Papers
Topics
Authors
Recent
Gemini 2.5 Flash
Gemini 2.5 Flash
125 tokens/sec
GPT-4o
53 tokens/sec
Gemini 2.5 Pro Pro
42 tokens/sec
o3 Pro
4 tokens/sec
GPT-4.1 Pro
47 tokens/sec
DeepSeek R1 via Azure Pro
28 tokens/sec
2000 character limit reached

Viscoelasticty with physics-augmented neural networks: Model formulation and training methods without prescribed internal variables (2401.14270v1)

Published 25 Jan 2024 in cs.CE

Abstract: We present an approach for the data-driven modeling of nonlinear viscoelastic materials at small strains which is based on physics-augmented neural networks (NNs) and requires only stress and strain paths for training. The model is built on the concept of generalized standard materials and is therefore thermodynamically consistent by construction. It consists of a free energy and a dissipation potential, which can be either expressed by the components of their tensor arguments or by a suitable set of invariants. The two potentials are described by fully/partially input convex neural networks. For training of the NN model by paths of stress and strain, an efficient and flexible training method based on a recurrent cell, particularly a long short-term memory cell, is developed to automatically generate the internal variable(s) during the training process. The proposed method is benchmarked and thoroughly compared with existing approaches. These include a method that obtains the internal variable by integrating the evolution equation over the entire sequence, while the other method uses an an auxiliary feedforward neural network for the internal variable(s). Databases for training are generated by using a conventional nonlinear viscoelastic reference model, where 3D and 2D plane strain data with either ideal or noisy stresses are generated. The coordinate-based and the invariant-based formulation are compared and the advantages of the latter are demonstrated. Afterwards, the invariant-based model is calibrated by applying the three training methods using ideal or noisy stress data. All methods yield good results, but differ in computation time and usability for large data sets. The presented training method based on a recurrent cell turns out to be particularly robust and widely applicable and thus represents a promising approach for the calibration of other types of models as well.

