水稻制繁种联合收获脱出物离散元仿真参数标定与试验

Calibration and Experiment on DEM Parameters for Threshing Materials in Seed-production Rice Combine Harvesting

摘要:

良种是实现作物稳产增收的根本基础，制繁种水稻是保障国家粮食安全的重要支撑，实现低损、高净度的收获是保障其战略价值得以实现的必然要求。制繁种水稻较常规稻结实率低、瘪谷比例高，易导致清选环节的损失率与含杂率偏高，由于缺乏准确的脱出物离散元仿真模型参数，仿真结果与实际作业存在较大差异，限制了清选装置的设计与优化。采用离散元法对水稻制繁种联合收获脱出物的接触参数进行系统标定，利用质构仪和自制试验台分别测定各制繁种水稻脱出物的本征参数、接触参数与悬浮速度;以制繁种水稻混合脱出物物理堆积角为参照，开展了 Placket-Burman 试验、最陡爬坡试验和 Box-Behnken 试验，确定了对制繁种水稻混合脱出物堆积角影响显著的参数及取值范围;利用 Design-Expert 软件中的优化模块，以制繁种水稻混合脱出物物理堆积角为目标值进行寻优，确定了制繁种水稻混合脱出物最优接触参数组合。标定结果表明：显著性参数最优组合为籽粒 - 瘪谷静摩擦因数为 0.87、瘪谷 - 瘪谷碰撞恢复系数为 0.45、瘪谷 - 瘪谷滚动摩擦因数为 0.037。在接触参数最优组合条件下，开展了堆积角和悬浮速度仿真验证试验。试验结果表明：制繁种水稻混合脱出物仿真堆积角与物理堆积角误差为 0.57%; 水稻籽粒、茎秆、瘪谷的悬浮速度仿真值与物理值的相对误差分别为 1.37%、1.95%、3.73%。在标定参数组合下，基于 CFD-DEM 开展了水稻制繁种联合收获清选装置气固耦合仿真试验，并进行了台架验证试验。仿真结果表明：清选环节籽粒损失率为 0.17%、含杂率为 0.58%, 与台架试验相对误差分别为 5.56% 和 4.92%, 验证了标定参数的准确性，可为水稻制繁种联合收获离散元仿真研究提供基础参数。

Abstract:

High-quality seeds are fundamental to achieving stable crop yields, and seed-production rice plays a critical role in ensuring national food security. Compared with conventional rice, seed-production rice is characterized by a lower seed-setting rate and a higher proportion of unfilled grains, which often leads to increased grain loss and impurity rates during combine harvesting. However, the lack of accurate discrete element method (DEM) parameters for threshed materials results in discrepancies between simulation predictions and actual harvesting performance, thereby restricting the effectiveness of simulation-based cleaning device design and optimization. To address this issue, the contact parameters of threshed materials from seed-production rice combine harvesting were systematically calibrated by using DEM. The intrinsic parameters, contact parameters, and suspension velocity of filled grains, straw segments, and unfilled grains were experimentally determined. Taking the physical repose angle of mixed threshed materials as the calibration response indicator, Plackett - Burman, steepest ascent, and Box - Behnken experimental designs were successively employed to screen and optimize the significant contact parameters. The optimal parameter combination was obtained by using the optimization module of DesignExpert software. The optimal parameter set included a static friction coefficient of 0.87 between filled and unfilled grains, a restitution coefficient of 0.45 between unfilled grains, and a rolling friction coefficient of 0.037 between unfilled grains. The relative error between simulated and physical repose angles was 0.57%, while the relative errors of simulated suspension velocities for filled grains, straw segments, and unfilled grains were 1.37%, 1.95%, and 3.73%, respectively. Based on the calibrated parameters, CFD - DEM coupled simulations of the cleaning process in a seed-production rice combine harvester were conducted and validated by bench tests. The simulated grain loss rate and impurity rate were 0.17% and 0.58%, respectively, with relative errors of 5.56% and 4.92% compared with experimental results. These findings confirmed the accuracy and applicability of the calibrated parameters and provided reliable fundamental parameters for CFD - DEM coupled simulation and optimization of seed-production rice combine harvesting.