journal article·2026-10Open access

Modeling isotropic polyconvex hyperelasticity by neural networks – sufficient and necessary criteria for compressible and incompressible materials

Gian-Luca Geuken, Patrick Kurzeja, David Wiedemann, Martin Zlatić, Marko Čanađija, Jörn Mosler

Abstract

This work investigates different sufficient and necessary criteria for hyperelastic, isotropic polyconvex material models, focusing on neural network implementations for compressible and incompressible materials. Furthermore, the expressiveness, accuracy, simplicity as well as the efficiency of those models is analyzed. This also enables an assessment of the practical applicability of the models. Convex Signed Singular Value Neural Networks (CSSV-NNs) are applied to compressible materials and tailored to incompressibility (inc-CSSV-NNs), resulting in a universal approximation for frame-indifferent, isotropic polyconvex energies for the compressible as well as incompressible case. While other existing approaches also guarantee frame-indifference, isotropy and polyconvexity, they impose too restrictive constraints and thus limit the expressiveness of the model. This is further substantiated by numerical examples of several, well-established classical models (Neo-Hooke, Mooney-Rivlin, Gent and Arruda-Boyce) and Treloar's experimental data. Moreover, the numerical examples include an explicitly constructed energy function that cannot be approximated by neural networks constrained by Ball's criterion for polyconvexity. This substantiates that Ball's criterion, though sufficient, is not necessary for polyconvexity.
Source: openaire · Quality: candidate unverified · Source record ↗

Research context

OpenAlex topics: Elasticity and Material Modeling · Model Reduction and Neural Networks · Composite Material Mechanics

SciVX discovery tags are terminology-based. OpenAlex topics are source-provided classifications.

References (56)

