Deep material network via a quilting strategy: visualization for explainability and recursive training for improved accuracy
Source abstract
Abstract
Abstract Recent developments integrating micromechanics and neural networks offer promising paths for rapid predictions of the response of heterogeneous materials with similar accuracy as direct numerical simulations. The deep material network is one such approaches, featuring a multi-layer network and micromechanics building blocks trained on anisotropic linear elastic properties. Once trained, the network acts as a reduced-order model, which can extrapolate the material’s behavior to more general constitutive laws, including nonlinear behaviors, without the need to be retrained. However, current training methods initialize network parameters randomly, incurring inevitable training and calibration errors. Here, we introduce a way to visualize the network parameters as an analogous unit cell and use this visualization to “quilt” patches of shallower networks to initialize deeper networks for a recursive training strategy. The result is an improvement in the accuracy and calibration performance of the network and an intuitive visual representation of the network for better explainability.
Source: crossref · Quality: baseline
Research context
OpenAlex topics: Composite Material Mechanics · Model Reduction and Neural Networks · Machine Learning in Materials Science
SciVX discovery tags are terminology-based. OpenAlex topics are source-provided classifications.
References (64)
Show source-provided references
1. Capuano, G. & Rimoli, J. J. Smart finite elements: a novel machine learning application. Comput. Methods Appl. Mech. Eng. 345, 363–381 (2019).10.1016/j.cma.2018.10.046 ↗
2. Raissi, M., Perdikaris, P. & Karniadakis, G. E. Physics-informed neural networks: a deep learning framework for solving forward and inverse problems involving nonlinear partial differential equations. J. Comput. Phys. 378, 686–707 (2019).10.1016/j.jcp.2018.10.045 ↗
3. Lu, L., Jin, P., Pang, G., Zhang, Z. & Karniadakis, G. E. Learning nonlinear operators via deeponet based on the universal approximation theorem of operators. Nat. Mach. Intell. 3, 218–229 (2021).10.1038/s42256-021-00302-5 ↗
4. Montes de Oca Zapiain, D., Stewart, J. A. & Dingreville, R. Accelerating phase-field-based microstructure evolution predictions via surrogate models trained by machine learning methods. npj Comput. Mater. 7, 3 (2021).10.1038/s41524-020-00471-8 ↗
5. Cuomo, S. et al. Scientific machine learning through physics–informed neural networks: where we are and what’s next. J. Sci. Comput. 92, 88 (2022).10.1007/s10915-022-01939-z ↗
6. Hu, C., Martin, S. & Dingreville, R. Accelerating phase-field predictions via recurrent neural networks learning the microstructure evolution in latent space. Comput. Methods Appl. Mech. Eng. 397, 115128 (2022).10.1016/j.cma.2022.115128 ↗
7. Oommen, V., Shukla, K., Goswami, S., Dingreville, R. & Karniadakis, G. E. Learning two-phase microstructure evolution using neural operators and autoencoder architectures. npj Comput. Mater. 8, 190 (2022).10.1038/s41524-022-00876-7 ↗
8. Feyel, F. & Chaboche, J.-L. FE2 multiscale approach for modelling the elastoviscoplastic behaviour of long fibre SiC/Ti composite materials. Comput. Methods Appl. Mech. Eng. 183, 309–330 (2000).10.1016/S0045-7825(99)00224-8 ↗
9. Moulinec, H. & Suquet, P. A numerical method for computing the overall response of nonlinear composites with complex microstructure. Comput. Methods Appl. Mech. Eng. 157, 69–94 (1998).10.1016/S0045-7825(97)00218-1 ↗
10. Michel, J.-C., Moulinec, H. & Suquet, P. Effective properties of composite materials with periodic microstructure: a computational approach. Comput. Methods Appl. Mech. Eng. 172, 109–143 (1999).10.1016/S0045-7825(98)00227-8 ↗
11. Lebensohn, R. A., Kanjarla, A. K. & Eisenlohr, P. An elasto-viscoplastic formulation based on fast fourier transforms for the prediction of micromechanical fields in polycrystalline materials. Int. J. Plast. 32, 59–69 (2012).10.1016/j.ijplas.2011.12.005 ↗
12. Rastkar, S., Zahedi, M., Korolev, I. & Agarwal, A. A meshfree approach for homogenization of mechanical properties of heterogeneous materials. Eng. Anal. Bound. Elem. 75, 79–88 (2017).10.1016/j.enganabound.2016.12.004 ↗
13. Conti, S., Müller, S. & Ortiz, M. Data-driven finite elasticity. Arch. Ration. Mech. Anal. 237, 1–33 (2020).10.1007/s00205-020-01490-x ↗
14. Minh Nguyen-Thanh, V., Trong Khiem Nguyen, L., Rabczuk, T. & Zhuang, X. A surrogate model for computational homogenization of elastostatics at finite strain using high-dimensional model representation-based neural network. Int. J. Numer. Methods Eng. 121, 4811–4842 (2020).10.1002/nme.6493 ↗
15. Oliver, J., Caicedo, M., Huespe, A. E., Hernández, J. & Roubin, E. Reduced order modeling strategies for computational multiscale fracture. Comput. Methods Appl. Mech. Eng. 313, 560–595 (2017).10.1016/j.cma.2016.09.039 ↗
16. Mozaffar, M. et al. Deep learning predicts path-dependent plasticity. Proc. Natl Acad. Sci. USA 116, 26414–26420 (2019).10.1073/pnas.1911815116 ↗
17. Ferreira, B. P., Pires, F. M. A. & Bessa, M. A. Adaptivity for clustering-based reduced-order modeling of localized history-dependent phenomena. Comput. Methods Appl. Mech. Eng. 393, 114726 (2022).10.1016/j.cma.2022.114726 ↗
18. Yang, H., Qiu, H., Xiang, Q., Tang, S. & Guo, X. Exploring elastoplastic constitutive law of microstructured materials through artificial neural network - A mechanistic-based data-driven approach. J. Appl. Mech. 87, 091005 (2020).10.1115/1.4047208 ↗
19. Ford, E., Maneparambil, K., Rajan, S. & Neithalath, N. Machine learning-based accelerated property prediction of two-phase materials using microstructural descriptors and finite element analysis. Comput. Mater. Sci. 191, 110328 (2021).10.1016/j.commatsci.2021.110328 ↗
20. Masi, F. & Stefanou, I. Multiscale modeling of inelastic materials with thermodynamics-based artificial neural networks (TANN). Comput. Methods Appl. Mech. Eng. 398, 115190 (2022).10.1016/j.cma.2022.115190 ↗
21. Li, X. et al. A transfer learning approach for microstructure reconstruction and structure-property predictions. Sci. Rep. 8, 13461 (2018).10.1038/s41598-018-31571-7 ↗
22. Rao, C. & Liu, Y. Three-dimensional convolutional neural network (3D-CNN) for heterogeneous material homogenization. Comput. Mater. Sci. 184, 109850 (2020).10.1016/j.commatsci.2020.109850 ↗
23. Liu, Z., Bessa, M. A. & Liu, W. K. Self-consistent clustering analysis: an efficient multi-scale scheme for inelastic heterogeneous materials. Comput. Methods Appl. Mech. Eng. 306, 319–341 (2016).10.1016/j.cma.2016.04.004 ↗
24. Bessa, M. A. et al. A framework for data-driven analysis of materials under uncertainty: countering the curse of dimensionality. Comput. Methods Appl. Mech. Eng. 320, 633–667 (2017).10.1016/j.cma.2017.03.037 ↗
25. Liu, Z., Fleming, M. & Liu, W. K. Microstructural material database for self-consistent clustering analysis of elastoplastic strain softening materials. Comput. Methods Appl. Mech. Eng. 330, 547–577 (2018).10.1016/j.cma.2017.11.005 ↗
26. Marshall, A. & Kalidindi, S. R. Autonomous development of a machine-learning model for the plastic response of two-phase composites from micromechanical finite element models. JOM 73, 2085–2095 (2021).10.1007/s11837-021-04696-w ↗
