Computational Study of the Industrial Synthesis of Tungsten Powders

Discrete Element Method (DEM) is a highly employed Lagrangian technique to represent particulate systems. When DEM techniques are extended by adding thermochemical conversion of solid particles as well as their interaction with the surrounding fluid, numerous challenging applications can be numerically studied. Nevertheless, industrial applications with large number of particles, such as powder synthesis or blast furnaces, are often time or size limited due to the high computational efforts that these simulations demand. This contribution introduces the Agglomerated Particle Method (APM) as a numerical technique aiming to reduce the computational costs of coupled Discrete Element Method and Computational Fluid Dynamics (DEM-CFD) approaches for the thermochemical conversion of powder beds.

From experimental and numerical investigation on thermochemical conversion of packed beds has been observed that the temperature or composition of particles in a small spatial domain do not vary significantly. Consequently, one single numerical solution may be representative for all the particles on such a domain. Thus, a collection of neighbor particles are represented by one single agglomerated particle solved by eXtended Discrete Element Method (XDEM) techniques.

The proposed model is firstly assessed with classic benchmark problems for heating and drying of packed beds. Later, the model approach is employed for predicting the industrial synthesis of metallic tungsten powder.

The comparison of APM predictions with resolved XDEM predictions and experimental data shows the proposed model as a viable technique to solve large scale powder applications, such as tungsten powder production, at feasible time.