mlsanityData to 14 Oct 2026

TRACE: A spatiotemporal contact memory graph network simulator for granular dynamics

arXiv 2609.02991 ↗First posted 2 Sep 2026cs.LG cs.NA math.NAcited by 0

Not on the map. Why?

Abstract (arXiv)

Learned graph simulators provide an efficient alternative to high-fidelity solvers for granular dynamics. However, granular motion depends strongly on inter-granular contact history, which is difficult to preserve when particle contacts form, break, and rearrange. Existing simulators mainly store temporal information in node features or node-level memory. Here we introduce TRACE, a graph-network simulator that stores interaction history directly on contact edges. Each edge maintains a persistent memory updated by attention-based message passing and a gated recurrent unit, while an edge-identity dictionary preserves this memory as the contact graph changes. A physics-structured decoder predicts inter-granular normal and tangential contact forces, enforces the Coulomb friction limit, and applies equal-and-opposite internal forces. The model is trained with single-step pretraining followed by autoregressive rollout fine-tuning. We evaluate TRACE on 2D and 3D granular column-collapse benchmarks. In both cases, TRACE produces stable, physically consistent long-horizon rollouts, closely reproducing the final deposit geometry and the kinetic energy released during collapse. Compared with graph network simulator (GNS) and node-memory graph neural simulator (NMGNS), TRACE reduces long-rollout position error by 31-62% and final-deposit error by 58-89% across the two benchmarks, while using fewer parameters and maintaining near-zero particle interpenetration. TRACE also achieves 12.2× and 8.9× speedups over the material point method (MPM) reference solver in 2D and 3D, respectively. Our code is available at https://github.com/Data-Driven-Computational-Geotechnics/TRACE.

Key facts. No citing papers in the map yet (data through 14 Oct 2026). Builds on 43 of its 83 references that are in the map. First posted on arXiv 2 Sep 2026.

Builds on

43 of 83 references.

References · 43

  1. 2017Attention Is All You Needcited by 60,361
  2. 2017Graph Attention Networkscited by 8,923
  3. 2014Learning Phrase Representations using RNN Encoder-Decoder for Statistical Machine Translationcited by 4,057
  4. 2018Relational inductive biases, deep learning, and graph networkscited by 1,360
  5. 2021E(n) Equivariant Graph Neural Networkscited by 965
  6. 2020Learning to Simulate Complex Physics with Graph Networkscited by 880
  7. 2020Learning Mesh-Based Simulation with Graph Networkscited by 868
  8. 2015Scheduled Sampling for Sequence Prediction with Recurrent Neural Networkscited by 803
  9. 2019Hamiltonian Neural Networkscited by 712
  10. 2020Temporal Graph Networks for Deep Learning on Dynamic Graphscited by 530
  11. 2016Interaction Networks for Learning about Objects, Relations and Physicscited by 509
  12. 2020Inductive Representation Learning on Temporal Graphscited by 462
31 more
  1. 2019Predicting Dynamic Embedding Trajectory in Temporal Interaction Networkscited by 448
  2. 2020Lagrangian Neural Networkscited by 398
  3. 2022Message Passing Neural PDE Solverscited by 386
  4. 2019Representation Learning for Dynamic Graphs: A Surveycited by 265
  5. 2018Graph networks as learnable physics engines for inference and controlcited by 244
  6. 2015Structural-RNN: Deep Learning on Spatio-Temporal Graphscited by 234
  7. 2021Geometric and Physical Quantities Improve E(3) Equivariant Message Passingcited by 228
  8. 2018Learning Particle Dynamics for Manipulating Rigid Bodies, Deformable Objects, and Fluidscited by 220
  9. 2020Solver-in-the-Loop: Learning from Differentiable Physics to Interact with Iterative PDE-Solverscited by 205
  10. 2022Equivariant Graph Mechanics Networks with Constraintscited by 68
  11. 2024AdaptiGraph: Material-Adaptive Graph-Based Neural Dynamics for Robotic Manipulationcited by 54
  12. 2024GIC: Gaussian-Informed Continuum for Physical Property Identification and Simulationcited by 49
  13. 2022Learning rigid dynamics with face interaction graph networkscited by 43
  14. 2022Learning Physical Dynamics with Subequivariant Graph Neural Networkscited by 39
  15. 2024Neural Material Adaptor for Visual Grounding of Intrinsic Dynamicscited by 32
  16. 2022Guaranteed Conservation of Momentum for Learning Particle-based Fluid Dynamicscited by 30
  17. 2021Constraint-based graph network simulatorcited by 27
  18. 2021Boundary Graph Neural Networks for 3D Simulationscited by 24
  19. 2023Dynamic-Resolution Model Learning for Object Pile Manipulationcited by 22
  20. 2023Graph Neural Network-based surrogate model for granular flowscited by 16
  21. 2022Learning Large-scale Subsurface Simulations with a Hybrid Graph Network Simulatorcited by 15
  22. 2022GNS: A generalizable Graph Neural Network-based simulator for particulate and fluid modelingcited by 12
  23. 2021Manipulation of Granular Materials by Learning Particle Interactionscited by 7
  24. 2023Accelerating Particle and Fluid Simulations with Differentiable Graph Networks for Solving Forward and Inverse Problemscited by 6
  25. 2024Inverse analysis of granular flows using differentiable graph neural network simulatorcited by 5
  26. 2021Learning 3D Granular Flow Simulationscited by 3
  27. 2025Sim2Dust: Mastering Dynamic Waypoint Tracking on Granular Mediacited by 3
  28. 2026WorldParticle: Unified World Simulation of Lagrangian Particle Dynamics via Transformercited by 3
  29. 2025Parameter-Efficient Conditioning for Material Generalization in Graph-Based Simulatorscited by 2
  30. 2023Three-dimensional granular flow simulation using graph neural network-based learned simulatorcited by 1
  31. 2025From images to properties: a NeRF-driven framework for granular material parameter inversioncited by 1

Built on by

Not cited yet.