pith. sign in

arxiv: 2604.06519 · v1 · submitted 2026-04-07 · 💻 cs.CE

Multiscale topology optimization of compressible and nearly incompressible anisotropic hyperelastic structures using physics-augmented neural networks

Asghar A. Jadoon , Aryan Tyagi , L. River Spencer , Reese E. Jones , Manuel K. Rausch , Ryan Alberdi , D. Thomas Seidl , Jan N. Fuhg This is my paper

Pith reviewed 2026-05-10 17:54 UTC · model grok-4.3

classification 💻 cs.CE
keywords multiscaleoptimizationanisotropicneuraltopologyhyperelasticmaterialmicroscale
0
0 comments X

The pith

Physics-augmented neural networks enable simultaneous optimization of macroscale material distribution and microscale descriptors for hyperelastic structures in nonlinear finite strain.

A machine-rendered reading of the paper's core claim, the machinery that carries it, and where it could break.

This paper presents a concurrent multiscale topology optimization framework for compressible and nearly incompressible anisotropic hyperelastic structures. It uses physics-augmented neural networks as surrogate constitutive models that are trained to enforce physical principles like convexity and material symmetry. The approach allows optimizing both the macroscale material layout and the microscale descriptors at once in a nonlinear finite strain setting, which was previously limited by the cost of repeatedly solving microscale problems.

Core claim

By constructing input-specific neural networks with invariant-based representations and structural tensors, physics-augmented neural networks accurately represent homogenized anisotropic hyperelastic responses while ensuring thermodynamic consistency and numerical stability. These networks replace the microscale boundary value problem and supply stresses and consistent tangent moduli through analytical derivatives, making large-scale multiscale optimizations tractable, as demonstrated on transversely isotropic, cubic anisotropic, and nearly incompressible isotropic microstructures.

What carries the argument

Input-specific neural networks within physics-augmented architectures that use invariants and structural tensors to enforce key physical constraints during training and evaluation.

Load-bearing premise

The input-specific neural networks maintain sufficient accuracy, convexity, and thermodynamic consistency during the optimization iterations without introducing instabilities.

What would settle it

Compare the optimized material distributions and performance metrics obtained from the PANN-based method against those from a direct FE2 simulation on a representative benchmark problem to check for significant differences in the final designs.

Figures

Figures reproduced from arXiv: 2604.06519 by Aryan Tyagi, Asghar A. Jadoon, D. Thomas Seidl, Jan N. Fuhg, L. River Spencer, Manuel K. Rausch, Reese E. Jones, Ryan Alberdi.

Figure 1
Figure 1. Figure 1: Overview of the proposed multiscale topology optimization framework. The bottom panel illustrates the of [PITH_FULL_IMAGE:figures/full_fig_p003_1.png] view at source ↗
Figure 2
Figure 2. Figure 2: Fiber-reinforced representative volume element (RVE) exhibiting transverse isotropy (left), training loss [PITH_FULL_IMAGE:figures/full_fig_p018_2.png] view at source ↗
Figure 3
Figure 3. Figure 3: Topology Optimization results for the anisotropic, simply-supported beam. The preferred fiber directions in [PITH_FULL_IMAGE:figures/full_fig_p018_3.png] view at source ↗
Figure 4
Figure 4. Figure 4: Compliance convergence for the simply supported beam with topology snapshots shown at iterations 39, 76, [PITH_FULL_IMAGE:figures/full_fig_p019_4.png] view at source ↗
Figure 5
Figure 5. Figure 5: Topology Optimization results for the anisotropic cantilever beam. [PITH_FULL_IMAGE:figures/full_fig_p020_5.png] view at source ↗
Figure 6
Figure 6. Figure 6: Representative volume element (RVE) with a spherical inclusion exhibiting cubic anisotropy (left), training [PITH_FULL_IMAGE:figures/full_fig_p020_6.png] view at source ↗
Figure 7
Figure 7. Figure 7: Topology optimization results for the compressible case with a cubic-anisotropic RVE. The problem setup, [PITH_FULL_IMAGE:figures/full_fig_p021_7.png] view at source ↗
Figure 8
Figure 8. Figure 8: Representative volume element (RVE) with randomly distributed inclusions exhibiting isotropic and nearly [PITH_FULL_IMAGE:figures/full_fig_p022_8.png] view at source ↗
Figure 9
Figure 9. Figure 9: Topology optimization results for the nearly incompressible isotropic case. The problem setup and compli [PITH_FULL_IMAGE:figures/full_fig_p023_9.png] view at source ↗
read the original abstract

Multiscale topology optimization (TO) of hyperelastic materials remains computationally prohibitive due to the repeated solution of microscale boundary value problems. In this work, we present a concurrent multiscale topology optimization framework that overcomes this limitation by leveraging physics-augmented neural networks (PANNs) as surrogate constitutive models. The proposed approach enables the simultaneous optimization of macroscale material distribution and microscale descriptors, within a unified nonlinear finite strain setting. The surrogate models are constructed using input-specific neural networks (ISNNs) that enforce key physical principles directly within the architecture, including convexity and material symmetry through invariant-based representations and structural tensors. This ensures thermodynamic consistency and numerical stability while accurately representing homogenized anisotropic hyperelastic responses. The trained PANNs replace the microscale boundary value problem and provide efficient evaluations of stresses and consistent tangent moduli using analytical first and second derivatives of the neural network, enabling tractable large-scale multiscale optimization. The framework is demonstrated on representative microstructures exhibiting transversely isotropic, cubic anisotropic, and nearly incompressible isotropic behavior. The results show that the proposed method captures complex multiscale interactions and enables physically meaningful spatial tailoring of material properties, while significantly reducing computational cost compared to classical FE$^2$ approaches. These findings establish PANNs as a powerful tool for high-fidelity multiscale topology optimization of nonlinear anisotropic materials.

