Parameter calibration and testing for discrete element simulation of walnut kernels

作者：Jianian Li, W. Zhang, Dianbo Li, Yingjie Yu · 发表于：Computational Particle Mechanics · 年份：2025 · DOI：10.1007/s40571-025-00995-0 · 被引用次数：5 · 研究领域：Soil Mechanics and Vehicle Dynamics、Agricultural Engineering and Mechanization、Mechanical and Thermal Properties Analysis

To determine the discrete element simulation parameters for the walnut kernel grading process, the stacking angle response of walnut kernels obtained from bench tests and simulation tests was used. The discrete element simulation parameters were calibrated using response surface optimization. The physical parameters required for the simulation test of the walnut kernel grading process were measured through bench testing. Discrete element simulation models of walnut kernels with varying degrees of completeness were established using 3D scanning technology and EDEM software. The walnut stacking angle was measured as 23.55° using the injection method, combined with MATLAB image processing for boundary fitting. The Plackett–Burman test, the steepest ascent test, and the Box–Behnken test were designed using Design–Expert software, with the stacking angle as the response value, to construct the regression model and optimize the parameters. The optimal combination of significant parameters is as follows: walnut kernel–walnut kernel rolling friction coefficient of 0.03, walnut kernel–walnut kernel static friction coefficient of 0.136, and walnut kernel–walnut kernel collision recovery coefficient of 0.26. The simulated and measured values of the natural angle of repose for walnut kernels under the optimal parameter combinations were compared. A two-sample t test ( p > 0.05) confirmed that there was no significant difference between the two, verifying the reliability of the simulation...