THE DESIGN AND CONSTRUCTION-OF A ’ powen TAKE—OFF DYNAMOMETER ‘ Thai. lorthn' Dbgmoi'MAg ,1 MICHIGAN .STATE. CQLLEGE . ' Harrison C. ‘Fiséh 195.2- ' '. ‘ _ r . II . . . -, ~ ' . -. n '- ' ”‘9“ " llllllllllllllllllllllllllllllllllllllllllmll"WI * A " ' 3 1293 10491 6139 This is to certifg that the thesis entitled "The Design and Construction of a Power Takeoff Dynamomoter" 3 presented bg Harrison C. Fisoh has been accepted towards fulfillment of the. requirements for _M._$_._ degree in Agricultural Engineering Major professor Inna December 25. 1952 lV1£3I_J RETURNING MATERIALS: P1ace in book drop to ”saunas remove this checkout from ”— your record. FINES win be charged if book is returned after the date stamped be10w. gm 2.4 «mi. 1 3 9 : THE DESIGU.AVD CONSTRUCTION CF‘A mm TAPE-OFF DYI~llI.-'O:‘E'ITR by Farrison C. Eiech ATIABSTRACT Submitted to the School of Graduate Studies of Kichigan State College of Agriculture and Applied Science in partial fulfillment of the requirements for the degree of (‘3‘; AT STE; CT Two variables are involved in rotative horsepower measurement—- torque and s need . Dynanometers suitable for measurement of power take-off horsepower consumed by farm machines fall into five classes afith regard to the ef— fects of torque thich they measure. Those effects are: l. Shaft torsional deflection 2. Belt or chain tension 3. A force between two coaxial rotating members 4. Gear reactions 5. Reactions resulting from self-furnished rotative effort. Almost any of the common methods of measuring speed may be coupled satisfactorily with any of the five classes of torque measurement. Important desirable qualities for a dynanometer to be used in re- search, teaching and extension are accuracy, simplicity of operation, quick adaptability to different tractors, sturdinoss, ability to record total energy and to give direct instantaneous readings of horsepower. To the writer‘s knowledge it is not possible to combine all of these qualities into one dynamometer, especially not when expense is a limitin: factor. It is necessary to decide which selection factors are dominant tuation and strive to satisfy those factors with the H0 in the existing 5 resources available. With the foregoing thoughts in mind, the author designed and built a power take-off dynanometer, using an automobile differential as a torque- measuring unit. The torque force is indicated by hydraulic pressure and is recorded on a roll cfiart. Speed and total revolutions are also recorded on the chart. The dynamometer mechanism is mounted on a specia two- wheel trailer which trails between the tractor and any drawn implement. The trailer also contains a drawbar dynamometer. m'r' 7 I A 7‘. (1 ‘~'~‘:Tv‘1“7' m-r "_ _" _;n D 816?. T) ,k e2; C lkl OF.A 0(“5-"7‘11 "‘7 1.1“-C'37 T“"” -, " ‘ * '7’??? ‘ \av} .¢.. . ‘ - .4 «A-.- .A. .a-, \' ... ~- by Harrison C. Fisch A ’3'" n n. T," STD 04 Submitted to the School of Craduate Studies of Yichijen (’1 State College of Agriculture and Afiplied Science in partial fulfillment of the requirements for the degree of NASTY} OF SCIEICB YEAR 1952 npyroved ;* '. f i 1 ' - ~ .‘(fffiffN‘f‘W Y‘T‘f“. ‘Y‘I ‘mfi ; AOL .IL'«£;»~~'U. ._... .l d 1- f ‘s The author gratefully acknowledges the guidance of Dr. -u tor h. Carleton in the preparation of this thesis and in the conduct of the pro- ject which this thesis concerns. Janes D Cawocd also owns the author's heartfelt thanks for his as- .J. sistance in the actual construction of the parts for the dynamometer and his practical sugrcstions for improvement. J Alnost every member of the staff of the Agricultural Engineering De- partment has been of some assistance to the author in connection with some phase of the project, for which the author is grateful. The author also wishes to express appreciation to tkm members of the . 1 reference department of the Xicnigan State College Library for their ready c00peration in obtaining interlihrary loans of theses concerning past dynamometer wor:. 1 ‘ n{-?.,a(:\u~p szf‘tvyg" ‘3 l I. II. III. TABLE OF COI-J‘TE NTS IIITRODIJCTION. O O O O O O O O O O O O O O O 0 RE QU IRELZE HTS OF A P O‘L'CR TAKE- OFF DY DIM'ZOIETER . REVIEW OF SlEED BEASUREZIENT . . . . . . . . . REVEW OF TORQUE IELASUREIEl-IT. . . . . . . . . A. B. C. Measurement of Shaft Torsional Deflection 1. 2. 3. 4. lbchanical Means. o o o o o o o o o 0 Optical Means o o o o o o o o o o o 0 Electrical means. 0 o o o o o o o o o a. Stroboscopically Illuminated Scales be Strain gages. o o o o o o o o o o 0 Calibration Methods . . . . . . . . . lieasurement of Belt or Chain Tension. . . 1. 2. Measurement of a Force Measurement of One Force and Two Angles Measurement of One Force Only . . . . Coaxial Kembers o o o o o o o o o o o o o 1. 2. 3. liechanical Weighing Linkage . . . . . Spring Deflection o o o o o o o o o o a. Indicated Mechanically. o o o o o b. Indicated Electrically. . . . . . Hydraulic Pressure. . . . . . . . . . D.. Measurement of Gear Reactions . . . . . . ii Transmitted Between Two PAGE 14 14 17 21 22 22 24 32 32 34 34 4O 44 48 V. VI. VII. VIII. X. TABLE OF CONTENTS (continued) IRGE E. Ebasurement of Self-Furnished Rotative Effort. . . . 54 1. Cradle Mountings . . . . . . . . . . . . . . . . 54 2. Power Input and Efficiency . . . . . . . . . . . 54 SELECTION OF DESIGN. . . . . . . . . . . . . . . . . . . 57 DETAILS OF COTSTRUCTION.AHD OPERATION. . . . . . . . . . 59 CALIBRATION RESULTS. . . . . . . . . . . . . . . . . . . 8O SUGGESTIOI‘IS FOR FURTHER DYNPLIIOETETER DEVELOPER . . . . . 86 A. Automatic 011 Replenishment. . . . . . . . . . . . . 86 B. Distance Recorder. . . . . . . . . . . . . . . . . . 87 C. Direct Power Readings. . . . . . . . . . . . . . . . 88 D. Indicator Positions. . . . . . . . . . . . . . . . . 89 E. Indicator Pencils. . . . . . . . . . . . . . . . . . 91 F. Recorder Ink well. . . . . . . . . . . . . . . . . . 91 G. Speed Indications. . . . . . . . . . . . . . . . . . 92 H. Self-Powered Dynamometer . . . . . . . . . . . . . . 93 REFERENCES CITED . . . . . . . . . . . . . . . . . .5. . 95 APPENDIX . . . . . . . . . . . .p. . . . . . . . . . . . 97 iii LIST OF FIGURES PAGE Fig. 1: Fottinger Torsionmeter . . . . . . . . . . . . . . . . . 8 Fig. 2 Collie's Torsionmeter. . . . . . . . . . . . . . . . . . 10 Fig. 3 Hepkinson-Thring Torsionmeter. . . . . . . . . . . . . . 12 Fig. 4 Bevis-Gibson Torsionmeter No. 1. . . . . . . . . . . . . 13 Fig. 5 Bevis—Gibson Torsionmeter No. 2. . . . . . . . . . . . . 15 Fig. 6 Amsler Torsionmeter (optical). . . . . . . . . . . . . . 16 Fig. 7 Amsler Torsionmeter (strobosc0pic) . . . . . . . . . . . 18 Fig. 8 Strain-gage Torquemeter. . . . . . . . .p. . . . . . . . 20 Fig. 9 Simplest Belt Dynamometer. . . . . . . . . . . . . . . . 23 Fig. 10 Thornycroft or Froude Belt Dynamometer . . . . . . . . . 26 Fig. 11 Batson and Hyde Belt Dynamometer . . . . . . . . . . . . 27 Fig. 12 Briggs Belt Dynamometer. . . . . . . . . . . . . . . . . 28 Fig. 13 Tatham Belt Dynamometer. . . . . . . . . . . . . . . . . 30 Fig. 14 McCall Chain Dynamometer . . . . . . . . . . . . . . . . 31 Fig. 15 Fischinger Dynamometer . . . . . . . . . . . . . . . . . 33 Fig. 16 Aryton.and Perry Torquemeter . . . . . . . . . . . . . . 35 Fig. 17 Ebrin Spring Dynamometer . . . . . . . . . . . . . . . . 36 Fig. 18 Van'Winkle Power Meter . . . . . . . . . . . . . . . . . 38 Fig. 19 Batson and Hyde Spring Dynamometer . . . . . . . . . . . 39 Fig. 20 Jones Spring Dynamometer . . . . . . . . . . . . . . . . 41 Fig. 21 Kagnetic Air-Gap Dynamometer . . . . . . . . . . . . . . 43 Fig. 22 Amsler Hydraulic Torquemeter . . . . . . . . . . . . . . 45 Fig. 23 Berry HyerUIIC Dynamometer. o o o o o o o o o o o o o O 47 iv LIST OF FIGURES (continued) PRGE Fig. 24 Batchelder Dynamometer . . . . . . . . . . . . . . . 49 Fig. 25 Rebinsan-Riehle Dynamometer. . . . . . . . . . . . . 51 Fig. 26 Planetary Gear Dynamometers. . . . . . . . . . . . . 52 Fig. 27 Cradle-Founted Power Source. . . . . . . . . . . . . 55 Fig. 28 Dynamometer Attached to Tractor. . . . ... . . . . . 59 Fig. 29 Ford Axle Torque Unit. . . . . . . . . . . . . . . . 60 Fig. 30 Torque Unit Parts. .h. . . . . . . . . . . . . . . . 60 Fig. 31 Torque Bar and Block . . . . . . . . . . . . . . . . 61 Fig. 32 Torque Unit Gears and Bearings . . . . . . . . . . . 61 Fig. 33 Torque Cylinder and Bose . . . . . . . . . . . . . . . 63 Fig. 34 Torque Cylinder Parts. . . . . . . . . . . . . . . . 63 Fig. 35 Torque Unit Gear Dimensiais. . . . . . . . . . . . . 66 Fig. 36 Reverse Unit and Chain . . . . . . . . . . . . . . . 70 Fig. 37 Reverse Unit Parts . . . . . . . . . . . . . . . . . 70 Fig. 38 Ground Drive and Speed Reducer . . . . . . . . . . . 72 Fig. 39 Pressure Indicators. . . . . . . . . . . . . . . . . 75 Fig. 40 Change Gears and Coupling. . . . . . . . . . . . . . 75 Fig. 41 Representative Chart Section . . . . . . . . . . . . 77 Fig. 42 Electric Timer . . . . . . . . . . . . . . . . . . . 78 Fig. 43 Wiring for Recorder. . . . . . . . . . . . . . . . . 78 Fig. 44 Calibration Setup. . . . . . . . . . . . . . . . . . 80 Fig. 45 Torque-PTO vs. Eddy-Current Dynamometer. . . . . . . 82 Fig. 46 Pressure-Gage vs. Indicator. . . . . . . . . . . . . 82 Fig. 47 lbthod.for Direct Power Readings . . . . . . . . . . 90 'V' INTRODUCTION' Ever since the first farm tractor was built, tractor users and de- signers have been searching for neW'ways to use rotational power to bet- ter advantage in farm.machines. The search has taken impetus from.the fact that a tractor is better fitted to deliver rotational power than linear power. The explanation for this lies chiefly in the loss in ef- ficiency which must be expected from the high gear ratios and consider- able amounts of slippage which accompany the pulling of heavy loads in the field. Whenever the PTO (power take-off) shaft can be used instead of the tractor drive wheels as a means of furnishing the necessary power, bet- ter efficiencies and other advantages are likely to arise. An example of this is the PTO (power take-off) manure spreader. Not only does it afford greater power efficiency by permitting the use of larger spreader loads with the same size of tractor, but it also permits better perform- ance under conditions where ground-drive spreaders would slip and refuse to spread, a larger choice of spreading speeds without additional apron drive complexity, and the privilege of unloading while standing still. It is not safe, however, to assume that greater efficiency and/br better performance always result from the use of PTO power to replace draft power. .An outstanding case in point is the rotary tiller, which has been in limited use ever since the first steam traction engines were put into service. In spite of more than 100 years of development effort, most recent tests indicate that the rotary tiller still takes more energy to prepare a satisfactory seedbed than does any of the conventional meth- ods of tillage using draft power. In order to adequately measure the power requirements of existing and preposed machines of this sort, some form of PTO dynamometer is re- quired. Wfithout a dynamometer, the power consumption of a machine must be arrived at by guesswork, and conclusions thus drawn are likely to be highly flavored by previous prejudices on the part of the investigator. Thus it behooves any organization engaged in research or develop- ment work concerning applications of PTO power to have a suitable dyna- mometer. It is the object of this thesis to discuss desirable construc- tion features for such a dynamometer, as well as specific details of one constructed by the writer for use at Michigan State College. III 1|" 1" I I' ‘ . , n. movemmns FOR A Pom TAKE-OFF DYNAj-FOiEER . J In about the middle of the 19th century, Morin (15) wrote these words concerning the general requirements of any dynamometer. ”l. 2. 