JAX-FEM: a differentiable GPU-accelerated 3D finite element solver for automatic inverse design and mechanistic data science

DOI10.1016/j.cpc.2023.108802zbMath1526.74068arXiv2212.00964MaRDI QIDQ6112664

Shuheng Liao, Wing Kam Liu, Tianju Xue, Chanwook Park, Jian Cao, Xiaoyu Xie, Zhengtao Gan

Publication date: 7 August 2023

Published in: Computer Physics Communications (Search for Journal in Brave)

Full work available at URL: https://arxiv.org/abs/2212.00964

zbMATH Keywords

neural networkautomatic differentiationtopology optimizationdata-driven multi-scale computationmachine learning librarytensile loading problem

Mathematics Subject Classification ID

Artificial neural networks and deep learning (68T07) Learning and adaptive systems in artificial intelligence (68T05) Finite element methods applied to problems in solid mechanics (74S05) Inverse problems in equilibrium solid mechanics (74G75) Topological methods for optimization problems in solid mechanics (74P15)

Related Items (2)

Isogeometric convolution hierarchical deep-learning neural network: isogeometric analysis with versatile adaptivity ⋮ Convolution hierarchical deep-learning neural network (C-HiDeNN) with graphics processing unit (GPU) acceleration

Cites Work

• Unnamed Item
• Unnamed Item
• Unnamed Item
• Automated solution of differential equations by the finite element method. The FEniCS book
• An incremental approach to the solution of snapping and buckling problems
• \textsc{tIGAr}: automating isogeometric analysis with \textsc{FEniCS}
• Adjoint optimization of pressurized membrane structures using automatic differentiation tools
• A kernel method for learning constitutive relation in data-driven computational elasticity
• Mechanistically informed data-driven modeling of cyclic plasticity via artificial neural networks
• Learning viscoelasticity models from indirect data using deep neural networks
• Non-commutative discretize-then-optimize algorithms for elliptic PDE-constrained optimal control problems
• On eigenvalues of matrices dependent on a parameter
• Review of options for structural design sensitivity analysis. I: Linear systems
• Inverse problems: A Bayesian perspective
• Optimal Solvers for PDE-Constrained Optimization
• Evaluating Derivatives
• deal.II—A general-purpose object-oriented finite element library
• The method of moving asymptotes—a new method for structural optimization
• Bi-CGSTAB: A Fast and Smoothly Converging Variant of Bi-CG for the Solution of Nonsymmetric Linear Systems
• Adjoint Sensitivity Analysis for Differential-Algebraic Equations: The Adjoint DAE System and Its Numerical Solution
• A Limited Memory Algorithm for Bound Constrained Optimization
• Computational Contact Mechanics
• JAX-fluids: a fully-differentiable high-order computational fluid dynamics solver for compressible two-phase flows

This page was built for publication: JAX-FEM: a differentiable GPU-accelerated 3D finite element solver for automatic inverse design and mechanistic data science