Development of a Three-Point-Linkage Dynamometer for Tillage Research

Transcription

1 J. ugric. Engng Res. (199) 5, Development of a Three-Point-Linkage Dynamometer for Tillage Research A. L. PALMER New South Wales Agriculture & Fisheries, Agricultural Research Centre. Trangie, NSW 83, Australia (Received 14 January 1991; accepfed in revised form 9 vember 1991) A three-point-linkage dynamometer for tillage research is described. The dynamometer was designed for ease of construction without special facilities. It is in the general form of a standard quick attachment coupler to suit tractors and implements having standard category 1, or 3 three-point-linkages. The instrumentation used consists of commercial load cells connected to commercial analogue input devices contained in an industry standard personal computer. The instrument system is powered from the tractor s electrical system. The dynamometer measures draught, vertical and lateral loads and associated moments. The instrument system contains provision for measuring ground speed, wheel slip, power-take-off speed and power-take-off torque. Force measurement errors are less than 3.9% full scale. 1. Introduction Engineers of NSW Agriculture have been involved in tillage research since Drawbar dynamometers for use with trailed implements were developed by Sanders and Holligan in 1979 and Brown m 1986 and a general purpose tillage dynamometer for testing implement components was developed by Kruger and Palmer in The research undertaken with these dynamometers was related mainly to large-scale dryland cereal farming. Agricultural engineering research in NSW Agriculture now encompasses irrigated row crops. Since most of the implements used in Australia s row crop industries are connected to tractors by three-point-linkages, the department s engineers saw a need for a three-point-linkage dynamometer. Because the nature of future research had only been defined in very general terms when the dynamometer was designed, application to as broad a range of tractors and implements as possible was a major design requirement. It was decided that standard linkage cateogries and 3 would be the most appropriate size for the dynamometer with provision for category 1 if possible.. Literature review.1. Mechanical design The published designs can be grouped into two major categories; one in which the force sensors are incorporated in the tractor and the other in which the force sensors are mounted on a detachable frame or frames. The latter category could be further split according to sensor type and number of frames. Some features of a number of published designs are summarized in Table Sensors in tractor approach A common approach was the incorporation of force-sensing elements into the tractor s linkage. This category of dynamometer has no additional frame between tractor and 157

2 158 TILLAGE RESEARCH Table 1 Frame and sensor configurations used by previous workers Ref.. of. frames Sensor rype Sensor location Quick attachment coupler Pro Frame allo wed shape Cantilever pins Cranked beam in double bending Cantilever pins Strain gauges Strain gauges Cantilever beams Bending beams Bending and torque tube Commercial cells Commercial cells Cantilever pins Cantilever beams Commercial cells Shear beams Commercial cells Bending and torsion tube Bending beams Tractor linkage pivots Tractor link ends Tractor link pivots Tractor links Tractor links Tractor linkage pivots Frame to implement Frame to implement Frame to implement Frame to tractor NO Inverted Inverted Frame to implement Inverted U U T Triangle Triangle Inverted U Inverted U Inverted U Inverted U Rectangle Inverted T implement. This approach was used by Reece,4 Morling,5 Scholtz, Luth et ~1.~ and Upadhyaya et al. Reece4 replaced the pins attaching the linkage arms to the tractor chassis by strain-gauged cantilever pins, with the gauges oriented in such a way that the output responded to forces in one direction only. This method of gauging a cantilever was described by Jensen, who also described a cantilever beam arrangement for use in a three-point-linkage dynamometer. In 1964, Scholtz used strain-gauged beams supported at both ends to re lace the original pins connecting the links to the tractor frame. Marling, Luth et al. Y and Upadhyaya et al.* used strain gauges to measure forces (strains) in the tractor s linkage system. These techniques had the advantage of minimum disruption to the tractor and implement system as there was no change to the tractor and implement geometry and little or no change to the mass and stiffness of any part of the system. This approach therefore appeared to be the method of choice for the measurement of dynamic and shock loads. The major disadvantage of this system was that, because the links containing the sensors were part of the tractor, they were not interchangeable between tractors, so that a specific sensor and link design was required for each tractor model. This restricted the application of a dynamometer to only one tractor, which did not fit well with the requirement for flexibility of application..1.. Force frame approach Another common approach was the use of a single instrumented frame attached between the implement and the tractor linkage. A variation of this approach was fhe use of two frames, one attached to the tractor linkage and the other to the implement in the manner of a quick attachment coupler. These frames were then connected to each other by force sensing elements. The use of a frame or frames was the approach adopted by Scholtz in 1966, Johnson and Voorhees, Barker et a[., Smith and Barker,13 Chung et al.,14 Reid et a1.,15

