[P1-78] Analysis of Friction Characteristics of Steel Powders using Parallelized Discrete Element Method
In order to increase the application of sintered metal parts, high dense and less pores parts have to be produced with stable quality. Since sintered metal parts cannot be fully densified during sintering, it is essential to understand the steel powders' characteristics in detail and to densify them in the compaction step. Sufficient powder characteristic data, however, cannot be evaluated by conventional versatile methods, and there are few cases where actual powder characteristics are applied to large scale particle simulation. In this study, we evaluated flowability and friction characteristics of steel powders experimentally in detail, and applied them to parallelized discrete element method (DEM).
We prepared steel powders with different kinds of solid lubricant and addition process. Their characteristics were evaluated by a method using a powder yield locus (PYL). The evaluation indexes are flowability (uniaxial collapse stress), inter-particle friction (internal friction angle), and friction between particles and mold (wall friction angle).
The PYL method, which can measure flowabilities under various vertical loads, enabled collection of powder characteristics corresponding to the actual powder behavior in the mold. Furthermore, inter-particle and wall friction coefficients were calculated using the friction angles obtained from these experiments. These coefficients were higher than those which were applied in previous DEM simulation reports.
We created a parallelized DEM code for powder behavior simulation, using FDPS (Framework for Developing Particle Simulator) developed by Iwasawa et al.[1] in order to develop into large scale parallelization in the future. FDPS is designed to reduce the calculation load at the time of region division and interaction calculation by constructing a tree structure. In this code, we evaluated flowabilities and friction characteristics by changing the model of the contact force between particles, especially the tangential component.
In the conference presentation, we will show the flowabilities and the friction characteristics of steel powders, which are evaluated by this parallelized DEM code.
[1] Iwasawa et al., Publ. Astron. Soc. Japan, 68, 54 (2016)
We prepared steel powders with different kinds of solid lubricant and addition process. Their characteristics were evaluated by a method using a powder yield locus (PYL). The evaluation indexes are flowability (uniaxial collapse stress), inter-particle friction (internal friction angle), and friction between particles and mold (wall friction angle).
The PYL method, which can measure flowabilities under various vertical loads, enabled collection of powder characteristics corresponding to the actual powder behavior in the mold. Furthermore, inter-particle and wall friction coefficients were calculated using the friction angles obtained from these experiments. These coefficients were higher than those which were applied in previous DEM simulation reports.
We created a parallelized DEM code for powder behavior simulation, using FDPS (Framework for Developing Particle Simulator) developed by Iwasawa et al.[1] in order to develop into large scale parallelization in the future. FDPS is designed to reduce the calculation load at the time of region division and interaction calculation by constructing a tree structure. In this code, we evaluated flowabilities and friction characteristics by changing the model of the contact force between particles, especially the tangential component.
In the conference presentation, we will show the flowabilities and the friction characteristics of steel powders, which are evaluated by this parallelized DEM code.
[1] Iwasawa et al., Publ. Astron. Soc. Japan, 68, 54 (2016)