THE UNIVERS I TY OF MI CHI GAN COLLEGE OF ENGINEERING Department of Mechanical Engineering Final Report EVALUATION OF SYNCRO DRIVE MECHANISM W. 0. Hermanson UMRI Project 04146 under contract with: EVANS PRODUCTS COMPANY PLYMOUTH, MICHIGAN administered by: THE UNIVERSITY OF MICHIGAN RESEARCH INSTITUTE ANN ARBOR September 1960

TABLE OF CONTENTS Page LIST OF FIGURES v ABSTRACT vii OBJECTIVE vii I. GENERAL DESCRIPTION OF EQUIPMENT AND TEST PROCEDURE 1 A. Description of Mechanism Evaluated 1 B. Test Equipment and Instrumentation 1 C. Test Procedure II. DISCUSSION OF TESTS AND RESULTS 5 A. Efficiency 5 B. Operation Over Full Travel of Ram 7 C. Displacement of Output Ram While Operating 7 III. CONCLUSIONS 13 IV. RECOMMENDATI ONS 15 iii

LIST OF FIGURES Figure Page 1 Over-all view of test setup. 2 2 Efficiency vs. ram load. 9 3 Efficiency vs. ram speed. 10 4 Oscillographic record, of torque and force. fL 5 Oscillographic record of ram displacement vs. time. y

ABSTRACT This report covers the evaluation of the Syncro Drive Mechanism as designed by the Maroth Engineering Co. of Wilton, Conn. The mechanical efficiency varies from a high of 64% to a low of 4Po%. This unit was operated at loads ranging from 12,000 to 100,000 lb and from speeds of 4.5 to 16.5 ram inches per minute. The output motion is not steady; the displacement vs. time curve dips three times per revolution indicating excessive deflection in one lifting unit. If the unit is to operate more smoothly, it must be redesigned to allow the load to be carried without excessive deflections which hinder its operating characteristics. OBJECTIVE The objective of this study was to evaluate the over-all performance of the Syncro Drive Mechanism. vii

I. GENERAL DESCRIPTION OF EQUIPMENT AND TEST PROCEDURE A. DESCRIPTION OF MECHANISM EVALUATED The Syncro Drive Mechanism converts rotary motion to linear motion of an output ram with a high mechanical advantage through the use of balls rolling on inclined planes. The inclined planes are formed in'the raceways so that the inclined planes are spaced over certain angular segments of the races and depressions or clearance spaces are found over the remaining angular segments of the raceways. Two pairs of raceways are used so that one may be raising the load while the other is being adjusted in its depressed segment. In operation, one of each pair of raceways is held stationary with respect to the output ram, while the other raceway of each pair is rotated by the driving mechanism. As the maximum climb is reached on a pair of raceways, the load is taken on the other pair. The first pair must then be adjusted so that it will be ready to accept the load from the second pair when they attain their maximum lift position. The adjustments are accomplished through an intermittent motion device which drives a nut along a screw cut on the outer diameter of the output ram. The load therefore is taken alternately on first one pair of raceways and then the other. The raceway positioning adjustments also take place alternately while the raceways are not loaded and in their clearance positions. The raceway cams are designed so that each pair of raceways lift the load 3 times during a revolution of the housing. Since there are two pairs of raceways, the load is transferred between pairs of raceways six times during a revolution of the Syncro Drive housing. B. TEST EQUIPMENT AND INSTRUMENTATION The Syncro Drive Mechanism was mounted on a large beam which in turn was mounted on a cast-iron bedplate for rigidity. The load was supplied by a hydraulic piston and cylinder arrangement which was also mounted on the large beam. The hydraulic pressure was supplied by a electrically driven pump and controlled by an adjustable relief valve. Both the Syncro Drive Mechanism and the fixed end of the cylinder fitted against end plates which were connected with two tie rods. The housing of the Syncro Drive was rotated through a chain drive by an electric motor equipped with an adjustable ratio drive system. Figure 1 is an over-all view of the test setup. The output force was measured using a strain-gage load cell which was located between the output ram and the hydraulic cylinder ram. The data output of the load cell was recorded on one channel of a dual-channel Sanborn recorder. 1

