Pith. sign in

REVIEW 5 major objections 6 minor 30 references

Predicting Stress and Damage in Carbon Fiber-Reinforced Composites Deformation Process using Composite U-Net Surrogate Model

T0 review · 5 major / 6 minor · reviewed 2026-08-16 · deepseek-v4-flash

Challenge this review Re-run · record.json Download PDF Read on arXiv ↗ Review JSON

Pith's one-line read Composite-Net simultaneously predicts the full stress and damage history of a carbon-fiber composite under tension, with per-case stress errors below 20 MPa and a speed-up of more than 60 times over the high-fidelity FEM solver.

desk verdict A promising modular surrogate architecture whose central damage-accuracy claim is undermined by two identically-zero equations as printed; worth refereeing after mandatory corrections. read the letter →

arxiv 2504.14143 v1 pith:V7OCC3EV submitted 2025-04-19 cs.LG

Zeping Chen , Marwa Yacouti , Maryam Shakiba , Jian-Xun Wang , Tengfei Luo , Vikas Varshney This is my paper · ORCID
classification cs.LG
keywords carbonfiber-reinforcedcompositesU-NetsurrogatemodelstressfieldpredictiondamageevolutionIGFEMauto-regressivedeeplearning
verification ladder T0 review T1 audit T2 compute T3 formal

The pith

A machine-rendered reading of the paper's core claim, the machinery that carries it, and where it could break.

The reading

Composite-Net is a deep-learning surrogate that takes a 2D microstructure image of a carbon-fiber-reinforced composite and predicts, strain step by strain step, the von Mises stress field (a scalar measure of stress used to judge material yielding) and the damage field as the material is pulled to failure. The paper claims this is the first model to cover the whole deformation history in this setting, including crack initiation and propagation, rather than only the initial state, the final damage, or an averaged stress-strain curve. If the claim is right, failure analysis that currently takes about 490 seconds per microstructure with the high-fidelity IGFEM solver would take about 8 seconds, making large design sweeps over fiber arrangements practical. The accuracy target is stated relative to the simulator that generated the labels: stress RMSE below 20 MPa in all test cases and below 10 MPa in half, with most predicted crack paths within a few pixels of the reference.

What carries the argument

The load-bearing object is the auto-regressive composite U-Net, an ensemble of three U-Nets that specialize by deformation phase. A U-Net is a symmetric convolutional encoder-decoder: the encoder compresses $256\times256$ input fields into a $2048\times1\times1$ code, and the transposed-convolution decoder expands the code back to full resolution, while skip connections between matching encoder and decoder blocks preserve fine spatial features. Damage-Net chains two U-Nets: the first predicts the stress components $\sigma_{11}$, $\sigma_{22}$, and $\sigma_{12}$ at the ultimate tensile stress using a hybrid loss of mean squared error and a physics residual enforcing $\nabla\cdot\sigma=0$, and the second maps those stress components to the final damage pattern using binary cross-entropy. UTS-Net and Necking-Net both take the microstructure, current strain, von Mises stress, and damage state and output increments $d\sigma_V$ and $dD$, which are added to the current fields to advance one strain step; this increment-prediction scheme is what makes long rollouts stable. A hand-off rule switches from UTS-Net to Necking-Net when the macro-stress increase drops below $0.1$ MPa, and Necking-Net stops at the final strain of 1.2%, so the three networks cover loading, peak stress, and post-peak localization in one continuous rollout.

What would settle it

Take a physical CFRC coupon with the same 54-µm square microstructure, 7-µm fibers, and 1.2% tensile loading, record strain fields with digital image correlation and crack growth with acoustic emission or micro-CT, and compare the measured crack path and stress-strain curve to Composite-Net's rollout; if the predicted crack location deviates from the measured one by more than the few-pixel tolerance used in the paper, or the predicted peak stress misses the measured ultimate tensile stress, the surrogate is matching its simulator rather than the material. A cheaper computational falsifier: retrain on IGFEM data generated with a different matrix damage law and check whether the same architecture, without retraining, keeps stress RMSE below 20 MPa.

