Multiscale topology optimization using neural network surrogate models
Abstract
In this study, we are concerned with optimization of macroscale elastic structures that are designed utilizing spatially varying microscale metamaterials. The macroscale optimization is accomplished using gradient-based nonlinear topological optimization. But instead of using density as the optimization decision variable, the decision variables are the multiple parameters that define the local microscale metamaterial. This is accomplished using single layer feedforward Gaussian basis function networks as a surrogate models of the elastic response of the microscale metamaterial. The surrogate models are trained using highly resolved continuum finite element simulations of the microscale metamaterials and hence are significantly more accurate than analytical models e.g. classical beam theory. Because the derivative of the surrogate model is important for sensitivity analysis of the macroscale topology optimization, a neural network training procedure based on the Sobolev norm is described. Since the SIMP method is not appropriate for spatially varying lattices, an alternative method is developed to enable creation of void regions. Lastly, the efficacy of this approach is demonstrated via several examples in which the optimal graded metamaterial outperforms a traditional solid structure.
- Authors:
-
- Lawrence Livermore National Lab. (LLNL), Livermore, CA (United States)
- Publication Date:
- Research Org.:
- Lawrence Livermore National Lab. (LLNL), Livermore, CA (United States)
- Sponsoring Org.:
- USDOE National Nuclear Security Administration (NNSA)
- OSTI Identifier:
- 1557076
- Alternate Identifier(s):
- OSTI ID: 1636038
- Report Number(s):
- LLNL-JRNL-760619
Journal ID: ISSN 0045-7825; 948858
- Grant/Contract Number:
- AC52-07NA27344
- Resource Type:
- Accepted Manuscript
- Journal Name:
- Computer Methods in Applied Mechanics and Engineering
- Additional Journal Information:
- Journal Volume: 346; Journal Issue: C; Journal ID: ISSN 0045-7825
- Publisher:
- Elsevier
- Country of Publication:
- United States
- Language:
- English
- Subject:
- 42 ENGINEERING; 97 MATHEMATICS AND COMPUTING; Topology optimization; Multiscale analysis; Neural networks; Material models
Citation Formats
White, Daniel A., Arrighi, William J., Kudo, Jun, and Watts, Seth E. Multiscale topology optimization using neural network surrogate models. United States: N. p., 2018.
Web. doi:10.1016/j.cma.2018.09.007.
White, Daniel A., Arrighi, William J., Kudo, Jun, & Watts, Seth E. Multiscale topology optimization using neural network surrogate models. United States. https://doi.org/10.1016/j.cma.2018.09.007
White, Daniel A., Arrighi, William J., Kudo, Jun, and Watts, Seth E. Wed .
"Multiscale topology optimization using neural network surrogate models". United States. https://doi.org/10.1016/j.cma.2018.09.007. https://www.osti.gov/servlets/purl/1557076.
@article{osti_1557076,
title = {Multiscale topology optimization using neural network surrogate models},
author = {White, Daniel A. and Arrighi, William J. and Kudo, Jun and Watts, Seth E.},
abstractNote = {In this study, we are concerned with optimization of macroscale elastic structures that are designed utilizing spatially varying microscale metamaterials. The macroscale optimization is accomplished using gradient-based nonlinear topological optimization. But instead of using density as the optimization decision variable, the decision variables are the multiple parameters that define the local microscale metamaterial. This is accomplished using single layer feedforward Gaussian basis function networks as a surrogate models of the elastic response of the microscale metamaterial. The surrogate models are trained using highly resolved continuum finite element simulations of the microscale metamaterials and hence are significantly more accurate than analytical models e.g. classical beam theory. Because the derivative of the surrogate model is important for sensitivity analysis of the macroscale topology optimization, a neural network training procedure based on the Sobolev norm is described. Since the SIMP method is not appropriate for spatially varying lattices, an alternative method is developed to enable creation of void regions. Lastly, the efficacy of this approach is demonstrated via several examples in which the optimal graded metamaterial outperforms a traditional solid structure.},
doi = {10.1016/j.cma.2018.09.007},
journal = {Computer Methods in Applied Mechanics and Engineering},
number = C,
volume = 346,
place = {United States},
year = {Wed Oct 10 00:00:00 EDT 2018},
month = {Wed Oct 10 00:00:00 EDT 2018}
}
Web of Science
Figures / Tables:
Works referenced in this record:
Optimal shape design as a material distribution problem
journal, December 1989
- Bendsøe, M. P.
- Structural Optimization, Vol. 1, Issue 4
Aims, scope, methods, history and unified terminology of computer-aided topology optimization in structural mechanics
journal, April 2001
- Rozvany, G. I. N.
- Structural and Multidisciplinary Optimization, Vol. 21, Issue 2
Sufficiency of a finite exponent in SIMP (power law) methods
journal, April 2001
- Rietz, A.
- Structural and Multidisciplinary Optimization, Vol. 21, Issue 2
Topology optimization of non-linear elastic structures and compliant mechanisms
journal, March 2001
- Bruns, Tyler E.; Tortorelli, Daniel A.
- Computer Methods in Applied Mechanics and Engineering, Vol. 190, Issue 26-27
An element removal and reintroduction strategy for the topology optimization of structures and compliant mechanisms
journal, January 2003
- Bruns, T. E.; Tortorelli, D. A.