Editorial analysis

A structured set of objections, weighed in public.

Desk editor's note, referee report, simulated authors' rebuttal, and a circularity audit. Tearing a paper down is the easy half of reading it; the pith above is the substance, this is the friction.

Referee Report

1 major / 2 minor

Summary. The manuscript proposes a concurrent multiscale topology optimization framework for compressible and nearly incompressible anisotropic hyperelastic structures under finite strains. Physics-augmented neural networks (PANNs) based on input-specific neural networks serve as surrogates for microscale boundary value problems, with invariants and structural tensors enforcing convexity, material symmetry, and thermodynamic consistency. Analytical first and second derivatives of the networks supply stresses and consistent tangents, enabling simultaneous optimization of macroscale material distribution and microscale descriptors. Demonstrations on transversely isotropic, cubic, and nearly incompressible cases claim to capture multiscale interactions while reducing cost relative to classical FE² approaches.

Significance. If the accuracy and stability of the PANN surrogates hold across optimization trajectories, the work would meaningfully advance tractable multiscale topology optimization for nonlinear anisotropic materials by replacing repeated microscale solves with physics-consistent, differentiable surrogates. The architecture's direct embedding of convexity and symmetry constraints, together with the provision of analytical derivatives, is a concrete strength that supports numerical stability in large-scale problems.

major comments (1)
  1. Abstract: the central claim that the trained PANNs 'maintain sufficient accuracy, convexity, and thermodynamic consistency throughout the optimization iterations' and 'significantly reduc[e] computational cost' rests on demonstrations whose quantitative support is not described. No error metrics (e.g., stress or energy errors versus direct FE²), side-by-side comparisons of final topologies or objective values, or stability indicators (e.g., convergence rates or non-physical designs) are reported, leaving the load-bearing assertion of surrogate fidelity during concurrent macro-micro optimization unverified.
minor comments (2)
  1. The precise definition of 'input-specific neural networks' and how training data are sampled across the joint space of strains and microscale descriptors should be clarified to allow readers to assess coverage of regimes encountered during optimization.
  2. Notation for the structural tensors and invariant sets used to enforce material symmetry could be consolidated in a single table or appendix for easier cross-reference with the network architecture.

Simulated Author's Rebuttal

1 responses · 0 unresolved

We thank the referee for the thoughtful review and for recognizing the potential of our concurrent multiscale topology optimization framework using physics-augmented neural networks. We address the single major comment below and will incorporate revisions to strengthen the manuscript.

read point-by-point responses

  1. Referee: Abstract: the central claim that the trained PANNs 'maintain sufficient accuracy, convexity, and thermodynamic consistency throughout the optimization iterations' and 'significantly reduc[e] computational cost' rests on demonstrations whose quantitative support is not described. No error metrics (e.g., stress or energy errors versus direct FE²), side-by-side comparisons of final topologies or objective values, or stability indicators (e.g., convergence rates or non-physical designs) are reported, leaving the load-bearing assertion of surrogate fidelity during concurrent macro-micro optimization unverified.

    Authors: We appreciate this observation. The current manuscript presents visual results of optimized topologies, objective histories, and deformed configurations for the transversely isotropic, cubic, and nearly incompressible cases, along with qualitative indications that the optimization proceeds stably without non-physical artifacts. However, we acknowledge that explicit quantitative error metrics (such as relative L2 errors in stress and strain energy between PANN predictions and direct FE² evaluations at sampled optimization steps) and direct side-by-side numerical comparisons of final objective values are not tabulated or plotted. In the revised version we will add these quantitative assessments in the numerical results section, including error evolution plots during optimization, tabulated objective values for PANN versus reference FE² runs on selected problems, and convergence-rate indicators. These additions will directly support the abstract claims without altering the overall conclusions. revision: yes

Circularity Check

0 steps flagged

No circularity; PANN surrogates trained on independent microscale data and used directly in optimization

full rationale

The paper trains input-specific neural networks (ISNNs) within the PANN framework on data from separate microscale boundary value problems for representative microstructures. These networks enforce convexity and symmetry via invariants and structural tensors in their architecture, then replace the microscale BVP during concurrent macro/micro topology optimization by supplying stresses and consistent tangents via analytical derivatives. This is a standard surrogate workflow: training data is generated independently via FE homogenization, and the optimization loop consumes the network outputs without any fitted parameter being redefined as a prediction or any load-bearing step reducing to self-citation. The abstract and framework description present the method as a computational replacement for repeated FE² solves, with numerical demonstrations on transversely isotropic, cubic, and nearly incompressible cases serving as validation rather than a closed derivation. No self-definitional loop, fitted-input prediction, or ansatz smuggled via citation appears in the central claim.