Watch

Extended reading notes

Core claim

On its own terms, the paper claims that one deep-learning architecture can replace the expensive finite-element step for a 2D unidirectional carbon-fiber composite under uniaxial tension. The model, Composite-Net, consists of three U-Nets: Damage-Net maps a microstructure image to the final damage pattern through two chained U-Nets, the first predicting the stress components $\sigma_{11}$, $\sigma_{22}$, and $\sigma_{12}$ at the ultimate tensile stress with a loss that combines mean squared error with an equilibrium residual enforcing $\nabla\cdot\sigma=0$, and the second converting those stress components into a binary damage field with binary cross-entropy loss. UTS-Net rolls forward from the initial state to the ultimate tensile stress, and Necking-Net rolls from the peak to the final strain of 1.2%. Both rollout networks autoregressively predict increments $d\sigma_V$ and $dD$ at each strain step and add them to the current fields, with a switch triggered when the macro-stress increase falls below 0.1 MPa. On 100 unseen microstructures, the paper reports von Mises stress RMSE below 20 MPa for all cases and below 10 MPa for half, at 8 seconds per case versus 490 seconds for IGFEM.

Load-bearing premise

The central claim stands or falls on whether the IGFEM simulations used to make the training labels are a faithful stand-in for real deformation and failure of this composite, and whether the randomly generated fiber arrangements represent the microstructures the model will meet in use; the paper itself notes in Section 4.3 that the data come from FEM simulations and that generalization to experiments or other material systems would require additional training and validation.

Editorial extensions

If this is right

  • A single microstructure case drops from about 490 seconds of IGFEM simulation to about 8 seconds of network inference, so parametric studies over thousands of fiber arrangements become feasible.
  • Engineers get the complete stress and damage history, not just the final crack, so damage initiation sites and the sequence of propagation are available for failure analysis.
  • The model's accuracy is tied to the fixed constitutive model and uniaxial loading path in the training data; changing loading mode or material system is stated by the authors to require additional training and validation.
  • Damage-Net is the bottleneck: the paper states that improving final-damage prediction accuracy will directly improve the whole Composite-Net, since the downstream networks rely on it.

Reading between the lines

Editorial extensions of the paper, not claims the author makes directly.

  • The paper does not claim this, but the increment-prediction scheme is a natural template for other history-dependent material laws: any constitutive model that advances by small strain steps could be wrapped by the same UTS/Necking split, provided training data exists.
  • The 8-second versus 490-second comparison mixes GPU inference against CPU simulation; a controlled benchmark on identical hardware would isolate how much of the speed-up is algorithmic.
  • A testable extension is to feed the same architecture microstructures with different fiber volume fractions or fiber diameters; the current fixed-fraction, fixed-diameter training set leaves open whether the learned mapping is geometric or material-specific.
  • The crack-path evaluation metric could be applied to experimental micro-CT images of actual cracks, which would test whether the surrogate's failure trajectories match physical damage rather than only simulator damage.
Share X Bluesky LinkedIn Reddit HN

Signed reviews

No signed human review yet.

Editorial analysis

A structured set of objections, weighed in public.

Desk editor's note, referee report, and a circularity audit.

Referee Report

5 major / 6 minor

Summary. The paper proposes Composite-Net, an ensemble of three U-Net models trained on IGFEM simulation data to predict von Mises stress and damage fields in a 2D unidirectional carbon-fiber-reinforced composite under uniaxial tension. Damage-Net maps microstructure to the final damage pattern; UTS-Net and Necking-Net autoregressively predict increments of stress and damage before and after the ultimate tensile stress, respectively. On 100 held-out IGFEM cases, the authors report stress RMSE below 5 MPa for UTS-Net, below 20 MPa for Necking-Net and Composite-Net, and a per-case runtime of 8 seconds versus 490 seconds for IGFEM. The stated contribution is simultaneous spatiotemporal prediction of stress and damage with a speed-up of over 60 times.

Significance. If substantiated, the framework is a useful surrogate for IGFEM-based microstructure analysis and extends prior U-Net work by explicitly modeling pre-UTS, post-UTS, and final-failure stages. Strengths of the paper include the explicit microstructure-generation procedure, the detailed constitutive and material-property descriptions in Appendix A, a held-out test set of 100 cases, and per-case best/average/worst visualizations. However, the quantitative damage evidence currently rests on an equation that is identically zero as printed, the physics-informed loss intended to enforce equilibrium is also vacuous as written, and the abstract reports accuracy numbers that do not appear in the results. These issues must be resolved before the central simultaneous stress-and-damage claim can be accepted. The paper does not release code or data, so the actual training and evaluation configuration cannot be independently checked.

