Calibration of discrete element model parameters for a multicomponent rice seedling substrate using response surface methodology and experimental validation

Seedling substrate is essential in industrialized rice seedling cultivation, with its physical properties critically affecting tray-filling machinery performance. This study developed a calibrated discrete element method (DEM) model for a representative substrate formulation consisting of peat, coconut coir, and vermiculite in a 6:3:1 ratio to enable accurate simulation of equipment-substrate interactions. Through systematic laboratory testing and numerical calibration, key contact parameters including static and rolling friction coefficients were determined using the angle of repose as the calibration criterion. Significant parameters were identified through Plackett-Burman design and optimized via steepest ascent and Box–Behnken experimental methodologies. The results showed that the optimal parameter combination, comprising a static friction coefficient of 0.459 between peat and coconut coir, 0.525 between peat and vermiculite, and a rolling friction coefficient of 0.409 between coconut coir and vermiculite, produced a simulated angle of repose of 38.70°. This value demonstrated only 2.95% deviation from experimental measurements. Additional validation through uniaxial compression simulations revealed a 2.9% relative error in maximum compressive displacement compared to physical tests. The calibrated DEM model accurately represents the substrate’s mechanical behavior, providing a reliable foundation for optimizing tray-filling equipment design.