Construction of a Soil Clods Recognition Bench-Scale Experiment for Discrete Element Method Modeling of Tilling Phenomena

The development of rotary claws in tiller machines has been evaluated using actual equipment and analysis based on the discrete element method (DEM), which is an effective method for modeling the movement of granular materials. The required functions of the rotary claw are to break the soil into small pieces and plow the soil flat. However, evaluation by testing using actual equipment is difficult for precise measurements, and a considerable amount of time is required for one test. Additionally, DEM simulations have difficulty reproducing the soil behavior owing to the large number of parameters that determine their accuracy. Therefore, a comparison with the experimental results is necessary when determining DEM parameters. This research aims to develop a soil clod recognition method and an evaluation system for evaluating tractor performance. The evaluation system was designed to measure the soil shape before and after tilling with a 1/4-scale tillage claw. To capture the changes in the soil shape and distribution and size of the soil clods, we developed an image processing method. This method uses pointcloud data obtained from a depth camera to calculate the difference before and after tilling. The experimental parameters were the soil moisture content and claw rotation speed. Their effects on the formation and decomposition of the soil clods were evaluated. To confirm the feasibility of the results obtained from the proposed evaluation system, DEM simulations were performed under identical conditions to compare the distribution and size of the soil clods and changes in soil shape. The experimental results showed that the location, number, and size of scattered soil clods varied with the moisture content of the soil, and rotational speed of the claw. Based on the experimental results, a comparison was made with the results of the DEM analysis to clarify the differences between the two. The developed model experiment system for soil clods and the image processing method made it possible to quantitatively compare soil dispersal between the experiment and the DEM, which may accelerate the search for parameters for the DEM to reproduce the tilling.