3. 4. The sensibility of the instrument should be preportioned to the intensity of the efforts to be measured, and should not be liable to alterations by use. The indications should be obtained by methods independent of the attendance, fancies or pre- possessions of the observer, and should conse- quently be furnished by the instrument itself, by means of tracings, or material results, re- maining after the experiments. 'we should be able to ascertain the effort exerted at each point of the path described by the point of application of the effort, and in certain cases, at each instant in the period of observations. If the experiment from.its nature must be continued a long time, the apparatus should be such as can easily determine the total quantity of work expended by the motor." These general statements by Korin still hold true today in every re- spect. However, there are certain additional requirements which need to be met by a dynamometer intended to measure the PTO output of a modern farm tractor under conditions anticipated in college research, extension, and teaching. Some of these are as follows: 1. 2. The dynamometer should be compact enough to serve as a shaft coupling in the PTO shaft between the tractor and the implement, with no disturbance of the original construction of either ma- chine. The dynamometer should be capable of quick and easy change from one tractor to another, so that it might be used in field studies 3. 4. 5. 6. 4 far removed from the college where a specially adapted tractor would not be available. The dynamometer should be simple and rugged enough to be used safely, easily, and accurately by untrained Operators. The dynamometer should be capable of transmitting the power cone sumed by the largest PTObdriven machines likely to be encounter- ed. For farm machines this would likely never exceed 50 hp, and very seldom exceed 25 hp. Instantaneous shock loads equivalent to 100 hp should be expected (IO). A dynamometer to be used for non-technical teaching and extensiOn purposes should give direct readings of horsepower without need for any calculation on the part of the observer. The dynamometer should be adaptable to three-point hitch imple- ments. HEVIEW OF SPEED HEASUREEENT In any given dynamometer, only two variables are involved in rota- tive horsepower measurement--torque and speed. In this discussion the mechanisms for measuring torque will receive far more attention than those for speed measurement. This is because speed measurements are confined essentially to the two methods listed below: 1. Use of tachometers or speed counters. These may be mechanical, electric or conceivably hydraulic. 2. Driving the recorder shaft at a speed preportional to that of the power shaft and making marks on the chart at even time intervals, average speed being calculated from the distance between time marks. Either of these two general methods can'be easily combined with al- most any method of torque measurement of which the writer is aware. Com- mercially manufactured speed or time measuring devices can be convenient- ly adapted to these methods. Almost always the primary concern of the dynamometer designer is that of Obtaining a torque unit suitable for his requirements. That done, the speed measurement can be added without dif- ficulty. The mechanical speed counter or tachometer is usually the most suit- able type for a dynamometer where the main objective is satisfactory per- formance with minimum.complications and expense. Electric tachometers of the generator-voltmeter or voltage pulsation types lend themselves to greater accuracy and ease of continuous record- 6 ing, especially when the moving parts are not readily accessible for me- chanical speed measurements. When any connection, electrical or otherwise, is inconvenient, the stroboscOpe may be useful. "Strobotacs" which read directly in rpm.may be used not only to determine speed but also to illu- minate torque scales (see.Amsler Torquemeter in following pages). The writer does not happen to knOW'Of any instances where a hydraul- ic tachometer has been used, but one could easily be designed around some type of positive displacement pump, the delivery of which would be propor- tional to its speed. A simple method of measuring the delivery of the pump would be to run the fluid through a rotameter similar to those used for measuring fuel flow. This would give a speed indication directly from the height of the rotor. For applications where a record of total energy is desired, some meth- od of driving the chart in preportion to shaft speed is desirable. This Idll produce a chart with abscissa preportional to total revolutions. If torque is recorded as the ordinate, the area beneath the torque trace will be proportional to total energy. When desired, a clock:mechanism.can.be installed to make marks on the chart at timed intervals. The average shaft speed for any interval may then be calculated from.the distance between time marks on the chart. REVIEW OF TORQUE HEASUREKENT It is often the case that reviews of literature are confined only to items which the reviewer feels at first glance have direct application to the specific set of circumstances that he must meet. The writer feels that best understanding will be obtained when the salient features of each method used in the past are carefully examined. With this in mind, the writer has cited at least one application of each torque measuring or indicating principle which he has encountered in the literature. A. Dynamometers which measure Shaft Torsional Deflection 1. Measurement by Mechanical Linkages The Fottinger Torsionmeter (3) is one example of an instrument which uses mechanical linkage to indicate twist. A sketch of this instrument is shown in Fig. l. Shaft D is continuous and unbroken. Disc A is fas- tened to shaft D. Sleeve C is also keyed to shaft D. Disc B is fastened to sleeve C, but is free to turn on shaft D. Length L is the distance over which the twist is to be measured. The L-shaped lever at J and its connecting strings magnify and transmit any relative movement between the plates to the stylus F, which records this movement as a trace on drum E. Drum.E remains stationary while the stylus mechanism rotates with the shaft. Drum.E may be slipped endwise on rods K to permit changing the paper which is wrapped around the drums .A light spring at G takes the slack out of the linkage and screw H permits positioning of the stylus for any desired zero load position. . .tmkmxzobmmn. tmeztkokr \ 6Q 9 It will be noted that, because the stylus moves repeatedly around the drum without movement of the paper, a jumble of lines is likely to occur under changing loads. Thus this instrument would be best suited to measuring a constant load or one which varies in the same pattern for each shaft revolution. Collie's Torsionmeter (3) is another example of measuring torsion- al strain by mechanical means. Fig. 2 shows the plan of this instrument. A and B are light countershafts mounted on a common axis and driven by chains from opposite ends of the main shaft. The free ends of the count- ershafts are connected by a sleeve. One end of the sleeve is threaded to take a multiple-thread worm which terminates one of the countershafts; the other end of the sleeve is loosely keyed to the other countershaft. Twisting of the main shaft causes relative rotation of the two count- ershafts, which in turn causes the sleeve to move axially. iovement of the sleeve actuates a rack and pinion, producing a movement of the point- er over the torque dial. Zero adjustment of the pointer can be made with the screw at S. This dynamometer would be difficult to apply on shafts of short length because the torsional movement of the main shaft would have to be consid- erable to exceed the slack movement in the mechanical linkages. 2. Measurement by Optical Means The Hepkinson-Thring Torsionmeter (3) demonstrates the application of a beam of light to measurement of torsional deflection. .An.important ad- vantage of its design is its compactness, which is possible because of its ability to accurately magnify any small strain. Fig. 3 shows the construc- tion principles of this dynamometer. www.mézoimmfln shadow N at manna-IN m It? ”I 4 IF I {dimly} 23. T” .< .. a ” Q . o . . _ . . u . a, II. 1141 11.! J! I l s Idlltllwnltr‘“ '10»! x a 4 LL56 3e.“ _. , . . m 11 The flanged sleeve D and flange E are both clamped to shaft C. The twist in the shaft over length L when.torque is applied produces a relative movement between the flanges. The small mirror at A is mounted in pivots on flange E and fastened to flange D by means of a light flat spring. The mirror pivots are mounted radially with respect to shaft C, so that if a ' lamp G directs a beam of light at mirror A it will be deflected back to the scale F at an angle H directly proportional to the movement of the mir- ror on its pivots. This pivotal movement is in turn directly proportional to the relative movement of flanges D and E caused by torsional deflection of shaft C. The fixed mirror at B reflects the beam from lamp G directly back to scale F, providing a zero point. The movable mirror should be sil- vered on both sides, so during each revolution it will send a reflection to both sides of zero. Thus if during a prolonged test the zero point given by the fixed mirror should be disturbed, accurate reading can still be obtained by taking the mean of the readings on either side of zero. The Bevis-Gibson Torsionmeter (3) sacrifices compactness for simpli- city and accuracy. Fig. 4 illustrates its operation. The opaque discs A and B are fixed to the shaft as far apart as convenient. Each disc car- ries a narrow radial slit near its periphery. The slits are arranged so that under zero torque they will lie in the same radial plane and the light beam from the lamp can be seen in the finder. When torque is applied, disc A will rotate slightly with respect to disc B. This necessitates a move- nwnt of the finder to bring the light into view again. This movement of the finder corresponds to the angular deflection of the shaft under the ap- plied torque. Thus the scale may be constructed to read torque directly. When it is desirable to read the torque at several successive portions 12' mwfizzoinmmr oz..m£....