Extrusion process of corn stalk powder in single orifice die processing based on discrete element method

Abstract: Mechanical behavior in the densification of biomass material is closely related to pellet quality. In order to explore the forming mechanism of typical biomass material from loose state to consolidation, the discrete element method (DEM) was introduced to investigate the movement and interaction of the milled corn stalk particles in the compacting process, and the verification experiments were carried out to test the effectiveness of the DEM simulation in this study. Firstly, the three-dimensional (3D) particle contact model of corn stalk powder based on the soft-sphere model of DEM was established, and the constraining walls in DEM model were completely consistent with the compressing cavity boundary conditions in geometric shape and dimension of experimental tests conducted in December, 2014; the loading speed in simulation was also set as the same value as the DEM model. Secondly, the diameter range of simulated particles was configured to 0.4-1.0 mm in accordance to the particle size distribution acquired through the screening experiment and calculation, and the generated particles were fully filled into the whole cavity at the original state before the compressing force was loaded. The mechanical parameters of the particles, such as normal stiffness, shear stiffness and friction coefficient between the 2 contact particles, were set to the values generated at random in specific range which was determined according to compacting experimental data. Thirdly, the comparison of compression stress relaxation data between tests and simulation was carried out and the validity of the simulation was verified by the hypothesis test. It was found that the force data with time from the hypothesis tests and DEM simulation followed the similar tendency, and the absolute error was not higher than 100 N in both initial loading stage and 20 seconds after stress relaxation. In the first 20 seconds of stress relaxation course, the values of absolute error were obviously higher than other time quantum. The consistency of the experimental and simulated data was fairly good on the whole, because there was little statistical significance between 2 group of data at 5% level in the significant difference analysis. The optimal numerical ranges of the mechanical parameters of the simulated particles in DEM model were also obtained. Namely, the normal stiffness was 1.2×104-1.8×104 N/m, the shear stiffness was 0.8×104-1.3×104 N/m and the friction coefficient was 0.10-0.12. Then, the compressive force was analyzed in DEM model at different compressive displacements, diameters and cone angles using the optimal mechanical parameters of particles. The result showed that the residual forces in stress relaxation were about 600 and 1 300 N respectively when the compressive displacements were set to 26 and 50 mm, which indicated that the compressive displacement had a great influence on pellet morphological stability while other parameters kept constant. When the diameter of single-hole die varied from 8 to 12 mm in DEM model, all the compression forces peaked near 1 100 N as compressing time went on, but the residual stress with 8 mm diameter was much higher than that with 12 mm, and in consideration of the consolidation degree, the recommended diameter was 8 mm compared with the mechanical behaviors of the diameter of 10 and 12 mm. The cone angle had a remarkable effect on the compression force, and the cone angle of 45° was suggested to get a reasonable balance between compression force and pellet density. The study indicates that the discrete element method provides an efficient and effective tool to address some engineering problems in biomass densification, and the soft-sphere model is appropriate to describe the mechanical behavior in the compression process of corn stalk powder.