3 A. L. PALMER 159 Kendall et af.,16 Thomson and Harris, Chaplin et al., Garner et af., and Bowers. These workers used a variety of frame and sensor configurations which are summarized in Table 1. Of the 11 reports of dynamometers including a frame, or frames, seven describe dynamometers modelled on quick attachment couplers to ASAE S78.6 (Ref. 1). The other four reports describe dynamometers which are dimensionally similar to the standard quick attachment coupler but do not use hooks to attach to the implement. The majority of reports describe custom-made sensing elements which require careful manufacture and calibration. The author s past experience suggests that the reliability of such sensors is rarely as good as that of commercial load cells. The most popular frame shape reported was the inverted U which allows the use of the tractor power-take-off (p.t.) whether or not torque and speed transducers are used for the measurement of p.t.o. power. Although Scholtz pointed out the desirability of being able to use the p.t.o. in 1966, many later workers have chosen to ignore it. Of the designs in the literature, only that of Kendall et a1.,16 used commercial load cells and was of an inverted U shape to permit use of the tractor p.t.o. It was designed to fit tractors and implements complying with Category of ASAE S17.1 (Ref. ). The force frame approach had the advantages of permitting easy resolution of the forces into horizontal draught, vertical force and sideways force components and their respective moments, as well as being able to be easily fitted to any standard tractor and implement combination. Against this were the disadvantages of substantially changing the tractor and implement geometry by moving the implement backwards and vertically relative to the tractor and adding additional mass and resillience to the system. These changes rendered this approach less suitable for the investigation of tractor behaviour and shock loads. It is significant and Scholtz,6, presented a design for incorporating the sensors into the tractor linkage in 1964, but abandoned that system, in 1966, in favour of a force-measuring frame attached between tractor and implement. While some workers have developed their own sensing elements, many have opted to use commercial load cells. The only difficulty with the latter approach was the need for accurate alignment of the cells to minimize cross-sensitivity between the force components, as noted by Barker er ~1. ~ Chaplin et al. devised a simple method of providing adjustment of the load cell alignment. It was decided that a quick attachment coupler design in an inverted U shape, sized to fit category 1, or 3 linkages, would best meet the unpredictable future needs of the department s agricultural engineering research group by providing maximum flexibility of application. This design would permit use of the tractor p.t.o. if required. The use of commercial load cells in a similar way to that devised by Chaplin et ~1. was adopted... Data acquisition system Various devices have been used to provide a display or read and print out of the electrical signals produced by the load cells. The trend appeared to be towards some form of computer-based data logger, either carried on-board the tractor, (Bowers ) or connected by radio link, (Luth et a/. ). 3. Design criteria From the literature review and consideration of the future use of the dynamometer, the following design criteria were developed. The dynamometer was to be capable of resolving the implement load into three force and three moment components and to be interchangeable between tractors fitted with category or 3 three-point-linkage systems.

