Abstract:Reliable evaluations of geotechnical hazards like landslides and debris flow require accurate simulation of granular flow dynamics. Traditional numerical methods can simulate the complex behaviors of such flows that involve solid-like to fluid-like transitions, but they are computationally intractable when simulating large-scale systems. Surrogate models based on statistical or machine learning methods are a viable alternative, but they are typically empirical and rely on a confined set of parameters in evaluating associated risks. Due to their permutation-dependent learning, conventional machine learning models require an unreasonably large amount of training data for building generalizable surrogate models. We employ a graph neural network (GNN), a novel deep learning technique, to develop a GNN-based simulator (GNS) for granular flows to address these issues. Graphs represent the state of granular flows and interactions, like the exchange of energy and momentum between grains, and GNN learns the local interaction law. GNS takes the current state of the granular flow and estimates the next state using Euler explicit integration. We train GNS on a limited set of granular flow trajectories and evaluate its performance in a three-dimensional granular column collapse domain. GNS successfully reproduces the overall behaviors of column collapses with various aspect ratios that were not encountered during training. The computation speed of GNS outperforms high-fidelity numerical simulators by 300 times.

What problem does this paper attempt to address?

The problems that this paper attempts to solve are as follows: In large - scale granular flow simulations, although traditional numerical methods (such as the discrete element method (DEM) and the material point method (MPM)) can accurately simulate the complex behaviors of granular flows, the computational cost is too high, making it difficult to conduct large - scale system simulations. In addition, surrogate models based on statistical or traditional machine - learning methods usually rely on empirical parameters and have limited generalization ability, and cannot directly consider the dynamic physical mechanisms of granular flows. Specifically, the paper aims to: 1. **Improve computational efficiency**: By developing a graph neural network (GNN) - based learning simulator (GNS), the computational time can be significantly reduced, enabling it to quickly evaluate large - scale granular flow dynamics. 2. **Enhance model generalization ability**: Ensure that the model can not only perform well within the training data range, but also accurately predict granular flow dynamics under unseen boundary conditions and geometries. 3. **Consider physical mechanisms**: Through GNN, learn the local interaction rules between particles, so as to better capture complex behaviors in granular flows such as nonlinearity, statics, collisions, and frictional dissipation. To solve these problems, the authors propose a GNN - based simulator (GNS), which uses a graph structure to represent the state of granular flows and learns the interactions between particles through a message - passing mechanism. The experimental results show that GNS exhibits good performance in three - dimensional granular column collapse experiments, can successfully predict granular flow dynamics under different initial aspect ratios, and is 300 times faster than high - fidelity numerical simulators. In summary, the main objective of this paper is to develop an efficient and generalized granular flow simulation tool to overcome the computational bottlenecks of traditional methods in large - scale simulations and provide a broader scenario evaluation ability.

Three-dimensional granular flow simulation using graph neural network-based learned simulator

Graph Neural Network-based surrogate model for granular flows

Inverse analysis of granular flows using differentiable graph neural network simulator

A Physical-Information-flow-constrained Temporal Graph Neural Network-Based Simulator for Granular Materials

Integrating Graph Neural Network-Based Surrogate Modeling with Inverse Design for Granular Flows

GNS: A generalizable Graph Neural Network-based simulator for particulate and fluid modeling

Learning 3D Granular Flow Simulations

Subspace graph networks for real-time granular flow simulation with applications to machine-terrain interactions

Simulating multiphase flow in fractured media with graph neural networks

Graph Convolutional Networks for Simulating Multi-phase Flow and Transport in Porous Media

Accelerating Particle and Fluid Simulations with Differentiable Graph Networks for Solving Forward and Inverse Problems

Subspace Graph Physics: Real-Time Rigid Body-Driven Granular Flow Simulation

Neural Graph Simulator for Complex Systems

Fluid Simulation System Based on Graph Neural Network

A Multigrid Graph U-Net Framework for Simulating Multiphase Flow in Heterogeneous Porous Media

A Gradient-based Deep Neural Network Model for Simulating Multiphase Flow in Porous Media

Graph Neural Networks (gnns) Based Accelerated Numerical Simulation

Three-dimensional Modeling of Granular Flow Impact on Rigid and Deformable Structures

Three-dimensional Numerical Simulation of Granular Flow with a GPU-accelerated SPH Model

Multi-Hierarchical Surrogate Learning for Structural Dynamical Crash Simulations Using Graph Convolutional Neural Networks

Learning to Simulate Crowd Trajectories with Graph Networks.