On the Reduction of Computational Costs for Tungsten Powder Bed Processes

During the Discrete Element Method (DEM) representation of powder bed processes (e.g. tungsten oxide reduction, tungsten carbide synthesis, selective laser sintering) a numerical solution for each single particle is impractical due to the extremely high number of particles (e.g. 10^12). However, in such processes, particles in the vicinity of each other observe low gradients concerning their thermodynamic state. This characteristic can be exploited to avoid solving repeatedly numerically equivalent equation systems.
This contribution presents two numerical methods aiming to reduce the computational costs of DEM approaches for the thermochemical conversion of powder beds. In the two methods after an appropriated numerical treatment, a group of particles under similar boundary conditions is substituted by a single-effective-entity. Consequently, the entire powder space is divided into sub-domains to be solved. The methods result in considerable lower number of equations that increase computational efficiency and enable feasible time simulations. The applications of the industrial synthesis of tungsten powders and the selective laser sintering (SLS) of powder metals are presented and discussed.