Show source-provided references
1. Constitutive Kolmogorov–Arnold Networks (CKANs): combining accuracy and interpretability in data-driven material modeling10.1016/j.jmps.2025.106212 ↗
2. Semicontinuity problems in the calculus of variations10.1007/BF00275731 ↗
3. Input convex neural networks
4. A three-dimensional constitutive model for the large stretch behavior of rubber elastic materials10.1016/0022-5096(93)90013-6 ↗
5. A mechanics-informed artificial neural network approach in data-driven constitutive modeling10.1002/nme.6957 ↗
6. Balazi, L., Neumeier, T., Peter, M. A., Peterseim, D., 2025. Neural network enhanced polyconvexification of isotropic energy densities in computational mechanics. https://arxiv.org/abs/2504.06425.
7. Convexity conditions and existence theorems in nonlinear elasticity10.1007/BF00279992 ↗
8. On the calculus of variations and sequentially weakly continuous maps
9. Constitutive inequalities and existence theorems in nonlinear elastostatics
10. Numerical Methods for Nonlinear Partial Differential Equations10.1007/978-3-319-13797-1 ↗
11. Buttazzo · 1994
12. Polyconvex neural networks for hyperelastic constitutive models: a rectification approach10.1016/j.mechrescom.2022.103993 ↗
13. Chen, Y., Shi, Y., Zhang, B., 2019. Optimal control via neural networks: a convex approach.https://arxiv.org/abs/1805.11835.
14. Direct Methods in the Calculus of Variations
15. On the different notions of convexity for rotationally invariant functions10.5802/afst.762 ↗
16. When invariants matter: the role of I1 and I2 in neural network models of incompressible hyperelasticity10.1016/j.mechmat.2025.105443 ↗
17. Thermodynamic relations for high elastic materials10.1039/tf9615700829 ↗
18. A novel neural network for isotropic polyconvex hyperelasticity satisfying the universal approximation theorem10.1016/j.jmps.2025.106209 ↗
19. Incorporating sufficient physical information into artificial neural networks: a guaranteed improvement via physics-based Rao-Blackwellization10.1016/j.cma.2024.116848 ↗
20. Theory and implementation of inelastic constitutive artificial neural networks10.1016/j.cma.2024.117063 ↗
21. A generalized dual potential for inelastic constitutive artificial neural networks: a JAX implementation at finite strains10.1016/j.jmps.2025.106337 ↗
22. Convex potential flows: universal probability distributions with optimal transport and convex optimization
23. Kalina, K. A., Brummund, J., Kästner, M., 2025. A physics-augmented neural network framework for finite strain incompressible viscoelasticity. https://arxiv.org/abs/2511.02959.10.1016/j.cma.2026.118892 ↗
24. Parametrized polyconvex hyperelasticity with physics-augmented neural networks10.1017/dce.2023.21 ↗
25. Kumar · 2022 · 275
26. Mechanical in-silico modeling of orthodontic tooth movement: a review of the boundary value problem10.1177/10812865251369422 ↗
27. Neural network based constitutive model for elastomeric foams10.1016/j.engstruct.2007.12.021 ↗
28. Neural networks meet hyperelasticity: a guide to enforcing physics10.1016/j.jmps.2023.105363 ↗
29. Constitutive artificial neural networks: a fast and general approach to predictive data-driven constitutive modeling by deep learning10.1016/j.jcp.2020.110010 ↗
30. Liu, Z., Wang, Y., Vaidya, S., Ruehle, F., Halverson, J., Soljačić, M., Hou, T. Y., Tegmark, M., 2025. KAN: Kolmogorov-Arnold networks. https://arxiv.org/abs/2404.19756.10.1103/4t7t-v19l ↗
31. Approximation of quasiconvex functions, and lower semicontinuity of multiple integrals10.1007/BF01168345 ↗
32. Quasi-convexity and lower semi-continuity of multiple variational integrals of any order10.1090/S0002-9947-1965-0188838-3 ↗
33. Necessary and sufficient conditions for polyconvexity of isotropic functions
34. Quasi-convexity and the lower semicontinuity of multiple integrals10.2140/pjm.1952.2.25 ↗
35. Multiple Integrals in the Calculus of Variations
36. Moseley · 2022
37. Hypo-elasticity, cauchy-elasticity, corotational stability and monotonicity in the logarithmic strain10.1016/j.jmps.2025.106074 ↗
38. Computational polyconvexification of isotropic functions10.1137/23M1589773 ↗
39. Multiscale modeling meets machine learning: what can we learn?10.1007/s11831-020-09405-5 ↗
40. Elastic modulus of the periodontal ligament10.1016/S0142-9612(97)00021-5 ↗
41. Characterization of convex isotropic functions10.1023/A:1007468902439 ↗
42. On the relation between polyconvexity and rank-one convexity in nonlinear elasticity10.1007/BF00040941 ↗
43. Neural network based constitutive model for rubber material10.5254/1.3547822 ↗
44. Šilhavý · 2002 · 513
45. Principal-stretch-based constitutive neural networks autonomously discover a subclass of Ogden models for human brain tissue10.1016/j.brain.2023.100066 ↗
46. On isotropic, frame-invariant, polyconvex strain-energy functions10.1093/qjmam/56.4.483 ↗
47. Hyperelastic models for rubber-like materials: consistent tangent operators and suitability for Treloar’s data10.1007/s00419-012-0610-z ↗
48. Can KAN CANs? input-convex Kolmogorov-Arnold networks (KANs) as hyperelastic constitutive artificial neural networks (CANs)10.1016/j.cma.2025.118089 ↗
49. Stress-strain data for vulcanised rubber under various types of deformation10.1039/tf9444000059 ↗
50. Truesdell · 1965
51. Consistent machine learning for topology optimization with microstructure-dependent neural network material models10.1016/j.jmps.2024.106015 ↗
52. Wiedemann, D., Peter, M. A., 2023. Characterization of polyconvex isotropic functions. https://arxiv.org/abs/2304.08385.
53. Characterization of polyconvex isotropic functions10.1007/s00526-025-03222-z ↗
54. In search of constitutive conditions in isotropic hyperelasticity: polyconvexity versus true-stress-true-strain monotonicity10.1016/j.jmps.2025.106465 ↗
55. Zaheer, M., Kottur, S., Ravanbakhsh, S., Poczos, B., Salakhutdinov, R., Smola, A., 2018. Deep sets. https://arxiv.org/abs/1703.06114.
56. Recovering Mullins damage hyperelastic behaviour with physics augmented neural networks10.1016/j.jmps.2024.105839 ↗