Machine learning materials physics: multi-resolution neural networks learn the free energy and nonlinear elastic response of evolving microstructures
From MaRDI portal
Publication:2020954
Abstract: Many important multi-component crystalline solids undergo mechanochemical spinodal decomposition: a phase transformation in which the compositional redistribution is coupled with structural changes of the crystal, resulting in dynamically evolving microstructures. The ability to rapidly compute the macroscopic behavior based on these detailed microstructures is of paramount importance for accelerating material discovery and design. Here, our focus is on the macroscopic, nonlinear elastic response of materials harboring microstructure. Because of the diversity of microstructural patterns that can form, there is interest in taking a purely computational approach to predicting their macroscopic response. However, the evaluation of macroscopic, nonlinear elastic properties purely based on direct numerical simulations (DNS) is computationally very expensive, and hence impractical for material design when a large number of microstructures need to be tested. A further complexity of a hierarchical nature arises if the elastic free energy and its variation with strain is a small-scale fluctuation on the dominant trajectory of the total free energy driven by microstructural dynamics. To address these challenges, we present a data-driven approach, which combines advanced neural network (NN) models with DNS to predict the homogenized, macroscopic, mechanical free energy and stress fields arising in a family of multi-component crystalline solids that develop microstructure. The hierarchical structure of the free energy's evolution induces a multi-resolution character to the machine learning paradigm: We construct knowledge-based neural networks (KBNNs) with either pre-trained fully connected deep neural networks (DNNs), or pre-trained convolutional neural networks (CNNs) that describe the dominant characteristic of the data to fully represent the hierarchically evolving free energy.
Recommendations
- A deep material network for multiscale topology learning and accelerated nonlinear modeling of heterogeneous materials
- Multiscale modeling of inelastic materials with thermodynamics-based artificial neural networks (TANN)
- Computational homogenization of nonlinear elastic materials using neural networks
- Machine learning materials physics: integrable deep neural networks enable scale bridging by learning free energy functions
- Efficient multiscale modeling of heterogeneous materials using deep neural networks
Cites work
- A data-driven computational homogenization method based on neural networks for the nonlinear anisotropic electrical response of graphene/polymer nanocomposites
- A framework for data-driven analysis of materials under uncertainty: countering the curse of dimensionality
- A multidimensional data-driven sparse identification technique: the sparse proper generalized decomposition
- A multiscale multi-permeability poroplasticity model linked by recursive homogenizations and deep learning
- An overview of the Trilinos project
- Computational homogenization of nonlinear elastic materials using neural networks
- Deep learning
- Discovering governing equations from data by sparse identification of nonlinear dynamical systems
- scientific article; zbMATH DE number 1069512 (Why is no real title available?)
- Knowledge-based artificial neural networks
- Machine learning materials physics: integrable deep neural networks enable scale bridging by learning free energy functions
- Machine learning materials physics: surrogate optimization and multi-fidelity algorithms predict precipitate morphology in an alternative to phase field dynamics
- Meta-modeling game for deriving theory-consistent, microstructure-based traction-separation laws via deep reinforcement learning
- Multi-scale computational homogenization: trends and challenges
- Numerical implementation of a neural network based material model in finite element analysis
- PetIGA: a framework for high-performance isogeometric analysis
- Predicting the effective mechanical property of heterogeneous materials by image based modeling and deep learning
- Theories of elasticity with couple-stress
- Three-dimensional isogeometric solutions to general boundary value problems of Toupin's gradient elasticity theory at finite strains
- Unconditionally stable, second-order accurate schemes for solid state phase transformations driven by mechano-chemical spinodal decomposition
- Unconditionally stable, second-order schemes for gradient-regularized, non-convex, finite-strain elasticity modeling martensitic phase transformations
- Variational system identification of the partial differential equations governing the physics of pattern-formation: inference under varying fidelity and noise
Cited in
(28)- Response classification of simple polycrystalline microstructures
- Machine learning materials physics: surrogate optimization and multi-fidelity algorithms predict precipitate morphology in an alternative to phase field dynamics
- Machine learning materials physics: integrable deep neural networks enable scale bridging by learning free energy functions
- Scale bridging materials physics: active learning workflows and integrable deep neural networks for free energy function representations in alloys
- Finite electro-elasticity with physics-augmented neural networks
- Towards out of distribution generalization for problems in mechanics
- Inside the black box: a physical basis for the effectiveness of deep generative models of amorphous materials
- Learning finite element convergence with the multi-fidelity graph neural network
- Multiscale modeling of inelastic materials with thermodynamics-based artificial neural networks (TANN)
- Microstructure-guided deep material network for rapid nonlinear material modeling and uncertainty quantification
- A heteroencoder architecture for prediction of failure locations in porous metals using variational inference
- Data-driven tissue mechanics with polyconvex neural ordinary differential equations
- Numerical analysis of non-local calculus on finite weighted graphs, with application to reduced-order modeling of dynamical systems
- Supervised learning through physical changes in a mechanical system
- Approximation of correlation functions in phase-field crystal model by machine learning approach
- A three-dimensional prediction method of stiffness properties of composites based on deep learning
- Accelerated offline setup of homogenized microscopic model for multi‐scale analyses using neural network with knowledge transfer
- Solving nonconvex energy minimization problems in martensitic phase transitions with a mesh-free deep learning approach
- Automated model discovery for skin: discovering the best model, data, and experiment
- Error estimates and physics informed augmentation of neural networks for thermally coupled incompressible Navier Stokes equations
- On the micromechanics of deep material networks
- Deep learning and multi-level featurization of graph representations of microstructural data
- Efficient multiscale modeling of heterogeneous materials using deep neural networks
- Embedding physical knowledge in deep neural networks for predicting the phonon dispersion curves of cellular metamaterials
- Neural cellular automata for solidification microstructure modelling
- The anisotropic graph neural network model with multiscale and nonlinear characteristic for turbulence simulation
- Peridynamic neural operators: a data-driven nonlocal constitutive model for complex material responses
- Microstructural inelastic fingerprints and data-rich predictions of plasticity and damage in solids
This page was built for publication: Machine learning materials physics: multi-resolution neural networks learn the free energy and nonlinear elastic response of evolving microstructures
Report a bug (only for logged in users!)Click here to report a bug for this page (MaRDI item Q2020954)