major comments (5)
  1. [Section 2.3.2, Eq. (4)] The physics residual loss as printed is identically zero: the numerator is the squared norm of the difference between the predicted stress divergence and itself, i.e., ||∇·σhat − ∇·σhat||^2 ≡ 0. Therefore the claimed physics-informed training of the first U-Net does not impose any equilibrium constraint. The residual should be defined as ||∇·σhat||^2 (or equivalently as a difference between the predicted divergence and a zero target), and if the quoted loss was actually used in training, the reported stress results may not reflect physics-informed training.
  2. [Section 3, Eq. (8)] The crack-path error metric as printed is also identically zero: the numerator contains \hat{X}_i − \hat{X}_i, i.e., the predicted x-index minus itself, so the reported %RMSE is 0% for every prediction regardless of quality. The surrounding text states that the metric compares predicted and ground-truth x-indices, so the equation should read (X_i − \hat{X}_i)^2. Please correct Eq. (8) and clarify whether the Damage-Net histogram in Fig. 8 was computed with Eq. (8) or with a pixel-wise RMSE; if Eq. (8) was used, the damage-accuracy statements in Section 3.1 are vacuous as printed.
  3. [Abstract and Section 3.4] The abstract claims that the model achieves a stress RMSE below 15 MPa and a damage RMSE below 40% in over 90% of cases, but these numbers do not appear in the results. Section 3.4 reports that 50% of Composite-Net cases have RMSE below 10 MPa and 100% have RMSE below 20 MPa, with no quantitative damage RMSE reported for the full pipeline. Please reconcile the abstract with the results, or add the supporting histograms and percentile values that substantiate the abstract's claim.
  4. [Section 3.4] The central claim is the simultaneous prediction of stress and damage, yet Section 3.4 reports quantitative accuracy only for the von Mises stress fields of Composite-Net. The damage fields of the full Composite-Net rollout are only described qualitatively. Please provide a quantitative damage metric for the full pipeline using the corrected Eq. (8) or another clearly defined metric, so that the 'stress and damage' claim is supported by evidence rather than by visual inspection.
  5. [Section 4.1] The reported 60x speed-up compares IGFEM on CPU (490 s per case) with Composite-Net on GPU (8 s per case). The speed-up claim in the abstract and conclusion should be qualified as a cross-hardware comparison, or same-hardware timings should be reported, because the current comparison conflates algorithmic speed-up with hardware differences.
minor comments (6)
  1. [Section 2.3.4] The text describing Necking-Net's input states that the repeated strain contour is included 'as part of the input for UTS-Net'; this should refer to Necking-Net.
  2. [Figure B.15 caption] The caption contains a typo: 'Composite-Nodel' should be 'Composite-Net'.
  3. [Section 2.3.2, Eq. (6)] The BCE loss expression has a stray hat symbol and unbalanced brackets; please rewrite it in standard form to avoid ambiguity.
  4. [Figure 2 caption] The notation x_epsilon, N_epsilon, x_{epsilon+d epsilon}, and N_{epsilon+d epsilon} is used in the caption but never defined in the surrounding text; please define these symbols.
  5. [Sections 3.1 and 3.4] The statements that worst cases constitute 'less than 5% of the testing dataset' should be quantified explicitly (e.g., 4 of 100 cases) to make the robustness claim precise.
  6. [Data and Code Availability] Saying that data and code are 'provided upon request' without a repository or access mechanism limits reproducibility; a public repository or a detailed training configuration would strengthen the paper.

Circularity Check

2 steps flagged · score 6.0 of 10

No derivation is self-citational, but the damage-path metric (Eq. 8) and the physics residual (Eq. 4) are self-referential by construction, making the damage-accuracy and physics-informed claims empty as printed.

  1. self definitional [Section 3, Eq. (8), crack-path %RMSE definition]
    "% RMSE = sqrt(1/L \sum_{i=1}^{L} ( \hat{X}_i - \hat{X}_i / 256 )^2 ) \times 100% (8) ... X_i and \hat{X}_i represent the ground truth and predicted x-index of the main crack at the y-index i."

    The formula subtracts the predicted x-index from itself, so the numerator is identically zero for every row and every test case. The reported crack-path error is therefore 0% by definition and cannot measure whether the predicted crack matches the ground-truth crack. This is a self-definitional evaluation: the damage-accuracy quantity is defined in terms of the predicted field alone, so the claimed damage-path accuracy is forced to be perfect regardless of model quality.

