Enhancing <scp>DEM</scp> Prediction of Maize Kernel Breakage: A Polyhedral Particle Based Calibration Framework

ABSTRACT As a major global food crop, maize kernel hardness significantly influences breakage during harvesting and post‐production processing. However, accurately predicting the breakage behavior of maize kernels remains challenging due to multiple influencing factors, such as kernel shape, moisture content, and variety. Based on the Tavares breakage theory within the discrete element method (DEM), this work combined experimental and simulation approaches to calibrate the parameters of a maize kernel breakage model. A polyhedral discrete element model of maize kernels was established, and the particle size distribution data of fragments were obtained through uniaxial compression experiment. By fitting the governing equations of the Tavares breakage model to experimental data, calibrated model parameters were determined as follows: E ∞ , d 0 , φ , A , and were 181.3, 8.22, 10, 67.7, and 0.033, respectively. Finally, the internal consistency of the calibrated parameters was validated by comparing re‐simulation and experimental results across several metrics: cumulative fragment size distribution, mass‐based proportion of fragments on each screen, force–displacement curves, and critical fracture force. The results showed that when the particle size distribution follows the incomplete beta function (IBF) law, the simulation results agree well with experimental values. Furthermore, this work demonstrated that the calibration parameters of the maize breakage model were reliable and could accurately predict and analyze the maize crushing process.