...........................~i~i...........iiiiiiiiiii ~~ ~~~iiiiiiiliiiil~~ ~~~~~.......... ~~~~~~Iw::_~~~~~~~~~~~~~~~~~~~........ ~ ~;'~~~~ ~~iiiiii~~~~~~~iiiiiiiii _:-I.......... ciiiiiiiiiiiiiiiiiii'::~~~~~~~~.................~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~i:'i'i'i' lir;~~~~~~~~~~~~~...... ~~~~~~~-~~~~:iiiiii~~~~~~~~~~~~, ~ ~....... ii~~~~i i~~~iiiiiiiiiiiiiiiiiiiiiiiiiiiiiii.................................. B~~~~~~~~~~~i~~~~~~~~~~~~~~~~~~~~~~i~~~~~~~~~~ B~.......::::::,.n~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~Pnslai *::::'Iliiiiii~~~~~~~~~~~~~~~ilil: T~~~~~~;~~~? ~ ~ ~ ~ ~ I:I:: ili.......... ~ ~ iii~~~iiili~~~iiiiiii~~~i~~i-iiiii ~~~~~B~~iiiiiiii............................................~ ~ ~ ~ ~ ~~ ~~~~~i.............iiiiiiiiiiiiQ.............. ~ ~ ~ ~ a................iilili?.......................... 111.1..........~ ~ ~ ~ ~ ~ ~ ~~ ~~~~ni'.....................................~Q..................::: rii~~~~~la~~~jiiiiiiiiiiiiiliiiiiiiiiiiliiilii 111118~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~......... i~~~~~~~~. " ~~~~~~~~~~~~~ h~~~~~~~~a~~~~~~~~~~~.~..... ~~~ii~~~~~~~~ii~~~~~~i~~~~~~~iiird~~~~~~~~~~i~~~~.iilliiii -i~~~~~~~~~~~~~~~~~~~~~~ii Q............................ ~ - Pk$8 ~1-: C ~:::~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~......................~ ~ ~ ~ ~ ~ ~ ~ ~~~~~::.............~ ~~~~~~~...........~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~~.:................................................... ~ r s ~~s~P ~.................................~ r ~ s ~~~ 8sa~ ppar~iii..................................~~~~~~~~~~~~~~~~~~~~~~ii............ ~ ~ ~ ~ iiii~r.............~~~~~~~~~~~~~~~~~~~~~~~~~R~iiriiiiiiiiil i jssa~ 1 ~~iiriiidiiiiiiiiiiiiiiiP 4........................~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~iiiiiiiiiiiiiiiiiii-..............~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~iiiiiiiiiiiiiiiiiiiii..............~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~iiiiiiiiiiiiiiiiiiii~................~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~iiiiiiiiiiiiiiiiii~...................iiii.............~al~as ~ j~............~~~~~~~~~~~~~~~~~~~~~~~~~~~:....................................................~................~~~~~~~~~~~~~~*::::::::::::::::::::::::::::::::::: ~~i~~i'iiiii:::-:-l. —~::~iiiiiiiiiiiii4:::::::::: ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~~~Q Si~~~~~~i~~~~iII~~~~~~~~iiB~~~C::': ~~~~~~~~~~~~~~' i:::~~~....... HtiI:il~~~~~~~~~~~~~~~~~~~~~t:: I::~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~..................

