A physics-informed deep learning framework for inversion and surrogate modeling in solid mechanics

DOI10.1016/j.cma.2021.113741zbMath1506.74476arXiv2003.02751OpenAlexW3137392741WikidataQ111462161 ScholiaQ111462161MaRDI QIDQ2021893

Ehsan Haghighat, Adrian Moure, Ruben Juanes, Maziar Raissi, Héctor J. Gómez

Publication date: 27 April 2021

Published in: Computer Methods in Applied Mechanics and Engineering (Search for Journal in Brave)

Full work available at URL: https://arxiv.org/abs/2003.02751

zbMATH Keywords

linear elasticityelastoplasticityinversionartificial neural networktransfer learningphysics-informed deep learning

Mathematics Subject Classification ID

Finite element methods applied to problems in solid mechanics (74S05) Finite element, Rayleigh-Ritz and Galerkin methods for boundary value problems involving PDEs (65N30) Isogeometric methods applied to problems in solid mechanics (74S22)

Related Items

Force density-informed neural network for prestress design of tensegrity structures with multiple self-stress modes ⋮ A general neural particle method for hydrodynamics modeling ⋮ Physics informed neural networks for continuum micromechanics ⋮ CAN-PINN: a fast physics-informed neural network based on coupled-automatic-numerical differentiation method ⋮ Physics-informed neural network simulation of multiphase poroelasticity using stress-split sequential training ⋮ Residual-based adaptivity for two-phase flow simulation in porous media using physics-informed neural networks ⋮ Physics-informed neural networks for inverse problems in supersonic flows ⋮ Scientific machine learning through physics-informed neural networks: where we are and what's next ⋮ Physics-informed PointNet: a deep learning solver for steady-state incompressible flows and thermal fields on multiple sets of irregular geometries ⋮ ReF-nets: physics-informed neural network for Reynolds equation of gas bearing ⋮ A physics-informed variational DeepONet for predicting crack path in quasi-brittle materials ⋮ A-PINN: auxiliary physics informed neural networks for forward and inverse problems of nonlinear integro-differential equations ⋮ A physics-informed neural network technique based on a modified loss function for computational 2D and 3D solid mechanics ⋮ A neural network-based approach for bending analysis of strain gradient nanoplates ⋮ Deep learning-accelerated computational framework based on physics informed neural network for the solution of linear elasticity ⋮ One-dimensional ice shelf hardness inversion: clustering behavior and collocation resampling in physics-informed neural networks ⋮ Integrated finite element neural network (I-FENN) for non-local continuum damage mechanics ⋮ Enhanced physics‐informed neural networks for hyperelasticity ⋮ Deep learning‐based reduced order models for the real‐time simulation of the nonlinear dynamics of microstructures ⋮ Thermodynamically consistent machine-learned internal state variable approach for data-driven modeling of path-dependent materials ⋮ Deep capsule encoder–decoder network for surrogate modeling and uncertainty quantification ⋮ A nonlocal energy‐informed neural network for isotropic elastic solids with cracks under thermomechanical loads ⋮ A stepwise physics‐informed neural network for solving large deformation problems of hypoelastic materials ⋮ StressD: 2D stress estimation using denoising diffusion model ⋮ On the use of graph neural networks and shape‐function‐based gradient computation in the deep energy method ⋮ BINN: a deep learning approach for computational mechanics problems based on boundary integral equations ⋮ Deep Ritz method with adaptive quadrature for linear elasticity ⋮ Physics-informed radial basis network (PIRBN): a local approximating neural network for solving nonlinear partial differential equations ⋮ Spiking recurrent neural networks for neuromorphic computing in nonlinear structural mechanics ⋮ A New Certified Hierarchical and Adaptive RB-ML-ROM Surrogate Model for Parametrized PDEs ⋮ Solving multi-material problems in solid mechanics using physics-informed neural networks based on domain decomposition technology ⋮ A class of mesh-free algorithms for some problems arising in finance and machine learning ⋮ Learning Markovian Homogenized Models in Viscoelasticity ⋮ Error estimates and physics informed augmentation of neural networks for thermally coupled incompressible Navier Stokes equations ⋮ CPINNs: a coupled physics-informed neural networks for the closed-loop geothermal system ⋮ Error convergence and engineering-guided hyperparameter search of PINNs: towards optimized I-FENN performance ⋮ HomPINNs: homotopy physics-informed neural networks for solving the inverse problems of nonlinear differential equations with multiple solutions ⋮ Variable linear transformation improved physics-informed neural networks to solve thin-layer flow problems ⋮ Adaptive task decomposition physics-informed neural networks ⋮ Adversarial deep energy method for solving saddle point problems involving dielectric elastomers ⋮ Isogeometric neural networks: a new deep learning approach for solving parameterized partial differential equations ⋮ Transfer learning based physics-informed neural networks for solving inverse problems in engineering structures under different loading scenarios ⋮ Physics-based self-learning spiking neural network enhanced time-integration scheme for computing viscoplastic structural finite element response ⋮ SeismicNET: physics-informed neural networks for seismic wave modeling in semi-infinite domain ⋮ Transfer learning of recurrent neural network‐based plasticity models ⋮ Physics-integrated neural differentiable (PiNDiff) model for composites manufacturing ⋮ Hybrid thermal modeling of additive manufacturing processes using physics-informed neural networks for temperature prediction and parameter identification ⋮ Improved training of physics-informed neural networks for parabolic differential equations with sharply perturbed initial conditions ⋮ Automatic boundary fitting framework of boundary dependent physics-informed neural network solving partial differential equation with complex boundary conditions ⋮ PINN-FORM: a new physics-informed neural network for reliability analysis with partial differential equation ⋮ Exact Dirichlet boundary physics-informed neural network EPINN for solid mechanics ⋮ Distributed PINN for Linear Elasticity — A Unified Approach for Smooth, Singular, Compressible and Incompressible Media ⋮ Adaptive transfer learning for PINN ⋮ Inverse modeling of nonisothermal multiphase poromechanics using physics-informed neural networks ⋮ A neural network-based enrichment of reproducing kernel approximation for modeling brittle fracture ⋮ AONN: An Adjoint-Oriented Neural Network Method for All-At-Once Solutions of Parametric Optimal Control Problems ⋮ Data-driven nonparametric identification of material behavior based on physics-informed neural network with full-field data ⋮ A deep learning model to predict the failure response of steel pipes under pitting corrosion ⋮ Physically recurrent neural networks for path-dependent heterogeneous materials: embedding constitutive models in a data-driven surrogate ⋮ Deep energy method in topology optimization applications ⋮ Learning constitutive models from microstructural simulations via a non-intrusive reduced basis method ⋮ Physics-informed neural network for modelling the thermochemical curing process of composite-tool systems during manufacture ⋮ A nonlocal physics-informed deep learning framework using the peridynamic differential operator ⋮ Latent map Gaussian processes for mixed variable metamodeling ⋮ Data-driven reduced order modeling of poroelasticity of heterogeneous media based on a discontinuous Galerkin approximation ⋮ Local approximate Gaussian process regression for data-driven constitutive models: development and comparison with neural networks ⋮ PhyCRNet: physics-informed convolutional-recurrent network for solving spatiotemporal PDEs ⋮ A representative volume element network (RVE-net) for accelerating RVE analysis, microscale material identification, and defect characterization ⋮ CENN: conservative energy method based on neural networks with subdomains for solving variational problems involving heterogeneous and complex geometries ⋮ A mixed formulation for physics-informed neural networks as a potential solver for engineering problems in heterogeneous domains: comparison with finite element method ⋮ Physics-informed neural networks for shell structures ⋮ Optimal control of PDEs using physics-informed neural networks ⋮ A robust unsupervised neural network framework for geometrically nonlinear analysis of inelastic truss structures ⋮ Nonlinear input feature reduction for data-based physical modeling