Axiom & Free-Parameter Ledger

1 free parameters · 1 axioms · 0 invented entities

The central claim rests on fitted neural network parameters learned from microscale simulations plus the domain assumption that invariant-based architectures can faithfully surrogate homogenized responses without degrading optimization quality.

free parameters (1)
  • Neural network weights and biases
    Fitted during training to microscale simulation data to approximate the homogenized constitutive law and its derivatives.
axioms (1)
  • domain assumption Homogenized anisotropic hyperelastic responses admit accurate representation by invariant-based neural networks that enforce convexity, material symmetry via structural tensors, and thermodynamic consistency.
    Invoked to justify replacing the microscale BVP with the PANN surrogate while preserving numerical stability in the optimization.

pith-pipeline@v0.9.0 · 5576 in / 1386 out tokens · 39309 ms · 2026-05-10T17:54:05.225251+00:00 · methodology

discussion (0)

Sign in with ORCID, Apple, or X to comment. Anyone can read and Pith papers without signing in.

Lean theorems connected to this paper

Citations machine-checked in the Pith Canon. Every link opens the source theorem in the public Lean library.

  • IndisputableMonolith/Cost/FunctionalEquation.lean washburn_uniqueness_aczel echoes
    ?
    echoes

    ECHOES: this paper passage has the same mathematical shape or conceptual pattern as the Recognition theorem, but is not a direct formal dependency.

    The surrogate models are constructed using input-specific neural networks (ISNNs) that enforce key physical principles directly within the architecture, including convexity and material symmetry through invariant-based representations and structural tensors.

  • IndisputableMonolith/Foundation/BranchSelection.lean branch_selection unclear
    ?
    unclear

    Relation between the paper passage and the cited Recognition theorem.

    PANNs rely on specialized architectures with embedded constraints to give thermodynamically consistent, physically admissible, and numerically stable representations of the free energy function.

What do these tags mean?
matches
The paper's claim is directly supported by a theorem in the formal canon.
supports
The theorem supports part of the paper's argument, but the paper may add assumptions or extra steps.
extends
The paper goes beyond the formal theorem; the theorem is a base layer rather than the whole result.
uses
The paper appears to rely on the theorem as machinery.
contradicts
The paper's claim conflicts with a theorem or certificate in the canon.
unclear
Pith found a possible connection, but the passage is too broad, indirect, or ambiguous to say the theorem truly supports the claim.

Reference graph

Works this paper leans on

95 extracted references · 95 canonical work pages · 1 internal anchor

  1. [1]

    Bell-Navas, D

    E. Bell-Navas, D. Portillo, and I. Romero. A fully variational numerical method for structural topology optimiza- tion based on a cahn-hilliard model.Computer Methods in Applied Mechanics and Engineering, 445:118233, 2025

    work page 2025

  2. [2]

    Chi Ma, Mengyuan Li, Jialan Liu, Mingming Li, Jialong He, Giovanni Totis, Chunlei Hua, Liang Wang, Gangwei Cui, Ruijuan Xue, Zhi Tan, Jun Yang, Kuo Liu, Yuansheng Zhou, Jianqiang Zhou, Xiaolei Deng, and Shengbin Weng. High-efficiency topology optimization method for thermal-fluid problems in cooling jacket of high-speed motorized spindle.International Com...

    work page 2025

  3. [3]

    Naoki Murai, Yuki Noguchi, Kei Matsushima, and Takayuki Yamada. Multiscale topology optimization of electromagnetic metamaterials using a high-contrast homogenization method.Computer Methods in Applied Mechanics and Engineering, 403:115728, 2023

    work page 2023

  4. [4]

    Topology opti- mization for nonlocal elastoplasticity at finite strain.Computer Methods in Applied Mechanics and Engineering, 435:117678, 2025

    Jike Han, Kozo Furuta, Tsuguo Kondoh, Kazuhiro Izui, Shinji Nishiwaki, and Kenjiro Terada. Topology opti- mization for nonlocal elastoplasticity at finite strain.Computer Methods in Applied Mechanics and Engineering, 435:117678, 2025

    work page 2025

  5. [5]

    Effect of formulation method for plastic deformation rate on topology optimization considering elastic-plastic behavior.European Journal of Mechanics - A/Solids, 106:105347, 2024

    Kyeong-Bin Kim, Tae-Hyun Kim, and Eun-Ho Lee. Effect of formulation method for plastic deformation rate on topology optimization considering elastic-plastic behavior.European Journal of Mechanics - A/Solids, 106:105347, 2024

    work page 2024

  6. [6]

    Guodong Zhang and Kapil Khandelwal. Gurson–tvergaard–needleman model guided fracture-resistant structural designs under finite deformations.International Journal for Numerical Methods in Engineering, 123(14):3344– 3388, 2022

    work page 2022

  7. [7]

    Topology optimization of pressure dependent elastoplastic energy absorb- ing structures with material damage constraints.Finite Elements in Analysis and Design, 133:42–61, 2017

    Ryan Alberdi and Kapil Khandelwal. Topology optimization of pressure dependent elastoplastic energy absorb- ing structures with material damage constraints.Finite Elements in Analysis and Design, 133:42–61, 2017

    work page 2017

  8. [8]

    Guodong Zhang, Ryan Alberdi, and Kapil Khandelwal. Topology optimization with incompressible materials under small and finite deformations using mixed u/p elements.International Journal for Numerical Methods in Engineering, 115(8):1015–1052, 2018

    work page 2018

  9. [9]

    Topology optimization of incompressible materials based on the mixed sbfem.Com- puters & Structures, 165:24–33, 2016

    Chao Li and Liyong Tong. Topology optimization of incompressible materials based on the mixed sbfem.Com- puters & Structures, 165:24–33, 2016

    work page 2016

  10. [10]

    An alternative truly-mixed formulation to solve pressure load problems in topology optimization.Computer Methods in Applied Mechanics and Engineering, 198(17):1500–1512, 2009

    Matteo Bruggi and Carlo Cinquini. An alternative truly-mixed formulation to solve pressure load problems in topology optimization.Computer Methods in Applied Mechanics and Engineering, 198(17):1500–1512, 2009

    work page 2009

  11. [11]

    Kochmann

    Li Zheng, Siddhant Kumar, and Dennis M. Kochmann. Data-driven topology optimization of spinodoid meta- materials with seamlessly tunable anisotropy.Computer Methods in Applied Mechanics and Engineering, 383:113894, 2021

    work page 2021

  12. [12]

    Enhancing energy absorption in printed pla components through post-processing and topo- logical optimization.Results in Engineering, 26:104972, 2025

    Moises Jimenez-Martinez, Julio Varela-Soriano, Jose Martinez-Trinidad, Mariel Alfaro-Ponce, and Manuel Coca- Gonzalez. Enhancing energy absorption in printed pla components through post-processing and topo- logical optimization.Results in Engineering, 26:104972, 2025

    work page 2025

  13. [13]

    Topology optimization of fiber-reinforced structures with discrete fiber orientations for additive manufacturing.Computers & Structures, 301:107468, 2024

    Md Mohaiminul Islam and Ling Liu. Topology optimization of fiber-reinforced structures with discrete fiber orientations for additive manufacturing.Computers & Structures, 301:107468, 2024

    work page 2024

  14. [14]

    Materials with prescribed constitutive parameters: An inverse homogenization problem.Interna- tional Journal of Solids and Structures, 31(17):2313–2329, 1994

    Ole Sigmund. Materials with prescribed constitutive parameters: An inverse homogenization problem.Interna- tional Journal of Solids and Structures, 31(17):2313–2329, 1994

    work page 1994

  15. [15]

    Tailoring materials with prescribed elastic properties.Mechanics of Materials, 20(4):351–368, 1995

    Ole Sigmund. Tailoring materials with prescribed elastic properties.Mechanics of Materials, 20(4):351–368, 1995

    work page 1995

  16. [16]

    Jun Wu, Ole Sigmund, and Jeroen P. Groen. Topology optimization of multi-scale structures: a review.Struct. Multidiscip. Optim., 63(3):1455–1480, March 2021

    work page 2021

  17. [17]

    Architected materials for mechanical compression: Design via simulation, deep learning, and experimentation

    Andrew Lew, Kai Jin, and Markus Buehler. Architected materials for mechanical compression: Design via simulation, deep learning, and experimentation. 12 2022

    work page 2022

  18. [18]

    Topology and ma- terial optimization in ultra-soft magneto-active structures: Making advantage of residual anisotropies.Advanced Materials, n/a(n/a):e18489

    Carlos Perez-Garcia, Rogelio Ortigosa, Jes ´us Mart´ınez-Frutos, and Daniel Garcia-Gonzalez. Topology and ma- terial optimization in ultra-soft magneto-active structures: Making advantage of residual anisotropies.Advanced Materials, n/a(n/a):e18489

    work page

  19. [19]

    Microstruc- tures to control elasticity in 3d printing.ACM Trans

    Christian Schumacher, Bernd Bickel, Jan Rys, Steve Marschner, Chiara Daraio, and Markus Gross. Microstruc- tures to control elasticity in 3d printing.ACM Trans. Graph., 34(4), July 2015

    work page 2015

  20. [20]

    A review of topology optimization for additive manufacturing: Status and challenges.Chinese Journal of Aeronautics, 34(1):91–110, 2021

    Jihong ZHU, Han ZHOU, Chuang W ANG, Lu ZHOU, Shangqin YUAN, and Weihong ZHANG. A review of topology optimization for additive manufacturing: Status and challenges.Chinese Journal of Aeronautics, 34(1):91–110, 2021. 34 Multiscale topology optimization using physics-augmented neural networksA PREPRINT

    work page 2021

  21. [21]

    Design concepts for generating optimised lattice structures aligned with strain trajectories.Computer Methods in Applied Mechanics and Engineering, 354:689– 705, 2019

    Stephen Daynes, Stefanie Feih, Wen Feng Lu, and Jun Wei. Design concepts for generating optimised lattice structures aligned with strain trajectories.Computer Methods in Applied Mechanics and Engineering, 354:689– 705, 2019

    work page 2019

  22. [22]