27. Fuhg, J. N., Marino, M. & Bouklas, N. Local approximate Gaussian process regression for data-driven constitutive models: development and comparison with neural networks. Comput. Methods Appl. Mech. Eng. 388, 114217 (2022).10.1016/j.cma.2021.114217 ↗
28. Teichert, G. H. & Garikipati, K. Machine learning materials physics: surrogate optimization and multi-fidelity algorithms predict precipitate morphology in an alternative to phase field dynamics. Comput. Methods Appl. Mech. Eng. 344, 666–693 (2019).10.1016/j.cma.2018.10.025 ↗
29. Hashemi, S. & Kalidindi, S. R. A machine learning framework for the temporal evolution of microstructure during static recrystallization of polycrystalline materials simulated by cellular automaton. Comput. Mater. Sci. 188, 110132 (2021).10.1016/j.commatsci.2020.110132 ↗
30. Liu, Z., Wu, C. T. & Koishi, M. A deep material network for multiscale topology learning and accelerated nonlinear modeling of heterogeneous materials. Comput. Methods Appl. Mech. Eng. 345, 1138–1168 (2019).10.1016/j.cma.2018.09.020 ↗
31. Liu, Z. & Wu, C. T. Exploring the 3D architectures of deep material network in data-driven multiscale mechanics. J. Mech. Phys. Solids 127, 20–46 (2019).10.1016/j.jmps.2019.03.004 ↗
32. Gajek, S., Schneider, M. & Böhlke, T. An FE–DMN method for the multiscale analysis of short fiber reinforced plastic components. Comput. Methods Appl. Mech. Eng. 384, 113952 (2021).10.1016/j.cma.2021.113952 ↗
33. Gajek, S., Schneider, M. & Böhlke, T. On the micromechanics of deep material networks. J. Mech. Phys. Solids 142, 103984 (2020).10.1016/j.jmps.2020.103984 ↗
34. Nguyen, V. D. & Noels, L. Micromechanics-based material networks revisited from the interaction viewpoint; robust and efficient implementation for multi-phase composites. Eur. J. Mech. A Solids 91, 104384 (2022).10.1016/j.euromechsol.2021.104384 ↗
35. Huang, T., Liu, Z., Wu, C. T. & Chen, W. Microstructure-guided deep material network for rapid nonlinear material modeling and uncertainty quantification. Comput. Methods Appl. Mech. Eng. 398, 115197 (2022).10.1016/j.cma.2022.115197 ↗
36. Wu, L., Adam, L. & Noels, L. Micro-mechanics and data-driven based reduced order models for multi-scale analyses of woven composites. Compos. Struct. 270, 114058 (2021).10.1016/j.compstruct.2021.114058 ↗
37. Nguyen, V. D. & Noels, L. Interaction-based material network: a general framework for (porous) microstructured materials. Comput. Methods Appl. Mech. Eng. 389, 114300 (2022).10.1016/j.cma.2021.114300 ↗
38. Gajek, S., Schneider, M. & Böhlke, T. An FE-DMN method for the multiscale analysis of thermomechanical composites. Comput. Mech. 69, 1087–1113 (2022).10.1007/s00466-021-02131-0 ↗
39. Liu, Z. Deep material network with cohesive layers: multi-stage training and interfacial failure analysis. Comput. Methods Appl. Mech. Eng. 363, 112913 (2020).10.1016/j.cma.2020.112913 ↗
40. Beniwal, D. & Ray, P. Learning phase selection and assemblages in high-entropy alloys through a stochastic ensemble-averaging model. Comput. Mater. Sci. 197, 110647 (2021).10.1016/j.commatsci.2021.110647 ↗
41. Scardapane, S. & Wang, D. Randomness in neural networks: an overview. Wiley Interdiscip. Rev. Data Min. Knowl. Discov. 7, e1200 (2017).10.1002/widm.1200 ↗
42. Guo, C., Pleiss, G., Sun, Y. & Weinberger, K. Q. On calibration of modern neural networks. In International Conference on Machine Learning 1321–1330 (PMLR, 2017).
43. Frankle, J. & Carbin, M. The lottery ticket hypothesis: finding sparse, trainable neural networks. Preprint at arXiv:1803.03635 (2018).
44. Minderer, M. et al. Revisiting the calibration of modern neural networks. Adv. Neural Inf. Process. Syst. 34, 15682–15694 (2021).
