a smooth mare and at speedsabove 4 km/hr, the P M brush less motor consumes 30 percent less energythan the series mo tor , resu lting in an increa sed driving rang e of about40 percent. Note also at spee ds below 3 km/h r, when the P M brushlessmotor is in the low-speed winding configuration, and at low-torque levels, the

12


17/41

energy consumed by the series motor system is almost twice that of the P Mbrushless motor system. It should be indicated here that if the winding turnsof the P M brushless motor were increased so that its top speed is reduced tobe comparable to the series motor, the current for a given torque would bereduced by 20 percent. This re duces the power loss in the elect roni c con-tr ol le r and, although it w i l l not significantly i nc re as e the peak efficiency of themotor, it wil l ra is e the efficiency at lower spe eds and result in an even gre ate rincrease i n driving range.

Brak ing

The performance of a series motor when used as a generator is unpre-dictable. Hence, th e mot ors cannot conveniently be used for brak ing and , inth e L RV, mechanical drum-type b rak es had to be added t o each wheel. TheP M brushless motor serves equally w e l l as a motor or as a gen erat or. Simplyby rev ers ing the input command, the motor provides controlled regener ativ ebraking to z ero speed. The controlled braking feature is inherent in the designof the con trol ler and req uir es ess enti ally no inc rea se in electron ic complexity.Experien ce gained on two exp erim enta l vehicles built at MSFC ndicates that theaver age energy re turn ed to the so urc e under normal driv ing and braking condi-t ions is l es s than 2 per cen t of the total energy consumed. Prim ary bat terysys tems , such as thos e used on the L RV, can easi ly accept thi s amount ofrec har ge e nergy with no detrime ntal effects.

Elec t ron ic Cont ro l le r

The outstanding advantage of the series motor over the P M brush lessmotor is the simplicity of the electronic controller. The armature cur ren t iscontrolled by a single power tran si st or plus four power contacts of a double-pole, double-throw rela y used fo r rev ers ing action. Although no rel ays arerequi red to rev erse the P M brushless motor, eight power transistors arerequired for controlling the two-phase sinusoidal sta tor curren ts. The set rans i s tors , however, also provide reversing and regenerative braking. Sincethese functions are performed electronically , a compariso n with a mechanicalsystem is difficult and may be meaningless.

Much of the s impl ic it y advantage of the series motor control ler w a slo st because of a require ment t o control braking and re ver sin g by pivoting thehand controller. Considerable logic cir cu itr y has been added to prevent inad-vertent re ve rs al of the mo tor s while the vehicle is in motion, to prevent drive

13


18/41

power f ro m being applied while bra king, and to inhibit motor cur ren t duringswitching of the reversing relay. Also, cir cui try had to be added to preventplugging2 of the motors which could occur if the vehicle were allowed to ro llin a direct ion opposite to that in which the direction al rel ay is set.

In actual parts count, the series motor controller has approximately30 percent fewer piece pa rts than the P M bru shl ess motor controller. Thepa rts count for the P M brushless motor controller is based on an MSFC design,which is being furni shed along with two LRV motors to the Lockheed Co. t o beused as the propulsion system f or an experimental vehicle program in whichthey ar e engaged under contract to MSFC. The con trol ler provides al l thefunctions relative to the P M brush less motors described in this rep ort andweighs approximately i. 27 kg. Figure 14 shows the P M brushless motorelectronics.

Grow th Capability

Consi derab le thought ha s been given to the possibility of an add-on kitwhich would convert the LRV t o a remotely controlled vehicle that would beteleoperated from earth . The kit would be installed by the astro nauts at th eend of the final LRV mission. The use of the series' motor in this applicationis questionable. The mo tor s are qualified only fo r th e sh or t life of th e LRV andfo r the relatively narrow tempe rature range expected to be encountered duringthe lun ar morning. The life and lower temper atu re range of the moto r islimited by oil vapor which may pene trate the seal between the harmonic ge ar

and the motor. A t cold tem per atu res, the motors may fail t o start becaus e ofcontamination of the bru she s. The te mp er atu re at which oil contaminationaffects the brushes is w e l l below the range anticipated for the manned missions.A remot ely o perate d vehicle would be subjected to a much greater temperaturerange. Hence, requalification of the mo tors would be req uired .

Other problems of teleoperation a r e create d by the mechanical brak esin controlling the speed of the vehicle while desce nding slopes . Th is wouldreq uir e an electromechanical ser vo in addition to the drive-control servo . TheP M bru shl ess motor ha s the advantage that speed control, direction control,and brakin g are all performed by a single input command. Remote drivin g and

2. Uncontrolled dynamic brakin g

14


19/41

navigation have been demon stra ted at MSFC with an experimental vehicleequipped with P M (brush-type) motors [ 23 .