    E. D. Sanders, A. Pereira, and G. H. Paulino. Optimal and continuous multilattice embedding.Science Advances, 7(16):eabf4838, 2021

    work page 2021

  23. [23]

    Oxford University Press, 2010

    Jacob Fish.Multiscale methods: bridging the scales in science and engineering. Oxford University Press, 2010

    work page 2010

  24. [24]

    Frederic Feyel. A multilevel finite element method (fe2) to describe the response of highly non-linear structures using generalized continua.Computer Methods in Applied Mechanics and Engineering, 192:3233–3244, 07 2003

    work page 2003

  25. [25]

    Miehe and A

    C. Miehe and A. Koch. Computational micro-to-macro transitions of discretized microstructures undergoing small strains.Archive of Applied Mechanics (Ingenieur Archiv), 72:300–317, 07 2002

    work page 2002

  26. [26]

    Oxford University Press, 10 2005

    L R G.Treloar.The Physics of Rubber Elasticity. Oxford University Press, 10 2005

    work page 2005

  27. [27]

    Balzani, P

    D. Balzani, P. Neff, J. Schr ¨oder, and G.A. Holzapfel. A polyconvex framework for soft biological tissues. adjustment to experimental data.International Journal of Solids and Structures, 43(20):6052–6070, 2006

    work page 2006

  28. [28]

    P. A. L. S. Martins, R. M. Natal Jorge, and A. J. M. Ferreira. A comparative study of several material models for prediction of hyperelastic properties: Application to silicone-rubber and soft tissues.Strain, 42(3):135–147, 2006

    work page 2006

  29. [29]

    A finite element formulation for nonlinear incompressible elastic and inelastic analysis.Computers & Structures, 26(1):357–409, 1987

    Theodore Sussman and Klaus-J ¨urgen Bathe. A finite element formulation for nonlinear incompressible elastic and inelastic analysis.Computers & Structures, 26(1):357–409, 1987

    work page 1987

  30. [30]

    Tortorelli, and Daniel A

    Seth Watts, William Arrighi, Jun Kudo, Daniel A. Tortorelli, and Daniel A. White. Simple, accurate surrogate models of the elastic response of three-dimensional open truss micro-architectures with applications to multiscale topology design.Struct. Multidiscip. Optim., 60(5):1887–1920, November 2019

    work page 1920

  31. [31]

    Senhora, Emily D

    Fernando V . Senhora, Emily D. Sanders, and Glaucio H. Paulino. Optimally-tailored spinodal architected mate- rials for multiscale design and manufacturing.Advanced Materials, 34(26):2109304, 2022

    work page 2022

  32. [32]

    E. D. Sanders, A. Pereira, and G. H. Paulino. Optimal and continuous multilattice embedding.Science Advances, 7(16):eabf4838, 2021. [33]{Jeroen P.}Groen, Jun Wu, and Ole Sigmund. Homogenization-based stiffness optimization and projection of 2d coated structures with orthotropic infill.Computer Methods in Applied Mechanics and Engineering, 349:722– 742, 2019

    work page 2021

  33. [33]

    A unified framework for nonlinear path- dependent sensitivity analysis in topology optimization.International Journal for Numerical Methods in En- gineering, 115(1):1–56, 2018

    Ryan Alberdi, Guodong Zhang, Lei Li, and Kapil Khandelwal. A unified framework for nonlinear path- dependent sensitivity analysis in topology optimization.International Journal for Numerical Methods in En- gineering, 115(1):1–56, 2018

    work page 2018

  34. [34]

    Multiscale simulation of spatially correlated microstructure via a latent space representation.International Journal of Solids and Structures, 301:112966, 2024

    Reese E Jones, Craig M Hamel, Dan Bolintineanu, Kyle Johnson, Robert Buarque de Macedo, Jan Fuhg, Niko- laos Bouklas, and Sharlotte Kramer. Multiscale simulation of spatially correlated microstructure via a latent space representation.International Journal of Solids and Structures, 301:112966, 2024

    work page 2024

  35. [35]

    Topology optimization of multiscale elasto- viscoplastic structures.International Journal for Numerical Methods in Engineering, 106(6):430–453, 2016

    Felix Fritzen, Liang Xia, Matthias Leuschner, and Piotr Breitkopf. Topology optimization of multiscale elasto- viscoplastic structures.International Journal for Numerical Methods in Engineering, 106(6):430–453, 2016

    work page 2016

  36. [36]

    Junji Kato, Daishun Yachi, Takashi Kyoya, and Kenjiro Terada. Micro-macro concurrent topology optimization for nonlinear solids with a decoupling multiscale analysis.International Journal for Numerical Methods in Engineering, 113(8):1189–1213, 2018

    work page 2018

  37. [37]

    Enhancing multiscale simulations with constitutive relations-aware deep operator networks.PAMM, 24(2):e202400123, 2024

    Hamidreza Eivazi, Mahyar Alikhani, Jendrik-Alexander Tr ¨oger, Stefan Wittek, Stefan Hartmann, and Andreas Rausch. Enhancing multiscale simulations with constitutive relations-aware deep operator networks.PAMM, 24(2):e202400123, 2024

    work page 2024

  38. [38]

