Introduction

Yield monitor accuracy has been the focus of numerous field investigations over the last several years. Accurate crop yield data must be obtained so that producers can make informed crop production decisions. Significant errors in crop yield prediction can adversely affect any attempts to manage crop production. Perez-Munoz and Colvin (1994) noted that yield monitors tend to overestimate crop yield in regions of low yield. Development of the Yield Monitor Test Facility was motivated by the need for producers to have confidence in yield monitor technology.

  Literature Review

A study was conducted by Shearer et al. (1997) in which four combines (two John Deere 9500's, a Gleaner R70 and a Case IH 1680) were used to harvest wheat with in the same field. Each combine was equipped with yield monitors, the John Deere’s used GreenStar® and the other combines used the Ag Leader 2000. The data from each combine was imported into a popular mapping package to compare the yield estimates. They found that significant differences in measured yield existed between adjacent harvest swaths. These differences were attributed to potential machine/operator variability. In response to these differences, they developed data filtering techniques to help minimize the influence of calibration and machine/operator differences. They concluded that machine calibration techniques were critical in multiple-combine harvest systems.

In a study conducted by Kettle and Peterson (1998), it was observed that the accuracy of combine yield monitors were significantly effected by hill-side conditions and variation in harvest rates. Test were conducted on two John Deere combines equipped with Greenstar® in the hilly Palouse region of northern Idaho. They observed that the yield monitor’s response to variation in inflow was exponential rather than linear and suggested that calibration should been done at several operating points rather than the traditional two point method. The observed yield monitor error was 20.3% when operating the combine at 1/3 the calibration through-put and 5.7% at half the calibration through-put. They also found that the error in yield monitor estimates were as high as 18.2% when harvesting uphill and 60.7% when harvesting downhill, on 6 to 9% slopes. Finally, they found a very weak correlation between the yield monitor estimates and hand samples (R² =0.203) indicating that the yield monitor is not as sensitive to local variation in crop yield.

Arslan and Colvin (1998) reported the development of a test stand for combine grain yield monitors. They constructed a facility to compare the accuracy of a yield monitor to an electronic scale. The found strong correlations between the measured flow in the yield monitor and the electronic scale (R² =0.99). They observed stronger agreement when testing at higher flow rates over longer durations.

A test facility has been developed by Kormann et al. (1998) to test multiple clean grain elevator yield monitors simultaneously. A system was built using a Massey Ferguson 38/40 elevator with three reference yield systems installed, the RDS Ceres II volumetric meter, and the MF Flow-Control and the Ag Leader 2000 mass flow sensors. A second elevator was placed in series with the first elevator and was equipped with the CLASS Quantimeter II. It should be noted that this system did not use the clean grain elevator fountain auger on either of the elevators. The system was designed so that the elevators could be tilted up to 15° forward or aft with the potential for a simultaneous tilt of 15° to port or starboard. In addition, provision was made to vary mass flow rate from 1 ton/hr to 35 ton/hr. They conducted test varying the inflow rate from 10 to 35 ton./hr and found mean calibration errors less than 3%. Only at the lowest level did they experience error of up to 7%. The volumetric sensors tended to perform better under low flow conditions then did the mass flow sensors. A second set of test were conducted by varying the tilt of the elevator system. They found mean errors up to 6% with standard deviations of up to 6%. The radiometric MF Flowcontrol device performed the best of all sensors with mean errors less than 2%, while the other sensors performed as follows from most to least accurate; CLASS Quantimeter II (<3%), AG Leader 2000 (<4%) and RDS Ceres II (<6%).

Facility and Research Objectives

The objectives for the Yield Monitor Test Facility development were as follows:

1.To develop a material handling and weigh tank system capable of storing 17.5 cubic meters (500 Bu.) of grain and provide grain flow to and from the clean grain elevator at rates up to 158 cubic meters per hour (4500 Bu/hr).

2. To develop a grain metering system capable of accurately varying grain flow from 2 to 158 cubic meters per hour ( 50 to 4500 Bu/hr) to the clean grain elevator with feedback measurement of the volumetric delivery.

3. Develop an elevator support fixture capable of simulating hill-side conditions typically experienced by grain combines (forward to aft, and port to starboard pitch of  12 degrees or 21% slope).

4. To develop a control and data logging system capable of generating a range of flow conditions and gathering data from the grain meter volumetric flow, the clean grain elevator yield monitor’s measured flow and the weigh tank’s measured accumulated mass at a one second sampling rate.

The research objectives for this paper were primarily associated with calibrating and characterizing system performance, and comparing mass flow measurement capability of the test facility with that of a commercially available yield monitor system under variation in grain flow at static set points. The specific research objectives were:

1. To calibrate weigh scale and grain metering system;

2. To quantify flow characteristic of system; i.e. transport delay and sensitivity to flow levels; and

3. To compare the performance of the test facility with a commercially available yield monitor under variation in static grain flow levels ranging from 1.3 to 21.1 kg/s.  

 Component and System Design

Conceptual Design

There were several key features which guided the development of the yield monitor test facility. First, it was desired to build a laboratory system that could handle up to 17 m³ of grain in batch storage and be housed in a laboratory. The system needed to be able to transfer grain from a supply tank to the clean grain elevator and then on to a receiving tank equipped with weigh scales. The grain must then be re-circulated to the supply tank for further test runs. One of the facilities design criteria was to dynamically change the elevator inflow rates. The flow rate design range was established at approximately 2 to 159 m³/hr for dry grain with a reasonable limit of 95 m³/hr for high moisture grain. This wide flow regime is in response to studies by Kormann et al. (1998) where errors up to 7% were observed in force impetus yield monitors under low inflow conditions of 10 metric tons per hour. Finally, as reported in the works of Kettle and Peterson (1998), yield monitor performance under hillside conditions has been identified as a source of up 60% yield error when combining in downhill conditions of 6 to 9% slope. As a result, the test facility was designed to accommodate up to 21% slopes (12 degree) in the forward, aft, port and starboard directions as well as a combination.

 Grain Handling System

The grain handling system was designed to hold up to 17.6 m³ of grain and is capable of flow rates up to 159 m³/hr of dry grain. The system plan view is shown in Figure 1 and digitized slide of the over all system is shown in Figure 2. The system consists of two 2.8 m diameter, two ring tall Brock hopper tanks. The supply tank was used to hold all of the grain prior to the start of each test. The receiving tank is support on a frame which is in turn equally supported by four 4500 kg load cells. Grain leaves the supply through a 30 cm diameter supply auger which is powered by a 2.2 kW Allen-Bradley AC variable speed motor drive. The grain is discharged into a volumetric grain metering device which consist of a surge hopper and two flow control turrets. The first turret is approximately 46 cm long and runs at 1.8 to 15.9 m³/hr through the use of a 0.75 kW Allen-Bradley AC variable speed drive. The second turret is 91 cm long, powered by a 1.5 kW Allen-Bradley AC variable speed drive and runs at 15.9  to 159 m³/hr. The grain meter is equipped with level control so that the 30 cm supply auger is operated intermittently as grain is metered out of the supply tank. The grain is transferred from the meter into a 25 cm horizontal auger which feeds the supply elevator. The supply elevator operates at a constant speed of approximately 400 rpm using a 5.6 kW AC motor and is capable of capacities of up to 159 m³/hr. Grain discharges from the supply auger into the 25 cm clean grain elevator infeed auger. The spouting connection in between the supply elevator and the clean grain elevator infeed auger is constructed of 25 cm diameter wire re-inforced flexible PVC duct to allow for hillside simulation motion of the clean grain elevator.

Once the grain enters the clean grain elevator it travels through the normal path of grain and discharges from the top of the fountain auger into a suspended catch hopper. The clean grain elevator is powered by a variable speed 15 kW hydraulic motor provided by White Hydraulics. The speed of the hydraulic motor is approximately 400 rpm and is controlled by a laboratory hydraulic test bench. This hopper discharges the grain into another 25 cm flexible spout, which discharges into a 30 cm U-trough. The U-trough operates at approximately 190 rpm and is powered by a 2.2 kW electric motor. The U-trough moves the grain up to a discharge point above the grain pump. The 25 cm grain pump is manufactured by Hutchinson and is capable of flow rates up 211 m³/hr of dry grain. It is powered by a 15 kW electric motor with a chain speed of 100 m/s. The grain pump elevates and discharges the grain into the receiving tank where weight data is being sampled once per second. The grain pump is also used to move the grain from the receiving tank back to the supply tank for repeat testing. 

