A comparative study on different neural network architectures to model inelasticity

Abstract

The mathematical formulation of constitutive models to describe the path-dependent, i.e., inelastic, behavior of materials is a challenging task and has been a focus in mechanics research for several decades. There have been increased efforts to facilitate or automate this task through data-driven techniques, impelled in particular by the recent revival of neural networks (NNs) in computational mechanics. However, it seems questionable to simply not consider fundamental findings of constitutive modeling originating from the last decades research within NN-based approaches. Herein, we propose a comparative study on different feedforward and recurrent neural network architectures to model inelasticity. Within this study, we divide the models into three basic classes: black box NNs, NNs enforcing physics in a weak form, and NNs enforcing physics in a strong form. Thereby, the first class of networks can learn constitutive relations from data while the underlying physics are completely ignored, whereas the latter two are constructed such that they can account for fundamental physics, where special attention is paid to the second law of thermodynamics in this work. Conventional linear and nonlinear viscoelastic as well as elastoplastic models are used for training data generation and, later on, as reference. After training with random walk time sequences containing information on stress, strain, and, for some models, internal variables, the NN-based models are compared to the reference solution, whereby interpolation and extrapolation are considered. Besides the quality of the stress prediction, the related free energy and dissipation rate are analyzed to evaluate the models. Overall, the presented study enables a clear recording of the advantages and disadvantages of different NN architectures to model inelasticity and gives guidance on how to train and apply these models.

References in corpus (14)

• Thermodynamics-based Artificial Neural Networks for constitutive modeling
• A deep material network for multiscale topology learning and accelerated nonlinear modeling of heterogeneous materials
• Physics informed neural networks for continuum micromechanics
• Geometric deep learning for computational mechanics Part I: Anisotropic Hyperelasticity
• Exploring the 3D architectures of deep material network in data-driven multiscale mechanics
• Multiscale modeling of inelastic materials with Thermodynamics-based Artificial Neural Networks (TANN)
• FE${}^{\text{ANN}}$ $-$ An efficient data-driven multiscale approach based on physics-constrained neural networks and automated data mining
• Data-driven Tissue Mechanics with Polyconvex Neural Ordinary Differential Equations
• A hybrid approach to simulate the homogenized irreversible elastic-plastic deformations and damage of foams by neural networks
• Thermodynamically Consistent Machine-Learned Internal State Variable Approach for Data-Driven Modeling of Path-Dependent Materials
• Variational Onsager Neural Networks (VONNs): A thermodynamics-based variational learning strategy for non-equilibrium PDEs
• Modular machine learning-based elastoplasticity: generalization in the context of limited data
• Finite electro-elasticity with physics-augmented neural networks
• Data-driven anisotropic finite viscoelasticity using neural ordinary differential equations

Cited by in corpus (10)

• Deep Learning in Deterministic Computational Mechanics
• Reduced and All-at-Once Approaches for Model Calibration and Discovery in Computational Solid Mechanics
• Recurrent neural networks and transfer learning for elasto-plasticity in woven composites
• Multiscale topology optimization of functionally graded lattice structures based on physics-augmented neural network material models
• Experimental Learning of a Hyperelastic Behavior with a Physics-Augmented Neural Network
• Hybrid data-driven and physics-informed regularized learning of cyclic plasticity with Neural Networks
• Precise, efficient and flexible modeling of crystallizing elastomers based on physics-augmented neural networks
• Empirical Hyper Element Integration Method (EHEIM) with Unified Integration Criteria for Efficient Hyper Reduced FE${}^2$ Simulations
• Unsupervised Material Fingerprinting: Ultra-fast hyperelastic model discovery from full-field experimental measurements
• A physics-augmented neural network framework for finite strain incompressible viscoelasticity