    Mozaffar, R

    M. Mozaffar, R. Bostanabad, W. Chen, K. Ehmann, J. Cao, and M. A. Bessa. Deep learning predicts path- dependent plasticity.Proceedings of the National Academy of Sciences, 116(52):26414–26420, 2019

    work page 2019

  39. [39]

    Tensor basis gaussian process models of hyperelastic materials

    Ari L Frankel, Reese E Jones, and Laura P Swiler. Tensor basis gaussian process models of hyperelastic materials. Journal of Machine Learning for Modeling and Computing, 1(1), 2020

    work page 2020

  40. [40]

    Jan N Fuhg and Nikolaos Bouklas. On physics-informed data-driven isotropic and anisotropic constitutive mod- els through probabilistic machine learning and space-filling sampling.Computer Methods in Applied Mechanics and Engineering, 394:114915, 2022. 35 Multiscale topology optimization using physics-augmented neural networksA PREPRINT

    work page 2022

  41. [41]

    A review on data-driven constitutive laws for solids.Archives of Computational Methods in Engineering, pages 1841–1883, 2025

    Jan Niklas Fuhg, Govinda Anantha Padmanabha, Nikolaos Bouklas, Bahador Bahmani, WaiChing Sun, Niko- laos N Vlassis, Moritz Flaschel, Pietro Carrara, and Laura De Lorenzis. A review on data-driven constitutive laws for solids.Archives of Computational Methods in Engineering, pages 1841–1883, 2025

    work page 2025

  42. [42]

    Data-driven multiscale nonlinear topology optimization of functionally graded soft composites under large deformations.Computational Mechanics, pages 1–30, 2025

    Shiguang Deng, Horacio D Espinosa, and Wei Chen. Data-driven multiscale nonlinear topology optimization of functionally graded soft composites under large deformations.Computational Mechanics, pages 1–30, 2025

    work page 2025

  43. [43]

    Neural networks meet hyperelasticity: A guide to enforcing physics.Journal of the Mechanics and Physics of Solids, 179:105363, 2023

    Lennart Linden, Dominik K Klein, Karl A Kalina, J ¨org Brummund, Oliver Weeger, and Markus K¨astner. Neural networks meet hyperelasticity: A guide to enforcing physics.Journal of the Mechanics and Physics of Solids, 179:105363, 2023

    work page 2023

  44. [44]

    Kalina, Lennart Linden, J ¨org Brummund, and Markus K ¨astner

    Karl A. Kalina, Lennart Linden, J ¨org Brummund, and Markus K ¨astner. Feann: an efficient data-driven mul- tiscale approach based on physics-constrained neural networks and automated data mining.Comput. Mech., 71(5):827–851, February 2023

    work page 2023

  45. [45]

    Kalina, Philipp Gebhart, J ¨org Brummund, Lennart Linden, WaiChing Sun, and Markus K¨astner

    Karl A. Kalina, Philipp Gebhart, J ¨org Brummund, Lennart Linden, WaiChing Sun, and Markus K¨astner. Neural network-based multiscale modeling of finite strain magneto-elasticity with relaxed convexity criteria.Computer Methods in Applied Mechanics and Engineering, 421:116739, 2024

    work page 2024

  46. [46]

    Inverse design of anisotropic microstructures using physics-augmented neural networks.Journal of the Mechanics and Physics of Solids, page 106161, 2025

    Asghar A Jadoon, Karl A Kalina, Manuel K Rausch, Reese Jones, and Jan Niklas Fuhg. Inverse design of anisotropic microstructures using physics-augmented neural networks.Journal of the Mechanics and Physics of Solids, page 106161, 2025

    work page 2025

  47. [47]

    A complement to neural networks for anisotropic inelasticity at finite strains

    Hagen Holthusen and Ellen Kuhl. A complement to neural networks for anisotropic inelasticity at finite strains. Computer Methods in Applied Mechanics and Engineering, 450:118612, 2026

    work page 2026

  48. [48]

    Zico Kolter

    Brandon Amos, Lei Xu, and J. Zico Kolter. Input convex neural networks. In Doina Precup and Yee Whye Teh, editors,Proceedings of the 34th International Conference on Machine Learning, volume 70 ofProceedings of Machine Learning Research, pages 146–155. PMLR, 06–11 Aug 2017

    work page 2017

  49. [49]

    Input specific neural networks.Computa- tional Mechanics, pages 1–23, 2026

    Asghar A Jadoon, D Thomas Seidl, Reese E Jones, and Jan N Fuhg. Input specific neural networks.Computa- tional Mechanics, pages 1–23, 2026

    work page 2026

  50. [50]

    Russ, Glaucio H

    Harikrishnan Vijayakumaran, Jonathan B. Russ, Glaucio H. Paulino, and Miguel A. Bessa. Consistent machine learning for topology optimization with microstructure-dependent neural network material models.Journal of the Mechanics and Physics of Solids, 196:106015, 2025

    work page 2025

  51. [51]