-20wz§aol m 6E m1.lt,.....s I a .9 \.o\< mmbszoemmch 28.90.. 3mm . 1v. k LSW ka ES; uQQ whim, Eda was: oQQ III/l/ V” ~umxu>t quawxwoA s¢3$0«\.k510\qVIunuw>a%)\ .LK / ‘11.- ‘ nl ..-I .11- ., 1 .-. 1H- .II. maoko DmmooemO .Eej 10m3&< L. I . . .I. -..-I.. . I .- . MIMI. ”111-1.H....1I. -11.! - WINK: mlozt 7: mqmawS L194 1 mnekmm. o\< “I'd - 1 - 1 IIIIWF 1 , .1: .11- Sfim I503 . - _ max/Eu 14 of each revolution, this may be conveniently accomplished by the use of discs containing several slits arranged in the form of a spiral. The lamp and the finder can be designed to slide radially toward or away from.the shaft in order to align preperly with each set of holes. In situations where a long section of shaft is not available for ap- plication of the Bevis—Gibson Torsionmeter, its construction may be modi- fied to permit excellent accuracy with a shorter shaft. A lamp is placed close to the shaft inside two concentric drums, as shown in Fig. 5. The drums are slotted to permit the light to pass through, and a finder is placed at some distance from the shaft. The principles upon which this torsionmeter Operates are exactly the same as described for the longashaft form. Fig. 6 shows still another application of the light beam principle of measuring torque, in a dynamometer produced by Amsler (3). This con- tstruction.makes use of two discs, 0 and N, which carry radial slits near their periphery, and a disc M which carries a transparent scale U. Discs 0 and N are fixed to sleeve A and serve as an eyepiece through which scale U is viewed for an instant during each revolution. When collar M moves with respect to sleeve A as the shaft torque changes, the change is read directly from scale U. In order for the image to appear continuous it is desirable for the instrument to be rotating at speeds in excess of 200 rpm. 3. Measurement by Electrical Means a. Stroboscopically Illuminated Scales. Extremely simple construc- tion of the torque unit is possible when Amsler's stroboscopic illumina- tion of the torque scale (3) is applied to torsion—measuring dynamometers. 15 N .O>\ tEMSZOEQmNN 200Q$VIM<>W® W .OIxux \ -rIIrI..I III, -IIIIIIIIJII. m ”205% he m4o?€1-- I I 11.1.1 I I 111- MI omit? 020V - .LrIII I.IIIIIIIII. I.. 1 I. I 11.11.1111 1 - -II-. . I I .I -11... I I 111111.; I- -I $.0on _ owes 0% SEQ ENSO. wagon 1 1V 7 11.11W 1 IIWIDYIIP.HH1451a%4WJIV .. ~. .1 0286 mnfizx \I\. QHLEOQQSW paste thzotwkml. 16 tubmzzoiwmmm 4W. m .Qt nodkomzyou Inseam , -. . zotcuw mete. zEw 1114141. 0 KIWIV . . I..\. T . av... 025301 Ladiw 1 I.. n. . .\.1s . _ _ . _ . . m. i . . . . _ .1. . : oz.m ||.|.P rL ‘ VXNML “Y‘s V1 ‘ \V m o u M .HQSO Q with. 13?: \S Loon. Ltwatwte 1”? c. wzefzoe U Fatima .m xmjaa to not 2902MLxM 2\ L39 kOoe quJUI 37 plication of power thus causes relative movement between B and C, causing link F‘to act on lever G to produce lateral movement of link H and sliding collar D. E is a fixed collar. The scale assembly does not turn.with the shaft, but rides on the collars D and E. O is a ring which follows sleeve D under torque variations, producing movement of pointer P through link H. The direct-reading quality of the scale mechanism.is accomplished as follows. Scale 8 is a split scale. AAbove its center it has markings which correspond to the moduli of different springs which may be used to vary the torque capacity. Below the center the markings indicate different shaft speeds at which the dynamometer may be used. Scale S is movable ver- tically. In operation, scale 8 is moved vertically so that the marking cor- responding to the spring in use comes opposite the stationary mark X. Then scale plate J is moved vertically in its slots N until its upper edge rests on the marking on scale S indicating the shaft speed at which the dynamo- meter is Operating. The curved lines on scale plate J are so laid out that horsepower may then be read directly at the end of pointer P. In laying out the curved lines on scale plate J the angularity effects of the link- age at different torques must be considered along with the speed factor. Batson and Hyde (3) mention a small dynamometer which uses a pair of clock-type springs. Because of the size limitation of commercial springs of this type, this dynamometer would likely have little application in siz- es of more than five horsepower, but it is mentioned as a design which might prove useful for separate measurement of the power requirement of a small unit within a larger agricultural machine, as for example the fan on a combine. Fig. 19 shows the construction of this dynamometer. The input flange 38 DYNAMOMETEFT mm SCALE Asses/Maw REMOVED O --- » .v SPRING SET‘TWGS -- V L, {_H_L.. - _ _ _ EA 0 ‘ J SPEED SETTINGS -: N 3 3 d SCALE ASSEMBLY F76./5 l/x-I/v WNKLE Fen/ER METER 39 kmum\ . n ) J :eESEQ tZEQV teal sz chkamw @ket 40 B is fixed on the shaft A and to the clock-type springs inside the case C. The case C and the output flange D are driven by the outer ends of the springs. The relative movement of shaft.A and flange D which results from torque causes the collar B to follow the helical slots cut in shaft AA and move lengthwise in the axial slots in the sleeve C. The sharp edge of the collar E thus moves along the scale F, which can be calibrated in torque units. b. Indicated by Electrical Means. Jones (11) described the design of a torquemeter which employed a helical spring as a coupling between the tractor PTO shaft and the main power shaft of the driven.machine. Jones built the torquemeter into and around a standard slip clutch which was placed in the main power shaft of most International Harvester PTO-driven machines during the 1920's. Fig. 20 shows some of its operating principles. Shaft.A is fastened to collar B and sleeve 0. An electrode is placed along the cutaway helical edge D of sleeve C. Shaft E can turn freely in collar B but is coupled to collar B by means of the torsion spring S, so that when torque is applied, a limited amount of relative movement is per- mitted between shafts A and E. Cam.J on.shaft E opens the breaker points K which are fixed to the stationary frame M. This induces a high voltage in the secondary winding of coil H. This high voltage is introduced to electrode G beneath the chart F which is supported between two rolls as shown, and jumps the small gap between electrode G and D, as D passes by. Each time the spark jumps it makes a small hole in the chart. The chart is advanced a small amount each revolution of the shaft by means of the pawl P’and the ratchet R. Pawl P may be actuated by the breaker cam. It kthm. Seat mmwja 35K Emu mEeE «siQ zct chm, aged limit! - 1.- llllll "IIIII .U .Q Whig 0 ezEQmJ ”2‘4 OENN 3205 SN .eQ T. a my“ -u-..--m;n..mWn---uuun Q LW 5... 10. 42 ‘will be noted that the spark will occur at the same instant during each revolution. Thus as torque is applied it will cause a different portion of electrode D to be opposite electrode G when the spark occurs, produc- ing a trace of punctures on the chart which will indicate the torque. The area of the chart may be measured with a planimeter to determine the total energy expended during any test run. Different torque ranges for this instrument may be produced by using a different helix angle for elec- trode D or by using a spring S of different modulus. Another type of instrument which measures spring deflection electric- ally was referred to by Shields (18) in an article in which he described a draft dynamometer. About the torque dynamemeter he said only that it served as the hub for the rear wheel of the tractor (only one wheel was used in his experiment) and that its principles of Operation were similar to that of the draft dynamometer. From these remarks and the detailed de- scription of the draft dynamometer the writer has concluded that a possible plan for the torque dynamometer might be that shown.in.Fig. 21. The writer has no reason to believe that this plan is the one used by Shields, but it will serve to illustrate the principles. .A is a stiff, flat strip of spring steel. It carries the power from the axle and inner hub section B to the outer hub section C, which is at- tached to the driving wheel. The writer visualizes that A might fit into snug slots in.the inner and outer hubs, being removable for replacement by springs of different moduli. As A bends, permitting slight rotation of B with respect to C, the armature D, which is a strap of soft iron fixed to the outer hub, will move with respect to the slip rings, which are fixed to the axle. If coils E are attached to the slip rings on either side of 1:3 wwwmzeeiizaQ Q \ \ Z a .' . ‘ \\ ..‘ i . _ \\ ‘ ~ 5 | \ ~\ '- \ FORCE Fe.) “\ DIAGRAM . I \ \ r3.- \\\ \\ I r r: 1/ ' \H\ $2 “P W LIE) , r7 .b“ \\ 1", '\ \y ." \\ \X‘ // ~ \ I [—__1 r “I l:___] \ ’J \\ / \‘ -r/l F76- 25 Hos/Nsorv- RIEHLE DYNAMOMETER NOTE: TOOTH NUMBERS MUST ALWAYS FIT THE RELATIONSHIP A 42C a D. pLA/VET GEM? F-OQCE DIAERA M REMEMBER w EXAM/NING ‘n-«E TABULAT’IONS BELOW THAT THE MDT/0N5 SUN GEAR SPIDER PLANET GEAR R/Nc GEAR pawn/ww- m A 5 C 0 SPEED PAT/o ,. A A ,A FIXED O l+f HZ— \/+-——5 ALL FIXED -/ —/ —/ -/ rm CAeEI -/ //VP' 7' .: x5 4-.3— +A O T'PUT -A :l o L, ( u ) 0( / D) L 734 u 3 :5. — A .EL 13. B FIXED I W 0 + ‘C‘ + 9....9 ALL Fuzz: +1 4! +/' +/ -- ‘ ‘ A A s T’OTAL,CnsEZ O (F/rED/ +/(/NPUT) /+ EL /+.D..(0wpu-r) +(/¢L—).l D B FIXED 4 T 0 :g— —-l “1% ALL FIXED + I +/ +1 4/ 75TAL, C4553 It; (denser) 4-] {/mpur) /- % 0037150) +(7+g—)-'l NOTE: SUN— PLANET RAT’IOS USUALLY RANGE FRO/VI /:I To 3:2. AUTOMOBILE OVERDR/VES, A TYP/C/JL APPLICATION OF CASE 2 sac-v5, Cow/.t-romx HAVE A suxv: PLAT/Ef RAT/0 or 3.1! 2. 72/5 READER MAY sHon/ FROM 614552 Tum- ‘n/IS mac/L0 PRODUCE AN OUTF’UT-INPUT‘ SPEED RATIO or: (H; :/ on Io: 7, Hex-26 PLANETARY GEAR DYNAMOMETER 53 produces a change in speed and sometimes a change in direction. See the tabulations which accompany Fig. 26. The ratios suggested in this tabula- tion are representative of common ratios recommended for a planetary sys- tem-in continuous Operation. Sunsplanet ratios of more than 3:2 mean that the planet gears will have an appreciably lower tooth strength than the teeth with.which they mesh on the sun and ring gears, encouraging wear and breakage. This strength relation would hold true regardless of the number of planet gears used to distribute the load, although a larger total load could be handled by using more planet gears. In any case, high sunzplanet ratios will mean a less compact, more expensive torque unit. An example of a circumstance where a specific ratio was considered more important than expense is described by Shedenhelm (l7) in.connection with a planetary dynamometer used for horsepower measurements of aircraft engines in flight. In this aircraft dynamometer the flow of power is the reverse of that shown in case two of the tabulation accompanying Fig. 26. The ring gear is driven by the engine, the spider drives the propeller, and the sun gear restraining force is measured by a hydraulic cylinder. The planet gears in this case are very small and numerous (20 in all), which makes possible an overall speed ratio more closely approaching 1:1 than.the usual arrangements. In aircraft construction a planetary speed reducer is often used to produce a propeller speed below crankshaft speed, so that the conversion of these reducers to dynamometers should require no fundamental construc- tion changes. If this circumstance existed in farm'tractors, the problem of Obtaining inexpensive and effective PTO dynamometers would be consider- ably simplified. 