水稻机械直播的土壤表层泥浆离散元参数标定

Calibration of discrete element parameters for soil surface mud in mechanical direct seeding of rice

【目的】针对水稻机械直播稻田播前土壤与触土部件的相互作用机理不明确等问题,可利用离散元分析法研究影响水稻精量穴直播质量的因素。由于播前土壤的结构复杂,其表层泥浆与耕层存在一定差异,因此,需分层构建复合土壤模型。为获取精确的水稻机械直播的表层泥浆离散元参数,开展试验研究和仿真分析,并标定表层泥浆的离散元仿真参数。【方法】按照水稻精量穴直播的整地要求处理,并在播种前去田间获取表层泥浆,采用漏斗法进行泥浆堆积角试验,利用数显倾角仪多次测量泥浆堆积角并取平均值。选择EDEM离散元软件中Hertz-Mindlin with JKR接触模型,开展泥浆堆积角仿真试验。以堆积角为响应值,通过Plackett-Burman试验筛选出对堆积角影响显著的3个参数,进一步开展最陡爬坡试验缩小显著性参数的取值范围。采用Box-Behnken试验建立表层泥浆堆积角与筛选的显著性参数的回归模型,并以物理试验测得的泥浆堆积角为目标值,对显著性参数寻优得到最佳参数组合。将最优参数代入仿真软件验证表层泥浆颗粒离散元参数的准确性。【结果】泥浆堆积角物理试验获取表层泥浆堆积角为40.20°。Plackett-Burman试验结果表明,泥浆剪切模量、泥浆-泥浆表面能和泥浆-钢表面能对表层泥浆堆积角有显著性影响,其余试验因素对堆积角的影响均不显著。最陡爬坡试验将显著性参数的最优区间缩小为:泥浆剪切模量1.6~4.6 MPa、泥浆-泥浆表面能0.25~0.35 J/m^(2)、泥浆-钢表面能0.02~0.04 J/m^(2)。通过堆积角回归模型,获取最优解为:泥浆剪切模量1.839 MPa、泥浆-泥浆表面能0.25 J/m^(2)、泥浆-钢表面能0.029 J/m^(2)。将最优参数仿真软件验证的结果为39.10°,与泥浆堆积角物理试验结果的相对误差为2.74%。【结论】基于Hertz-Mindlin with JKR接触模型标定的水稻机械直播的表层泥浆颗粒离散元参数准确可靠,可为构建水稻机械直播稻田土壤的离散元模型提供理论依据。

[Objective]In order to solve the problem that the interaction effect between the mechanical direct seeder and the preceding seeding soil was not very clear,the discrete element method was applied to study the factors affecting the quality of precision rice hill-drop drilling.Due to the complex structure of the soil before sowing and a certain difference between the surface mud and the plough layer,it is necessary to construct a composite soil model in layers.For obtaining accurate discrete element parameters of surface mud for mechanical direct seeding,the experiments and simulations analysis were conducted,the discrete element model of the soil surface mud was built by calibrating the key parameters.[Method]The tillage requirement of field was the same as the precision rice hill-drop drilling,the surface mud was collected before sowing,the mud accumulation angle test was conducted by funnel method,The mud accumulation angle was measured many times by digital display inclinometer and the average value was calculated.Hertz-Mindlin with JKR contact model in the EDEM discrete element software was selected to simulate the soil surface mud.Taking the accumulation angle as the response value,the Plackett-Burman experiments were applied to screen out the three parameters which had a significant impact on the accumulation angle,and the steepest ascent experiments were conducted to narrow down the range of three significant parameters.The regression equation of the accumulation angle was established with the selected significant parameters by using Box-Behnken experiments,Taking the mud accumulation angle measured by physical test as the target value,the optimal parameter combination was obtained by optimizing the significant parameters.Then the optimal parameters were substituted into the EDEM to verify the accuracy of discrete element parameters of surface mud particles.[Result]The physical test of mud accumulation angle showed that the surface mud accumulation angle was 40.20°,and Plackett-Burman experimental results showed that the mud shear modulus,the mud-mud surface energy and the mud-steel surface energy had significant effects on the surface mud accumulation angle,the other test factors had no significant effect on it.The steepest ascent experimental results showed that the optimal range of significant parameters were as follows:the mud shear modulus was 1.6-4.6 MPa,the mud-mud surface energy was 0.25-0.35 J/m^(2),the mud-mud surface energy was 0.02-0.04 J/m^(2).Through the regression model of accumulation angle,the optimal parameters were obtained:the mud shear modulus 1.839 MPa,the mud-mud surface energy 0.25 J/m^(2),and the mud-steel surface energy 0.029 J/m^(2).The results verified by the optimal parameter simulation software were 39.10°,and the relative error with the physical test results of the mud accumulation angle was 2.74%.[Conclusion]The above tests showed that the discrete element parameters of surface mud particles calibrated by the Hertz-Mindlin with JKR contact model for mechanical direct seeding of rice were accurate and reliable.This study showed that the surface mud of preceding seeding soil was accuracy,and the results would be used for building the compound discrete element model of the mechanical direct seeding of rice.