- International Journal for Numerical Methods in Engineering, Vol. 57, Issue 10
Filters in topology optimization based on Helmholtz-type differential equations
journal, December 2010
- Lazarov, B. S.; Sigmund, O.
- International Journal for Numerical Methods in Engineering, Vol. 86, Issue 6
Free material optimization via mathematical programming
journal, October 1997
- Zowe, Jochem; Kočvara, Michal; Bendsøe, Martin P.
- Mathematical Programming, Vol. 79, Issue 1-3
Free material optimization: recent progress†
journal, February 2008
- Kočvara, Michal; Stingl, Michael; Zowe, Jochem
- Optimization, Vol. 57, Issue 1
Shape optimization by the homogenization method
journal, March 1997
- Allaire, Grégoire; Bonnetier, Eric; Francfort, Gilles
- Numerische Mathematik, Vol. 76, Issue 1
A geometric projection method for designing three-dimensional open lattices with inverse homogenization: A geometric projection method
journal, July 2017
- Watts, Seth; Tortorelli, Daniel A.
- International Journal for Numerical Methods in Engineering, Vol. 112, Issue 11
Error estimates and condition numbers for radial basis function interpolation
journal, April 1995
- Schaback, Robert
- Advances in Computational Mathematics, Vol. 3, Issue 3
Universal Approximation Using Radial-Basis-Function Networks
journal, June 1991
- Park, J.; Sandberg, I. W.
- Neural Computation, Vol. 3, Issue 2
Multilayer feedforward networks are universal approximators
journal, January 1989
- Hornik, Kurt; Stinchcombe, Maxwell; White, Halbert
- Neural Networks, Vol. 2, Issue 5
Interpolation and rates of convergence for a class of neural networks
journal, March 2009
- Cao, Feilong; Zhang, Yongquan; He, Ze-Rong
- Applied Mathematical Modelling, Vol. 33, Issue 3
Approximation capability of interpolation neural networks
journal, December 2010
- Cao, Feilong; Lin, Shaobo; Xu, Zongben
- Neurocomputing, Vol. 74, Issue 1-3
On learning the derivatives of an unknown mapping with multilayer feedforward networks
journal, January 1992
- Gallant, A. Ronald; White, Halbert
- Neural Networks, Vol. 5, Issue 1
Neural networks as material models within a multiscale approach
journal, October 2009
- Unger, Jörg F.; Könke, Carsten
- Computers & Structures, Vol. 87, Issue 19-20
Accurate cyclic plastic analysis using a neural network material model
journal, March 2004
- Furukawa, Tomonari; Hoffman, Mark
- Engineering Analysis with Boundary Elements, Vol. 28, Issue 3
Artificial Neural Networks in numerical modelling of composites
journal, May 2009
- Lefik, M.; Boso, D. P.; Schrefler, B. A.
- Computer Methods in Applied Mechanics and Engineering, Vol. 198, Issue 21-26
Neural network constitutive modelling for non-linear characterization of anisotropic materials
journal, August 2010
- Man, H.; Furukawa, T.
- International Journal for Numerical Methods in Engineering, Vol. 85, Issue 8
Neural network based constitutive model for elastomeric foams
journal, July 2008
- Liang, G.; Chandrashekhara, K.
- Engineering Structures, Vol. 30, Issue 7
A neural network-based surrogate model for carbon nanotubes with geometric nonlinearities
journal, January 2018
- Papadopoulos, V.; Soimiris, G.; Giovanis, D. G.
- Computer Methods in Applied Mechanics and Engineering, Vol. 328
On the implementation of an interior-point filter line-search algorithm for large-scale nonlinear programming
journal, April 2005
- Wächter, Andreas; Biegler, Lorenz T.
- Mathematical Programming, Vol. 106, Issue 1
Optimal lattice-structured materials
journal, November 2016
- Messner, Mark C.
- Journal of the Mechanics and Physics of Solids, Vol. 96
Approximation of a function and its derivative with a neural network
journal, January 1992
- Cardaliaguet, Pierre; Euvrard, Guillaume
- Neural Networks, Vol. 5, Issue 2
Approximation of function and its derivatives using radial basis function networks
journal, March 2003
- Mai-Duy, Nam; Tran-Cong, Thanh
- Applied Mathematical Modelling, Vol. 27, Issue 3
On the limited memory BFGS method for large scale optimization
journal, August 1989
- Liu, Dong C.; Nocedal, Jorge
- Mathematical Programming, Vol. 45, Issue 1-3
Data dimensionality estimation methods: a survey
journal, December 2003
- Camastra, Francesco
- Pattern Recognition, Vol. 36, Issue 12
Polynomial Bounds for VC Dimension of Sigmoidal and General Pfaffian Neural Networks
journal, February 1997
- Karpinski, Marek; Macintyre, Angus
- Journal of Computer and System Sciences, Vol. 54, Issue 1
Almost Linear VC-Dimension Bounds for Piecewise Polynomial Networks
journal, November 1998
- Bartlett, Peter L.; Maiorov, Vitaly; Meir, Ron
- Neural Computation, Vol. 10, Issue 8
Works referencing / citing this record:
Simple, accurate surrogate models of the elastic response of three-dimensional open truss micro-architectures with applications to multiscale topology design
journal, May 2019
- Watts, Seth; Arrighi, William; Kudo, Jun
- Structural and Multidisciplinary Optimization, Vol. 60, Issue 5
Figures / Tables found in this record: