Momentum-Conserving Graph Neural Networks for Deformable Objects
Provides a physically consistent GNN for simulating deformable materials, addressing a key limitation of existing models for physics simulations.
MomentumGNN introduces a graph neural network architecture that predicts per-edge impulses to guarantee linear and angular momentum conservation, outperforming baselines in deformable object dynamics.
Graph neural networks (GNNs) have emerged as a versatile and efficient option for modeling the dynamic behavior of deformable materials. While GNNs generalize readily to arbitrary shapes, mesh topologies, and material parameters, existing architectures struggle to correctly predict the temporal evolution of key physical quantities such as linear and angular momentum. In this work, we propose MomentumGNN -- a novel architecture designed to accurately track momentum by construction. Unlike existing GNNs that output unconstrained nodal accelerations, our model predicts per-edge stretching and bending impulses which guarantee the preservation of linear and angular momentum. We train our network in an unsupervised fashion using a physics-based loss, and we show that our method outperforms baselines in a number of common scenarios where momentum plays a pivotal role.