制种玉米离散元柔性模型构建与试验

Development and Experimental Validation of Flexible Discrete Element Model for Seed Maize Ears

摘要:

针对制种玉米剥皮过程籽粒损失率和破损率较高及建模精度不足等问题，本文基于离散元法构建制种玉米果穗柔性复合体模型。通过物理试验测定制种玉米果穗各部分本征参数及接触参数，基于Hertz-Mindlin with Bonding V2接触模型构建苞叶-籽粒-穗芯柔性复合体离散元模型，采用Plackett-Burman试验、最陡爬坡试验和Box-Behnken响应面试验相结合的方法对玉米芯轴和籽粒等关键粘结参数进行标定优化，建立了能够准确表征剥皮过程籽粒脱落和破损行为的离散元模型。结果表明，玉米芯轴三点弯曲仿真试验与物理试验最大载荷相对误差为3.89%，籽粒压缩仿真试验与物理试验相对误差为8.79%，籽粒脱落试验与物理试验轴向和切向相对误差为7.58%、6.89%，参数标定精度较高。在制种玉米果穗剥皮仿真试验验证中籽粒脱落率和破损率为1.774%、0.457%，台架试验籽粒脱落率和破损率分别1.697%和0.432%，相对误差为4.54%、5.79%，均满足制种玉米剥皮质量标准，表明模型能够较真实反映剥皮过程的籽粒脱落与破损特征，为制种玉米果穗精细化建模和剥皮装置优化设计提供参考依据。

Abstract:

Aiming to address high kernel loss and breakage rates during seed maize husking and insufficient modeling precision, a flexible composite discrete element method (DEM) model for seed maize ears was developed. The intrinsic parameters and contact parameters of different components of seed maize ears were determined through physical experiments. A flexible composite DEM model comprising bract-kernel-cob was constructed based on the Hertz-Mindlin with Bonding V2 contact model. The combination of Plackett-Burman design, steepest ascent method, and Box-Behnken response surface methodology was employed to calibrate and optimize the key bonding parameters of cob shaft and kernels, establishing a DEM model capable of accurately characterizing kernel detachment and breakage during husking. The simulation results showed high calibration accuracy, with relative errors of 3.89% for cob shaft three-point bending, 8.79% for kernel compression, and 7.58% and 6.89% for kernel detachment in axial and tangential directions respectively compared with physical experiments. In husking simulation validation, kernel detachment and breakage rates were 1.774% and 0.457% respectively, compared with bench test results of 1.697% and 0.432%, with relative errors of 4.54% and 5.79%. All results met seed maize husking quality standards, confirming the model realistically reflects kernel detachment and breakage characteristics during husking. The research result can provide reliable reference for refined modeling of seed maize ears and optimization design of husking devices.