A revisit of common normal method for discrete modelling of non-spherical particles

Kamyar Kildashti, Kejun Dong, Bijan Samali

Research output: Contribution to journalArticlepeer-review

17 Citations (Scopus)

Abstract

Discrete element method (DEM) is prominent for studying granular materials at particle scale. However, how to model non-spherical particles in DEM is still challenging. In light of the present contact detection algorithms in the literature, common normal (CN) and geometric potential (GP) are two methods used for particles with smooth surfaces. Yet it has been long believed that CN gives erroneous results while GP is more preferable since they were firstly proposed for ellipsoidal particles decades ago. A revisit of CN in this work identifies two problems in the original CN, and then a new CN is proposed which can overcome these problems. Based on the comparison to sub-particle scale finite element analyses, the new CN has been further shown to be able to predict the contact plane more accurately than the original CN and GP. Such an advantage is found for the modelling of ellipsoidal and superquadric particles. The study not only proposes an improved CN algorithm but also demonstrates that CN should receive more attentions in DEM, though GP is now much more widely used.
Original languageEnglish
Pages (from-to)1-6
Number of pages6
JournalPowder Technology
Volume326
DOIs
Publication statusPublished - 2018

Keywords

  • computer simulation
  • discrete element method
  • granular materials

Fingerprint

Dive into the research topics of 'A revisit of common normal method for discrete modelling of non-spherical particles'. Together they form a unique fingerprint.

Cite this