Skip to main navigation Skip to search Skip to main content

Tensor-basis neural network modelling of solid stresses in granular material

  • Baptiste Hardy
  • , Sankaran Sundaresan
  • , Ali Ozel

Research output: Contribution to conference › Abstract › peer-review

96 Downloads (Pure)

Abstract

This study presents a novel approach for modelling the solid phase stress in granular flows using Galilean in- variant tensor basis neural networks (GiTBNN). The ground truth for NN training has been generated using Discrete Element Method (DEM) simulation results of dry and cohesive particles flowing over inclined planes. The solid granular pressure for dry particles predicted by NN model are in excellent agreement with the ground truth.
Original languageEnglish
Publication statusPublished - May 2025
Event12th International Conference on Multiphase Flows 2025 - Toulouse, France
Duration: 11 May 2025 → 16 May 2025
https://www.icmf2025.com

Conference

Conference12th International Conference on Multiphase Flows 2025
Country/TerritoryFrance
CityToulouse
Period11/05/25 → 16/05/25
Internet address

Fingerprint

Dive into the research topics of 'Tensor-basis neural network modelling of solid stresses in granular material'. Together they form a unique fingerprint.

Cite this