The input torque was measured between the output shaft of the electric motor gearbox and the small sprocket of the chain drive by a commercial straingage torque cell equipped with slip rings The data output of the torque cell was recorded on the second channel of the dual-channel Sanborn recorder. The torque cell when installed in this position also measures the losses due to the inefficiency of the chain drive and the bearings which support the torque cell and drive sprocket. A torque loss of approximately 5* may be attributed to the chain and bearings. It may be noted that most applications of the mechanism would require a similar input drive arrangement. A Micro-Switch actuated by a cam on the Syncro-Drive housing closed a circuit that imprinted a timing mark for each revolution of the housing on the Sanborn charts. The same Micro-Switch controlled a Standard Electric timer which indicated the time required for one (or more) revolutions of the housing. The core of a linear variable differential transformer (LVDT) was connected to the output ram. The case of the LVDT was attached to the stationary tie rods The displacement data from the LVDT Mere recorded on a single-channel Sanborn recorder. C. TEST PROCEDURE A representative of the Maroth Engineering Coo adjusted the location of the load bearing nuts to provide the smoothest operation of the unit before final data were taken. The torque cell was calibrated using a fixed torque arm and a Chatillon Dynomometer spring scale. The output load cell was calibrated against a commercial load cell of known accuracy. The load and torque cells were calibrated prior to each major series of runs. The adjustable speed drive on the motor was set at an appropriate ratio for each series of runs. Naturally the Syncro Drive required more torque at the higher loads; hence the electric drive motor slowed down according to its speed torque characteristic. No attempt was made to account for this speed drop by a change in drive ratio. The unit was operated at seven different speed settings ranging from approximately 4.5 to 16.5 inches per minute of output ram speed. At each speed setting the load was increased from 5 to 50 tons in approximately 5-ton steps. When operating at some of the higher speeds, the load was limited to a figure below 50 tons by the representative of Maroth Engineering Co. On three occasions the ram was operated from its fully retracted position to its fully extended position to determine the effects (if any) of binding due to eccentric column loading. 5

At the minimum speed (4.5 inches per minute) the Syncro Drive was operated at from 0 to 50 tons output force with the ram displacement transformer installed to monitor the displacement motion. if

II. DISCUSSION OF TESTS AND RESULTS A. EFFICIENCY 1. Description.-The Syncro Drive Mechanism was operated at various speeds and loads and the following data were recorded. 1. Torque (in. lb) vs. time - Sanborn Recorder. 2. Output force (lb) vs. time - Sanborn Recorder. 3. Time for 5 revolutions of Syncro Drive - Standard Electric Timer and marks on Sanborn charts. The torque and force values used in the computation of the efficiency are the average values taken from the Sanborn charts during the third of five timed revolutions of the Syncro Drive. The average values were obtained from approximation of the actual curves. The approximated averages were spot-checked several times with the average obtained by actual integration of the curves and the approximations were found to be within 1% of the actual values. The following computations were used to obtain the efficiency values: Input Power (HP) HP (in) = TN where T = torque cell torque (ft-lb) 5252 N = torque cell speed (rpm) - x 5 x 45 x 1 where T1 = torque cell data (in. lb) 12 Ml 1 5252 ^M1 = time for 5 revolutions of Syncro Drive (min) T! x 2.549 x 10-4 45 = ratio of chain drive M1 1 Output Power (HP) HP (out) = FD where F = force (lb) 33000 D = distance moved per minute (ft) F x 5 x 300x 1 M1 12 33500 Ml 12 33000 where F1 = load cell force (lb) M1 = time for 5 revolutions of Syncro Drive (min).300 = output stroke per revolution of Syncro Drive (in.) 5

