Papers
Topics
Authors
Recent
Gemini 2.5 Flash
Gemini 2.5 Flash
156 tokens/sec
GPT-4o
7 tokens/sec
Gemini 2.5 Pro Pro
45 tokens/sec
o3 Pro
4 tokens/sec
GPT-4.1 Pro
38 tokens/sec
DeepSeek R1 via Azure Pro
28 tokens/sec
2000 character limit reached

Multiscale graph neural networks with adaptive mesh refinement for accelerating mesh-based simulations (2402.08863v1)

Published 14 Feb 2024 in cs.CE

Abstract: Mesh-based Graph Neural Networks (GNNs) have recently shown capabilities to simulate complex multiphysics problems with accelerated performance times. However, mesh-based GNNs require a large number of message-passing (MP) steps and suffer from over-smoothing for problems involving very fine mesh. In this work, we develop a multiscale mesh-based GNN framework mimicking a conventional iterative multigrid solver, coupled with adaptive mesh refinement (AMR), to mitigate challenges with conventional mesh-based GNNs. We use the framework to accelerate phase field (PF) fracture problems involving coupled partial differential equations with a near-singular operator due to near-zero modulus inside the crack. We define the initial graph representation using all mesh resolution levels. We perform a series of downsampling steps using Transformer MP GNNs to reach the coarsest graph followed by upsampling steps to reach the original graph. We use skip connectors from the generated embedding during coarsening to prevent over-smoothing. We use Transfer Learning (TL) to significantly reduce the size of training datasets needed to simulate different crack configurations and loading conditions. The trained framework showed accelerated simulation times, while maintaining high accuracy for all cases compared to physics-based PF fracture model. Finally, this work provides a new approach to accelerate a variety of mesh-based engineering multiphysics problems

