PARAMETER CALIBRATION FOR DISCRETE ELEMENT SIMULATION OF CUTTING DECK CLEANING IN SMALL PLOT WHEAT COMBINE HARVESTER
小区小麦联合收获机割台清理离散元仿真参数标定
DOI : https://doi.org/10.35633/inmateh-72-31
Authors
(*) Corresponding authors:
Abstract
The analysis of the clearing process of the cutting deck of a small plot wheat combine harvester requires the use of discrete element simulation methods. However, the current simulation test lacks the contact parameters such as wheat stalk and stalk-seed. In this paper, the wheat stalks and seeds at harvest time were taken as the research objects, and the calibration study of the discrete element simulation model parameters of stalks and stalk-seeds was carried out by means of mechanical test determination and EDEM software simulation. The stiffness coefficients of wheat stalks were determined by mechanical tests; the average values of wheat stalk stacking angle of 39.22° and wheat stalk-seed stacking angle of 44.41° were obtained by stacking angle tests. By the steepest climb test and binary regression test, the stalk normal stiffness coefficient was determined to be 5e+08N/m2 and tangential stiffness was determined to be 6.35e+08N/m2; the stalk-stalk collision recovery coefficient was obtained to be 0.551, static friction coefficient was obtained to be 0.797, and rolling friction coefficient was obtained to be 0.079 by the two-level analytical factorization test, the steepest climb test, and the three-factor response surface test. Based on this, the average value of wheat stalk-seed stacking angle was obtained to be 39.22° and the average value of wheat stalk-seed stacking angle was obtained to be 44.41° by the stacking angle test. On this basis, the coefficient of recovery of stalk-stalk collision was 0.434, the coefficient of static friction was 0.884, and the coefficient of rolling friction was 0.339 obtained by the three-factor response surface test. Three validation experiments were carried out by substituting the obtained parameters into the simulation test, and the error values were close to the error value %0.255 in the model, which proved that the experimental data were reliable.
Abstract in Chinese