Transmissions

A s previously stated, the baseline and alternate gear reducers developedfo r the LRV are the harmonic d riv e and the planeta ry-spur gea r combination.The units w e r e both developed by United Shoe Machinery Corp. The ha rmon icdrive ha s the advantage of a high-gear rat io in a lightweight package and, inaddition, allows for a sealed system so that the high-speed ge ar s may belubricated with oil and an atmosphere inserted to tra ns fe r heat. The systemdeveloped fo r the LRV weighs 0.7 kg compared to 1. 1 3 kg for the planetary-spu r gear. The planetary-spur gea r is mor e rugged, mor e efficient, canopera te in vacuum, and ha s an established performance r ecord .

A third type trans missi on which w a s developed through a programsponsored by MSFC is the rol ler-gear drive [ 31. This transmission has areduc tion ra ti o of 15:l and w a s developed as a low-speed backup syst em fo r thetwo high-speed units. The prog ra m included the development of a l a rge rbrushless-type motor si mi la r to the one desc ribed above but with torque andspeed sc ale d to match the lower ratio . The rol ler -ge ar approach evolved fromthe roller-friction drive and basically consists of a roller-friction drive withgears mounted on all rolling contacts. The ro ll er ge ar retai ns many of theadvantages of the r oller -fric tion dri ve while eliminating its high-preloadrequirements. In effect, the g ea rs now tra ns mi t the high-torque loads while

the ro lle rs act a s integral bearings supporting the ge ar s, transmitting low-torque loa ds, and providing overall s table ge ar alignment and res ultan t highefficiency. The efficiency of the roll er -gea r deve loped fo r th e LRV is betterthan 97 percent and, because of th i s high value, is very difficult t o mea sureaccurately. The curves of Figure 15 compare the maximum efficiency v er su storque of the ser ies -mo tor har monic drive , th e I. 6-N-m , brushless-motor ,spur-gear drive, and the 8.0-N-m, brushless-motor, roller -gear drive. Thela rg er brushless motor had a higher torque-to-weight ra ti o and, henc e, w a s2 t o 3 percent less efficient than the 1. 56-N-m br us hl es s motor .

15


20/41

Other Motor Systems

The discu ssion thus far ha s been limited mainly to a comparison of theseries motor and the brush less motor developed for the LRV. Two other types

of mot ors which have been proposed f o r lunar vehicle application ar e the homo-pola r inductor moto r and the variable-frequency ac induction moto r, Thefollowing comments are bas ed on exper ienc e gained in working with exper imen-ta l mode ls of the se motor s.

The homopolar inductor motor is a brus hles s dc motor with a variable-field r ot or ins tead of a PM . The obvious advantage her e is the broad torque-speed-range capability. Also, i n la rg e high-speed mo tor s, t he use of PM'smay be prohibite d beca use of l imit atio n in the size of the magnetic materialsand in centrifuga l stresses which they can withstand. Two views of an exp eri -mental homopolar inductor rotor are shown in Figure 16. The field winding(low er view) is not rigidly mounted on the s haft and, t her ef ore , does notrotate. The tw o excitation w i r e s are brought out through one of the st at or sl ots.This winding is normally wound under the stator winding in a void in the yokeof the s tat or to facilitate heat removal.

An excitation current in the field winding produces a magnetic flux witha path that leaves the sa lient poles at one end of the sha ft, cr os se s the air gap,trav els through the sta tor , and returns ac ros s the air gap back into the polesat the opposit e end of the shaft. A l l the poles at one end of th e s haf t have alike polarity and are opposite in polarity to those at the oth er end. One dis -

advantage of this concept is that only a portion of the r ot or sp ans the length ofthe stator winding and, therefore, less than half of the stator iron and w i r eare producing torq ue at any instant. Hence, thi s motor would be sev er al ti me sheavier than a P M brushless motor with equivalent torque and efficiency.

Another disadvanta ge of th e homopolar moto r is the effect of armaturereaction caused by the soft iron rotor. Part of the flux leaving the arm atu re( the s ta tor ) is attracted by the rotor and retu rns to the arm ature through thesoft iron pole piec es of the rotor. This dis tor ts the flux generated in the roto rand effectively sh ift s the ce nt er of the salie nt pole. In a brush-type motor thisis compensated by mechanically shifting the brushes. Since bru shl ess mo tor sare commutated by shaft position sen so rs , th ese se ns or s eith er have to beshifted mechanically o r th ei r output signal shifted electronically. The amountof shift is proportional to the differe nce in the ar ma tu re and field flux and ispolarity sensitive. Providing this function adds significant complexity to the system.The P M brushless motor, however, is not sensi tive to thi s problem. Since thepermea bility of a permanent magnet is essential ly the s a m e as that of air, theeffect of armature reac tion in a P M brushless motor is negligible.