54 The reader may observe from the free-body diagram of a single planet gear in Fig. 26 that the force on the hub bearing will always be twice the circumferential component of the force acting at the pitch line on op- posite ends of the gear diameter. Thus the force indicated for a given torque and radius from sun center will be twice as large if the spider is restrained as it will if either the sun or ring gear is restrained. E. Dynamometers which measure Self-Furnished Rotative Effort l. Cradle Mounting, with Direct Force Readings Cradle-mounted engines or motors constitute a much-used and very sat- isfactory type of transmission dynamometer for stationary indoor use. 'With adaptations, they might also prove satisfactory for field use as a PTO dy- namometer. Fig. 27 illustrates the principle on which such a dynamometer ‘works. The motor is usually cradled on large ball bearings in the frame A. The output shaft of the motor turns in the center of the cradle bearings. As the motor turns the shaft against a resistance force, an equal and op- posite force tends to rotate the motor in its bearings. This torque may be resisted by the scales, which indicate torque force directly. Linkages are fewer and calculations simpler with this type of dynamometer than any other type with which the writer is acquainted. 2. Power Input Coupled with Efficiencies When.desired, the horsepower output of a hydraulic or electric motor may be determined without cradle mountings, by measuring the power input to the motor and allowing for the efficiency, which.nmq*be determined by calibration against an absorption dynamometer. MGWDOW Twioq QMEPSQE quthw RNCQ MESeW Eotwbmxdu 4¢>NM§V N 0 r _ .T 2%.”..ch mac ulbmcfixi 56 The power input to a hydraulic motor can be readily determined from the pressure differential between the inlet and outlet sides and the rate of oil circulation. The latter factor is directly proportional to the speed, and may be calculated as some constant times the tachometer readings Fbr example, if a hydraulic motor with an efficiency of 85% is being oper- ated at 2000 rpm, at which speed it takes 20 gpm of oil with a pressure differential of 500 psi, then the horsepower“ [Zia aJL‘gggj] [13: m '1 [#:1- I mam: 13(35 #1] a 495 HP I B c MN m‘ GAL. z. m 33,000 FT-L .A convenient, although expensive, means of measuring the power con, sumption of different rotating components of a farm machine such as a corn! picker would be to drive each subassembly with a small hydraulic motor, all the motors being driven from one large pump. Electric motors might be applied in similar fashion, their power re- quirement being indicated by a wattmeter. Electric motors are in general less compact and more expensive for similar horsepowers than are hydraulic motors, so their use'would be even more limited. SELECTION OF‘DYNAMOEETER DESIGN The writer was assigned the problem of constructing a PTO dynamometer for use by Michigan State College. It was desired that the dynamometer be suitable for use in all three phases of college endeavor--teaching, extenr sion, and research. However, at the time construction of the dynamometer was first considered, the immediate pressure was for a dynamometer suit- able for measuring maximum power and total energy expended by rotary till- age machines being used in connection with college tillage projects. With that immediate application in view, the writer set about selection of the design, keeping in mind the desirable requirements for a dynamometer as set forth in Chapter II of this thesis. Under the existing administration of the tillage projects, a very highly desired dynamometer quality was simplicity and durability, since in general any highly trained personnel would be limited to supervisory capacity and would not always be with the dynamometer in the field to in- sure its preper operation. The sensitivity to torque variations which would be possible with electric indication of stresses, as by means of a strain gage, would have been an unnecessary luxury for this application where measurement of power consumed rather than the improvement in design of specific machine elements was the main objective. Expense was also a barrier to the selection of a design capable of extreme sensitivity, since under the existing condition of finances it was out of the question to request a large single expenditure for parts from which a dynamometer could be fabricated. A possible means of circumvent- 58 ing this situation would have been to accumulate parts and instruments over a period of several years, then assemble them. Since the delay in- volved would have been a highly uncertain factor, the writer decided to go ahead with plans for a.dynamometer which could be assembled from parts available at relatively low cost from industrial hardware suppliers and used parts from former machines. The development by Otto and Hall (8) of a cart-type draft dynamometer at the same time as the writer was considering a PTO dynamometer afforded van opportunity to mount a PTO dynamometer on the same cart and circumvent the difficulty which was encountered in designing a dynamometer compact enough to serve as a short, easily removed, PTO shaft coupling from the materials available. On the basis of the existing situation, then, the writer assembled the dynamometer which is described in detail in the following section. As will be pointed out, it falls considerably short of meeting all the requirements set forth in section II. It is, however, quite satisfactory for many purposes. DETAILS OF CONSTRUCTION AND CREATION The torque dynamometer completely assembled, mounted on the draft dynamometer cart and attached to a tractor is shown in Fig. 28. —‘H Fig. 28 Dynamometer Attached to Tractor The unit occupying the full length of the center of the cart is a remodeled rear axle and differential assembly from a Model A Ford light delivery truck. Figs. 29 through 32 show the construction details of this axle assembly. It will be noted that a heavy steel plate A, as shown in the figures, has replaced the ring gear on the differential. The pinion and drive shaft mounting has been removed entirely, and through the hole D designed for the drive pinion has been inserted a torque arm.B, which 6O Fig. 30 Torque Unit Parts 61 Fig. 31 Torque Bar and Block Fig. 32 Torque Unit Gears and Bearings 62 is fastened in the curved block C. The block C is held firmly to the cage of the differential by three bolts, the center one of which serves also as a pin to keep the torque arm from slipping out of its hole in the block. The block 6 serves not only as a mounting for the torque arm, but also as a splash shield to prevent oil being thrown out through the opening D in- to which the drive pinion was formerly fitted. The size of this opening is sufficient to allow the torque arm about five inches of movement at its usual operating radius, which is more than adequate to allow for the slight movement of the piston in the torque- measuring cylinder (see Figs. 35-34) which is attached at a right angle to the torque arm. It also will permit considerable piston displacement resulting from oil leakage without the arm striking the side of the hole. The only damage which would occur if the arm did strike ‘the edge of the opening D would be cessation of proper torque readings, which could be rem- edied by replenishing the oil in the. system. Replenishment is accomplished rather easily by forcing oil in through a grease fitting placed in the hy- draulic line. This method of replenishment is simple, but not automatic, and the writer feels that an automatic replenishment device as described on page 86 would be more satisfactory for extended tests. The method by which torque force is produced on the torque arm is exactly the same as that described for the Batchelder Dynamometer on page 48, and will not be described again here. As in the Batchelder Dynamometer, torque is determined as one-half the indicated force times its radius of ‘T'n application, and horsepower is determined by the standard formula “P36? 000 .. a where T is torque in in.-lb. and n is rpm. 63 Fig. 34 Torque Cylinder Parts 64 Torque may be calculated from the following formula: 1' a (P)(.785)(r)(6/5)(l/2) a .4712? where: P'I gage pressure produced by the torque .785 I»piston area in square inches r s’torque radius 6/5areverse gear reduction factor (see page 69) 1/28 factor required because with any differential dynamometer indicated force is twice trans- mitted force (see page 50) The torque radius r is adjustable through a fairly wide range, so that it is possible when operating at a nearly constant speed to choose a radius which will permit simplified readings of torque from a convene tional pressure gage. For example, at a speed of 535 rpm, one horsepower will require 118 ins-1b. of torque in the drive shaft and will produce 236 in.-lb. of torque on the torque arm. The torque-measuring cylinder shown in Figs. 33-34 wasadapted from parts of a 1946 Plymouth automobile shock absorber. It has a diameter of one inch and a cross-sectional area of 0.785 sq. in. Thus each pound of force on the piston will produce a pressure of 1.275 psi. The radius at which each 20 psi will equal one horsepower at 535 input rpm is then calculated from the following rela- or tionship, using unit cancellation as a check: . was .3 2.36 in-Ib ' zofm/h? g l u. 1‘. 1 “:9 r in] ‘63 r9!“ y- g (23‘) ("275.1— 2. [6.0 inches i, 20. Because of friction losses in the torque unit, this relation of pres- sure to horsepower will not hold exactly true, as discussed in.the followb ing section concerning calibration results. This calculation does furnish a very close approximation of the proper torque arm length adjustment, howh 65 over, and the final adjustment can then be made while calibrating against an accurate absorption dynamometer. Perhaps the most serious inadequacy of this dynamometer is its limit- ed safe power capacity. As can be seen from Fig. 31, the gears in the torque unit consist of two end gears having 24 teeth, and three bevel pin- ions with half that number of teeth. The approximate dimensions and other data concerning the gears are shown in Fig. 35. Using a text on machine design by Faires (6), we may calculate the probable tooth strengthof the pinions as follows 8 _ sYb (L-b) Fla-Tar" where: F8 .7. Endurance strength of tooth s : Endurance limit (estimated) 8 58,000 psi Y 8 Outline factor 30.308 b a Face width all/16 inch L 8 Cone distance 12 inches Pd 3 Diametral pitch a 6 2/3 (approx.) 