Grain Metering

The grain metering device consist of a surge hopper and two flow control turrets. The grain meter surge bin is capable of holding up 1 m³ of grain, refer to Figure 3. The first turret is approximately 46 cm (M1), runs between 1.8 to 15.9 m³/hr, and is powered by a 0.75 kW AC variable speed drive. The second turret is 91 cm long (M2), powered by a 1.5 kW AC variable speed drive and runs between 15.9 and 159 m³/hr. Each drive system consists of a 1750 RPM motor, a direct coupled 90° gear reducer, and a chain and sprocket drive linkage to provide the final speed reduction. M1 runs at approximately 7 RPM to give 15.9 m³ /hr and M2 runs at approximately 36 RPM to give 159 m³/hr.  The flow meter motor is powered by an Allen-Bradley AC variable speed inverter drive which is in turn controlled by a 0 to 10 V speed reference signal provided by the SLC-500 programmable controller. The variable speed drive is capable of  a 10:1 speed reduction as the motor frequency is varied from 60 to 6 Hz, with constant torque throughout the speed range.

 Each meter has a low level sensor, while a single sensor serves to detect high level in the surge hopper. Grain level is maintained approximately in between the high and low levels. All three level sensors are 18 mm capacitive proximity switches manufactured by Carlo Gavazzi. In addition, each turret motor shaft speed is monitored using a 12 mm inductive proximity switch (Altech Corp) which picks up the rotation of a shaft coupler fitted with two lugs located 180 degrees apart. Each revolution of the turret motor shaft generates two 24 volt DC pulses. This pulse train is conditioned by an optical isolator and a de-bouncing circuit to provide a clean TTL signal. The shaft speed is measured using a Keithley CTM05 counter card which runs in the background on the system PC. The shaft speed is used to confirm the volumetric delivery of grain from the meter. Meter flow is calibrated for each specific grain type.

Elevator Support and Hillside Simulation Fixture

The clean grain elevator from a John Deere 9600 combine was installed in a fixture which was designed for simulating hillside conditions. The fixture consists of the following components; a pitch and yaw pivoting support base, hydraulic cylinder mounting brackets and supports; hydraulic motor assembly to drive the elevator; elevator support tower; elevator infeed hopper; discharge grain collection hopper and flexible spouting. The base consists of a steel floor support stand which is rigidly attached to a 10 cm x 10 cm x 2.5 cm steel base plate which is in turn fixed to the concrete floor using epoxy adhesive anchor bolts. A fixed 5 cm diameter shaft is located at the top of the floor support stand. A rectangular steel tube support frame is mounted horizontally on the pitch shaft using a sleeve bearing at each end of the shaft. This shaft allows the elevator to pitch in order to simulate forward to aft motion. Another shaft is mounted on top of the support frame (rotated 90° from the pitch shaft) using two additional sleeve bearings to provide yaw rotation in order to simulate port to starboard motion. The elevator support tower sets on top of the yaw shaft and is attached to the sleeve bearings. The support tower serves to hold the clean grain elevator, the infeed hopper, the hydraulic motor, and the discharge grain collection hopper in place. It is constructed out of structural steel tubing.

The hillside simulation fixture is capable of pitch and yaw of 15° in either directions and can accommodate simultaneous pitch and yaw. In its present configuration, the axes are rotated and held in position by turnbuckles. However, it is intended that the system will be enhanced with the addition of hydraulic cylinders which will allow for dynamic testing of hillside conditions for the clean grain elevator.