Definition Search Book Streamline Icon: https://streamlinehq.com
References (69)
  1. M. Ambati, T. Gerasimov, and L. De Lorenzis, “A review on phase-field models of brittle fracture and a new fast hybrid formulation,” Computational Mechanics, vol. 55, pp. 383–405, 2015.
  2. M. Ambati, R. Kruse, and L. De Lorenzis, “A phase-field model for ductile fracture at finite strains and its experimental verification,” Computational Mechanics, vol. 57, pp. 149–167, 2016.
  3. F. Ernesti, M. Schneider, and T. Böhlke, “Fast implicit solvers for phase-field fracture problems on heterogeneous microstructures,” Computer Methods in Applied Mechanics and Engineering, vol. 363, p. 112793, 2020.
  4. S. Goswami, C. Anitescu, and T. Rabczuk, “Adaptive fourth-order phase field analysis for brittle fracture,” Computer Methods in Applied Mechanics and Engineering, vol. 361, p. 112808, 2020.
  5. G. Zhang, T. F. Guo, K. I. Elkhodary, S. Tang, and X. Guo, “Mixed graph-fem phase field modeling of fracture in plates and shells with nonlinearly elastic solids,” Computer Methods in Applied Mechanics and Engineering, vol. 389, p. 114282, 2022.
  6. G. A. Francfort and J.-J. Marigo, “Revisiting brittle fracture as an energy minimization problem,” Journal of the Mechanics and Physics of Solids, vol. 46, no. 8, pp. 1319–1342, 1998.
  7. A. Egger, U. Pillai, K. Agathos, E. Kakouris, E. Chatzi, I. A. Aschroft, and S. P. Triantafyllou, “Discrete and phase field methods for linear elastic fracture mechanics: a comparative study and state-of-the-art review,” Applied Sciences, vol. 9, no. 12, p. 2436, 2019.
  8. J. G. Ribot, V. Agrawal, and B. Runnels, “A new approach for phase field modeling of grain boundaries with strongly nonconvex energy,” Modelling and Simulation in Materials Science and Engineering, vol. 27, no. 8, p. 084007, 2019.
  9. B. Runnels and V. Agrawal, “Phase field disconnections: A continuum method for disconnection-mediated grain boundary motion,” Scripta Materialia, vol. 186, pp. 6–10, 2020.
  10. W. Xu, H. Yu, J. Zhang, C. Lyu, Q. Wang, M. Micheal, and H. Wu, “Phase-field method of crack branching during sc-co2 fracturing: A new energy release rate criterion coupling pore pressure gradient,” Computer Methods in Applied Mechanics and Engineering, vol. 399, p. 115366, 2022.
  11. S. A. Vajari, M. Neuner, P. K. Arunachala, A. Ziccarelli, G. Deierlein, and C. Linder, “A thermodynamically consistent finite strain phase field approach to ductile fracture considering multi-axial stress states,” Computer Methods in Applied Mechanics and Engineering, vol. 400, p. 115467, 2022.
  12. W. Li, M. Ambati, N. Nguyen-Thanh, H. Du, and K. Zhou, “Adaptive fourth-order phase-field modeling of ductile fracture using an isogeometric-meshfree approach,” Computer Methods in Applied Mechanics and Engineering, vol. 406, p. 115861, 2023.
  13. J. Han, S. Matsubara, S. Moriguchi, and K. Terada, “Variational crack phase-field model for ductile fracture with elastic and plastic damage variables,” Computer Methods in Applied Mechanics and Engineering, vol. 400, p. 115577, 2022.
  14. V. Agrawal and B. Runnels, “Robust, strong form mechanics on an adaptive structured grid: efficiently solving variable-geometry near-singular problems with diffuse interfaces,” Computational Mechanics, pp. 1–19, 2023.
  15. V. Agrawal and B. Runnels, “Block structured adaptive mesh refinement and strong form elasticity approach to phase field fracture with applications to delamination, crack branching and crack deflection,” Computer Methods in Applied Mechanics and Engineering, vol. 385, p. 114011, 2021.
  16. Y. Chen and Y. Shen, “A “parallel universe” scheme for crack nucleation in the phase field approach to fracture,” Computer Methods in Applied Mechanics and Engineering, vol. 403, p. 115708, 2023.
  17. S. Brach, E. Tanné, B. Bourdin, and K. Bhattacharya, “Phase-field study of crack nucleation and propagation in elastic–perfectly plastic bodies,” Computer Methods in Applied Mechanics and Engineering, vol. 353, pp. 44–65, 2019.
  18. O. Gültekin, H. Dal, and G. A. Holzapfel, “Numerical aspects of anisotropic failure in soft biological tissues favor energy-based criteria: A rate-dependent anisotropic crack phase-field model,” Computer methods in applied mechanics and engineering, vol. 331, pp. 23–52, 2018.
  19. M. Marulli, A. Valverde-González, A. Quintanas-Corominas, M. Paggi, and J. Reinoso, “A combined phase-field and cohesive zone model approach for crack propagation in layered structures made of nonlinear rubber-like materials,” Computer Methods in Applied Mechanics and Engineering, vol. 395, p. 115007, 2022.
  20. B. Yin and M. Kaliske, “An anisotropic phase-field model based on the equivalent crack surface energy density at finite strain,” Computer Methods in Applied Mechanics and Engineering, vol. 369, p. 113202, 2020.