58 000 (0.308 (0.6875 (2.0-0.6875 Thus F. s ( 1 (52567K220) ) l = 1210 pounds If we assume that each of the three pinions will carry an equal load, the total transmitted load would be 3, 630 pounds. This force would be trans- mitted at an approximate radius of 1.8 inches. The horsepower which would be transmitted at 535 rpm by a force of this magnitude would be: Tn _ (3630){l.8)(535) 6370736 " 63,000 3 55°55 hP- The reader, if acquainted with gear design, will remember that tooth strength always gives a highly optimistic view of the power which gears can transmit. The shocks caused by inaccuracies of tooth manufacture or loose assembly greatly decrease the safe load for gears. The faster the —— - Pun/v BEARING 799057” PA 0 / .' I 5' =_ " ._ . . l.’ 32. ‘30: '1 H - Y . . it ’4 ‘ T 3 I I g - _ -_— ' U v 3': ’7 7" T: : (SPHERICAL f mpuerUPrAcs) 3 I ‘95 - '.8 RD. .:\:6 efiko '. R $- : a") GEAR: Z4 TEETH P/NION: /2 7'5er ’0 .4th5 THAT ’JC‘TH FORM /5 MI}- INVOLUTE Gquson/ aremcm'r: BEVEL. p/N/QN err-CH ANGLE 30°. Omit/£77231. PITCH”: ég, ALL ABC/E m..uE.‘: 84.550 ON MEASUREMEVTS 4CCUA’QT’E To I . CI '“ Finesse; MAT-ERIA‘IL v.4: l045 STEEL. Evouenrvca LIMIT .55, 06".) P51, Ems/ELL. .800. He. 35 Becca U/wr GEAR D/MENS/ONS 67 speed and the more errors in tooth profile, the lower the safe transmitted load. The pinions in this assembly will be turning at speeds as high as 1100 rpm, and the assembly was manufactured with considerable slack (about 0.025 inch) between mating gear teeth. Following the instructions of Faires for computing the dynamic load, we may solve the equation'below by trial and error to find the saf;.transmittedt load. 4. F; 3 0.05Vm (‘C ‘2) o. 05 Van + (BC + Fag—*- where: F6 3 Dynamic load (should not exceed FE above) : l210 lbs. max. Vm s. Pitch-line velocity in fpm a 505 fpm @ 535 rpm b : Face width . 11/15 inch C : Tooth error factor (0.025 inch error assumed) : 40,000 Ft : Transmitted load d If in the first trial zero is substituted for Ft, the following re- sult is obtained: 0.05) 506) °'“75)k°'°°°)+°] ... o = 3600 funnel. (o..svs)(40.oooFT-E This trial shows that according to theory these gears should break from running free, with no load. Actually, the formula used here was de- veloped to be reasonably accurate for errors of less than 0.005 inch. The large tooth error here does not therefore lend itself to calculation by this formula, and the writer knows from.observation that the gears are more durable than thereby indicated. Thus it is very difficult to forecast how long these gears will last. Calculations for limiting wear load, also made according to Eaires (6), give a value of 6.1 hp. Since this is based primarily on shape and size of teeth instead of strength or profile errors, it would seem the most logical method for rating gears of this sort. 68 In addition to the questionable strength of the gears, another prdb- lem.presents itself in the fact that all the bearings within the differen- tial cage are plain (steel against steel). This includes both the radial and thrust surfaces for the end gears and pinions. It can be shown that a total radial load of 131 pounds on the pinion bearings is necessary to transmit one horsepower at 535 rpm. If it is assumed that the three pinions divide the load equally, then each would carry 44 pounds per horsepower. The projected area of each pinion'bear- ing (see Fig. 35) is 0.562 sq. in. Thus the bearing pressure will average 77.5 psi/hp. Marks (12) suggests that for hardened alloy steel on.hard- ened steel allowable bearing pressure would be 2100 psi. If it is assumed that this represents the conditions in the pinion bearings, the maximum safe horsepower for the pinion bearings would be $252; a 27.1 hp. It is the feeling of the writer that this high allowable pressure is optimistic for the type of steel used in these bearings, although no definite informr ation could be obtained regarding the material. To reduce danger of gelling these bearings the writer recommends use of an extreme-pressure gear lubricant of’the type suitable for hypoid gearing. The additives in this lubricant make it suitable for high rub- bing pressures. A viscosity specification of SKE 140 is to be preferred because the shock-absorbing qualities of a heavy lubricant are superior to a light lubricant in loose bearings and gears, both of which exist in this assembly. The cage assembly was disassembled after the original test and no evidence of additional wear was noticeable on the gear teeth, so it is the hope of the writer that the gears and bearings will last longer than 69 suggested by some of the calculations herein presented. Because the torque unit produced a reversal of direction in the pow- er train, it was necessary to install a reverse mechanism of some sort to restore the proper rotation direction for standard farm machines. For this purpose a reversing gear was salvaged from the PTO assembly of a 'Willys Universal Jeep which had been long unused as a source of PTO power. .A close-up view of it is shown in Figs. 36-37. It consists of twoihelical gears with a face width of 0.75 inches. One has 20 teeth, the other 24, giving a speed reduction of 6:5. This means that the output shaft of the dynamometer makes only five revolutions for each six revolutions of the tractor PTO shaft. These gears are carried on tapered roller bearings. A calculation of capacity for these gears by methods suggested by Faires (6) indicated that they should be safe for 25 hp at 535 rpm. It will be noticed from Fig. 36 that a roller chain is used to drive the reverse gear. This is to permit both the input and output shafts of the dynamometer to lie along the center of the cart, eliminating trouble- some misaligned universal joints which would occur if one of the shafts was off center. The chain, being slightly flexible, also absorbs the slight misalignment resulting from.a bent input shaft on the reverse gear assembly. Two 15-tooth No. 50 (five-eighths inch pitch) sprockets were used on four-inch centers. Cullman's Chain Catalog (5) suggests that a five-eighths inch pitch roller chain running under the conditions existing here should not carry continuously more than about four horsepower. This recommendation is based primarily on wear considerations. The average ultimate strength of this chain is 6100 pounds, which on a 15-tooth sprocket at 535 rpm would handle 86 horsepower. 7O Fig. 37 Reverse Unit Parts 71 Another justification for using this under-rated chain is that stock sprockets of larger pitch which would fit within the space limitations were not available. A three-fourths inch pitch 15-tooth sprocket, the smallest commonly available sprocket in that size, would have a pitch diameter of almost exactly four inches. 0n four-inch centers this would cause interference. The flexible shaft from'which the recorder chart is driven is shown in Fig. 37. It fits into a slotted adapter which screws into the center of the gear shaft. This adapter turns in a sevenpeighths inch spark plug body, which serves as a means of mounting the flexible shaft to the re- verse gear housing. The small speed reducer which is used to drive the recorder chart is shown in Fig. 38. It has a reduction ratio of 10031 in a single worm.and sector gear. The sprocket ratio between it and the input shaft of the re- ducer was chosen to be 6336, because this proved convenient for permitting quick change of chains to drive from.aither the ground wheels of the cart or the flexible shaft attached to the reverse gear without need for sprock- et changes. The chart drive mechanism itself has a built-in reduction ratio of 332 which cannot be changed. It also has three pairs of change gears (see Fig. 40). Change gears not in use are stored inside the bottom of the chart drive housing. The ratios obtainable with these change gears are 131, 2:1, 1:2, 1:4, and 4:1. Each revolution of the final drive (chart roll) advances the chart 4.5 inches. The flexible shaft which drives the chart is attached to the smaller of the two gears in the reverse gear housing. This means that it turns 72 Fig. 38 Ground Drive and Speed Reducer 73 faster than the PTO output shaft by a ratio of 6:5. The overall ratio from the PTO output shaft up to the change gears in the chart drive thus becomes (5:6)(100:1)(6:1)(3:2) 3750:l. By combining the chart drive change gear ratios with this 750:1 ratio, the reader may obtain 187.5:1, 375:1, 750:1, 1500:l and 3000:l as ratios between the PTO output shaft and the chart roll. These ratios will permit the following PTO revolu- tions per inch of chart: 41.7, 83.3, 167, 334, and 667. The writer has found that a chart travel of one inch per 167 revolu- tions is necessary to catch almost all torque fluctuations as separate humps on the chart. For a more extended test, where average torque is the point of interest, a slower chart Speed would be desirable to save on chart expense and chart—reading time. The ground drive may be used interchangeably with the PTO drive for the chart by moving the chain shown in.Fig. 38 from the speed reducer sprocket down to the final sprocket of the ground drive. In Fig. 38 this sprocket is visible as the only one which does not carry a chain. The mountings of the speed reducer and chart drive housing are flexible enough to permit removal of this chain without difficulty. The ground drive consists simply of a series of five sprockets plus the sprocket which is mounted on the chart drive housing, which is used for both PTO and ground drive. The ground drive is engaged by tightening a nut which secures the sprocket to the wheel hub. .A handle shown in the figure facilitates this operation. The sprocket on the wheel hub has 12 teeth, and the others have sizes which produce the following ratio up through the chart drive sprocket (36 teeth): (12:21)(12:17)(22:36) a 1.4.05 74 The dynamometer tires have a loaded circumference of 89 inches. Using a 2:1 reduction in the chart drive change gears, as mentioned earlier for the PTO drive, a chart movement of one inch for 20 feet of ground travel results. It follows from the change gear ratios previously discussed that chart movements of one inch for 2.5, 5, 10, 20, or 40 feet can be quickly obtained. Two steam-cylinder indicators were adapted for use in recording torque and draft on the chart. Fig. 39 shows the indicators installed on the chart drive housing. Different springs and pistons were available with both indicators to provide different pressure ranges. The springs and pistons for the Lehman and Michele indicator at the left are arranged on top of the indicator housing in Fig. 39. Those for the Robertson indicat- or are leaning against the lower portion of the chart. ‘A description of both of the indicators with a tabulation of spring relationships will be found in the Appendix. It is usually more satisfactory to use the R0bertson indicator with the lower pressure ranges for the drawbar cylinder, using the high pres- sure (Lehman-Michels) indicator for the torque cylinder. The indicators may be exchanged to register either draft or torque by exchanging hose lines. Hydraulic quick-couplers, one of which is shown at C in Fig. 40, facilitate this Operatione At a PTO load of 20 horsepower a pressure of about 400 psi would be expected, while with a 3.5 mph draWbar load of the same magnitude only about 120 psi would be likely. The writer chose a small cylinder which would develop high pressures for PTO readings with the thought that piston friction would have a less significant effect on the readings. --~.a—————-__.. -M- ‘—.- ~‘\\\\\\\\\ -\\\\\\\\\\\ ~‘ \ \\\\\\\\\\\\\ Fig. 40 Change Gears and Coupling 76 The chart drive housing which is used with this dynamometer was in- tended for a stationary operation recorder Of the on-off type, used to record the running time of refrigerator motors, coolant pumps, and simil- ar things. It has room for several pens to be mounted simultaneously on the chart. The pens have a sidewise movement of about 1/8 inch, being pulled to the side by an electromagnet when a six-volt circuit is closed, then returned to their original position.by spring pressure when the cir- cuit is broken. Two of these on-off pens were retained with the assembly for use with the PTO-Draft Dynamometer. They are visible in Fig. 40. One is operated by a timing mechanism, the other manually. The clock-operated pen produces marks on the chart which are one-half second apart in time. The six-volt electric timer, shown in Fig. 42, closes the circuit for one-half second, Opens it for one-half second, then closes it again. This produces a square- cornered zig-zag trace as shown in Fig. 41. This trace can be used to determine either average speed or total time expended in a given test. The manually operated pen is controlled by a button which may be mounted on the tractor or any other convenient place. The Operator may produce by means of a code number Of depressions of the button an instan- taneous record on the chart of the time when some event takes place, such as passing a marker stake or striking an obstruction. Upon reading the chart, these code markings will be helpful in interpreting the indicator traces. The wiring method for operation of the two pens is shown in Fig. 43. A six-volt battery, shown at thexight, supplies the current to Operate the pens and the electric timer. The timer is mounted in the box which.carries 77 “— I .‘- -‘. .I -‘ 9&1"rtiiillllllllll'fllal‘ii" '\ .n :nwu’ .‘t'ti'l ,. Fig. 41 Representative Chart Section 78 Fig. 43 'Wiring for Recorder 79 the miles per hour meter. This box is visible in the foreground. Beside each of these terminals a small white number or letter is faintly visible. The top terminal, marked C, is the common ground or return terminal for all the pens. The actuating current for each of the pens is introduced at the numbered terminal which corresponds to the position of the pen in the recorder. The pens are numbered from left to right as they are viewed from the front of the recorder. The positive terminal of the battery is attached to the "POS" terminal of the timer. The terminal marked "LT" on the timer furnishes timed current impulses to a recorder pen. The button for manual pen control is shown resting on the torque unit, with its wires coiled around it. This button acts as a switch in a line from the battery (via the "PCS" terminal of the timer) to the pen. The common ground on.the chart housing is grounded to the timer housing, which is in turn grounded to the negative terminal of the battery. In its present form, the PTObDraft Dynamometer does not give direct readings of horsepower. 'With either the PTO or drawbar sections of the dynamometer, it is necessary to wait until the dynamometer stops and the chart can be read to determine horsepower. Torque or draft may be deter- mined directly from pressure gages. When necessary, a portable tachometer may be attached to the tractor engine to give a basis for calculation of PTO speed under load. Suggestions for more convenient instantaneous horse- power readings will be found in the last section of this thesis. CALIBRATION RESULTS For calibration, the PTO dynamometer was attached as shown in Fig. 44 to the eddy-current absorption dynamometer used for belt horsepower tests by the Agricultural Engineering Department. ‘A standard 1.375 inch six-spline shaft was attached to the absorption dynamometer, making it pos- sible to connect the two dynamometers with any universal joint shaft hav- ing a standard spline in both end-yokes. At this writing, both New Hol- land and Fbx hay choppers are equipped wdth joints of this sort, so ar- rangements were made to borrow one from a dealer for the calibration test. A Ford-Ferguson tractor owned by the Agricultural Engineering Department was used to power the calibration test. Fig. 44 calibration Setup ‘81 The dynamometers were allowed to run continuously under part load for two hours before the final readings were taken, to allow time for the lubricants in the various Operating parts to come up to equilibrium temper- ature and to make sure that everything was working properly. The field current through the absorption dynamometer was then increased from the min- imum (which produced a reading of six pounds four ounces on the scale) to produce load increments of five pounts up to 50 pounds. Output speed was maintained at about 520 rpm. The gage pressures produced by the loads on the PTO dynamometer as well as the pressure indicator readings were record- ed for each increment of absorption load. The results are summarized graphically in Figs. 45-46. Fig. 45 shows that the torque calculated from.PTO indicator readings consistently parallels the torque absorbed by the eddy-current dynamometer, but exceeds it by about 25 in.-lb. at all loads except idle, where the dif- ference is greater. This difference in torque results from the friction and lubricant drag in the PTO dynamometer. As indicated by the dotted lines in the PTO tOrque curve in Fig. 45, the variation in torque between a scale weight of one pound (which indicates the friction and windage in the eddy-current dynamometer) and six and one-quarter pounds is uncertain, because the load rose to the higher figure as soon as the field generator was started, even.with the rheostat in the maximum.resistance position. Thus it is likely, but hard to prove, that the approximate 25 in.-lb. dif- ference between indicated and actual torque comes into effect at some load between one and six pounds. The writer believes on the basis of this calibration test that rather accurate indications of power will be Obtained from this dynamometer through— HYDRAULIC P95534195 — PSI 82 800 , . ' / ,x 700» .. /,/ / /' 2 'Jj 2% a”. eff/"[34- f’ q 6" ,// {O .XX)’ O‘A’( ,1 «(9 VA,» /m/ “1.0 Never/ON Loss 4J0 0433/ 394“ wPI’O Dm/AMOMEV’ER / f , / ,« \ . ’ fiche“ ' . . f , +’ Boo e0 /’/£IU6 . , ’. ‘40“ 2(1) {4/- , f /00 - f?" I O ,a/‘I' '_ -_-- _ 0 5 /0 I5 20 25 3O 35 ‘90 45 ha} HO i- 1"va 80* 70 - (ad 50 NET SCALE W‘s/or” - £3. ‘ F76. 45 Vo'mvus -—P.7'0 rs. [DOYeCURRENT DYNAMOME‘QER DIFFERENCE av Hesse URE HEAD/Nos OF GAGE AND /lVD/CATOR /O /5 3O 35 4O 45 50 ”57' SCALE WEIGHT - L a. 56.46 PRESSURE -- GAGE vs. IND/(Arm? 20 25 83 out its working range by the simple process of reducing torque values cal- culated from indicator readings by 25 in.-1b. For example, suppose that under a certain load the indicator pencil deflects a distance which cor- responds to 92.5 psi. (For moduli of indicator springs see Appendix.) On page 64 the following torque formula was developed for this dynamometer: Torque = .471rP where: r 8 effective length dynamometer torque arm P 8 indicated load pressure If we assume an.erm.length of 15 inches, the torque becomes .47l(15)(92.5): 655 in.-lb. In Fig. 45 this corresponds to the PTO torque reading at 40 pounds not scale weight. Subtracting 25 in.-lb. from our calculated value, we Obtain as the true torque 630 in.-lb. In Fig. 45 this corresponds to the eddy-current dynamometer torque at 40 pounds not scale weight. The latter torque was calculated as the product of the scale weight and its effective radius, or (40)(15.756) :5 630 in.-1b. Note that the 25 in.-lb. factor in this calculation was obtained as the vertical distance between the upper and lower curves in Fig. 45. In Fig. 46 gage pressures and pressures calculated from.indicator readings are plotted against net scale weight at the absorption dynamometer. It can be seen from this figure that gage pressures were consistently four to seven psi higher than indicator pressures. The writer believes that this was at least in part a result of the method used for damping the pres- sure pulsations at the gage to Obtain steady readings. .A partially Opened globe valve was used for this purpose, and it was observed that it acted as a check valve, permitting pressure to rise but not to fall without furth- er opening Of the valve. The type of seat used in the construction of this 84 valve was responsible for this action. The writer was careful always to approach the load from.the low side, but he believes that pressure pulsa- tions resulting from vibration may have exceeded the average pressure, and that the gage indicated the upper level of these pulsations. Future operators should therefore be cautioned not to regard gage readings as exactly accurate, unless a different method is used to dampen vibration. The Lehman and Michele pressure indicator was used throughout the test, but springs were exchanged frequently to permit maximum.sensitivity Of readings. In spite of five spring changes during the ten load changes, the calculated indicator pressure shown in Fig. 46 approaches a straight line and parallels the gage pressure. This indicates to the writer that the moduli of the different springs have been accurately maintained dur- ing service. If’the moduli of the springs did not correspond accurately wdth the labels,‘we would expect that calculations based on these moduli would produce a stepped indicator pressure line instead of the straight line seen in Fig. 46. ' The only dependable gage available for the test was limited to 100 psi, so that tests above that were limited to indicator readings only. As it happened, only one reading was taken above this level, because the writer was forced to discontinue the test at a load slightly above 50 pounds on the absorption scale as a result of slippage of the chain which drives the reverse gears. It was looser than may have been necessary at the start of the test, and the fact that both of its sprockets are out- board of its bearings permitted them to move somewhat under load, further loosening the chain. The writer could have tightened the chain by remov- ing one link or by moving the reverse gear housing in its adjusting slots 85 and could have continued the test to higher loads, but decided against it for the following reasons: 1. 