    Andre Luis Ferreira da Silva, Ruben Andres Salas, Emilio Carlos Nelli Silva, and J.N. Reddy. Topology opti- mization of fibers orientation in hyperelastic composite material.Composite Structures, 231:111488, 2020

    work page 2020

  52. [52]

    Topology optimization of fiber reinforced structures considering stress constraint and optimized penalization.Composite Structures, 316:117006, 2023

    Andre Luis Ferreira da Silva, Ruben Andres Salas, and Emilio Carlos Nelli Silva. Topology optimization of fiber reinforced structures considering stress constraint and optimized penalization.Composite Structures, 316:117006, 2023

    work page 2023

  53. [53]

    Courier Corporation, 1997

    Raymond W Ogden.Non-linear elastic deformations. Courier Corporation, 1997

    work page 1997

  54. [54]

    Structural tensors for anisotropic solids.Arch

    JM Zhang and J Rychlewski. Structural tensors for anisotropic solids.Arch. Mech, 42(3):267–277, 1990

    work page 1990

  55. [55]

    On the representation of constitutive relations using structure tensors.International journal of engineering science, 32(12):1889–1892, 1994

    Bob Svendsen. On the representation of constitutive relations using structure tensors.International journal of engineering science, 32(12):1889–1892, 1994

    work page 1994

  56. [56]

    Academic press, 1982

    Morton E Gurtin.An introduction to continuum mechanics, volume 158. Academic press, 1982

    work page 1982

  57. [57]

    A polyconvex model for materials with cubic symmetry.Modelling and Simulation in Materials Science and Engineering, 15(5):451, 2007

    Nayden Kambouchev, Javier Fernandez, and Raul Radovitzky. A polyconvex model for materials with cubic symmetry.Modelling and Simulation in Materials Science and Engineering, 15(5):451, 2007

    work page 2007

  58. [58]

    Anisotropic polyconvex energies on the basis of crystallographic motivated structural tensors.Journal of the Mechanics and Physics of Solids, 56(12):3486–3506, 2008

    J ¨org Schr¨oder, Patrizio Neff, and Vera Ebbing. Anisotropic polyconvex energies on the basis of crystallographic motivated structural tensors.Journal of the Mechanics and Physics of Solids, 56(12):3486–3506, 2008

    work page 2008

  59. [59]

    Neu- ral networks meet hyperelasticity: A guide to enforcing physics.J

    Lennart Linden, Dominik K Klein, Karl A Kalina, J ¨org Brummund, Oliver Weeger, and Markus K¨astner. Neu- ral networks meet hyperelasticity: A guide to enforcing physics.J. Mech. Phys. Solids, 179(105363):105363, October 2023

    work page 2023

  60. [60]

    Viscoelasticty with physics-augmented neural networks: model formulation and training methods without prescribed internal vari- ables.Computational Mechanics, 74(6):1279–1301, 2024

    Max Rosenkranz, Karl A Kalina, J ¨org Brummund, WaiChing Sun, and Markus K ¨astner. Viscoelasticty with physics-augmented neural networks: model formulation and training methods without prescribed internal vari- ables.Computational Mechanics, 74(6):1279–1301, 2024

    work page 2024

  61. [61]

    Extreme sparsification of physics-augmented neural networks for interpretable model discovery in mechanics.Computer Methods in Applied Mechanics and Engineering, 426:116973, 2024

    Jan Niklas Fuhg, Reese Edward Jones, and Nikolaos Bouklas. Extreme sparsification of physics-augmented neural networks for interpretable model discovery in mechanics.Computer Methods in Applied Mechanics and Engineering, 426:116973, 2024. 36 Multiscale topology optimization using physics-augmented neural networksA PREPRINT

    work page 2024

  62. [62]

    Polyconvex physics- augmented neural network constitutive models in principal stretches.International Journal of Solids and Struc- tures, page 113469, 2025

    Adrian Buganza Tepole, Asghar Arshad Jadoon, Manuel Rausch, and Jan Niklas Fuhg. Polyconvex physics- augmented neural network constitutive models in principal stretches.International Journal of Solids and Struc- tures, page 113469, 2025

    work page 2025

  63. [63]

    Parametrized polyconvex hyperelasticity with physics-augmented neural networks.Data-Centric Engineering, 4:e25, 2023

    Dominik K Klein, Fabian J Roth, Iman Valizadeh, and Oliver Weeger. Parametrized polyconvex hyperelasticity with physics-augmented neural networks.Data-Centric Engineering, 4:e25, 2023

    work page 2023

  64. [64]

    Finite electro-elasticity with physics-augmented neural networks.Computer Methods in Applied Mechanics and Engineering, 400:115501, 2022

    Dominik K Klein, Rogelio Ortigosa, Jes ´us Mart´ınez-Frutos, and Oliver Weeger. Finite electro-elasticity with physics-augmented neural networks.Computer Methods in Applied Mechanics and Engineering, 400:115501, 2022

    work page 2022

  65. [65]

    Polyconvex neural network models of thermoelasticity.Journal of the Mechanics and Physics of Solids, 192:105837, 2024

    Jan N Fuhg, Asghar Jadoon, Oliver Weeger, D Thomas Seidl, and Reese E Jones. Polyconvex neural network models of thermoelasticity.Journal of the Mechanics and Physics of Solids, 192:105837, 2024

    work page 2024

  66. [66]