  21. A. Hunter, B. A. Moore, M. Mudunuru, V. Chau, R. Tchoua, C. Nyshadham, S. Karra, D. O’Malley, E. Rougier, H. Viswanathan, et al., “Reduced-order modeling through machine learning and graph-theoretic approaches for brittle fracture applications,” Computational Materials Science, vol. 157, pp. 87–98, 2019.
  22. A. J. Lew, C.-H. Yu, Y.-C. Hsu, and M. J. Buehler, “Deep learning model to predict fracture mechanisms of graphene,” npj 2D Materials and Applications, vol. 5, no. 1, p. 48, 2021.
  23. B. Euser, E. Rougier, Z. Lei, E. E. Knight, L. P. Frash, J. W. Carey, H. Viswanathan, and A. Munjiza, “Simulation of fracture coalescence in granite via the combined finite–discrete element method,” Rock Mechanics and Rock Engineering, vol. 52, pp. 3213–3227, 2019.
  24. D. Montes de Oca Zapiain, J. A. Stewart, and R. Dingreville, “Accelerating phase-field-based microstructure evolution predictions via surrogate models trained by machine learning methods,” npj Computational Materials, vol. 7, no. 1, p. 3, 2021.
  25. Z. Yang, C.-H. Yu, and M. J. Buehler, “Deep learning model to predict complex stress and strain fields in hierarchical composites,” Science Advances, vol. 7, no. 15, p. eabd7416, 2021.
  26. D. Sharma, V. Pandey, I. V. Singh, S. Natarajan, J. Kumar, and S. Ahmad, “A polygonal fem and continuum damage mechanics based framework for stochastic simulation of fatigue life scatter in duplex microstructure titanium alloys,” Mechanics of Materials, vol. 163, p. 104071, 2021.
  27. L. Zhang and X. Wei, “Prediction of fatigue crack growth under variable amplitude loading by artificial neural network-based lagrange interpolation,” Mechanics of Materials, vol. 171, p. 104309, 2022.
  28. Y. Wang, D. Oyen, W. Guo, A. Mehta, C. Scott, N. Panda, M. Fernández-Godino, G. Srinivasan, and X. Yue, “Stressnet—deep learning to predict stress with fracture propagation in brittle materials. npj mater,” Degrad, vol. 5, no. 1, pp. 1–10, 2021.
  29. S.-i. Amari, “Backpropagation and stochastic gradient descent method,” Neurocomputing, vol. 5, no. 4-5, pp. 185–196, 1993.
  30. L. Bottou, “Large-scale machine learning with stochastic gradient descent,” in Proceedings of COMPSTAT’2010: 19th International Conference on Computational StatisticsParis France, August 22-27, 2010 Keynote, Invited and Contributed Papers, pp. 177–186, Springer, 2010.
  31. N. Ketkar and N. Ketkar, “Stochastic gradient descent,” Deep learning with Python: A hands-on introduction, pp. 113–132, 2017.
  32. N. Black and A. R. Najafi, “Learning finite element convergence with the multi-fidelity graph neural network,” Computer Methods in Applied Mechanics and Engineering, vol. 397, p. 115120, 2022.
  33. C. W. Park, M. Kornbluth, J. Vandermause, C. Wolverton, B. Kozinsky, and J. P. Mailoa, “Accurate and scalable graph neural network force field and molecular dynamics with direct force architecture,” npj Computational Materials, vol. 7, no. 1, p. 73, 2021.
  34. A. Mayr, S. Lehner, A. Mayrhofer, C. Kloss, S. Hochreiter, and J. Brandstetter, “Boundary graph neural networks for 3d simulations,” in Proceedings of the AAAI Conference on Artificial Intelligence, vol. 37, pp. 9099–9107, 2023.
  35. N. N. Vlassis, R. Ma, and W. Sun, “Geometric deep learning for computational mechanics part i: Anisotropic hyperelasticity,” Computer Methods in Applied Mechanics and Engineering, vol. 371, p. 113299, 2020.
  36. N. N. Vlassis and W. Sun, “Geometric deep learning for computational mechanics part ii: Graph embedding for interpretable multiscale plasticity,” arXiv preprint arXiv:2208.00246, 2022.
  37. R. Perera, D. Guzzetti, and V. Agrawal, “Graph neural networks for simulating crack coalescence and propagation in brittle materials,” Computer Methods in Applied Mechanics and Engineering, vol. 395, p. 115021, 2022.
  38. Z. Li and A. B. Farimani, “Graph neural network-accelerated lagrangian fluid simulation,” Computers & Graphics, vol. 103, pp. 201–211, 2022.
  39. R. Bhattoo, S. Ranu, and N. Krishnan, “Learning articulated rigid body dynamics with lagrangian graph neural network,” Advances in Neural Information Processing Systems, vol. 35, pp. 29789–29800, 2022.
  40. Z. Jin, B. Zheng, C. Kim, and G. X. Gu, “Leveraging graph neural networks and neural operator techniques for high-fidelity mesh-based physics simulations,” APL Machine Learning, vol. 1, p. 046109, 11 2023.
  41. C. Jiang and N.-Z. Chen, “Graph neural networks (gnns) based accelerated numerical simulation,” Engineering Applications of Artificial Intelligence, vol. 123, p. 106370, 2023.
  42. J. C. Wong, C. C. Ooi, J. Chattoraj, L. Lestandi, G. Dong, U. Kizhakkinan, D. W. Rosen, M. H. Jhon, and M. H. Dao, “Graph neural network based surrogate model of physics simulations for geometry design,” in 2022 IEEE Symposium Series on Computational Intelligence (SSCI), pp. 1469–1475, 2022.
  43. R. Perera and V. Agrawal, “Dynamic and adaptive mesh-based graph neural network framework for simulating displacement and crack fields in phase field models,” Mechanics of Materials, vol. 186, p. 104789, 2023.
  44. X. Shao, Z. Liu, S. Zhang, Z. Zhao, and C. Hu, “Pignn-cfd: A physics-informed graph neural network for rapid predicting urban wind field defined on unstructured mesh,” Building and Environment, vol. 232, p. 110056, 2023.
  45. T. Pfaff, M. Fortunato, A. Sanchez-Gonzalez, and P. W. Battaglia, “Learning mesh-based simulation with graph networks,” arXiv preprint arXiv:2010.03409, 2020.
  46. J. Gasteiger, J. Groß, and S. Günnemann, “Directional message passing for molecular graphs,” arXiv preprint arXiv:2003.03123, 2020.
  47. F. Scarselli, M. Gori, A. C. Tsoi, M. Hagenbuchner, and G. Monfardini, “The graph neural network model,” IEEE transactions on neural networks, vol. 20, no. 1, pp. 61–80, 2008.
  48. J. Gilmer, S. S. Schoenholz, P. F. Riley, O. Vinyals, and G. E. Dahl, “Neural message passing for quantum chemistry,” in International conference on machine learning, pp. 1263–1272, PMLR, 2017.
  49. Q. Li, Z. Han, and X.-m. Wu, “Deeper insights into graph convolutional networks for semi-supervised learning,” Proceedings of the AAAI Conference on Artificial Intelligence, vol. 32, Apr. 2018.
  50. T. K. Rusch, M. M. Bronstein, and S. Mishra, “A survey on oversmoothing in graph neural networks,” arXiv preprint arXiv:2303.10993, 2023.
  51. C. Cai and Y. Wang, “A note on over-smoothing for graph neural networks,” arXiv preprint arXiv:2006.13318, 2020.
  52. M. Fortunato, T. Pfaff, P. Wirnsberger, A. Pritzel, and P. Battaglia, “Multiscale meshgraphnets,” arXiv preprint arXiv:2210.00612, 2022.
  53. K. Stüben et al., “An introduction to algebraic multigrid,” Multigrid, pp. 413–532, 2001.
  54. J. Xu and L. Zikatanov, “Algebraic multigrid methods,” Acta Numerica, vol. 26, pp. 591–721, 2017.
  55. M. Eliasof and E. Treister, “Diffgcn: Graph convolutional networks via differential operators and algebraic multigrid pooling,” Advances in neural information processing systems, vol. 33, pp. 18016–18027, 2020.
  56. Z. Yang, Y. Dong, X. Deng, and L. Zhang, “Amgnet: Multi-scale graph neural networks for flow field prediction,” Connection Science, vol. 34, no. 1, pp. 2500–2519, 2022.
  57. R. J. Gladstone, H. Rahmani, V. Suryakumar, H. Meidani, M. D’Elia, and A. Zareei, “Gnn-based physics solver for time-independent pdes,” arXiv preprint arXiv:2303.15681, 2023.
  58. W. Liu, M. Yagoubi, and M. Schoenauer, “Multi-resolution graph neural networks for pde approximation,” in Artificial Neural Networks and Machine Learning–ICANN 2021: 30th International Conference on Artificial Neural Networks, Bratislava, Slovakia, September 14–17, 2021, Proceedings, Part III 30, pp. 151–163, Springer, 2021.
  59. I. Luz, M. Galun, H. Maron, R. Basri, and I. Yavneh, “Learning algebraic multigrid using graph neural networks,” in International Conference on Machine Learning, pp. 6489–6499, PMLR, 2020.
  60. M. Lino, C. Cantwell, A. A. Bharath, and S. Fotiadis, “Simulating continuum mechanics with multi-scale graph neural networks,” arXiv preprint arXiv:2106.04900, 2021.
  61. M. Lino, S. Fotiadis, A. A. Bharath, and C. Cantwell, “Towards fast simulation of environmental fluid mechanics with multi-scale graph neural networks,” arXiv preprint arXiv:2205.02637, 2022.
  62. Y. Cao, M. Chai, M. Li, and C. Jiang, “Efficient learning of mesh-based physical simulation with bi-stride multi-scale graph neural network,” in International Conference on Machine Learning, pp. 3541–3558, PMLR, 2023.
  63. H. Gao and S. Ji, “Graph u-nets,” in international conference on machine learning, pp. 2083–2092, PMLR, 2019.
  64. S. Barwey, V. Shankar, V. Viswanathan, and R. Maulik, “Multiscale graph neural network autoencoders for interpretable scientific machine learning,” Journal of Computational Physics, vol. 495, p. 112537, 2023.
  65. J. Yosinski, J. Clune, Y. Bengio, and H. Lipson, “How transferable are features in deep neural networks?,” Advances in neural information processing systems, vol. 27, 2014.
  66. R. Perera and V. Agrawal, “A generalized machine learning framework for brittle crack problems using transfer learning and graph neural networks,” Mechanics of Materials, vol. 181, p. 104639, 2023.
  67. G. Francfort, “Variational fracture: twenty years after,” International Journal of Fracture, vol. 237, no. 1-2, pp. 3–13, 2022.
  68. Y. Shi, Z. Huang, S. Feng, H. Zhong, W. Wang, and Y. Sun, “Masked label prediction: Unified message passing model for semi-supervised classification,” arXiv preprint arXiv:2009.03509, 2020.
  69. A. Vaswani, N. Shazeer, N. Parmar, J. Uszkoreit, L. Jones, A. N. Gomez, Ł. Kaiser, and I. Polosukhin, “Attention is all you need,” Advances in neural information processing systems, vol. 30, 2017.
Citations (5)
View on Semantic Scholar

Summary

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

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