Draft force reduction by an electrification unit on implements

In all field trials and construction of the electrified slurry grubber, we followed the work of Wiecha et al. (2019b). The slurry-injecting prototype was designed with a traction roller and a continuous traction element. The traction elements were star packers. Because of its weight, we constructed the roller near the mounting of the implement. A set of seven furrow grubbers with slurry-injecting application unit was mounted to the following frame before recompacting the soil with a closing unit.

In accordance with Lindner et al. (2011), a generator mounted to the front power take-off (front p.t.o.) of the tractor was used for generating a voltage of 650 V. An inverter regulated this voltage to the needs of the ZF electric motor, which had the technical specifications of a maximum of 400 V, 230 A, 125 kW, and up to 10300 rpm with a gear ratio of 35.84. A radar sensor manufactured by DICKEY corporation model 4282089M2 measured the speed over ground. We calculated the rotation speed of the traction roller at every moment of driving from the values taken from the radar sensor. System control was done by the steering software on a laptop. The software package was ControlDesk 6.2 by dSPACE-embedded systems. The rotation speed was set to 20% slippage on the traction roller through ControlDesk software. The draft force sensor was manufactured with draft force cell “HBM U1” by Ingenieurbüro Rottmaier in Erding, Germany and had a draft force limitation to 200 kN. It had a working load of 240 kN and a compound error of <±0.03% from the final value (reference temperature was 23°C). The logging speed was set to 10 values per second for the draft force sensor and the radar sensor.

Figure 1 shows the setup of one of the field trials.

Figure 1

In Figure 1, from right to left, we see Fendt 818 (old type) with an additional front ballast of 1.2 t for better pulling. Afterward, the draft force unit was tied together with loops to the second Fendt Vario 818 (newer type). On the second Fendt to the front PTO, we placed the electric generator. The second tractor carried electrified slurry grubber.

Under very dry conditions between March and May 2019, three fields with heavy soil types were identified for receiving clear data in the draft force trials. The size of the fields near Freising in Germany ranged from 3 to 12 ha. With the soil types D = soil on Pleistocene sediments, L = loam, and SL = sandy loam, the fields are located under the following global positioning system coordinates:

– Field 1: 48°25′45.5″N 11°40′16.2″E; L4D, 59/52 (land value)

On every field, the area of plane ground was used for driving. Drive speed was set to 8 km h⁻¹ in all field trials in this study. In the data, we used a single driving lane with and without electrification near one another in the same direction of every field. Data logging started on reaching a speed of 8 km h⁻¹ and stopped at the moment the speed was reduced. We drove the whole lane with this speed until the end of the field. Turns and stops were not included in the results.

Speeding up between 1 and 8 km h⁻¹ was also discarded as changes of slope. The data were tracked in LabView software 2016 using an NI-USB 6001. All seven furrow grubbers were adjusted to a working depth of 22.5 cm. The depth was manually checked by hand before starting to drive. The slippage on the traction roller was set to 20%.

The final slurry grubber was designed to meet the case of cultivating maize in 75-cm row spaces. On the traction roller, a rotating coulter was mounted for cutting the upper soil layer in front of each single grubber furrow. Now, steering software was implemented in the ISOBUS terminal. An identical slippage of 20% on the traction roller was adjusted by using the terminal. In contrary to the prototype, two electric motors from NIDEC type PEV50-8-D011, one on each side of the implement, were used. With a voltage of 540 V DC, the motor was able to reach a peak power of 85 kW, a maximum torque of 320 N m, and a rotation speed of up to 6000 rpm. Figure 2 shows the setup on the field with the final implement. As described before, the sensor manufactured by DICKEY corporation model 4282089M2 and the draft force unit of Ingenieurbüro Rottmair (Erding, Germany) were used during the tests with the final implement.

Figure 2

The pulling tractor was a Claas Axion 870 with mounted draft force measuring unit, followed by the Fendt 818 Vario with front PTO generator and the mounted implement with 6-m working width (Figure 2). The fields with their soil types were located under the following coordinates:

– Field 4: 48°25′26.1″N 11°40′00.3″E; SL4D, 50/46

Conditions on the fields between February and April 2021 were humid in the early spring. The decision for using the fields and the driving area within the fields was made by the factor of driveability. On the fields 4–6, identical preceding crop and intercrop was ensured. As in the case of fields 1–3 from prototype tests, all fields were untilled before. Logging the data was done by LabView 2016 and the NI-USB 6001 box. Frequency of logging and description of cutting the data in the point of driving speed were also identical to the prototype test in 2019.

For the field tests in 2019 and 2021, we used the Welch’s t-test performed in R (R Core Team 2014). This test is used to find the means of two independent data groups. In our case, one group of data represents the draft force values without electrification and the second group of data represents the values with electrification.

From the six fields, we received totally N = 6521 draft force values. In Table 1 are shown the values per field, with and without electrification. The number of values per field depends on the length of the driving lanes during the field trials. With the final implement, the space requirement of turning around the field with two tractors tied together is bigger. Due to this, the n-values of these trials are lesser.

Figure 3 shows that in all three cases, the draft force reduction between the left boxplot with no electrification and the right boxplot with electrification is significant. Fields 1 and 2 show a higher variation of values under electrification, which was not expected. With an electrification assistance, a smoother run of the implement was anticipated. The next step is to evaluate the significance by Welch’s two-sample t-test using R software for each field.

Figure 3

Table 2 shows the output of every t-test done separately for every field. In all three cases, the p-value is lower than 0.05, which shows significance. Obviously, there is no explanation for the broader variance under electrification. One reason could be that the test vehicle naturally has additional sensors installed on the traction roller and the traction roller is, therefore, more flexibly mounted in order to deliver values during measuring. The measured values from the traction roller itself are not included in this investigation, as the traction force savings of the overall tractor–implement setup have to be shown. Table 3 presents the draft force consumption of the first prototype.

As seen in Figure 4, in the case of field 4 and field 6, the draft force reduction is clearly measurable. Regarding the middle of Figure 4, field 5 seems to be different. The received values of field 5 are close to each other and show a draft force reduction of nearly 5%. Looking at the field itself, the soil type is less heavy and dries faster during spring because of strong wind blowing in a largely free and exposed location. After laying down the heavy implement on the soil surface, the soil seemed to smear under the pressure of the rotating roller. Due to a long period of humidity, no better conditions could be used for this field during the tests. Again, two-sample t-test is used for verifying the data. The results are presented in Table 4.

Figure 4

On comparing Tables 3 and 5, it is observed that the received draft force reduction is smaller with the final implement than with the early prototype. This may be due to several reasons. First, the area of traction symbolized by the contact surface between soil and the roller is relatively big. The final implement with its 6-m working width has to pull eight grubber furrows (the prototype with its 3-m working width carried seven grubber furrows). Therefore, the draft force consumption by the final implement could be smaller.

Second, the traction elements on the prototype roller were designed to have a good formfit with the soil for perfect transfer of the traction force. In contrast, the coulter element on the electrified roller of the final implement was designed to cut into the soil for making an easier way for the following grubber furrows. The aim was to reduce the draft force not only through traction, but also through opening the soil before.

Third reason could be the amount of metal penetrating the soil and the amount of metal laying on top of the soil. Elements of flat metal rotating on the soil surface do not transfer traction. Only the traction elements that go down into the soil transport the soil to the rear. In case of the final implement, only a few centimeters of metal are intruding the soil. Therefore, traction support could not be as good as on the traction roller of the first prototype.