Definition Search Book Streamline Icon: https://streamlinehq.com
References (72)
  1. Peter Haupt. Continuum Mechanics and Theory of Materials. Springer Berlin Heidelberg, Berlin, Heidelberg, 2000. ISBN 978-3-662-04109-3. doi:10.1007/978-3-662-04775-0.
  2. Computational Methods for Plasticity. John Wiley & Sons, Ltd, Chichester, UK, 2008. ISBN 978-0-470-69452-7. doi:10.1002/9780470694626.
  3. A Review of the Application of Machine Learning and Data Mining Approaches in Continuum Materials Mechanics. Frontiers in Materials, 6:110, 2019. ISSN 2296-8016. doi:10.3389/fmats.2019.00110.
  4. Neural Networks for Constitutive Modeling: From Universal Function Approximators to Advanced Models and the Integration of Physics. Archives of Computational Methods in Engineering, 2023. ISSN 1886-1784. doi:10.1007/s11831-023-10009-y.
  5. Perceptrons: an introduction to computational geometry. The MIT Press, Cambridge (Massachusetts), 1972. ISBN 978-0-262-63022-1. doi:10.7551/mitpress/11301.001.0001.
  6. Knowledge-based modeling of material behavior with neural networks. Journal of Engineering Mechanics, 117(1):132–153, 1991. ISSN 0733-9399, 1943-7889. doi:10.1061/(ASCE)0733-9399(1991)117:1(132).
  7. Implicit constitutive modelling for viscoplasticity using neural networks. International Journal for Numerical Methods in Engineering, 43(2):195–219, 1998. ISSN 0029-5981, 1097-0207. doi:10.1002/(SICI)1097-0207(19980930)43:2<195::AID-NME418>3.0.CO;2-6.
  8. Multiscale methodology for bone remodelling simulation using coupled finite element and neural network computation. Biomechanics and Modeling in Mechanobiology, 10(1):133–145, 2011. ISSN 1617-7959, 1617-7940. doi:10.1007/s10237-010-0222-x.
  9. Artificial neural network modelling to predict hot deformation behaviour of as HIPed FGH4169 superalloy. Materials Science and Technology, 30(10):1170–1176, 2014. ISSN 0267-0836, 1743-2847. doi:10.1179/1743284713Y.0000000411.
  10. A data-driven computational homogenization method based on neural networks for the nonlinear anisotropic electrical response of graphene/polymer nanocomposites. Computational Mechanics, 64(2):307–321, 2019. ISSN 0178-7675, 1432-0924. doi:10.1007/s00466-018-1643-0.
  11. SO(3)-invariance of informed-graph-based deep neural network for anisotropic elastoplastic materials. Computer Methods in Applied Mechanics and Engineering, 363:112875, 2020. ISSN 00457825. doi:10.1016/j.cma.2020.112875.
  12. DNN2: A hyper-parameter reinforcement learning game for self-design of neural network based elasto-plastic constitutive descriptions. Computers & Structures, 249:106505, 2021. ISSN 00457949. doi:10.1016/j.compstruc.2021.106505.
  13. F. Ghavamian and A. Simone. Accelerating multiscale finite element simulations of history-dependent materials using a recurrent neural network. Computer Methods in Applied Mechanics and Engineering, 357:112594, 2019. ISSN 00457825. doi:10.1016/j.cma.2019.112594.
  14. Study on constitutive relation of nickel-base superalloy inconel 718 based on long short term memory recurrent neural network. Metals, 10(12):1588, 2020. ISSN 2075-4701. doi:10.3390/met10121588.
  15. On the importance of self-consistency in recurrent neural network models representing elasto-plastic solids. Journal of the Mechanics and Physics of Solids, 158:104697, 2022. ISSN 00225096. doi:10.1016/j.jmps.2021.104697.
  16. Learning internal representations by error propagation. In Readings in Cognitive Science, pages 399–421. Elsevier, 1988. ISBN 978-1-4832-1446-7. doi:10.1016/B978-1-4832-1446-7.50035-2.
  17. A comparative study on different neural network architectures to model inelasticity. International Journal for Numerical Methods in Engineering, 124(21):4802–4840, 2023. ISSN 0029-5981, 1097-0207. doi:10.1002/nme.7319.
  18. Long short-term memory. Neural Computation, 9(8):1735–1780, 1997. ISSN 0899-7667. doi:10.1162/neco.1997.9.8.1735.
  19. Learning phrase representations using rnn encoder-decoder for statistical machine translation. arXiv, 2014. doi:10.48550/arXiv.1406.1078.
  20. Neural networks meet hyperelasticity: A guide to enforcing physics. Journal of the Mechanics and Physics of Solids, 179:105363, 2023. ISSN 00225096. doi:10.1016/j.jmps.2023.105363.
  21. Thermodynamics-based artificial neural networks for constitutive modeling. Journal of the Mechanics and Physics of Solids, 147:104277, 2021. ISSN 00225096. doi:10.1016/j.jmps.2020.104277.
  22. Physics-informed neural networks: A deep learning framework for solving forward and inverse problems involving nonlinear partial differential equations. Journal of Computational Physics, 378:686–707, 2019. ISSN 00219991. doi:10.1016/j.jcp.2018.10.045.
  23. Physics informed neural networks for continuum micromechanics. Computer Methods in Applied Mechanics and Engineering, 393:114790, 2022. ISSN 00457825. doi:10.1016/j.cma.2022.114790.
  24. A mechanics-informed artificial neural network approach in data-driven constitutive modeling. International Journal for Numerical Methods in Engineering, 123(12):2738–2759, 2022. ISSN 0029-5981, 1097-0207. doi:10.1002/nme.6957.
