Umeå University's logo

umu.sePublications
Change search
CiteExportLink to record
Permanent link

Direct link
Cite
Citation style
  • apa
  • ieee
  • modern-language-association-8th-edition
  • vancouver
  • Other style
More styles
Language
  • de-DE
  • en-GB
  • en-US
  • fi-FI
  • nn-NO
  • nn-NB
  • sv-SE
  • Other locale
More languages
Output format
  • html
  • text
  • asciidoc
  • rtf
Towards a graph neural network solver for granular dynamics
Umeå University, Faculty of Science and Technology, Department of Physics. (Digital Physics)ORCID iD: 0000-0002-0787-4988
Umeå University, Faculty of Science and Technology, Department of Physics.
Algoryx Simulation AB, Umeå, Sweden.
Umeå University, Faculty of Science and Technology, Department of Physics. (Digital Physics)
2021 (English)In: VII International Conference on Particle-based Methods (PARTICLES 2021): Technical program, Hamburg, Germany, 2021Conference paper, Oral presentation with published abstract (Other academic)
Abstract [en]

The discrete element method (DEM) is a versatile but computationally intensive method for granular dynamics simulation. We investigate the possibility of accelerating DEM simulations using graph neural networks (GNN), which automatically support variable connectivity between particles. This approach was recently found promising for particle-based simulation of complex fluids [1]. We start from a time-implicit, or nonsmooth, DEM [2], where the computational bottleneck is the process of solving a mixed linear complementarity problem (MLCP) to obtain the contact forces and particle velocity update. This solve step is substituted by a GNN, trained to predict the MLCP solution. Following [1], we employ an encoder-process-decoder structure for the GNN. The particle and connectivity data is encoded in an input graph with particle mass, external force, and previous velocity as node attributes, and contact overlap, normal, and tangent vectors as edge attributes. The sought solution is represented in the output graph with the updated particle velocities as node attributes and the contact forces as edge attributes. In the intermediate processing step, the input graph is converted to a latent graph, which is then advanced with a fixed number of message passing steps involving a multilayer perceptron neural network for updating the edge and node values. The output graph, with the approximate solution to the MLCP, is finally computed by decoding the last processed latent graph.

Both a supervised and unsupervised method are tested for training the network on granular simulation of particles in a rotating or static drum. AGX Dynamics [3] is used for running the simulations, and Pytorch [4] in combination with the Deep Graph Library [5] for the learning. The supervised model learns from ground truth MLCP solutions, computed using a projected Gauss-Seidel (PGS) solver, sampled from 1200 simulations involving 50-150 particles. The unsupervised model learns to minimize a loss function derived from the MLCP residual function using particle configurations extracted from the same simulations but ignoring the approximate solution from the PGS solver. The simulation samples are split into training data (80%), validation data (10%), and test data (10%). Network hyperparameter optimization is performed. The supervised GNN solver reaches an error level of 1% for the contact forces and 0.01% on the particle velocities for a static drum. For a rotating drum, the respective errors are 10% and 1%. The unsupervised GNN solver reaches 1% velocity errors, 5% normal forces errors, but it has significant problems with predicting the friction forces. The latter is presumably because of the discontinuous loss function that follows from the Coulomb friction law and therefore we explore regularization of it. Finally, we discuss the potential scalability and performance for large particle systems.

Place, publisher, year, edition, pages
Hamburg, Germany, 2021.
National Category
Physical Sciences Computer Sciences Fluid Mechanics and Acoustics
Research subject
Physics; Computer Science
Identifiers
URN: urn:nbn:se:umu:diva-187950OAI: oai:DiVA.org:umu-187950DiVA, id: diva2:1597695
Conference
PARTICLES 2021: VIIth International Conference on Particle-Based Methods, Hamburg, Germany / Hybrid, online, October 4-5, 2021
Funder
Mistra - The Swedish Foundation for Strategic Environmental Research, 2017/14 #6Available from: 2021-09-27 Created: 2021-09-27 Last updated: 2021-11-29Bibliographically approved

Open Access in DiVA

No full text in DiVA

Other links

URL

Authority records

Servin, MartinWallin, Erik

Search in DiVA

By author/editor
Servin, MartinWallin, Erik
By organisation
Department of Physics
Physical SciencesComputer SciencesFluid Mechanics and Acoustics

Search outside of DiVA

GoogleGoogle Scholar

urn-nbn

Altmetric score

urn-nbn
Total: 326 hits
CiteExportLink to record
Permanent link

Direct link
Cite
Citation style
  • apa
  • ieee
  • modern-language-association-8th-edition
  • vancouver
  • Other style
More styles
Language
  • de-DE
  • en-GB
  • en-US
  • fi-FI
  • nn-NO
  • nn-NB
  • sv-SE
  • Other locale
More languages
Output format
  • html
  • text
  • asciidoc
  • rtf