Deep Learning Prediction of Crack Propagation in Ductile Materials using Surface Displacements Fields
Lahouari BENABOU
Abstract. This study presents a deep learning framework for predicting incremental crack propagation in ductile materials using only surface displacement measurements. A synthetic dataset is generated via the eXtended Finite Element Method (XFEM), enabling the simulation of diverse crack paths under various mixed-mode loading conditions on an Arcan specimen made of aluminum alloy 6061-T6. From each simulation, sequences of displacement fields (ux,uy) are extracted over a regular grid within the region of interest, together with the corresponding crack-path maps, to form the training dataset. Three convolutional-recurrent architectures are developed in Python using the Keras library to capture the spatiotemporal features of crack growth. Model 1 (CNN-BiLSTM) processes displacement fields through a single branch. Model 2 (Dual-Branch CNN-BiLSTM) incorporates an explicit damage branch fed with the current crack-path map (downsampled via MaxPooling2D). Model 3 (ConvLSTM) employs native spatiotemporal convolution. All models regress the next crack tip position, enabling iterative full-path prediction from displacement sequences alone. Comparative evaluation on unseen mixed-mode cases demonstrates the potential of the proposed approach for non-intrusive fracture assessment.
Keywords
Deep Learning, Crack Propagation, CNN, LSTM
Published online 11/8/2026, 4 pages
Copyright © 2026 by the author(s)
Published under license by Materials Research Forum LLC., Millersville PA, USA
Citation: Lahouari BENABOU, Deep Learning Prediction of Crack Propagation in Ductile Materials using Surface Displacements Fields, Materials Research Proceedings, Vol. 73, pp 270-273, 2026
DOI: https://doi.org/10.21741/9781644904459-52
The article was published as article 52 of the book Mechanics of Advanced Materials and Structures
Content from this work may be used under the terms of the Creative Commons Attribution 3.0 license. Any further distribution of this work must maintain attribution to the author(s) and the title of the work, journal citation and DOI.
References
[1] G.R. Irwin, Onset of fast crack propagation in high strength steel and aluminum alloys, Sagamore Research Conference Proceedings 2 (1956), 289-305. https://doi.org/10.21236/AD0099305
[2] D.S. Dugdale, Yielding in steel sheets containing slits, Journal of the Mechanics and Physics of Solids 8 (1960) 100-104. https://doi.org/10.1016/0022-5096(60)90013-2
[3] M. Schwarzer, B. Rogan, Y. Ruan, et al., Learning to fail: Predicting fracture evolution in brittle material models using recurrent graph convolutional neural networks, Computational Materials Science 162 (2019) 322-332. https://doi.org/10.1016/j.commatsci.2019.02.046
[4] S. Goswami, C. Anitescu, S. Chakraborty, T. Rabczuk, Transfer learning enhanced physics informed neural networks for phase-field modeling of fracture, Theoretical and Applied Fracture Mechanics 106 (2020) 102447. https://doi.org/10.1016/j.tafmec.2019.102447
[5] Y.C. Hsu, C.H. Yu, M.J. Buehler, Using deep learning to predict fracture patterns in crystalline solids, Matter 3 (2020) 197-211. https://doi.org/10.1016/j.matt.2020.04.019
[6] Y. Wang, D. Oyen, W. Guo, et al., StressNet – Deep learning to predict stress with fracture propagation in brittle materials, Materials Degradation 5 (2021). https://doi.org/10.1038/s41529-021-00151-y
[7] N. Moës, J. Dolbow, T. Belytschko, A finite element method for crack growth without remeshing, International journal for Numerical Methods in Engineering 46 (1999), 131-150. https://doi.org/10.1002/(SICI)1097-0207(19990910)46:1<131::AID-NME726>3.0.CO;2-J

