Calibration tests for the contact parameters of flour particle using discrete element method

Accurate engineering modeling of flour is often required for the calibration of its contact parameters in flour conveying systems. In this study, a typical calibration of contact model parameters was proposed to consider the effects of cohesion between flour particles and particle size, according to the static and dynamic angle of repose. Hertz-Mindlin with Johnson-Kendall-Roberts (JKR) contact model was also selected using the particle scaling theory. The orthogonal simulation tests were carried out to implement the sensitivity and variance analysis of the contact parameters using the discrete element method (DEM) model. The result showed that there were the most significant effects of JKR surface energy, the powder-powder coefficient of rolling friction, and the powder-steel coefficient of static friction on the static angle of repose. Since the contact parameters were calibrated only by static angle of repose, the cross-effects of multiple parameters were verified as well. Different combinations of contact parameters produced a similar effect of static angle of repose. Furthermore, five sets of contact parameters were selected to obtain a unique set of parameter values as the candidates. Their simulation from the static angle of repose was the closest related to the test. Then the dynamic stacking was further studied. The flour was filled in a cylindrical container with a transparent bottom and stainless-steel side. In the tests, the cylindrical container was set as 100 mm in diameter and 30 mm in height and then rotated around the axis at the given speed. The camera was used to record the dynamic angle of repose, which was defined as the maximum tilt angle between the slope and the horizontal plane before the powder collapsed. The contact parameters calibrated by the dynamic angle of repose were also verified under the condition of a single filling rate and a single rotational speed. Therefore, the dynamic angle of repose was further calibrated at the filling rates of 0.2 and 0.4, respectively, while the rotational speeds of 30, 40, 50, and 60 r/min. Optimal contact parameters were achieved for the flour with a powder-powder coefficient of restitution of 0.6, a powder-powder coefficient of static friction of 0.2, a powder-powder coefficient of rolling friction of 0.1, a powder-steel coefficient of restitution of 0.6, a powder-steel coefficient of static friction of 0.6, a powder-steel coefficient of rolling friction of 0.5 and a JKR surface energy of 0.12 J/m². The dynamic angle of repose after DEM calibration was the closest to the test. The calibration with the dynamic angle of repose under various conditions effectively reduced the uncertainty of the calibration. After that, a rectangular container test was designed to verify the accuracy of the contact parameters. Among them, the flour was poured into a cube container with a side length of 100 mm, where the top surface of the assembly of flour was kept flat, and then the cabin lid of the container was lifted away, leading to the powder naturally flowing out of the container. The ratio of the flour mass remained to calculate the total mass in the container. There was a minimal relative error between the numerical simulation with the calibrated parameters and the measurement from the test. The DEM contact parameters of flour were improved in the reliability, simplicity, and feasibility of calibration. The finding can also provide a strong reference for the DEM parameters calibration of powdery materials in engineering applications.