F1 x 3.788 x 10-e M1 Efficiency Power out Power in x 3.788 x10-6 T x 2.549 x 10-4 F x 1.486 x 10-2 T. 2. Results. —The efficiencies for various mean speeds (the average of the speeds for the smallest and largest loads for a given speed series) are given be. low: Ram Speed 4.5 inches per minute Ram Speed 6.95 inches per minute Output force Efficiency Output force Efficiency 11,800 lb 50.1% 12,400 lb 47.8* 21,800 55.8 22,400 64.0 31,000 56.-1 32,600 55.6 40,500 55 9 41,000 54.1 50,500 56.6 50,000 56.0 62,000 56.7 61,500 53.7 70,500 59o8 71,000 57.0 81,000 5355 82,000 54.1 89,000 55.1 91,000 51.0 98,000 52.9 100,000 48.7 Ram Speed 9.4 inches per minute Ram Speed 11.7 inches per minute Output force Efficiency Output force Efficiency 13,600 lb 45.4* 12,800 lb 43.7* 22,400 49.6 23,600 47.4 33,600 54.2 31,600 53.6 41,000 54.1 42,000 52.0 52,000 54.2 51,000 52.2 58,000 5358 61,100 47.1 73,000 54.2 6

Ram Speed. 13.6 inches per minute Ram Speed 1355 inches per minute Output force Efficiency Output force Efficiency 14,600: lb 45.2* 14,800 lb 40. 0 22,800 46.4 24,000 45.7 32,000 52.8 32,400 50.6 40,000 51.7 42,000 48.9 51,000 50.5 52,000 49.8 60,000 50.9 63,000 56.7 Ram Speed 16.5 inches per minute Output force Efficiency 14,400 lb 41.1% 23,800 45.9 32,800 51-3 42,000 50.9 52,000 49.8 63,000 51.3 A plot of efficiency vs. ram load at various speeds appears as Fig. 2. A plot of efficiency vs. ram speed at various loads appears as Fig. 5* A sample of the Sanborn torque input and force output records appears in Fig. 4. Note the variation in torque input. Six distinct torque pulses appear per revolution of the Syncro Drive Mechanism. Based upon the known location of the timing marks (with respect to the housing of the Syncro Drive), the three major torque peaks occur when the raceways nearest the chain drive end of the Syncro Drive are lifting the load. B. OPERATION OVER FULL TRAVEL OF RAM 1. Description.-At ram loads of 10, 20, and 30 tons, the ram was operated from its fully retracted to its fully extended position. 2. Results.-No chbrge in efficiency could be noted during the 8-1/2-inch extension of the ram. C. DISPLACEMENT OF OUTPUT RAM WHILE OPERATING 1. Description.-The Syncro Drive Mechanism was operated at an average output speed of 4.5 inches per minute and at loads ranging from 0 to 50 tons 7

with the ram displacement transformer installed to monitor the displacement motion. 2. Results. —The displacement-vs.-time curve was smooth and continuous at loads up to 25 tons. At loads from 30 to 50 tons the displacement curve dipped three times per revolution of the Syncro Driveo This dip is approximately.020 inch at 50-ton load on the output ram. Figure 5 shows the Sanborn record of the displacement at 50 tons. 8

70 |o/ N / \ 65 60 I _ _ 55 z 50 o w',,DA L1._SLL 45 MEAN SPEED - 4.50 /N./ MIN.= 6.95 9.49 40.. = /3.6/ = 16.5/ 35 0 10 20 30 40 50 60 70 80 90 100 RAM LOAD (1000'S LB) Fig. 2. Efficiency vs. ram load.

65 60 z 5C 45 RAM LOAD = 2O, 000 lb 40 = 40,000 -= 60,000 =- 80,000 35 2 4 6 8 10 12 14 16 18 20 RAM SPEED (IN./MIN) Fig. 3. Efficiency vs. ram speed.

TT ~* GOao pui an@bToq jo pzoooa oTdIjBioTTTOSO't *'T, ATJIG OJ3OUS Jo { L ^IT'^^i-^Tf^:^"^!''u?^'^^ ^-*i ii n o~JZ - (IT OOOOT{ t ii::i qT'u 00 T { i- IoJ - z Iii -. M | e:iljgp^^g^^^ill~^f1 u 111f 1I I[ 11 qLT *uT 000'1T {'T II: ~~~~~ojla~~~~~~~~~e~~z -~rI ~~~~~~~~~in'B~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ ~~~~~~~t IIIII:111 i I o~~~~~~~~~~~~~~~~~~~~xaZ~~~~~~~~~~~~~~~~~~~~~~~~~~~i

^- Note: Three Dips Per Revolution,,l -r..... ll.. 1, lllllllllll,.}1.............................0..,-,,94 in. |-< ------ 1 Revolution of - Syncro Drive Fig. 5 Oscillographic record of ram displacement vs. time. J~:J 7 I.'J ~.JJX jJJJJJJJ' JJJJJJJJJ' Jl J _ _.JJ J- _ J JJ J J.T. ] J, J~ J J J. I IJ~ _,.. j j jr j j j j j j I jl j X j-| j j.,. jX j j j 1. r j j 1.., j T1..,: L LH -.lIIII,, 111TlIIIl|I,,,,,.,,t 111 411L,, }..11rI — Tr rtr1..g~EII........11 1 IRevluI 1 1111111111111Itio Io'1bIrgaIf, Id. 1111J1r r |l_ l~ r l. -r r. X_ 1_.: $y ncr o D~r 1 lL iveIIsrrrIIIIII |[WW1Ig|IIII1II 7111IBr F XigIs1-_. 5. 1s i IIlLogLphi recrda of ra disp1 1:1Hlacemen It rs IL.L 1 tI 1r 1,1 imer |. 11 T { Jli__'5,||| rl... m ll ~l-L 1J lI 1uL~il I1:114W12l- 11l~

III. CONCLUSIONS The torque input to the Syncro Drive Mechanism pulsates six times per revolution. Three of the six pulses are much larger than the others. The larger pulses occur as the load is lifted on the set of raceways nearest- the input-chain drive sprocket. The magnitude of the pulses can be somewhat controlled at a given load by the adjustment which changes the axial distance between the two operating nuts. It does not appear that the torque input curve can be made smooth and free from pulses by adjustment of the nut distance. At higher operating speeds, the torque input pulses become less severe, probably due to the flywheel effect of the rotating housing. The torque pulse requirements are still present internally; however, some of the required energy can be drawn from the flywheel and therefore does not appear as sharply on the input torque curve. Deflections that occur in the normal operation of this particular unit are sufficient to cause improper synchronization of the two raceways resulting in jerky operation and relatively low efficiency. The output displacement at high loads is not smooth but rather undergoes a series of three dips per resolution of the unit. These dips may be traced to excessive deflections in one of the pairs of raceway systems. This irregular displacement is the cause of the fluctuations in output load as shown in Fig. 4. The relief valve system in the hydraulic set up was not sensitive enough to compensate for the changes in flow caused by the ram displacement. If the Syncro Drive were to be used to lift a dead load, rather than work against a hydraulic system, the inertia forces due to the acceleration of the ram and load might be quite large. It appears that larger load-carrying capacity units could be built; however, it must be realized that the bulk of the unit may also be increased considerably. 13

IV. REC OMMENDATIONS If additional units are to be built, the following design data must be reevaluated: 1. Deflections of ram screw between load-carrying nuts. (This matter assumes great importance since the nut adjustments are alternately made on a loaded and then on an unloaded screw.) 2. Deflections of all other load-carrying members in the unit including the raceways and balls themselves. 35. Means t-a posittlyely insure proper synchronization of thC. two raceway systems (changes in efficiency of approximately 71 were noted to be due to synchronization adjustments). If the unit is intended to be operated over a wide range of ambient temperatures, an investigation into the effects of thermal expansion and contraction should be made. The close tolerances required over large distance could be greatly affected by the use of dissimilar metals which may have slightly different thermal coefficients of expansion. If heavy dead loads (those loads that may present high inertia forces when accelerated) are intended to be lifted with this unit, then further testing under this type of load is advised. 15