Discrete Element Simulation Modeling Method and Parameters Calibration of Sugarcane Leaves

Abstract

Objective The construction method of the discrete element model and the setting of simulation parameters in the strip- and blade-shaped sugarcane leaf are unclear. The simulation model’s accuracy greatly influences the dynamic response characteristics between particles, and it is necessary to improve the accuracy of simulation parameters through parameter calibration. Method The discrete element parameters are optimized and calibrated based on the response surface methodology (RSM) with sugarcane leaf physical angle of repose as the response value. Firstly, the basic physical parameters and angle of repose of sugarcane leaves were measured by physical tests, and the simulation model of sugarcane leaf was established by the multi-sphere polymerization model and XML method. The effects of the sugarcane leaf model filled with different radii particles on the simulation angle of repose and simulation efficiency were analyzed to find the optimal filling particle size of the sugarcane leaf model. Then, a Plackett-Burman test was used to select the parameters that significantly influence the simulation angle of repose. Furthermore, the optimal value ranges of the three significant parameters were determined by a steepest ascent search test, and the second-order regression equation between the significant parameters and angle of repose was established based on the Box-Behnken test, the optimal combination of parameters was obtained with the physical angle of repose of 21.15° as the optimal target value. Finally, a gas-solid coupling simulation test was conducted with the trash content as the test index and compared with the field test. Result The optimal filling particle size of the sugarcane leaf simulation model was 2 mm. The optimal combination of significant parameters was as follows: the static and rolling friction coefficients between sugarcane leaves were 0.21 and 0.05, respectively, and the static friction coefficient between sugarcane leaves and steel was 0.30. There was no significant difference between the simulation value and the test value of trash content, and the maximum relative error between them was 8%, which further showed that the parameter calibration of the sugarcane leaf model was reliable. Conclusions The results showed that the modeling method and parameter calibration of the sugarcane leaf model was accurate and reliable and could be used for subsequent gas-solid coupling simulation research, as well as providing a reference for the calibration of the discrete element parameters of the strip-and blade-shape materials.