Control System and Instrumentation

The control system for the grain handling system was developed for two levels of operation. First, the system was designed to run under manual control from the motor control center, where the motor contactors are controlled through a traditional pushbutton switch and relay system. Each motor was provided with an individual circuit breaker and motor starter/heater unit. Emergency stop and motor overload protection were provided to disable the system if a motor overheats or if the emergency stop is pushed. This manual mode of operation was provided for maintenance and emergency operation. The second level of control was developed to provide automated control of the start up process, level and speed control for the grain meter, systems monitoring, data acquisition, dynamic grain inflow simulation and dynamic hillside condition simulation. A three position switch is used to select between manual and automatic system operation. The Yield Monitor Test Facility control system hierarchy is shown in Figure 4. The automated control system consists of a Allen-Bradley SLC-500 programmable logic controller for machine systems control, and a Gateway 2000 Pentium personal computer which serves as the User Interface Terminal (UIT) and as the Data Acquisition System (DAS). The UIT was programmed in Visual Basic 6.0 with the use of Allen-Bradley’s RSTools^TM, Computer Board’s Universal Library and Keithly’s DriverLINX ActiveX controls. This software were used to develop the user interface screens for both motor control and data acquisition control. A typical Visual basic interface screen is shown in Figure 5 for the automatic start control and grain flow parameter selection.The SLC-500 was programmed in ladder logic using Allen-Bradley’s RSLogix 500^TM to  control all of the grain handling system functions.

The primary components of the data acquisition system were the Pentium PC, a Computerboards’ CIO-DAS801 analog card and a Keithley CTM05 counter card. The PC was as a user interface to the data acquisition system and for data storage. The principal data variables were the four receiving tank load cell weights, the tare weight prior to grain flow,  the grain meter shaft speeds, and an interrupt driven running clock for assisting in data synchronization. A synchronization signal is sent to the yield monitor undergoing testing to initiate data acquisition simultaneously with the start of data collection in the PC based system. This signal assists in associating the weigh scale and meter data with the yield monitor data. However, future enhancements will incorporate the GPS time into the PC based data acquisition for precise synchronization.

The receiving tank was instrumented with four 4500 kg load cells which supported the bulk tank and its support frame. Each load cell was calibrated at the factory to a response range of 3.0 mV/volt of excitation at full scale loading. The load cells were matched and calibrated  in the laboratory with a load cell transmitter. The transmitter provided zero offset and gain adjustment. It also provides a stable 10 V DC excitation voltage to the load cell with line length compensation. Additionally, the transmitter provides a scaled 4 to 20 mA output signal which provides a more robust signal in the presence of potential RF noise produced by the AC inverters. The load cell signals are transmitted to the CIO-DAS801 for data acquisition. The four signals are combined to give a tank weight which can be used to monitor the flow of grain through the system.

The grain meter shaft speeds are used to determine the volumetric flow rate of grain being supplied to the clean grain elevator . Each turret motor has a shaft speed sensing unit, which consist of a special dual pick-up shaft coupling and a inductive proximity switch. The pick-up lugs are tapped into the shaft coupling at 180 degrees apart and rotates at the same RPM as the motor shaft. Therefore each revolution of the motor shaft produces two pulses on the inductive proximity switch. This signal is monitored by a Keithley CTM05 counter card in the PC. The CTM05 counts pulses per second, calculates turret shaft speed based on the gear ratio, and then determines the delivered volumetric flow.

 Experimental Methods

A series of calibration test were conducted on the Yield Monitor Test Facility system. Since this system is intended to access the accuracy of clean grain elevator yield monitors it must first be calibrated and validated to insure that the results that it measures are accurate. All calibration test reported in this document were conducted using corn at normal storage moisture content conditions.

Receiving Tank Weigh Scale System

The load cells and transmitters were individually calibrated in the laboratory. It was also necessary to calibrate the overall system as installed under the receiving tank. In order to do this, the DAS801 was used to collect a tare weight from the empty receiving tank. Grain was then transported on a flat bed fifth wheel trailer equipped with two 6.5 m³ hoppers. Approximately 5.3 m³ of grain were taken to a local feed mill to establish a net grain weight. The batch of grain was then loaded into the receiving tank and the batch weight was recorded using the DAS801. A second batch was also taken to the feed mill consisting of approximately 10.6 m³  and weighed. This method provided a reference weight for two batches independent of each other. As a result, a response curve for 5.3 and 10.6 m³ of grain was generated to confirm system calibration. The percentage difference between the mill scale and the receiving tank scale for the two test batches were 0.7% for the 5.3 m³ batch and 1% for the 10.6 m³ batch. These percent errors were considered to be reasonable for this type of system.

  Grain Meter Speed and Volumetric Delivery

The grain meter consist of two turrets, one 46 cm long (M1) and the other 91 cm long (M2). The two meters are operated separately, the first providing grain flow from 1.8 to 15.9 m³/hr and the second providing flow rates from 15.9 to 159 m³/hr. The primary strategy for calibrating the metering system was to: 1) confirm that the grain meter level control and 30 cm supply auger are maintaining level throughout the flow rate range and 2) determine the amount of grain moved through the meter to the weight scale at a preset flow rate and flow duration time. This was accomplished by selecting three discrete flow rates for M1 and six flow rates for M2. The total grain mass and bulk density of the grain was pre-determined and grain was circulated once through the system to ensure that all trap points were full. The grain batch was then transferred from the supply tank to the receiving tank with the meter running at the prescribed flow rates. The time was recorded for complete transfer, the DAS801 was used to record accumulated grain mass at one second intervals and the meter shaft speed was recorded. In addition, a visual check was performed to make sure that the minimum grain level was maintained above each turret during the entire process. This data made it possible to determine both instantaneous and overall variability in flow rate and shaft speed. This procedure was repeated for all flow set points for both M1 and M2. This information was then used to develop a linear flow set point adjustment relationship that would compensate for error and provide an actual flow with in 0.2% of set point at corn flows of 21.1 kg/s and 1.28% at corn flows of 1.3 kg/s. The meter shaft speeds over the flow range from 1.3 kg/s to 21.1 kg/s were found to be fairly stable with a mean of 636 rpm and a standard deviation of 12.1 rpm and mean of 1096 rpm and standard deviation of 15.1 rpm, respectively. It should be noted that in every observed case, the measured rpm varied between two points. For example, the measured turret speed for 21.1 kg/s flow oscillated between 1080 and 1110 rpm. This is likely due to the sampling rate of 1 hz being too low. A higher sampling rate should correct this and improve the accuracy of this measurement.

 Transport Delays

As a result of the transport distance which exist between the elevator and the receiving tank, it is necessary to establish a relationship between the time at which the grain leaves the clean grain elevator and the time at which it enters the receiving tank. The grain transport system which is down stream from the clean grain elevator runs at a constant speed. Consequently, for most higher grain flow rates, the transport time should  be fairly independent of the inflow rate to the elevator. However, at the lower flow rates there may be some internal surging as the grain builds up in the inclined section of the U-trough auger and vertical leg of the grain pump. In order to identify these relationships a series of test were conducted at discrete flow rates between 1.8  and 159 m³/hr. During each test the grain flow was started at the meter with the rest of the system running. The elapsed time required for grain to arrive at the GreenStar sensor and at the receiving tank weight scale were documented for the range of grain flows. The results of the transport delay test are presented in table 2.

Static Grain Flow Tests

The major impetus of the design of this test facility was to determine the response of the yield monitor sensor to variation in inflow rates. Grain meter calibration tests were conducted prior to evaluating the response of a GreenStar® yield monitor. Yield monitor data were collected simultaneously with some of  the calibration testing. If a calibration test failed, it was repeated after data analysis to complete all execution runs defined for the yield monitor inflow rate test. In general, recirculation of  10.6 m³ batch of grain was limited to seven runs. There does not appear to be any effect on the yield monitor performance as a result of the number of grain flow cycles, with in the range of cycles used in this test. A list of the static flow test runs are provided in Table 1.

The GreenStar® yield monitor system was equipped with the latest hardware and firmware version presently being marketed by John Deere as of June 2000. Likewise, the system was calibrated using the revised two point calibration. First the system was calibrated at a corn flow set point of 12.7 kg/s, and the recommended adjustments were made in the calibration factor. Then the system was set to “Low Flow Comp Mode” and the system was run at 6.3 kg/s and the flow comp number was adjusted. The remainder of the tests were run with this calibration.

An individual static flow test was run in the following manner. First, the test corn was transferred from the grain transport trailer into the supply tank. Samples of grain were taken at discrete times in the grain flow for use in an Dickey John GAC 2100 grain test unit to determine moisture content. The bulk density and broken corn and foreign material were evaluated using the recommended practice found in the USDA Grain Inspection Handbook. A small sample of grain was run through the system to fill all void spaces in the flow path from the supply tank to the receiving tank. The SLC-500 grain meter parameters were selected to set the mode of operation and the desired throughput. Data logging with the GreenStar® system was initiated and a new test farm and field number were defined. The system was started up working backward from the grain pump to the supply elevator. The grain meter was filled with grain. Consequently, starting the grain meter and supply auger would begin the flow of grain at the prescribed flow rate. Data was collected for a predetermined test run time which was based on the number of bushels in the system and the flow rate. Activating grain flow at the meter initiated data acquisition at the weigh scale, grain meter shaft speed sensor and the GreenStar® system. The grain flow continued until the preset runtime expired. The system was then systematically shut down to allow all grain to move from the grain meter through the clean grain elevator and on to the receiving tank weight scale. The accumulated data was then saved to file on the hard drive of the PC and the PC Card of the GreenStar® system.  This data was then used to evaluate the performance of the GreenStar® sensor under variation in inflow rate.

 Results and Discussion

The GreenStar® yield monitor sensor was evaluated for variation in measured flow ranging from 1.3 kg/s to 21.1 kg/s. It was found that the GreenStar® yield sensor performed well across this wide range of static flows as shown in Table 1. At calibration flow rates of 12.7 kg/s, the percent difference in total mass flow between the GreenStar® and the weight scale were below 1%, while the difference only increased to 3% at the extreme flow rates of 1.3 and 21.1 kg/s. These results were replicated during the second set of test runs confirming that the GreenStar® sensor is able to predict accumulated grain flow. This is further confirmed when examining Figures 6, 7 and 8, which compare the accumulated mass of the weight scale, GreenStar® and grain meter. At grain flows of 4.2, 12.7 and 21.1 kg/s, there is minimal observed deviation between the accumulated masses measured and/or predicted by the three devices. In all three cases, the weigh scale runs slightly higher than the grain meter during flow, but seems to home in on the meter value at the end of flow. This may be caused by momentum forces in the falling grain stream as it impacts the surface of the grain in the weigh tank and thus transmits a dynamic load into the load cells  The GreenStar® seems to more closely follow the weigh scale at the lower flows, while it more closely follows the meter’s accumulated mass at the high flow rate. However, the difference is still minimal when looking at the accumulated mass.

 The instantaneous flow rates recorded by GreenStar® exhibited a moderate amount of variability at the calibration flow level of 12.7 kg/s, while higher variation is experienced at the low flow rates. The instantaneous flow rates are presented in Table 2. The mean GreenStar® instantaneous flow rate was 13.0 kg with a standard deviation of 0.5kg. At two standard deviations from mean the variation is under 8%. Similar results were found for the high end flow of 21.1 kg/s. At this level, the instantaneous flow variation is under 7%. However, the variability in instantaneous flow becomes high at low flow rates. It is not possible to draw a firm conclusion on the cause of  variation at the low flow rates. It could be attributed to sensor error, or it could be the result of actual grain flow variability caused by surging within the material handling system, or perhaps within the clean grain elevator. However, the variability is a moderate 7% at high flow rate of 21.1 kg/s where it is not as likely to experience flow surging. It is clear that the instantaneous flow is not as stable across the full range of flow rates as the accumulated mass. This is for obvious reasons, when assuming a normal distribution for all flow conditions, the instantaneous flow variation tends to cancel each other out and leave an accumulated mass which is a function of the mean flow rate. As seen in Table 2, the mean flow rates of the three devices are very similar. Another interesting observation is that instantaneous flow rate standard deviation seems to be smaller near the calibration points and increases at the extreme flow rates. Although the GreenStar® sensor appears to do a good job of predicting total accumulated mass flow at the extremes flow rates, the instantaneous flow rates may be less reliable further away from the calibration point.

