Stratified DEM model construction and parameter optimization based on the cornstalk ontological traits

To establish an optimal discrete element model and simulation parameters that accurately represent the biomechanical characteristics of cornstalk nodes during harvest, comprehensive shear and compression tests were conducted to quantify the mechanical properties of node components. These experimental results served as evaluation metrics for calibrating the bonding parameters. A triple-bonded particle model was constructed utilizing particles with radii of 1.47 mm, 1 mm, and 0.75 mm to replicate the complex microstructure of the node. Statistical optimization techniques, including Plackett-Burman design, steepest ascent methodology, and response surface test, were employed to identify the significant factors influencing the interfacial bonding parameters between node pith-node pith, node pith-stem pith, and node pith-stem bark, thereby refining the parametric range and determining optimal bonding parameters. The optimized bonding parameters were subsequently validated through bending tests and a stem-pulling roller-like test of the intact cornstalk. Analysis of destructive forces and morphological deformations of the cornstalk with nodes demonstrates that the developed DEM model and optimized bonding parameters exhibit high fidelity in replicating the mechanical behavior of the actual cornstalk. The triple-bonded particle model and calibrated DEM parameters provide a robust theoretical foundation for developing a comprehensive corn plant-scale DEM model for future investigations of microscale interactions between the header and corn plants during mechanical harvesting processes.