16


21/41

Exp erie nce gained with the variable-freq uency ac-induction mo tor indi-cates performance character is t ics very s imila r to the series motor. The roto rfie ld of the ac motor is produced by cur ren t which is induced f rom the sta torwinding. A high-stator curr ent induces a stro ng rot or field and resu lts in high-torque at low speed. A s the s ta tor cur ren t is decreased, the rotor f ield cor-respondingly decr ea ses . The reduced counter emf now allows the motor todevelop a higher speed, but at a lower output torque. A s in the series motor,the tor que of a variable-frequency ac induction motor varies as square of thes ta tor cur ren t .

The broad torque-speed range of the variable-frequency ac inductionmotor is w e l l suited t o a vehicle propulsion system. The motor requ ire s noposition s ens ors and hence is one of the mos t rugged typ es of mot or s. Th edisadvantages of the LRV application ar e essentially the sa me a s those of theseries motor. The copper los s and contr oller loss es required t o generate theroto r f lux is significant and de tr ac ts from the system efficiency. Also themotor must be made la rg er to minimize the heat generated in the ro tor and toincr ease the radiat ing surface area to dissipate thi s heat.

In applications where output power and efficiency at the higher motorspeeds are important facto rs , the P M bru shl ess m oto r has an additional advan-tage over brushle ss moto rs with soft iron rotors. The frequency applied tothe stator windings of a brush less motor is equal to the speed multiplied by thepa ir s of poles. (Th is discuss ion is also applicable to the variable frequencyinduction motor but h er e the frequency is found by multiplying the rotor speedtime s the number of pole pai rs plus the sl ip frequency.) Fo r example, a six-pole motor running

at9000 rpm h as an applied frequency of

450 Hz.The induc-

tive reactance of a motor designed fo r high stall torque would be large whenrunning at a speed corresponding to this frequency because of the lar ge numberof turns. The voltage dropped across the reactance is in quadrature with thesys tem l os se s and with the counter emf. Although this does not in itself r ep r e -sent a d i r ec t l o s s , it can s eve rly lim it the output power of a motor in a lowvoltage system such as the LRV. This is illustrated by the vector diagram.

The applied battery voltage VA is the

Lecto r sum of the reactiv e dr opIX

and the in-phase components consisting

of the con trol ler voltage dro p V , th ewinding resist ance drop V

counter emf V. Since the output power

IXL

C, and theR

17


22/41

of a motor is equal to the counter emf t imes the a rmature ~ u r r e n t , ~ he effi-ciency is given by the equation

V-IN =V I + V c I + V I v + v c + v

where V I and V I repres ent the controller losse s and the copper los ses ,

respe ctive ly. If the qua dra tur e component is sma ll, the voltage available forcounter emf is large compared to the los se s, and the efficiency w i l l be high.If, however, the quadrature component becomes larg e, the loss es remainessen tially the sa me while the voltage available fo r coun ter emf dec re as es

C R

and li mi ts the efficiency and the output power of th e motor.

The soft iro n in the r ot or of th e induction motor and the homopolarmo to r significantly in cr ea se s the inductance of the st at or winding. The perma -nent magnet in the P M motor, as previously indicated, is equivalent to an aircore and has negligible effect on th e inductance of the st at or winding. Hence,all other fact ors being equal, the P M motor has m or e voltage available fo rdeveloping cpunter emf and can produce higher power and higher efficiencythan an equivalent motor with a soft iron rotor.

3. The output power of the induction motor is usually expressed in te rm s ofthe equivalent rotor resistance ra th er than the counter emf.

18


23/41

--J

I-0I-

a

-Ela>W

-H

a>W

-n

a>W

-

19


24/41


25/41

a0

a,c,

21


26/41

L

MOTOR ANDHROTTLE CONTROL ,HANDCONTROL LECTRONICS TRANSMISSION WHEEL

Figure 3. Mobility system components.

BATTERY

22

MOTORb AND bT EER ING CONTROLCONTROL LECTRONICS TRANSMISSIONHAND WHEEL *

r


27/41

@-m

kb)

23


28/41

FJI T C HHEADING NAV

POWER

0

0

I

SYST EMRESET

O f F

POWER

' U X 0

- t 15 voc-

COUUARD

8E A R

8"P

Documents

NASA TMX-64623 Rover Mobility Systems