Abstract
The application of discrete element modelling (DEM) to cohesive fine powders in industrial processes, such as additive manufacturing, requires accurate and efficient calculations of van der Waals interaction forces. In DEM community, it is a general practice to reduce particle stiffness to accelerate the simulations; however, this study shows that, for cohesive particles, there are many cases where previously proposed scaling methodologies fail to preserve the original particle behaviour. The reason was attributed to underestimated sliding and rolling resistances and a poorly resolved non-contact cohesive interaction, thus limiting the applicability of these scaling approaches for contact-dominated systems. To address these significant issues, a new stiffness scaling methodology is proposed for the modelling of cohesive fine powders, which includes an established scaling law for contact adhesion, modified sliding and rolling resistances, and a new force-estimation scheme for the calculation of non-contact van der Waals interaction. The new approach was verified with a series of simple cases; stiffness independent results were demonstrated for head-on particle-particle collisions, particle-wall collisions, and particle-agglomerate collisions. The predictions of stop distance of a particle sliding and/or rolling over a flat surface was preserved when the stiffness was scaled down almost four orders of magnitude, which was not possible with previous scaling approaches. The new approach was further validated by packing of cohesive fine particles. This work confirmed that not only was the packing density insensitive to the particle stiffness, but the details of the packing structure (coordination number and packing density distribution) were also maintained when the original particle stiffness was scaled down by three orders of magnitude. Finally, the applicability of the new approach was explored by simulations of homogeneous simple shearing, which was found to be controlled by system cohesiveness and inertial number. (C) 2020 The Author(s). Published by Elsevier Inc.