Search papers, labs, and topics across Lattice.
This paper addresses the computational challenges of solving the pressure-Poisson equation in incompressible flow solvers by introducing a data-driven algebraic multigrid (AMG) smoother that leverages a modified graph convolutional isomorphism network (GCIN). The GCIN predicts optimal polynomial coefficients for constructing a sparse pseudo-inverse operator, allowing the method to adapt to local anisotropies in unstructured grids while preserving linearity. The proposed approach achieves substantial performance improvements, reducing the number of V-cycles needed and providing wall-clock speedups of 4% to 37%, with robust generalization to significantly larger meshes and real-world datasets.
Achieving up to 37% faster convergence on complex flow problems by harnessing graph neural networks to optimize algebraic multigrid solvers.
Solving the pressure-Poisson equation remains the primary computational bottleneck in incompressible unstructured flow solvers primarily due to the inherent sensitivity of traditional linear solvers to mesh irregularities. This work introduces a data-driven algebraic multigrid (AMG) smoother that uses a modified graph convolutional isomorphism network (GCIN). The graph neural network predicts optimal polynomial coefficients to construct a sparse pseudo-inverse operator across diverse grid topologies. The coefficients are optimized to reduce the residual after each V-cycle iteration. By directly capturing the algebraic structure of the system from the sparse coefficient matrix, the proposed method maintains the solver's linearity while adapting to local anisotropies in unstructured grids. Our framework demonstrates significant performance gains by reducing the number of V-cycles required for a given tolerance and delivering wall-clock speedups from 4% to 37% across diverse benchmarks. Notably, the model exhibits robust generalization by maintaining efficiency on meshes up to 128 times larger than those seen in training, and by accelerating the solver's convergence on unseen industry-relevant problems such as the AirfRANS dataset.