45. Qin, Z., Yu, F., Liu, C. & Chen, X. How convolutional neural network see the world-A survey of convolutional neural network visualization methods. Math. Found. Comput. 1, 149–180 (2018).10.3934/mfc.2018008 ↗
46. Gao, T. & Zhang, W. A mass constraint formulation for structural topology optimization with multiphase materials. Int. J. Numer. Methods Eng. 88, 774–796 (2011).10.1002/nme.3197 ↗
47. Nguyen, L. M., Liu, J., Scheinberg, K. & Takáč, M. SARAH: a novel method for machine learning problems using stochastic recursive gradient. In International Conference on Machine Learning 2613–2621 (PMLR, 2017).
48. Zhang, W., Zhang, F., Zhang, J., Zhang, J. & Zhang, J. Optimization of identification structure parameters based on recursive maximum likelihood iteration. In 2018 International Computers, Signals and Systems Conference (ICOMSSC) 726–731 (IEEE, 2018).10.1109/ICOMSSC45026.2018.8941760 ↗
49. Irgens, F. Continuum Mechanics (Springer Science & Business Media, 2008).
50. Lebensohn, R. A. et al. Modeling void growth in polycrystalline materials. Acta Mater. 61, 6918–6932 (2013).10.1016/j.actamat.2013.08.004 ↗
51. Stewart, J. A. & Dingreville, R. Microstructure morphology and concentration modulation of nanocomposite thin-films during simulated physical vapor deposition. Acta Mater. 188, 181–191 (2020).10.1016/j.actamat.2020.02.011 ↗
52. Dingreville, R., Stewart, J. A., Chen, E. Y. & Monti, J. M. Benchmark Problems for the Mesoscale Multiphysics Phase Field Simulator (MEMPHIS). Report no. SAND2020-12852 (Sandia National Laboratories, 2020).10.2172/1729722 ↗
53. Drugan, W. J. & Willis, J. R. A micromechanics-based nonlocal constitutive equation and estimates of representative volume element size for elastic composites. J. Mech. Phys. Solids 44, 497–524 (1996).10.1016/0022-5096(96)00007-5 ↗
54. Dingreville, R., Robbins, J. & Voth, T. E. Wave propagation and dispersion in elasto-plastic microstructured materials. Int. J. Solids Struct. 51, 2226–2237 (2014).10.1016/j.ijsolstr.2014.02.030 ↗
55. Alberdi, R., Robbins, J., Walsh, T. & Dingreville, R. Exploring wave propagation in heterogeneous metastructures using the relaxed micromorphic model. J. Mech. Phys. Solids 155, 104540 (2021).10.1016/j.jmps.2021.104540 ↗
56. Michel, J., Moulinec, H. & Suquet, P. A computational scheme for linear and non-linear composites with arbitrary phase contrast. Int. J. Numer. Methods Eng. 52, 139–160 (2001).10.1002/nme.275 ↗
57. Stein, M. Large sample properties of simulations using Latin hypercube sampling. Technometrics 29, 143–151 (1987).10.1080/00401706.1987.10488205 ↗
58. Hill, R. The elastic behaviour of a crystalline aggregate. Proc. Phys. Soc. A 65, 349 (1952).10.1088/0370-1298/65/5/307 ↗
59. Agarap, A. F. Deep learning using rectified linear units (ReLU). Preprint at arXiv:1803.08375 (2018).
60. Paszke, A. et al. Pytorch: an imperative style, high-performance deep learning library. Adv. Neural Inf. Process Syst. 32 (2019).
61. Kingma, D. P. & Ba, J. Adam: a method for stochastic optimization. Preprint at arXiv:1412.6980 (2014).
62. Reddi, S. J., Kale, S. & Kumar, S. On the convergence of Adam and beyond. Preprint at arXiv:1904.09237 (2019).
63. Loshchilov, I. & Hutter, F. Sgdr: Stochastic gradient descent with warm restarts. Preprint at arXiv:1608.03983 (2016).
64. Ypma, T. J. Historical development of the Newton–Raphson method. SIAM Rev. 37, 531–551 (1995).10.1137/1037125 ↗