  25. Finite electro-elasticity with physics-augmented neural networks. Computer Methods in Applied Mechanics and Engineering, 400:115501, 2022a. ISSN 00457825. doi:10.1016/j.cma.2022.115501.
  26. FEANN: An efficient data-driven multiscale approach based on physics-constrained neural networks and automated data mining. Computational Mechanics, 71:827–851, 2023. ISSN 00457825. doi:10.1016/j.cma.2022.115501.
  27. Sobolev training of thermodynamic-informed neural networks for interpretable elasto-plasticity models with level set hardening. Computer Methods in Applied Mechanics and Engineering, 377:113695, 2021a. ISSN 00457825. doi:10.1016/j.cma.2021.113695.
  28. Automated constitutive modeling of isotropic hyperelasticity based on artificial neural networks. Computational Mechanics, 69(1):213–232, 2022. ISSN 0178-7675, 1432-0924. doi:10.1007/s00466-021-02090-6.
  29. Polyconvex anisotropic hyperelasticity with neural networks. Journal of the Mechanics and Physics of Solids, 159:104703, 2022b. ISSN 00225096. doi:10.1016/j.jmps.2021.104703.
  30. FE2 Computations with Deep Neural Networks: Algorithmic Structure, Data Generation, and Implementation. Mathematical and Computational Applications, 28(4):91, 2023. ISSN 2297-8747. doi:10.3390/mca28040091.
  31. Component-based machine learning paradigm for discovering rate-dependent and pressure-sensitive level-set plasticity models. Journal of Applied Mechanics, pages 1–13, 2021b. ISSN 0021-8936. doi:10.1115/1.4052684.
  32. Thermodynamically consistent neural network plasticity modeling and discovery of evolution laws. Journal of the Mechanics and Physics of Solids, 180:105416, 2023. ISSN 0022-5096. doi:10.1016/j.jmps.2023.105416.
  33. Modular machine learning-based elastoplasticity: Generalization in the context of limited data. Computer Methods in Applied Mechanics and Engineering, 407:115930, 2023a. ISSN 0045-7825. doi:10.1016/j.cma.2023.115930.
  34. Learning solutions of thermodynamics-based nonlinear constitutive material models using physics-informed neural networks. Computational Mechanics, 2024. ISSN 0178-7675, 1432-0924. doi:10.1007/s00466-023-02435-3.
  35. Extreme sparsification of physics-augmented neural networks for interpretable model discovery in mechanics. arXiv, 2023b. doi:10.48550/arXiv.2310.03652.
  36. Unsupervised learning of history-dependent constitutive material laws with thermodynamically-consistent neural networks in the modified Constitutive Relation Error framework. working paper or preprint, 2024. URL https://hal.science/hal-04368755.
  37. Data-driven anisotropic finite viscoelasticity using neural ordinary differential equations. Computer Methods in Applied Mechanics and Engineering, 411:116046, 2023. ISSN 00457825. doi:10.1016/j.cma.2023.116046.
  38. Viscoelastic Constitutive Artificial Neural Networks (vCANNs) – a framework for data-driven anisotropic nonlinear finite viscoelasticity. Journal of Computational Physics, 499:112704, 2023. ISSN 0021-9991. doi:10.1016/j.jcp.2023.112704.
  39. A mechanics-informed neural network framework for data-driven nonlinear viscoelasticity. AIAA SCITECH 2023 Forum, 2023. doi:10.2514/6.2023-0949.
  40. Physics-informed data-driven discovery of constitutive models with application to strain-rate-sensitive soft materials. arXiv, 2023. doi:10.48550/arXiv.2304.13897.
  41. Equivariant geometric learning for digital rock physics: estimating formation factor and effective permeability tensors from morse graph. International Journal for Multiscale Computational Engineering, 21(5):1–24, 2023. ISSN 1543-1649. doi:10.1615/IntJMultCompEng.2022042266.
  42. Constitutive artificial neural networks: A fast and general approach to predictive data-driven constitutive modeling by deep learning. Journal of Computational Physics, 429:110010, 2021. ISSN 00219991. doi:10.1016/j.jcp.2020.110010.
  43. Learning hyperelastic anisotropy from data via a tensor basis neural network. Journal of the Mechanics and Physics of Solids, 168:105022, 2022. ISSN 00225096. doi:10.1016/j.jmps.2022.105022.
  44. Data-driven tissue mechanics with polyconvex neural ordinary differential equations. Computer Methods in Applied Mechanics and Engineering, 398:115248, 2022. ISSN 00457825. doi:10.1016/j.cma.2022.115248.
  45. Geometric deep learning for computational mechanics Part I: anisotropic hyperelasticity. Computer Methods in Applied Mechanics and Engineering, 371:113299, 2020. ISSN 00457825. doi:10.1016/j.cma.2020.113299.
  46. Automatic differentiation in machine learning: a survey. arXiv, 2015. doi:10.48550/arXiv.1502.05767.
  47. Physics-guided neural networks (pgnn): An application in lake temperature modeling. arXiv, 2021. doi:10.48550/arXiv.1710.11431.
  48. NN-EUCLID: Deep-learning hyperelasticity without stress data. Journal of the Mechanics and Physics of Solids, 169:105076, 2022. ISSN 0022-5096. doi:10.1016/j.jmps.2022.105076.
  49. Input convex neural networks. arXiv, 2016. doi:10.48550/arXiv.1609.07152.
