Received: 16 July 2019 Revised: 27 November 2019 Accepted: 28 November 2019
DOI: 10.1002/nme.6283
RESEARCH ARTICLE
On Quasi-Newton methods in fast Fourier
transform-based micromechanics
Daniel Wicht Matti Schneider Thomas Böhlke
Karlsruhe Institute of Technology,
Institute of Engineering Mechanics,
Karlsruhe, Germany
Correspondence
Thomas Böhlke, Karlsruhe Institute of
Technology, Institute of Engineering
Mechanics, Karlsruhe 76131, Germany.
Email: thomas.boehlke@kit.edu
Funding information
German Research Foundation (DFG),
Grant/Award Number: “Integrated
engineering of continuous-discontinuous
long fiber reinforced polymer structures”
(GRK 2078); German Research
Foundation (DFG), Grant/Award
Number: “Lamellar Fe-Al in situ
composite materials” (BO 1466/12-2)
SUMMARY
This work is devoted to investigating the computational power of Quasi-Newton
methods in the context of fast Fourier transform (FFT)-based computational
micromechanics. We revisit FFT-based Newton-Krylov solvers as well as mod-
ern Quasi-Newton approaches such as the recently introduced Anderson accel-
erated basic scheme. In this context, we propose two algorithms based on the
Broyden-Fletcher-Goldfarb-Shanno (BFGS) method, one of the most power-
ful Quasi-Newton schemes. To be specific, we use the BFGS update formula
to approximate the global Hessian or, alternatively, the local material tangent
stiffness. Both for Newton and Quasi-Newton methods, a globalization tech-
nique is necessary to ensure global convergence. Specific to the FFT-based
context, we promote a Dong-type line search, avoiding function evaluations alto-
gether. Furthermore, we investigate the influence of the forcing term, that is,
the accuracy for solving the linear system, on the overall performance of inex-
act (Quasi-)Newton methods. This work concludes with numerical experiments,
comparing the convergence characteristics and runtime of the proposed tech-
niques for complex microstructures with nonlinear material behavior and finite
as well as infinite material contrast.
KEYWORDS
BFGS, composites, crystal viscoplasticity, FFT-based micromechanics, homogenization,
Quasi-Newton methods
1 INTRODUCTION
Fast Fourier transform (FFT)-based solution schemes
5,6
have become an established tool in computational microme-
chanics. Their popularity rests on their computational efficiency, the ability to treat nonlinear material behavior and
the compatibility to imaging techniques such as microcomputed tomography. Owing to these advantages, FFT-based
solvers have found widespread application in various fields, including the homogenization of polycrystals at small
7
and finite strains,
8
damage
9
and fracture mechanics,
10
fatigue prediction,
11
electromechanically coupled materials,
12
and concurrent multiscale simulations.
13,14
[Correction added on 14 February 2020, after first online publication: Funding information has been updated.]
This is an open access article under the terms of the Creative Commons Attribution-NonCommercial-NoDerivs License, which permits use and distribution in any medium,
provided the original work is properly cited, the use is non-commercial and no modifications or adaptations are made.
© 2019 The Authors. International Journal for Numerical Methods in Engineering published by John Wiley & Sons, Ltd.
Int J Numer Methods Eng. 2020;121:1665–1694. wileyonlinelibrary.com/journal/nme 1665