4 16 TILLAGE RESEARCH Interchangeability between tractors and implements was considered to be more important to future applications than the ability to measure dynamic or shock loads. The shape and dimensions of the dynamometer were to comply with ASAE S78.6 (Ref. 1) for quick-attachment couplers categories and 3 (3 to 168 kw drawbar power). If possible the design was also to meet the requirements for category 1 (15 to 35 kw drawbar power). It was considered unlikely that the dynamometer would lie within these dimensions because of the need for two frames, where a standard coupler has only one. It was therefore decided to allow the rearward displacement of the implement to be as much as 3 mm where ASAE S78.6 (Ref. 1) specifies 17 mm for categories and 3 and 13 mm for category 1. Clearance for use of the tractor p.t.o., possibly with torque and speed transducers, was considered essential. The force sensing and recording elements were to be commercially available load cells connected to a personal computer-based data logging system. The data logging system was to incorporate provision for the measurement of the true ground speed and the rotational speed of the tractor drive wheel(s). The main reason for this approach was that such a system could be used for general purpose data logging when the dynamometer was not required. The whole system was to be powered from the tractor s electrical system. This was considered to be more reliable and convenient than say engine-driven generators given that data and programs were to be held in the system s memory until copied to floppy disks. It was considered undesirable to use hard disks at all or to use floppy disks while the tractor was in motion. 4. Description of the dynamometer 4.1. Mechanical design The dynamometer consists of two inverted U shaped frames connected by six load cells as shown in Fig. 1. The load cells are arranged in an orthogonal configuration with E D C B A F D d Fig. 1. Three-point-linkage dynamometer showing disposition of load cells. A, Rear frame with quick attachment hooks to connect to implement; B, Front frame with pins to connect to tractor links; C, Housing for amplifier/multiplexer; D, Longitudinal (draught) load cells-one at the top and one at each side; E, Lateral load cell; F, Vertical load cell-one at each side

5 A. L. PALMER 161 provision for adjustment similar to that used by Chaplin et al. * The vertical and lateral load cells are contained within the thickness of the dynamometer frames. The longitudinal load cells protrude slightly forward of the dynamometer towards the tractor where they are least likely to cause interference with tractor components. Containment of the load cells within the envelope of the dynamometer frames minimized the overall size of the assembly and provided some mechanical protection against damage to the load cells. The frames are of box girder construction which aimed to minimize deflections and load cell misalignment under load. The load cells have a force capacity of approximately 5 kn which provides adequate sensitivity at the low end of the design tractor power range with sufficient strength for the high power range. The load capacities of the dynamometer assembly are limited by the load cell capacities and vary with the point of application of the load. As moments due to the application of vertical or lateral forces in planes other than the plane of the vertical and lateral load cells must be resisted by the longitudinal load cells, the horizontal draught force capacity will be affected by the magnitude and position of application of the vertical or lateral forces. As the vertical force will normally be directed downward and applied behind the plane of the vertical load cells, it will tend to counterbalance the moment due to draught forces applied below the lower hitch point, thus increasing the draught capacity. The capacities are summarized in Table. 4.. Load cell adjustment The load cells were fitted with spherical rod ends to eliminate off axis loading. Each load cell was mounted on pins, at 9 to each other and to the axis of the load cell, as described by Chaplin et al. * The position of each rod end on its pin was adjusted to align the load cell and fixed by collars on either side. A spirit level was used to measure alignment of the load cells. The dynamometer was supported on a level shaft passing through the lower tractor linkage pin holes so that the frames could be supported at a third point in either a horizontal or the normal vertical working position. The spirit level was attached in turn to the machined ends of the load cells by jigs. The lengths of the load cell/rod end assemblies were adjusted to align the two frames with 16 mm between their inner surfaces. The dynamometer was supported in a horizontal position while the alignment of the horizontal draught measuring load cells was adjusted in both directions and the alignment of each of the other three load cells was adjusted in one direction. The dynamometer was then rotated 9 about the level shaft to its normal vertical position. Table Force capacities of the dynamometer as limited by the load cells Applied load Position of load application Vertical lift In plane of implement hitch. (Rearward points of application will counterbalance draught loads applied below the lower hitch point.) Design limir, 1 kn Horizontal draught Lateral (sideways) force At lower hitch point in absence of vertical or lateral forces. 75 mm below lower hitch point. (Greater forces are possible if the moment is counterbalanced by a lifted weight behing plane of hitch.) In the plane of the hitch. (Moments due to application in other planes will reduce the horizontal draught capacity.) 1 5 5

6 16 TILLAGE RESEARCH The remaining ends of each of the vertical and lateral load cells were adjusted to complete the alignment Data acquisition The instrument system is based on a standard personal computer. The electronic signals from the load cells are multiplexed and amplified by a model PCLD-789 amplifier multiplexer board, manufactured by Advantech Co, Ltd of Taiwan, which is mounted in a box on the dynamometer. The amplified signals are digitized in the computer by a Model PCL-718 input/output card, also manufactured by Advantech. The computer will be carried in the tractor cabin and powered from the tractor s electrical system by a 1 Vd.c. to 4 Va.c. inverter with a nominal 3 W load capacity. The inverter was constructed from a commercial kit of parts to a design by Clarke.3 This inverter incorporates an auto-start circuit which activates the inverter only when a load is applied. This allows the inverter to be placed close to the tractor s batteries, to keep the high current cables as short as possible, and to draw current only when required, which minimizes battery drain. The forward voltage drop across the diodes in a computer s switch-mode power supply prevents the triggering of the inverter s auto-start circuit unless a resistive load is also applied. The plastic casing supplied in the kit was replaced with a fabricated metal box, including provision for restraint and an electrical ground to the tractor chassis. The standard power outlet was replaced with an outlet incorporating earth leakage circuit breaker protection against electric shock. The multiplexer and amplifier module is powered from the computer s regulated power supply and does not provide power for excitation of the strain gauge bridges in the load cells. The regulated load cell excitation supply, shown schematically in Fig., was connected in parallel with the computer s power supply unit. The d.c. path provided by the transformer s primary winding was sufficient to trigger the auto-start circuitry in the inverter. The circuit of the load cell excitation source is generally as described in National Semiconductor Application te 4 (Ref. 4). The excitation voltage is measured through one of the spare analogue inputs of the computer s input/output card. The excitation voltage was trimmed to slightly less than the full scale input of the input/output card so that upwards drift could be detected. During calibration, the excitation voltage remained stable with no measurable drift or noise. The use of this simple excitation D4 u Fig.. Schematic of regulated load cell excitation supply

7 A. L. PALMER 163 supply permitted the use of a single amplifier and multiplexer instead of multiple strain gauge bridge conditioners. A tare facility in the software eliminates the need to balance the strain gauge bridges to compensate for the mass of the dynamometer. The data is displayed on the computer screen and stored temporarily in the computer s memory as it is received then copied to floppy disks for permanent storage at the end of each measuring interval. An electroluminescent display was used for the computer to save space and power. This display draws its power from the computer s regulated power supply and has proven to be quite legible in all conditions except direct sunlight. The computer-based instrument system selected for this dynamometer can accept digital inputs from existing revolution counter? to measure ground speed, tractor drive wheel rotational speed or p.t.o. speed. 5. Calibration Calibration was conducted in two stages. It was first necessary to adjust and calibrate the computer s input/output card and the amplifier and multiplexer module. The manufacturer of these components recommends that the calibration be checked at regular intervals to maintain accuracy. A Memocal model 81B microprocessor-controlled calibrator, manufactured by Ero Electronic S.p.a. of Milan, was used as the precision voltage source for this calibration. This calibrator has a claimed accuracy of.% full scale for ranges of less than 1 V and.1% full scale for ranges up to 1 V. Jigs were constructed to allow the dynamometer to be loaded vertically or horizontally so that the overall calibration of the load cells, when mounted in the frames, and the Intercept =.59 R =.999 Vertical error 3.9% f.s. (4.6% load) 5 Applied vertically downward load, kn Fig. 3. Vertical load cell calibration

8 164 TILLAGE RESEARCH electronics could be confirmed. A 1 t universal testing machine was used to apply loads up to 1 kn vertically and 5 kn horizontally, i.e. to approximately 1 and 5%, respectively, of its capacity. The vertical loading jig was designed to apply a vertically downward load in a line behind the implement attachment point such that the upper horizontal load cell was loaded to capacity at approximately the same rate as the two vertical load cells. The two lower horizontal load cells were loaded at half this rate. This allowed cross-sensitivity between a vertical load and the sum of the outputs from the horizontal load ceils to be checked. The horizontal loading jig was designed to apply a load in a line below the lower implement attachment so as to load all three horizontal load cells at the same rate while applying zero load to the other load cells. Again, this system allowed the testing for cross-sensitivity between an horizontal load and the outputs from the vertical and lateral load cells. The dynamometer was subjected to both compressive, i.e. pushing the dynamometer frames toward each other, and tensile horizontal loads. The calibration data supplied with the load cells was initially used to give an indicated load for each load cell. Linear regressions were then performed to compare these indicated loadings with the applied load, both on an individual load cell basis and on the basis of the sums of the indicated loads from parallel load cells. Figs 3, 4 and 5 depict the indicated loads parallel to the applied load for the vertically downward, horizontal compression and horizontal tensile load tests respectively. In each test, the load was applied and released in steps of 1 kn. Ten readings were recorded, while the applied * + Unloading Loading- \ Intercept = -.33 R* = Horizontal error.3% f.s. (8.3% load) I I I I I 5 Applied horizontal forward load, kn J Fig. 4. Horizontal load cell calibration under forward horizontal load showing lateral and vertical cross sensitivity. Cl, total horizontal; +, total vertical (<6-g% applied horizontal);, lateral (< 1.5% applied horizontal)

9 A. L. PALMER 165 Applied horizontal rearward load, kn 5 Fig. 5. Horizontal load cell calibration under rearward horizontal load showing vertical and lateral cross sensitivity., total horizontal; +, total vertical (<lo.7 applied horizontal) Q, lateral ((4.5% applied horizontal) load was stable, at each step. All of these readings are plotted in Figs 3, 4 and 5. Little hysteresis was detected. It is interesting to note that for the vertical load calibration, the higher readings on the dynamometer were obtained during the unloading part of the cycle, which indicates hysteresis in the dynamometer, whereas for the horizontal calibration, the higher dynamometer readings were obtained during the loading part of the cycle indicating hysteresis in the testing machine. It should be noted that the zero errors, appearing as intercept values, are due to the weights of the calibration jigs and portions of the dynamometer frames. These zero offsets were not compensated for in the software set-up because of time constraints during the calibration. In all cases, the magnitude of the slope of the regression line is close to unity, demonstrating good agreement between indicated and applied loads. The applied loads had to be recorded manually from the analogue load indicator on the universal testing machine, the resolution of which was not as fine as that of the dynamometer instrument system. This, together with the small fraction of the universal testing machine s capacity used, may account for some, if not all, of the small error and hysteresis detected. Cross-sensitivity between the horizontal applied loads and the vertical and lateral load cells was generally small, as depicted in Figs 4 and 5, and could have been due to load misalignment caused by elastic flexing of the calibration jigs which was observed but not measured. As specified in the design criteria, flexing of the dynamometer itself was minimal thus maintaining load cell alignment. Flexing of the vertical loading jig caused obvious load misalignment leading to significant outputs from the horizontal load cells. These spurious horizontal and lateral readings have therefore been omitted from Fig. 3.

10 166 TILLAGE RESEARCH In use, the nominal calibration data will be adjusted to reflect the slope of the regression lines for each load cell produced during this calibration. When measuring the force exerted on the linkage by the implement, zero load readings will be taken without an implement attached to compensate for the weight of those parts of the dynamometer supported by the load cells. When measuring the forces exerted on the implement by the soil, the zero load readings will be taken with the implement attached and in a normal working attitude but without any part of it touching the ground or other support. 6. Conclusions The three-point-linkage dynamometer described in this paper provides for the measurement of loads imposed on agricultural tractors by implements mounted on a standard three-point-linkage conforming to categories 1, or 3. It could also be used with other linkage-mounted equipment, such as lifting or carrying devices, including p.t.o. driven equipment, such as hay-making machines or post-hole diggers. The use of six orthogonal load cells allows the applied load to be fully resolved into its components. The computer-based instrument system can be programmed to calculate and record any combination of the components of the applied load and speed. The 35 kg mass of the dynamometer limits its use with small tractors to lightweight implements. This mass and the rearward displacement of the implement by mm, which is slightly more than allowed by ASAE S78.6 (Ref. 1), may significantly alter the dynamic characteristics of the tractor/implement system limiting the usefulness of this dynamometer in the study of shock loads and other dynamic effects. The dynamometer will fit tractors fitted with three point linkage systems to ASAE S17.1 (Ref. ) categories 1, and 3 although the movement range of the p.t.o. may be slightly restricted on category 3 and clearance around the mudguards may be reduced on categories 1 and. Calibration showed that the system performed well in relation to the nominal specifications of its components and that the provision for accurate alignment of the load cells was successful in minimizing cross-sensitivity. The use of a single regulated load cell excitation supply and a standard amplifier multiplexer provides a cheaper and reliable alternative to expensive multiple strain gauge bridge signal conditioners. Acknowledgements The development of the dynamometer was funded by a grant from the Rural Credits Development Fund of the Reserve Bank of Australia. Mr K. Healey, Mr R. Lund, Mr D. Robbins and Mr P. Weldon assisted with the design. construction and calibration of the dynamometer. NSW State Rail, Scientific Services made their universal testing machine available for the calibrations. References Sanders, K. F.; Holligan, P.J. A self contained mobile drawbar dynamometer for tillage research. Agricultural Engineering Conference, Geelong, 3 September- October 198. (The Institution of Engineers, Australia, National Conference Publication. 8/7) 7-71 Brown, G. A. A portable vector force transducer for tractor drawbars. Conference on Agricultural Engineering, Adelaide, 4-8 August (The Institution of Engineers, Australia. National Conference Publication. 86/9) 17-3 Kruger, 1. R.; Palmer, A. L. Provision of a dynamometer for tillage energy research. End of grant report NERDDP/EG/85/379. (National Energy Research, Development and Demonstration Program. Dept of Resources and Energy, Canberra.) 1985

11 A. L. PALMER 167 Reece, A. R. A three-point-linkage dynamometer. Journal of Agricultural Engineering Research (l): Morling, R. W. Soil force analysis as applied to tillage equipment. ASAE Paper , Scholtz, D. C. A three-point linkage dynamometer for mounted implements. Journal of Agricultural Engineering Research (): 5-58 Luth, H. T.; Floyd, V. G.; Heise, R. P. Evaluating energy requirements of machines in the field. ASAE Paper , 1978 Upadhyaya, S. K.; Kemble, L. J.; Collins, N. E. Draft prediction of mounted implements from strain gage readings. ASAE Paper , Jensen, J. K. Experimental stress analysis. Agricultural Engineering (9): 65 Scholtz, D. C. A three-point linkage dynamometer for restrained linkages. Journal of Agricultural Engineering Research 1966, ll( 1): Johnson, C. E.: Voorhees, W. B. A force dynamometer for three-point hitches. Transactions of the ASAE 1979, (): 6-8 * Barker, G. L.; Smith, L. A.; Colwick, R. F. Three-point hitch dynamometer for directional force measurement. ASAE Paper , 1981 Smith, L. A.; Barker, G. L. Equipment to monitor field energy requirements. Transactions of the ASAE 198, 5(6): l4 Chung, T. G.; Marley, S. J.; Buchele, W. F. Development of a three-point hitch dynamometer. ASAE Paper , 1983 l5 Reid, J. T.; Carter, L. M.; Clark, R. L. Draft measurements with a three-point dynamometer. ASAE Paper , 1983 l6 Kendall, C. K.; Nachtigal, C. L.; Dooley, J. H. Three-point hitch dynamometer data acquisition system. ASAE Paper l7 Thomson, S. J.; Harris, H. D. Design, construction and testing of a three point linkage dynamometer. Conference on Agricultural Engineering, Adelaide, 4-8 August 1986 (The Institution of Engineers, Australia, National Conference Publication X6/9.) Chaplin, J.; Lueders, M.; Zhao, Y. Three-point hitch dynamometer design and calibration. Applied Engineering in Agriculture (l): IO-13 Garner, T. H.; Dodd, R. B.; Dan Wolf, D.; Peiper, U. M. Force analysis and application of a three-point hitch dynamometer. Transactions of the ASAE (4): * Bowers, C. G. Jr Tillage Draft and energy measurements for twelve southeastern soil series. Transactions of the ASAE 1989, 3(5): ASAE S78.6 Attachment of implements to agricultural wheel tractors equipped with quickattaching coupler. In ASAE Standards 36th ed u ASAE S17.1 Three-point free-link attachment for hitching implements to agricultural wheel tractors. In ASAE Standards 36th ed t3 Clarke, J. Compact 1/3OV 3OVA inverter. Electronics Australia 19x5. 47(9): 4-5 * Anon IC provides on-card regulation for logic circuits. National Semiconductor Application te 4 February 1971 in Linear Applications Handbook (National Semiconductor Corporation). 198 pp AN4-l-AN4-6