Uses Software

• COMSOL
• Theano
• SciPy
• TensorFlow
• Keras
• DiffSharp
• MXNet
• AdaGrad
• Adam
• SciANN
• F3DAM
• XPINNs
• PRMLT
• PINNsNTK

Cites Work

• Unnamed Item
• Unnamed Item
• Unnamed Item
• Unnamed Item
• Isogeometric analysis using T-splines
• Isogeometric analysis: CAD, finite elements, NURBS, exact geometry and mesh refinement
• Computational inelasticity
• The numerical solution of linear ordinary differential equations by feedforward neural networks
• SciANN: a Keras/Tensorflow wrapper for scientific computations and physics-informed deep learning using artificial neural networks
• Physics-constrained deep learning for high-dimensional surrogate modeling and uncertainty quantification without labeled data
• Physics-informed neural networks: a deep learning framework for solving forward and inverse problems involving nonlinear partial differential equations
• Machine Learning for Fluid Mechanics
• Isogeometric Analysis
• Solving high-dimensional partial differential equations using deep learning
• Extended Physics-Informed Neural Networks (XPINNs): A Generalized Space-Time Domain Decomposition Based Deep Learning Framework for Nonlinear Partial Differential Equations
• Learning data-driven discretizations for partial differential equations
• Data-Driven Identification of Parametric Partial Differential Equations
• Elasto‐plastic solutions of engineering problems ‘initial stress’, finite element approach