Fast 3D diffusion for scalable discrete granular media synthesis
Abstract
Discrete Element Method (DEM) simulations of granular media are computationally intensive, particularly during initialization phases dominated by large displacements and kinetic energy. This paper presents a novel generative pipeline based on 3D diffusion models that directly synthesizes arbitrarily large granular assemblies in DEM-compatible initial configurations. The approach employs a two-stage pipeline. First, an unconditional diffusion model generates independent 3D voxel grids representing granular media; second, a 3D inpainting model, adapted from 2D techniques using masked inputs and repainting strategies, seamlessly stitches these grids together. The inpainting model uses the outputs of the unconditional diffusion model to learn from the context of adjacent generations and creates new regions that blend smoothly into the context region. Both models are trained on binarized 3D occupancy grids derived from a database of small-scale DEM simulations, scaling linearly with the number of output voxels. DEM initialization cases that spanned over days can now be prepared in hours, practically enabling simulations containing more than 200k ballast particles. The pipeline remains fully compatible with existing DEM workflows as it post-processes the diffusion generated voxel grids into DEM compatible particle meshes. Being consistent with the original DEM simulations on selected geometrical and contact-network descriptors, the pipeline is also compatible with many other applications in the field of granular media. Showcased on two examples (railway ballast and lunar regolith), the pipeline reimagines the way initialization of granular media simulations is performed, enabling scales of generation previously unattainable with traditional DEM simulations.
Related
- Strongly non-linear continuous model for wave propagation in 2D granular media: derivation and hyperbolicity analysis
- Long-Time Large Deviation Asymptotics for Mean-Field Granular Media Equations
- Scalable persistence pairing graphs on scalar fields with an application to porous media
- Scalable Discrete-to-Continuous Channel Simulation for Compression and Privacy
- 1D Convection-Diffusion in Porous Media: A Unified Solution from Bounded Domain to Half-Space
BibTeX
@article{hassan2026fast,
title = {{Fast 3D diffusion for scalable discrete granular media synthesis}},
author = {Muhammad Moeeze Hassan and Régis Cottereau and Filippo Gatti and Patryk Dec},
journal = {Computers and Geotechnics},
year = {2026},
month = jun,
doi = {10.1016/j.compgeo.2026.108336},
eprint = {2508.19752},
archivePrefix = {arXiv},
url = {https://hal.science/hal-05652831v1},
note = {CC BY-NC-ND 4.0},
}
RIS
TY - JOUR
TI - Fast 3D diffusion for scalable discrete granular media synthesis
AU - Muhammad Moeeze Hassan
AU - Régis Cottereau
AU - Filippo Gatti
AU - Patryk Dec
PY - 2026
DA - 2026/06/10
JO - Computers and Geotechnics
DO - 10.1016/j.compgeo.2026.108336
UR - https://hal.science/hal-05652831v1
AB - Discrete Element Method (DEM) simulations of granular media are computationally intensive, particularly during initialization phases dominated by large displacements and kinetic energy. This paper presents a novel generative pipeline based on 3D diffusion models that directly synthesizes arbitrarily large granular assemblies in DEM-compatible initial configurations. The approach employs a two-stage pipeline. First, an unconditional diffusion model generates independent 3D voxel grids representing granular media; second, a 3D inpainting model, adapted from 2D techniques using masked inputs and repainting strategies, seamlessly stitches these grids together. The inpainting model uses the outputs of the unconditional diffusion model to learn from the context of adjacent generations and creates new regions that blend smoothly into the context region. Both models are trained on binarized 3D occupancy grids derived from a database of small-scale DEM simulations, scaling linearly with the number of output voxels. DEM initialization cases that spanned over days can now be prepared in hours, practically enabling simulations containing more than 200k ballast particles. The pipeline remains fully compatible with existing DEM workflows as it post-processes the diffusion generated voxel grids into DEM compatible particle meshes. Being consistent with the original DEM simulations on selected geometrical and contact-network descriptors, the pipeline is also compatible with many other applications in the field of granular media. Showcased on two examples (railway ballast and lunar regolith), the pipeline reimagines the way initialization of granular media simulations is performed, enabling scales of generation previously unattainable with traditional DEM simulations.
SN - 0266-352X
SN - 1873-7633
ER -