Konstantinos Krestenitis
A Contact Detection Code using Triangles for Non-Spherical Particle Simulations
Krestenitis, Konstantinos; Weinzierl, Tobias; Koziara, Tomasz
Abstract
We present a novel DEM (discrete element method) code with explicit time stepping. DEM codes simulate billions of small particles that interact with each other primarily through collisions. Different to state-of-the-art codes, we rely on triangulated non-spherical particles. This is computationally demanding, and we thus devise an asynchronous data exchange communication technique using MPI (Message Passing Interface) on manycore supercomputers, we discuss possible solutions to handle ghost particles that overlap multiple subdomains. At the compute node level, shared memory parallelism as well as vectorised SIMD executions are studied. On the algorithmic side, we explore a hybrid parallelisation approach and memory layouts that are suited to combine robust exact geometry checks with a fast penalty-based method.
Citation
Krestenitis, K., Weinzierl, T., & Koziara, T. (2016). A Contact Detection Code using Triangles for Non-Spherical Particle Simulations. In Proceedings of the 24th Conference on Computational Mechanics (ACME-2016): 31 March - 01 April 2016, Cardiff University, Cardiff (227-230)
Conference Name | 24th Conference on Computational Mechanics (ACME-UK 2016) |
---|---|
Conference Location | Cardiff, UK |
Start Date | Mar 31, 2023 |
End Date | Apr 1, 2016 |
Publication Date | Apr 1, 2016 |
Deposit Date | Sep 4, 2017 |
Publicly Available Date | Sep 5, 2017 |
Publisher | Cardiff University |
Pages | 227-230 |
Book Title | Proceedings of the 24th Conference on Computational Mechanics (ACME-2016): 31 March - 01 April 2016, Cardiff University, Cardiff. |
Publisher URL | https://acme2016.sciencesconf.org/resource/page/id/25 |
Files
Accepted Conference Proceeding
(395 Kb)
PDF
You might also like
Calculating the minimum distance between triangles on SIMD Hardware
(2015)
Conference Proceeding
Fast DEM collision checks on multicore nodes
(2018)
Conference Proceeding
Upscaling ExaHyPE – on each and every core
(2023)
Report