2. Sufficient data had already been Obtained to demonstrate the ac- curacy of the dynamometer. The load under which the chain slipped was in excess of seven horsepower at 520 rpm, which was greater than the limiting load for wear in the differential gears (6.1 hp) mentioned in the pre- ceding section. SUGGESTIONS FOR FUTURE DYMUECEETER DEVELOHS'IENT The PTO-Draft Dynamometer as discussed in.the preceding pages is as yet by no means a perfect instrument. The writer wishes to Offer here a number of suggestions regarding things which he feels should be added to the dynamometer to improve its performance and widen its usefulness. A. Automatic Oil Replenishment Perhaps the thing which is most urgently needed at the time Of this writing is a means of automatically replenishing the Oil supply in both the torque and draft cylinders. The problem is more acute in.the draft cylinder, because it has a large volume and seems more susceptible to leaks. The writer has installed a reserve oil tank for the draft cyl- inder which allows Oil to be drawn into the qylinder and air to be re- moved by the process Of moving the piston back and forth manually. This tank is visible at A in Fig. 43. This tank Offers an improvement in con- venience over'the former method of trying to pour oil into the line, but seems to be effective in removing the air only when the Operator has a clear understanding of the techniques required for the process--such things as not moving the cylinder too fast, having the proper valves Open, and similar details. The method used by the writer for filling the torque cylinder, that of forcing oil into the line under pressure with a grease gun, seems to have a more positive effect in expelling the air. This method is not practical for the draft cylinder, however, because of the relatively large volumes of oil required. The author feels that a rather 87 simple and effective method of filling both lines under pressure would be to construct or Obtain a tank of about five gallons water capacity which would be capable of withstanding 90 psi, which is the maximum air pressure obtainable in.the Agricultural Engineering Building. About two gallons of oil could be poured into the tank, the remainder of the tank volume being used for air pressure storage. The oil from.the bottom of the tank could be admitted by suitable valving or preferably hydraulic quick-couplers to either the torque or draft hydraulic circuit. This could be done quickly with little Operator effort whenever necessary during the course of an ex- tended test. A tank of such large volume should maintain sufficient air pressure to be effective even after several quarts of oil have escaped from the bottom of the tank. This method could be made entirely automatic by installing it with a system of check valves, so that the pressure of the air in the taim:would be acting to force Oil into the cylinder at all times when the load pres- sure in the cylinder was not sufficient to overcome the air pressure. A similar method of check valving was used satisfactorily in an.English dyna- mometer used by the National Institute of Agricultural Engineering (16). The check valves could be constructed from ball bearings mounted in such a way that gravity would tend to close the valve, as in piston-type spray pumps 0 B. Distance Recorder ‘At this “miting there is no means of recording the ground distance covered by the dynamometer while the chart is being driven from the PTO shaft unless the measurement is made manually and recorded by hand on the chart. A record of the ground distance is essential if it is desired to 1‘gflfli] 88 determine the total energy expended over a certain ground area with some sort of PTO implement.. The writer thinks of two methods which might be used satisfactorily to obtain a record of ground distance. One method would involve the use of a mechanical counter of some sort, which would be operated by the ground drive already used with the draft dynamometer. A simple hand revolution counter might be satisfactory for short runs. For extended tests a type of counter similar to that used in odometers or hourmeters would be better suited. The other method would involve the use of a third Operation pen on the chart, in addition to the ones used for time and event recording at this writing. The circuit which would Operate this pen could be opened and closed at regular distance intervals by a set of breaker points oper- ated from the dynamometer ground drive. The wmiter would favor the latter method because it would record the information permanently on the chart. C. Direct Readings of Power Both Hall (8) and.the writer originally planned to construct direct- reading attachments which would give a reading of power directly without computation. Time available for construction did not permit the develop- ment of this phase of the machine, but some of the parts are already on hand for the addition Of this feature. The plan agreed on by Hall and the waiter involved the use of a standard generator-voltmeter type of tachometer. This has been purchased. The voltmeter has already been mounted on the dynamometer, and is visible as the miles-per-hour meter in Fig. 43. The generator is stored in the Research Laboratory Instru- ment Room. The generator was to be operated either by the PTO or ground drive, as desired. Its output was to be regulated by a variable resistance, 89 which would change the voltmeter readings in proportion to the change in resistance. The variable resistance was to be attached to the pencil linkage of either pressure indicator, so that the resistance would be pro- -portional to either torque or draft pressure, as desired. A suggested wiring diagram for this apparatus is shown in Fig. 47. The direct read- ings of power would be based on the assumption.that the voltage output of the generator would be proportional to speed and the voltage drop across the resistance would be proportional to pressure. Thus the voltage across the voltmeter, being proportional to both pressure and speed, would also be proportional to power. The main obstacle to completion of this part of the dynamometer has been the construction of a suitable variable re- sistance. In order to avoid affecting the readings of the indicator to which it wdll be attached, it must be practically friction-free, yet it must have a resistance which varies in an accurate straight-line relation- ship with the pressure. The writer visualizes that a solenoid-wound re- sistance coil wdth a.center plunger carried in recirculating ball bushings and attached to the indicator might prove satisfactory, but at this writ- ing he has not devised what he considers a workable method of eliminating variable resistance within the sliding contact which would be used. D. Indicator Position Vfitlzthe pressure indicators installed as they now are, it is diffi- cult to coordinate time and pressure on the chart without considerable measurement. As shown in Fig. 39, the indicators are mounted on the front face of‘the chart, while the time pen is mounted on the top face. This means that the recording of time for a given torque variation.is performed about two inches earlier on.the chart in terms of chart travel than casual moioqmnw kmianfi Leonid ton QoEM§ mvcbfi. S MMQSESC . . .NMEME worioexwk 0 god Whig Longu > kgsfieo; /\/V‘ ' Likmomsmx HMQQOomE mmonmmwol Obi flak Lhtto OKQ KEESS . xm kaqokoq 0 mgommmmq . mokqmm «mo 91 observation of the chart would indicate. This leads to confusion in read- ing the chart by anyone not cognizant of this fact, and even if the reader Of’the chart is aware of the distance difference he may make mistakes in direction or magnitude of the measurements he makes to coordinate time and force. The writer can see no way to correct this difficulty without complete- ly rebuilding the chart drive and its housing. In order to have both in- dicators abreast of the time and event pens on the chart, a wider chart with an entirely different system of rolls and flat surfaces would be necessary. E. Indicator Pencils The pressure indicators are at this writing equipped with ordinary pencil leads, which give a readable mark only if the lead is soft. A soft lead wears rapidly, so that during the course of a day's work in the field it may be necessary to sharpen, reposition or renew the pencil point a dozen times. Each new pencil point fits into its holder in a slightly different position, so the zero point on the chart is changed. If ball- point pen units or some other method Of producing a visible mark with less maintenance time could be installed, test time and tempers would be con- served. F. Recorder Ink well The Operation recorder Which'was procured for use with the dynamometer was not designed for portable, bouncing use. This is Obvious from the fact that the inkwell which feeds the built-in pens is a partially covered trough about five inches long and three-fourths inch deep. In order for 92 the pens to work properly, a considerable amount Of ink must be in the reservoir, yet anything more than a trace will spill out when the dyna- mometer moves over rough ground. This produces large red blotches on the chart, blotting out the record, and necessitating a stop to refill the reservoir before proceeding. It would be possible to seal this ink- well by putting rubber seals around the shank Of each pen.and plugging all other holes, wdth a removable plug for ink addition. This would re- quire many hours Of delicate work, since the pen shanks are slender glass tubes, but the “miter believes on.the basis of one season's use Of the recorder that this work would be well worth while. An incidental con- venience resulting from a sealed ink reservoir would be that it would not be necessary to clean the ink from the well every night to prevent pens clogging the next day with dried ink. G. Speed Indications The clock-operated pen in the recorder now affords a record of ground or PTO shaft speed, but there is no means Of Observing this speed directly. 