Simulation and test on performance of soil-cassava jitter separation device of cassava harvester

Abstract: In order to obtain the soil-cassava separating mechanism and the effect laws of different factors on soil-cassava separating performance of spring type soil-cassava jitter separation device when the cassava tubers were lifted by cassava harvester, as well as the optimal values of factors, dynamics simulation model of soil-cassava jitter separation device system and regression mathematical model of separating performance of the device were established by using coupling method of FEM (finite element method) and SPH (smoothed particle hydrodynamics) and quadratic regression general rotation design method. The soil-cassava separating performance of spring type soil-cassava jitter separation device was studied. The effects of the length of long hole, spring stiffness and spring pre-tightening force on the soil-cassava separating cleanliness and the collision force between hanging scroll of spring below and long hole of sleeve above were discovered, respectively. Then, the influence factors of soil-cassava separating performance were optimized by using the fmincon function in the optimization toolbox of MATLAB software. Thus, the optimal combination of factors was obtained and the optimization result was verified by cassava tuber lifting experiment in the field. When the separation device was used in the field experiment, the average quality of the tuber-soil mixture was 5.835 kg, and the average quality of the residual soil on the tubers was 0.794 kg. When the separation device was not used in the field experiment, the average quality of the tuber-soil mixture was 7.998 kg, and the average quality of the residual soil on the tubers was 2.627 kg. The average clean degree in field experiment was 0.698. The results show that the regression mathematical models of soil-cassava separation performance are significant at 0.002 level. The regression coefficients are significant at the level from 0.001 to 0.028. The regression mathematical models are highly significant, which can be used in the study of soil-cassava separation performance of spring type soil-cassava jitter separation device. When the separation of the soil and the tuber-soil mixture nearly occurs, the spring tensile force is greater than the sum of the force binding tubers with soil and the gravity of the tuber-soil mixture, and the spring is contracted. The hanging scroll of spring below and the lower end of the long hole of sleeve above are separated. The cassava tubers are driven by the contraction of the spring, which results in dithering of soil and softening of the tuber-soil mixture. After the separation of the soil and the tuber-soil mixture, the spring rebounds under the function of the spring restoring force, and the contact collision occurs between the hanging scroll of the spring below and the top of the long hole of sleeve above, which results in the vibration of the tuber-soil mixture. The soils of the tuber-soil mixture become looser. Finally, the soils fall to ground, and the soil and the tubers are separated. With the increasing of the length of the long hole, cleaning degree and collision force are changing along a parabola pattern with a downward opening. With the increasing of spring stiffness, cleaning degree increases. With the increasing of spring stiffness, collision force decreases. With the increasing of spring pre-tightening force, cleaning degree and collision force increase. The optimal combination of factors is as follows: The length of the long hole is 2.68 cm, the spring stiffness is 20.04 kN/cm and the spring pre-tightening force is 335.2 N. And the corresponding cleaning degree and the collision force are 0.778 and 320 N, respectively. The relative error of cleanliness is about 10% compared with the results of field experiment. Optimization result of the factors is reasonable, which can be the design basis of the spring type soil-cassava jitter separation device.