Computational homogenization of nonlinear elastic materials using neural networks

In this work, a decoupled computational homogenization method for nonlinear elastic materials is proposed using neural networks. In this method, the effective potential is represented as a response surface parameterized by the macroscopic strains and some microstructural parameters. The discrete values of the effective potential are computed by finite element method through random sampling in the parameter space, and neural networks are used to approximate the surface response and to derive the macroscopic stress and tangent tensor components. We show through several numerical convergence analyses that smooth functions can be efficiently evaluated in parameter spaces with dimension up to 10, allowing to consider three‐dimensional representative volume elements and an explicit dependence of the effective behavior on microstructural parameters like volume fraction. We present several applications of this technique to the homogenization of nonlinear elastic composites, involving a two‐scale example of heterogeneous structure with graded nonlinear properties. Copyright © 2015 John Wiley & Sons, Ltd.

[1]  R. Hill.  Elastic properties of reinforced solids: some theoretical principles , 1963 .

[2]  R. L. Hardy.  Multiquadric equations of topography and other irregular surfaces , 1971 .

[3]  R. Goodrich,et al.  Weighted trigonometric approximation and inner—Outer functions on higher dimensional Euclidean spaces , 1981 .

[4]  J. Renard,et al.  Etude de l'initiation de l'endommagement dans la matrice d'un matériau composite par une méthode d'homogénéisation , 1987 .

[5]  J. Willis,et al.  On the overall properties of nonlinearly viscous composites , 1988, Proceedings of the Royal Society of London. A. Mathematical and Physical Sciences.

[6]  George Cybenko,et al.  Approximation by superpositions of a sigmoidal function , 1989, Math. Control. Signals Syst..

[7]  Bobby G. Sumpter,et al.  Theory and Applications of Neural Computing in Chemical Science , 1994 .

[8]  Joel M. Bowman,et al.  Vibrational self-consistent field method for many-mode systems: A new approach and application to the vibrations of CO adsorbed on Cu(100) , 1997 .

[9]  Ah Chung Tsoi,et al.  Universal Approximation Using Feedforward Neural Networks: A Survey of Some Existing Methods, and Some New Results , 1998, Neural Networks.

[10]  Manolis Papadrakakis,et al.  Structural optimization using evolution strategies and neural networks , 1998 .

[11]  J. Michel,et al.  Effective properties of composite materials with periodic microstructure : a computational approach , 1999 .

[12]  H. Rabitz,et al.  General foundations of high‐dimensional model representations , 1999 .

[13]  Ted Belytschko,et al.  A finite element method for crack growth without remeshing , 1999 .

[14]  Frédéric Feyel,et al.  Multiscale FE2 elastoviscoplastic analysis of composite structures , 1999 .

[15]  Roberto Guardani,et al.  Neural network based approach for optimization of industrial chemical processes , 2000 .

[16]  N. Kikuchi,et al.  A class of general algorithms for multi-scale analyses of heterogeneous media , 2001 .

[17]  Manolis Papadrakakis,et al.  Reliability-based structural optimization using neural networks and Monte Carlo simulation , 2002 .

[18]  V. Kouznetsova,et al.  Multi‐scale constitutive modelling of heterogeneous materials with a gradient‐enhanced computational homogenization scheme , 2002 .

[19]  Nicholas C. Handy,et al.  On the representation of potential energy surfaces of polyatomic molecules in normal coordinates , 2002 .

[20]  J. Michel,et al.  Nonuniform transformation field analysis , 2003 .

[21]  V. Kouznetsova,et al.  Multi-scale second-order computational homogenization of multi-phase materials : a nested finite element solution strategy , 2004 .

[22]  Sergei Manzhos,et al.  Using neural networks to represent potential surfaces as sums of products. , 2006, The Journal of chemical physics.

[23]  Sergei Manzhos,et al.  A random-sampling high dimensional model representation neural network for building potential energy surfaces. , 2006, The Journal of chemical physics.

[24]  Francisco Chinesta,et al.  A new family of solvers for some classes of multidimensional partial differential equations encountered in kinetic theory modeling of complex fluids , 2006 .