  50. Parametrized polyconvex hyperelasticity with physics-augmented neural networks. Data-Centric Engineering, 4:e25, 2023. ISSN 2632-6736. doi:10.1017/dce.2023.21.
  51. Thermodynamically consistent machine-learned internal state variable approach for data-driven modeling of path-dependent materials. Computer Methods in Applied Mechanics and Engineering, 402:115348, 2022. ISSN 00457825. doi:10.1016/j.cma.2022.115348.
  52. Physically enhanced training for modeling rate-independent plasticity with feedforward neural networks. Computational Mechanics, 72, 2023. ISSN 1432-0924. doi:10.1007/s00466-023-02316-9.
  53. A Hybrid Approach Employing Neural Networks to Simulate the Elasto-Plastic Deformation Behavior of 3D-Foam Structures. Advanced Engineering Materials, 24(2):2100641, 2021. ISSN 1527-2648. doi:10.1002/adem.202100641.
  54. Multiscale modeling of inelastic materials with Thermodynamics-based Artificial Neural Networks (TANN). Computer Methods in Applied Mechanics and Engineering, 398:115190, 2022. ISSN 00457825. doi:10.1016/j.cma.2022.115190.
  55. Geometric deep learning for computational mechanics Part II: Graph embedding for interpretable multiscale plasticity. Computer Methods in Applied Mechanics and Engineering, 404:115768, 2023. ISSN 0045-7825. doi:10.1016/j.cma.2022.115768.
  56. Variational onsager neural networks (VONNs): A thermodynamics-based variational learning strategy for non-equilibrium PDEs. Journal of the Mechanics and Physics of Solids, 163:104856, 2022. ISSN 00225096. doi:10.1016/j.jmps.2022.104856.
  57. Theory and implementation of inelastic constitutive artificial neural networks. arXiv, 2023. doi:10.48550/arXiv.2311.06380.
  58. An incremental variational formulation of dissipative magnetostriction at the macroscopic continuum level. International Journal of Solids and Structures, 48(13):1846–1866, 2011. ISSN 00207683. doi:10.1016/j.ijsolstr.2011.02.011.
  59. Homogenization of inelastic solid materials at ÿnite strains based on incremental minimization principles. Application to the texture analysis of polycrystals. Journal of the Mechanics and Physics of Solids, 50(10):2123–2167, 2002. doi:https://doi.org/10.1016/S0022-5096(02)00016-9.
  60. Christian Miehe. Strain-driven homogenization of inelastic microstructures and composites based on an incremental variational formulation. International Journal for Numerical Methods in Engineering, 55(11):1285–1322, 2002. ISSN 0029-5981. doi:10.1002/nme.515.
  61. Alexander Mielke. A Mathematical Framework for Generalized Standard Materials in the Rate-Independent Case. In Multifield Problems in Solid and Fluid Mechanics, volume 28, pages 399–428. Springer Berlin Heidelberg, 2006. ISBN 978-3-540-34959-4. doi:10.1007/978-3-540-34961-7_12.
  62. The Derivation of Constitutive Relations from the Free Energy and the Dissipation Function. In Advances in Applied Mechanics, volume 25, pages 183–238. Elsevier, 1987. ISBN 978-0-12-002025-6. doi:10.1016/S0065-2156(08)70278-3.
  63. J.R. Rice. Inelastic constitutive relations for solids: An internal-variable theory and its application to metal plasticity. Journal of the Mechanics and Physics of Solids, 19(6):433–455, 1971. ISSN 00225096. doi:10.1016/0022-5096(71)90010-X.
  64. Jean Jacques Moreau. On Unilateral Constraints, Friction and Plasticity. In Gianfranco Capriz and Guido Stampacchia, editors, New Variational Techniques in Mathematical Physics, pages 171–322. Springer Berlin Heidelberg, Berlin, Heidelberg, 2011. ISBN 978-3-642-10960-7. doi:10.1007/978-3-642-10960-7_7.
  65. Sur les matériaux standard généralisés. Journal de Mécanique, 14:39–63, 1975.
  66. Maurice A Biot. Mechanics of incremental deformations. 1965. URL https://hal.science/hal-01352219.
  67. Stress representations for tensor basis neural networks: alternative formulations to finger-rivlin-ericksen. arXiv, 2023c. doi:10.48550/arXiv.2308.11080.
  68. A generic physics-informed neural network-based constitutive model for soft biological tissues. Computer Methods in Applied Mechanics and Engineering, 372:113402, 2020. ISSN 00457825. doi:10.1016/j.cma.2020.113402.
  69. Automated discovery of generalized standard material models with EUCLID. Computer Methods in Applied Mechanics and Engineering, 405:115867, 2023. ISSN 00457825. doi:10.1016/j.cma.2022.115867.
  70. Physics-constrained symbolic model discovery for polyconvex incompressible hyperelastic materials. arXiv, 2023. doi:10.48550/arXiv.2310.04286.
  71. Neural network-based multiscale modeling of finite strain magneto-elasticity with relaxed convexity criteria. Computer Methods in Applied Mechanics and Engineering, 421:116739, 2024. doi:https://doi.org/10.1016/j.cma.2023.116739.
  72. Inelastic material behavior of polymers – Experimental characterization, formulation and implementation of a material model. Mechanics of Materials, 52:40–57, 2012. ISSN 01676636. doi:10.1016/j.mechmat.2012.04.011.
Citations (14)
View on Semantic Scholar

Summary

We haven't generated a summary for this paper yet.

Ai Sparkles Streamline Icon: https://streamlinehq.com Summarize Now