    A physics-augmented neural network frame- work for modeling and characterizing thermo-visco-plastic behavior.Computer Methods in Applied Mechanics and Engineering, 456:118954, 2026

    Reese E Jones, Asghar A Jadoon, D Thomas Seidl, and Jan N Fuhg. A physics-augmented neural network frame- work for modeling and characterizing thermo-visco-plastic behavior.Computer Methods in Applied Mechanics and Engineering, 456:118954, 2026

    work page 2026

  67. [67]

    Ordinary and strong ellipticity in the equilibrium theory of incompressible hyper- elastic solids.Archive for Rational Mechanics and Analysis, 83(1):53–90, 1983

    Layne Zee and Eli Sternberg. Ordinary and strong ellipticity in the equilibrium theory of incompressible hyper- elastic solids.Archive for Rational Mechanics and Analysis, 83(1):53–90, 1983

    work page 1983

  68. [68]

    An example of a quasiconvex function that is not polyconvex in two dimensions.Archive for rational mechanics and analysis, 117(2):155–166, 1992

    Jean-Jacques Alibert and Bernard Dacorogna. An example of a quasiconvex function that is not polyconvex in two dimensions.Archive for rational mechanics and analysis, 117(2):155–166, 1992

    work page 1992

  69. [69]

    G Aubert. On a counterexample of a rank 1 convex function which is not polyconvex in the case n= 2.Proceed- ings of the Royal Society of Edinburgh Section A: Mathematics, 106(3-4):237–240, 1987

    work page 1987

  70. [70]

    Convexity conditions and existence theorems in nonlinear elasticity.Archive for rational mechanics and Analysis, 63(4):337–403, 1976

    John M Ball. Convexity conditions and existence theorems in nonlinear elasticity.Archive for rational mechanics and Analysis, 63(4):337–403, 1976

    work page 1976

  71. [71]

    Cambridge university press, 2004

    Stephen P Boyd and Lieven Vandenberghe.Convex optimization. Cambridge university press, 2004

    work page 2004

  72. [72]

    Loss of polyconvexity by homogenization: a new example.Calculus of V ariations and Partial Differential Equations, 30(2):215–230, 2007

    Marco Barchiesi. Loss of polyconvexity by homogenization: a new example.Calculus of V ariations and Partial Differential Equations, 30(2):215–230, 2007

    work page 2007

  73. [73]

    Loss of polyconvexity by homogenization.Archive for rational mechanics and analysis, 127(2):183–190, 1994

    Andrea Braides. Loss of polyconvexity by homogenization.Archive for rational mechanics and analysis, 127(2):183–190, 1994

    work page 1994

  74. [74]

    Learning hyperelastic anisotropy from data via a tensor basis neural network.Journal of the Mechanics and Physics of Solids, 168:105022, 2022

    Jan N Fuhg, Nikolaos Bouklas, and Reese E Jones. Learning hyperelastic anisotropy from data via a tensor basis neural network.Journal of the Mechanics and Physics of Solids, 168:105022, 2022

    work page 2022

  75. [75]

    Neural networks meet anisotropic hypere- lasticity: A framework based on generalized structure tensors and isotropic tensor functions.Comput

    Karl A Kalina, J ¨org Brummund, Waiching Sun, and Markus K¨astner. Neural networks meet anisotropic hypere- lasticity: A framework based on generalized structure tensors and isotropic tensor functions.Comput. Methods Appl. Mech. Eng., 437(117725):117725, March 2025

    work page 2025

  76. [76]

    Pytorch: An imperative style, high-performance deep learning library.Advances in neural information processing systems, 32, 2019

    Adam Paszke, Sam Gross, Francisco Massa, Adam Lerer, James Bradbury, Gregory Chanan, Trevor Killeen, Zeming Lin, Natalia Gimelshein, Luca Antiga, et al. Pytorch: An imperative style, high-performance deep learning library.Advances in neural information processing systems, 32, 2019

    work page 2019

  77. [77]

    Adam: A Method for Stochastic Optimization

    Diederik P Kingma. Adam: A method for stochastic optimization.arXiv preprint arXiv:1412.6980, 2014

    work page internal anchor Pith review Pith/arXiv arXiv 2014

  78. [78]

    Filters in topology optimization.International journal for numerical methods in engineering, 50(9):2143–2158, 2001

    Blaise Bourdin. Filters in topology optimization.International journal for numerical methods in engineering, 50(9):2143–2158, 2001

    work page 2001

  79. [79]

    Bendsoe, m.p.: Optimal shape design as a material distribution problem

    Martin Bendsøe. Bendsoe, m.p.: Optimal shape design as a material distribution problem. structural optimization 1, 193-202.Structural Optimization, 1:193–202, 01 1989

    work page 1989

  80. [80]

    Zhou and G.I.N

    M. Zhou and G.I.N. Rozvany. The coc algorithm, part ii: Topological, geometrical and generalized shape op- timization.Computer Methods in Applied Mechanics and Engineering, 89(1):309–336, 1991. Second World Congress on Computational Mechanics

    work page 1991

Showing first 80 references.