[25]  Sergei Manzhos,et al.  Using redundant coordinates to represent potential energy surfaces with lower-dimensional functions. , 2007, The Journal of chemical physics.

[26]  Peter Wriggers,et al.  An adaptive method for homogenization in orthotropic nonlinear elasticity , 2007 .

[27]  A. Meher Prasad,et al.  High dimensional model representation for piece‐wise continuous function approximation , 2007 .

[28]  Tarek I. Zohdi,et al.  A numerical method for homogenization in non-linear elasticity , 2007 .

[29]  Julien Yvonnet,et al.  The reduced model multiscale method (R3M) for the non-linear homogenization of hyperelastic media at finite strains , 2007, J. Comput. Phys..

[30]  M B Kasiri,et al.  Modeling and optimization of heterogeneous photo-Fenton process with response surface methodology and artificial neural networks. , 2008, Environmental science & technology.

[31]  Sergei Manzhos,et al.  Using neural networks, optimized coordinates, and high-dimensional model representations to obtain a vinyl bromide potential surface. , 2008, The Journal of chemical physics.

[32]  Julien Yvonnet,et al.  Computational homogenization for nonlinear conduction in heterogeneous materials using model reduction , 2008 .

[33]  Richard Dawes,et al.  Interpolating moving least-squares methods for fitting potential energy surfaces: a strategy for efficient automatic data point placement in high dimensions. , 2008, The Journal of chemical physics.

[34]  Julien Yvonnet,et al.  Numerically explicit potentials for the homogenization of nonlinear elastic heterogeneous materials , 2009 .

[35]  Koichi Yamashita,et al.  Fitting sparse multidimensional data with low-dimensional terms , 2009, Comput. Phys. Commun..

[36]  Pierre Suquet,et al.  Nonuniform transformation field analysis of elastic–viscoplastic composites , 2009 .

[37]  R Komanduri,et al.  Accurate prediction of higher-level electronic structure energies for large databases using neural networks, Hartree-Fock energies, and small subsets of the database. , 2009, The Journal of chemical physics.

[38]  Fariborz Haghighat,et al.  Multiobjective optimization of building design using TRNSYS simulations, genetic algorithm, and Artificial Neural Network , 2010 .

[39]  Sergei Manzhos,et al.  A model for the dissociative adsorption of N2O on Cu(1 0 0) using a continuous potential energy surface , 2010 .

[40]  Julien Yvonnet,et al.  A simple computational homogenization method for structures made of linear heterogeneous viscoelastic materials , 2011 .

[41]  J. Yvonnet,et al.  A simple computational homogenization method for structures made of heterogeneous linear viscoelastic materials , 2011 .

[42]  David Néron,et al.  Multiparametric analysis within the proper generalized decomposition framework , 2012 .

[43]  Christian Soize,et al.  Computational nonlinear stochastic homogenization using a nonconcurrent multiscale approach for hyperelastic heterogeneous microstructures analysis , 2012, International Journal for Numerical Methods in Engineering.

[44]  Thomas Böhlke,et al.  Reduced basis homogenization of viscoelastic composites , 2013 .

[45]  Julien Yvonnet,et al.  COMPUTATIONAL HOMOGENIZATION METHOD AND REDUCED DATABASE MODEL FOR HYPERELASTIC HETEROGENEOUS STRUCTURES , 2013 .

[46]  Dansheng Yu,et al.  Approximation by neural networks with sigmoidal functions , 2013 .

[47]  A. Ammar,et al.  PGD-Based Computational Vademecum for Efficient Design, Optimization and Control , 2013, Archives of Computational Methods in Engineering.

[48]  Christian Soize,et al.  Uncertainty quantification in computational stochastic multiscale analysis of nonlinear elastic materials , 2013 .

[49]  Elías Cueto,et al.  Nonincremental proper generalized decomposition solution of parametric uncoupled models defined in evolving domains , 2013 .

[50]  P. Breitkopf,et al.  Multiscale structural topology optimization with an approximate constitutive model for local material microstructure , 2015 .

[51]  Matthew Mosby,et al.  Hierarchically parallel coupled finite strain multiscale solver for modeling heterogeneous layers , 2015 .