The variation in instantaneous flow rates at the meter are minimal, as shown in Table 2. This is to be expected, since the turret rpm is fairly stable. However, the higher variation in instantaneous flow of the weigh scale is somewhat to be expected. The hysteresis of the load cells is 0.02% of the full scale load of 4536 kg, which is 0.9 kg of potential load variation. On a per second flow basis this can be significant at low flows. Another potential source of variation in the weigh scale may be vibration from the grain pump. Although slip joints were provided to isolate the weigh scale tank from the grain pump, there appears to be some vibration transmitted by friction in these joints. Another potential cause of variation may be momentum force loading from the falling grain stream. It was found that the use of a running average filter, helps minimize the influence of these dynamics. The data presented in Table 1 and Figures 6, 7, and 8 are un-filtered accumulated flow values. However, the data presented in Table 2 for the weigh scales are instantaneous one second flows filtered with a five second averaging filter.

However, it should be remember that most flow experienced by an actual combine yield monitor system will be in the range of 4.2 and 16.9 kg/s. Consequently, the main issue to consider is whether an instantaneous flow rate variation of approximately 7% is acceptable at calibration flow levels. More importantly, is the sensed flow variation caused by the mass flow technique or the yield monitor, or are these flow variations an artifact of the grain metering and transport components of the test facility.

 Summary and Conclusions

Performance results of the Yield Monitor Test Facility were positive, although a few limitation have been identified. First, it was found that the grain metering system can accurately meter grain at varying flows to the clean grain elevator with excellent accuracy over a prescribed period of time. A potential concern is whether or not the grain flow to the GreenStar® sensor is instantaneously uniform, or are there flow ripples due to the grain handling system. Secondly, the weigh scale system appears to be very accurate in measuring the accumulated mass flow in the system. However, the potential for grain stream dynamics, tank vibration and load cell hysteresis confound attempts to measure instantaneous grain flow at small flow rates ( 1 to 2 kg/s). Thirdly, the GreenStar® sensor demonstrated high accuracy across a range of static flow rates ranging from 1.3 to 21.1 kg/s. When considering accumulated mass flow, the percent difference between GreenStar® and the weigh scale were less then 1% at calibration flow rates and approximately 3% for extreme flow rates. However, the accuracy of GreenStar® instantaneous grain flow is somewhat suspect. Variations in instantaneous one second grain flows were observed to be approximately 8% at calibration flow rates.

Although the results from these test indicate that the GreenStar® mass flow sensor accurately predicts total accumulated mass flow across a broad range of static inflow rates, one must consider the possibility that other sources of error may exist in the overall combine yield monitor system. For instance; error in ground speed, measured swath width, transport delays in the threshing unit and other issues may still influence the overall accuracy of the yield map.

Acknowledgments

The authors would like to extend their grateful appreciation to Ed Hutchens, Carl King, Lee Rechtin, and Ed Roberts for there help fabricating, assembling the system components, and transporting the grain, to Lloyd Dunn and Burl Fannin for there help in wiring the controls and instrumentation, to John Deere Corporation, Hutchinson, and White Hydraulics for technical and equipment support and to Mike Ellis of the Worth and Dee Ellis Farms for providing grain for the test.

 References

Arslan, S. and T.S. Colvin. 1998. Laboratory Test Stand for Combine Grain Yield Monitors. Applied Engineering in Agriculture, Vol. 14(4):369-371

 Kettle, L.Y. and C.L. Peterson. 1998. An evaluation of Yield Monitors and GPS Systems on Hillside Combines Operating on the Steep Slopes in the Palouse. ASAE Paper No. 98-1046. ASAE, St. Joseph, Michigan.

 Kormann G., M. Demmel, H. Auernhammer. 1998. Testing Stand for Yield Measurement Systems in Combine Harvesters. ASAE Paper No. 98-3102. ASAE, St. Joseph, Michigan.

 Perez-Munoz, F. And T.S. Colvin. 1994. Continuous grain yield monitoring. ASAE Paper No. 94-1053. ASAE, St. Joseph, Michigan.

 Shearer, S. A., S. G. Higgins, S. G. McNeill, and G. A. Watkins. 1997. Data Filtering and Correction Techniques for Generating Yield Maps from Multiple -Combine Harvesting Systems. ASAE Paper No. 97-1034. ASAE, St. Joseph, Michigan.