'With a small amount of additional linkage, a tachometer or speedometer or both could be mounted on the dynamometer. The writer has installed a suit- able fitting for driving a tachometer on the speed reducer. This fitting is visible at B in Fig. 43. It is mounted on the main worm shaft, which travels 1.2 times as fast as the dynamometer output shaft. A 5:6 reduc- tion would be necessary to Operate a direct-drive tachometer. Installation Of a speedometer would require the addition of suitable gearing to provide it with a fairly rapid shaft speed, since the speed Of the ground drive final shaft at ordinary field speeds is only about 20 rpm. 95 H. Self-Powered Dynamometer Unit After using a cart-type dynamometer for draft tests for one season, the writer is finnky convinced that a self-powered unit of some sort would be a definitely preferable form Of dynamometer. The cart dynamometer is about ten feet long from front pin to rear pin. Add this tO the length Of the implements being tested, and the overall size becomes so great as to make plot work very inconvenient. It is admitted that a cart dynamometer is usually easier to build and easier to transport to the place of test- ing, but from that point on its advantages begin to dim in the eyes Of the operator manipulating a 40-foot assembly of dynamometer and once-over till- age equipment through a field laid out in Latin square plots. A feature Of growing importance that is difficult to build into a cart-type dynamometer is the ability to test three—point hitch implements. It is true that suitable linkages can be mounted on the rear Of the cart to carry such tools, but the force relations and traction effects which may have importance in the test cannot be the same as if the dynamometer is built into the tractor itself. SO it is the recommendation Of the writer that it may be wise to spend available time in the design of a self—contained dynamometer and power unit which may better fill the needs of the Agricultural Engineer- ing Department rather than make all of the additions or changes which have been herein suggested to make the cart dynamometer more useful. The writer would suggest building the new dynamometer into some existing standard mod- el Of tractor. The model to be used would depend upon the type Of torque unit which is tO be used and the space available for this type of unit ‘Within the tractor transmission. .A rather careful preliminary investiga- 94 tion would be necessary befbre selecting the tractor in which to build the unit. The writer suggests that some Of the major implement companies would be willing to work out a long-term lease agreement fer the use Of a new tractor for this purpose, although there is no reason an Older trac- tor owned by the college could not be used if its transmission design is suitable. The torque unit which the wmiter feels would be most suitable for this tractor-contained dynamometer is some type similar in principle to the Berry Dynamometer (l4) herein earlier described. This is a compact unit which produces hydraulic pressure without the need for seals of any sort, eliminating all likelihood of leaks except in the pressure indicat- or itself. TO reduce leakage from this source, the writer would suggest that an.indicator be built from hydraulic valve parts, which are more suitable for Oil pressure than are steam cylinder indicators, or, if money is available, that a recording unit intended for recording of wide ranges of hydraulic pressure be purchased. To record three-point hitch draft, the writer feels that pressure- sensitive links to replace the solid links now used on three-point hitch tractors would be quite satisfactory. This would mean the use of at least two and possibly three separate pressure indicators for the draft portion Of the dynamometer, in addition to one for the torque portion. Methods for recording time and distance similar to those used or suggested for the cart dynamometer should be suitable for this dynamometer also. 1. 2. 3. 5. 6. 7. 8. 9. RE EEPJBTTCES CITED Ali, Syed A. "Theoretical and Experimental Stress Analysis Of Common Mechanisms in Farm Kachinery." Thesis, Michigan State College, 1952. Atwood, Stephen S. Electric and Yagnetic Fields. J. hfiley and Sons, N. Y., 1949. Batson, R. G. and Hyde, J. H. Mechanical Testing. E. P. Dutton & CO., . N. Y., 1923. p. 36-100. Burrough, D. E. "A Method for Determining Power and Torque in Farm machinery Under Field Conditions." Paper presented at.American Society Of Agricultural Engineers Meeting, Chicago, Illinois, December, 1951. Cullman Chain Catalog, No. 50, Cullman Chain CO., Chicago, Illinois, 1951. . Faires, Virgil Moring. Design of Machine Elements. hachillan CO., N. Y., 1948. Flather, John J. Dynamometers and ihasurement g£_Power. J. Wiley & Sons, N. Y., 1902. p. 74-165. Hall, Garth Omer. "The Design and Construction of a Drawbar Dyna- mometer for Use in.Agricultural Research." Thesis, Michigan State College, 1952. Ham, C. W}, and Crane, E. J. Eschanics Eglrachinezy. McGraw-Hill Book CO., N. Y., 1938. 54-1! 10. 11. 12. 13. 14. 15. 16. 17. "18. 96 Hansen, Nerlin. "Loads Imposed on Power Takeoff Shafts by Farm Im- plements." Paper presented at American Society of Agricultural Engineers Meeting, Chicago, Illinois, December, 1951. Jones, M. M. "The Direct Application of Mechanical Power to Tillage." Thesis, Iowa State College, 1927. Hanks, Lionel S. Mechanical Engineers Handbook. McGraw-Hill Book CO., N. Y., 1941. p. 1017. McCall, Rcbert J. "Design and Construction of a Tranemdssion Dyna- mometer." Thesis, Ohio State University, 1941. HcKibben, Eugene C. "1950 Annual Report." USDA Tillage Laboratory, Auburn,.Alabama. Morin, A. J. Enclycopedia Americana, Vol. 9, 1949. p. 464. National Institute of Agricultural Engineering, Silsoe, Bedfordshire, England. Technical Memorandum 22[109e/47/A. August 1950. Shedenhelm, L. E. Pilot's Powerplant Hanual. Civil Aeronautics Bulletin No. 28, October 1940. p. 27. Shields, J. W} "Electric Dynamometer for Testing Tractor Uses." Agricultural Engineering, September 1949, p. 433. APR-:3 ND IX LEHIJAN AND MICPELS PRESSURE INDICATOR (For Robertson indicator see last page) The indicator herein.described is now being used as one Of the indi- cators on the PTO-Draft Dynamometer. The following information has been prepared to simplify the use of the indicator by persons not already fa- miliar with it. The indicator came to us from the Hechanical Engineering Department and its box bears the inventony number KE 1296. It was made in Germany; hence, most infcrmation printed on it is in German. The indicator is similar in design to conventional steamrcylinder indicators, but has a much greater number of pressure ranges available than usually considered necessary for steam. Accompanying the indicator are nine Springs of increasing moduli and finree sizes of pistons. Printed around the base of each spring is a notation made up Of three parts. An example follows: 1 kg 3 30 n/m...0,75 kg...klb. 20,27 m/m "1 kg 8 30 m/h" means 1 kg/bmz Of pressure in the indicator produces a movement of 30 mm at the pencil with this spring and the proper size piston. "0,75 kg" means that the approximate Operating pressure for which this spring is intended is 0.75 kg/hmZ. This pressure would correspond ‘With a pencil movement Of (0.75) (30)/25.4 8 0.885 inches, which is rough- ly half’the possible pencil movement of 1.75 inches. "Klb. 20,27 m/m" refers to the piston size for which the first two 98 notations will hold true with this spring. Kolben means piston; 20.27 mm is the diameter. Any Of the nine springs may be used with any of the three pistons, but a conversion factor for piston area will be required if a piston other than that specified on.the spring is applied. The three piston sizes and their areas are as follows: Kolben 20.27 mm, 0.5 sq. in. Kolben 14.35 mm, 0.25 sq. in. Kolben 9.06 mm, 0.1 sq. in. Thus, if a spring bearing the label Klb. 20,27 m/m is used with the 9.06 mm kolben, the actual pressure will be five times that directly in- dicated. Similar relations can be drawn for any spring with any piston. Sleeves must be changed with a special wrench provided in.the box whenever pistons are changed. Accompanying each spring is a small ruler which may be used to scale the pressure directly fromlthe chart produced by the pencil. Suppose, for example, that we are using kolben.20.27 mm with the spring labeled 1 kg 3 1 mm and pencil movement of 0.75 inch is Observed when pressure is applied. Then use of the small ruler accompanying the spring will indicate a pres- sure Of 19 kg/bmg. Inspection of the complete spring label, however, will reveal that this spring was meant for 9.06 mm kolben which has 1/5 the area of the kolben we are using. Thus our true pressure is 3.8 kg/bmg. l kg/cmz is equivalent to 14.2 psi, so in English units our pressure is 54 psi. On page 100 is a tabulation of spring and piston relations which will simplify the choice of proper pistons or spring size wdthout extensive calculations. v a I . O u u D a . 99 For determination of draft or PTO loads the following information Will be useful: Diameter draft piston I 4.75 in. Area 3 17.72 sq. in. Diameter torque piston : 1.00 in. Area 8 0.785 sq. in. Torque radius :13.0 in. (Kay be varied if desired) Note: Torque indicated at torque cylinder is exactly twice that required to drive the tested machine. This is a result Of the inherent multiplication of the differential torque unit. ‘Spring Label kg/mm 1 1 1 kg 3 45mm kg 8 30mm kg 3 20mm kg 8 16mm kg : 12mm kg ll 7mm. kg 1: 3.5mm kg 3 2.25m 5 kg kg‘: 45mm kg 3 30mm kg 3 20mm kg 3 16mm 3 12mm kg :’ 7mm kg I: 3.5mm kg 3 2.25m kg a: 1mm PlSTON‘SPnING RELATIONSHIPS LEI—51.3414 AND HICHL‘LS momma 1. Using 20.27mm Piston 100 Actual Spring Eodulus kg/mm 1 kg 1 kg 1 kg 1 kg 1 kg 1 kg 1 kg 1 kg 1 kg 1 kg 1 kg 1 kg 1 kg 2. 45mm 30mm 20mm 16mm 12mm 7mm 3051111“ 405111111 5mm Using 14.35mm Piston psi/in 8.02 12.01 18.05 22.54 30.02 51.50 103.0 80.1 72.0 22.5mm 15mm 10mm 8mm 6mm 3.5mm 1075111111 2.25m 2.5mm 16.04 24.02 36.1 45.08 60.04 103.0 206.0 160.2 144 flax. Safe Pressure psi 13 19 32 38 51 90 177 141 123 26 38 64 77 102 177 354 281 246 101 3. Using 9.06mm Piston Sprig Label Actual Spring Modulus I48/1111“ kg/mm psi/in 1 kg 3 45mm 1 kg 0 9mm 40.1 1 kg 2 30mm 1 kg 3 6mm 60.6 1 kg ‘5 20mm 1 kg 3 4mm 90.25 1 kg 8 16mm 1 kg 8 3.2mm 112.7 1 kg 3 12mm 1 kg : 2.4mm 150.1 1 kg 3 7mm 1 kg 81.4mm 259.5 1 kg a 3.5mm 1 kg 3 0.7mm 515.0 1 kg 3 2.25m 1 kg : 0.9mm 400.0 1 kg 3 1mm 1 kg 1mm 360.2 I’ax. Safe Pressure psi 64 96 160 192 256 448 885 703 615 3"”. h V... .'i 102 ROBERTSON PRESSUEE INDICATCR This indicator has three springs and two pistons. The springs pro- duce a pencil displacement of one inch for 10, 40, or 50 psi respectively, with the large piston. The large piston has a diameter Of 0.799 inch and an area Of 0.50 square inch. The small piston has a diameter Of 0.565 inch ani an area of 0.25 square inch. Note that effective spring moduli will be doubled with this piston. This indicator is limited to about 200 psi because of limitations in the pencil linkage (2 inch maximum movement). Thus it is limited to ap- plications Where rotational horsepower does not exceed about 8 hp at 540 rpm and to a draft force of 3540 1b. This may be compared to the capacity Of the L 8: M indicator of 39 hp @. 540 rpm and 16,000 lb. draft. - . . APR 1919671! ' "JUM ' . I i rx. _ l 1 . 1’ ' A I .; X 5 '\ 1 1 1 I l 't I W I i F I '. ’ I t I f. k , I y I I ‘ ~ - e '- l I a 6 f )I - " ' 1 I ,| I ~0 ' 7" ' ‘ ‘7 l ‘ I. h .- .’_. no I. 4 ' ' I ' . , . x r. r - . . . ,' . - l r — O “1“ a ‘ ' ‘ -.' '. ' : \ -' x . e ‘3". . ~l ‘ 6 .. e. _ |' v o . ' , .. j - -.\'J "'P" "-' ‘ . J | 'v \ ' l ' I ‘I a . .. u .-.‘_‘ -x»¢~,~~‘ __ _ I -II .; 9‘ ,‘ -1". - ‘\_ . ,‘ ~_ ‘ a t \- I e , - i401. ' V,‘J. 3‘"- 3“? ~l .' '. .J -_ _ - . n. . ”R‘s“ H.- n _ ° , ' “\’( Y. ,‘ ' ‘. A . ' .7',‘ - e“ ‘x "I o ‘ — u I ' .‘J. x ‘ 1' ’- 1 . - v I ,_ .6 J“ . "7'1? 4414111444 TI“