Physics-informed neural network for modelling the thermochemical curing process of composite-tool systems during manufacture
From MaRDI portal
Publication:2237458
DOI10.1016/J.CMA.2021.113959zbMATH Open1506.74123arXiv2011.13511OpenAlexW3109749236MaRDI QIDQ2237458FDOQ2237458
Authors: Sina Amini Niaki, Ehsan Haghighat, Trevor Campbell, Anoush Poursartip, Reza Vaziri 
Publication date: 27 October 2021
Published in: Computer Methods in Applied Mechanics and Engineering (Search for Journal in Brave)
Abstract: We present a Physics-Informed Neural Network (PINN) to simulate the thermochemical evolution of a composite material on a tool undergoing cure in an autoclave. In particular, we solve the governing coupled system of differential equations -- including conductive heat transfer and resin cure kinetics -- by optimizing the parameters of a deep neural network (DNN) using a physics-based loss function. To account for the vastly different behaviour of thermal conduction and resin cure, we design a PINN consisting of two disconnected subnetworks, and develop a sequential training algorithm that mitigates instability present in traditional training methods. Further, we incorporate explicit discontinuities into the DNN at the composite-tool interface and enforce known physical behaviour directly in the loss function to improve the solution near the interface. We train the PINN with a technique that automatically adapts the weights on the loss terms corresponding to PDE, boundary, interface, and initial conditions. Finally, we demonstrate that one can include problem parameters as an input to the model -- resulting in a surrogate that provides real-time simulation for a range of problem settings -- and that one can use transfer learning to significantly reduce the training time for problem settings similar to that of an initial trained model. The performance of the proposed PINN is demonstrated in multiple scenarios with different material thicknesses and thermal boundary conditions.
Full work available at URL: https://arxiv.org/abs/2011.13511
Recommendations
- A mixed formulation for physics-informed neural networks as a potential solver for engineering problems in heterogeneous domains: comparison with finite element method
- A physics-informed deep learning framework for inversion and surrogate modeling in solid mechanics
- Error estimates and physics informed augmentation of neural networks for thermally coupled incompressible Navier Stokes equations
- Non-invasive inference of thrombus material properties with physics-informed neural networks
- Physics-informed neural networks for shell structures
deep learningsurrogate modellingphysics-informed neural networkscomposites processingexothermic heat transferresin reaction
Artificial neural networks and deep learning (68T07) Chemical and reactive effects in solid mechanics (74F25)
Cites Work
- DGM: a deep learning algorithm for solving partial differential equations
- SciANN: a Keras/Tensorflow wrapper for scientific computations and physics-informed deep learning using artificial neural networks
- The numerical solution of linear ordinary differential equations by feedforward neural networks
- The weighted sum method for multi-objective optimization: new insights
- Title not available (Why is that?)
- Advanced Lectures on Machine Learning
- Title not available (Why is that?)
- A semi-discrete shell finite element for textile composite reinforcement forming simulation
- 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
- Quantifying total uncertainty in physics-informed neural networks for solving forward and inverse stochastic problems
- Adaptive fourth-order phase field analysis for brittle fracture
- Title not available (Why is that?)
- DeepXDE: a deep learning library for solving differential equations
- Surrogate modeling for fluid flows based on physics-constrained deep learning without simulation data
- A deep energy method for finite deformation hyperelasticity
- \textit{hp}-VPINNs: variational physics-informed neural networks with domain decomposition
- A physics-informed deep learning framework for inversion and surrogate modeling in solid mechanics
- Conservative physics-informed neural networks on discrete domains for conservation laws: applications to forward and inverse problems
Cited In (30)
- Physics-informed deep learning for simultaneous surrogate modeling and PDE-constrained optimization of an airfoil geometry
- Prediction of optical solitons using an improved physics-informed neural network method with the conservation law constraint
- Physics-informed neural network simulation of multiphase poroelasticity using stress-split sequential training
- Theory and implementation of inelastic constitutive artificial neural networks
- Physics-informed neural network frameworks for crack simulation based on minimized peridynamic potential energy
- Physics-informed machine-learning model of temperature evolution under solid phase processes
- Inverse modeling of nonisothermal multiphase poromechanics using physics-informed neural networks
- Physics-informed \textsc{MeshGraphNets} (PI-MGNs): neural finite element solvers for non-stationary and nonlinear simulations on arbitrary meshes
- Variational temporal convolutional networks for I-FENN thermoelasticity
- Elasticity-mechanics-informed generative adversarial networks for predicting the thermal strain of thermal barrier coatings penetrated by CaO-MgO-\(\mathrm{Al_2O}_3\)-\(\mathrm{SiO}_2\)
- Exact enforcement of temporal continuity in sequential physics-informed neural networks
- Deep learning in computational mechanics: a review
- A symmetry group based supervised learning method for solving partial differential equations
- Adaptive transfer learning for PINN
- Physical information neural networks for 2D and 3D nonlinear Biot model and simulation on the pressure of brain
- CAN-PINN: a fast physics-informed neural network based on coupled-automatic-numerical differentiation method
- Residual-based adaptivity for two-phase flow simulation in porous media using physics-informed neural networks
- An Adaptive Physics-Informed Neural Network with Two-Stage Learning Strategy to Solve Partial Differential Equations
- BINN: a deep learning approach for computational mechanics problems based on boundary integral equations
- Transfer learning based physics-informed neural networks for solving inverse problems in engineering structures under different loading scenarios
- Physics-integrated neural differentiable (PiNDiff) model for composites manufacturing
- I-FENN for thermoelasticity based on physics-informed temporal convolutional network (PI-TCN)
- Deep energy method in topology optimization applications
- CENN: conservative energy method based on neural networks with subdomains for solving variational problems involving heterogeneous and complex geometries
- Physics informed neural networks for continuum micromechanics
- Scientific machine learning through physics-informed neural networks: where we are and what's next
- Enhanced physics‐informed neural networks for hyperelasticity
- Physics-informed deep learning for three-dimensional transient heat transfer analysis of functionally graded materials
- Physical informed neural network for thermo-hydral analysis of fire-loaded concrete
- The line rogue wave solutions of the nonlocal Davey-Stewartson I equation with \textit{PT} symmetry based on the improved physics-informed neural network
Uses Software
This page was built for publication: Physics-informed neural network for modelling the thermochemical curing process of composite-tool systems during manufacture
Report a bug (only for logged in users!)Click here to report a bug for this page (MaRDI item Q2237458)