  2. other [Section 2.3.2, Eq. (4), physics equilibrium residual]
    "Lphysics = 1/N \sum_{n=1}^{N} 1/P_n \sum_{p=1}^{P_n} \| \nabla \cdot \hat{\sigma}^{(n,p)} - \nabla \cdot \hat{\sigma}^{(n,p)} \|^2. (4)"

    The residual is the squared norm of the predicted stress divergence minus itself, hence identically zero for any prediction. The total loss Ltotal = 0.5 LMSE + 0.5 Lphysics reduces to 0.5 LMSE, so the claimed equilibrium constraint is vacuous by construction. The physics term is defined in terms of the same network output on both sides, so it cannot impose any independent physical condition; this is a self-referential loss, not a physics-informed constraint.

full rationale

The core machine-learning pipeline is not circular: Composite-Net, UTS-Net, Necking-Net, and Damage-Net are trained by supervised regression against IGFEM outputs, and the held-out stress RMSE values in Sections 3.2-3.4 compare predictions with independent test-simulation fields. The fiber generator and constitutive model are taken from the authors' prior work ([20], [24]), but those are externally published simulations and verifications, not assumptions that the current surrogate's outputs are true; self-citation here is not load-bearing in the circularity sense. The two genuine by-construction defects are Eq. (8) and Eq. (4): as printed, both are self-referential and vanish identically. Eq. (8) is the only quantitative crack-path metric proposed, so the damage-path accuracy claim is unsupported or definitionally zero; Eq. (4) makes the physics-informed training empty. These are correctness and validity defects in the manuscript rather than a self-citation chain, but they are 'predictions reducing by construction' and justify a partial-circularity score of 6. The Section 4.3 limitation about simulator dependence is explicit and appropriate; it does not create circularity because the paper's stated goal is to build an IGFEM surrogate.

Assumptions & free parameters 5 free parameters · 5 assumptions · 0 invented entities

The central claim rests on the fidelity of the IGFEM ground truth, the representativeness of the generated microstructures, the interpolation to a regular grid, and the trained network weights. The paper introduces no new physical entity; the only invented artifact is the network architecture and its hand-chosen thresholds.

free parameters (5)
  • Neural network weights (Damage-Net, UTS-Net, Necking-Net) = not disclosed (trained on 2000 and 200 FEM cases)
    The surrogate's predictions are entirely produced by weights fitted to FEM outputs; no weight count, architecture parameter count, or final weights are reported.
  • UTS-to-necking switch threshold = d sigma_M < 0.1 MPa
    Hand-chosen criterion for when UTS-Net hands over to Necking-Net; no sensitivity analysis is reported.
  • Crack-path binarization threshold = damage > 0.9
    Hand-chosen threshold for extracting the main crack path in evaluation; affects the reported crack RMSE.
  • Physics loss weight = 0.5
    Equal weighting of MSE and physics loss chosen without ablation; the physics loss itself is written as identically zero in Eq. 4.
  • Strain increment for auto-regressive rollout = not specified
    The step size in applied strain used by UTS-Net and Necking-Net is not reported; this controls the number of rollout steps and error accumulation.
assumptions (5)
  • domain assumption IGFEM simulation output is accurate ground truth for the deformation and failure of this CFRC system.
    Section 2.1 and Appendix A define the constitutive framework; Section 4.3 acknowledges that the model is trained on simulations and may not generalize to experiments without retraining.
  • domain assumption The random fiber generator produces microstructures representative of real CFRC microstructures.
    Section 2.2 states the algorithm was adapted from [20] and previously validated there; no new validation or comparison to experimental microstructures is given.
  • domain assumption The 256 by 256 structured grid interpolation preserves the FEM fields well enough for accurate learning.
    Section 2.2 describes PyVista point-wise interpolation and integration-based averaging, but no resolution or error convergence study is reported.
  • domain assumption A U-Net can learn the incremental mapping from microstructure, strain, stress, and damage to the next increment.
    This is the standard supervised learning assumption behind Sections 2.3.3 and 2.3.4; no formal guarantee exists and performance is empirical.
  • domain assumption Material constants and constitutive parameters in Tables A.1 and A.2 are correct for the CFRC system under study.
    These values are taken from prior literature, such as [24], and are not re-measured in this work.

how reviews work

0 comments
Cite this review

Pith. "Pith review of Predicting Stress and Damage in Carbon Fiber-Reinforced Composites Deformation Process using Composite U-Net Surrogate Model." pith.science (2026). https://pith.science/paper/V7OCC3EV

@misc{pith2026250414143,
  author       = {Pith},
  title        = {Pith review of: Predicting Stress and Damage in Carbon Fiber-Reinforced Composites Deformation Process using Composite U-Net Surrogate Model},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/V7OCC3EV}},
  note         = {Machine review of arXiv:2504.14143}
}
read the original abstract

Carbon fiber-reinforced composites (CFRC) are pivotal in advanced engineering applications due to their exceptional mechanical properties. A deep understanding of CFRC behavior under mechanical loading is essential for optimizing performance in demanding applications such as aerospace structures. While traditional Finite Element Method (FEM) simulations, including advanced techniques like Interface-enriched Generalized FEM (IGFEM), offer valuable insights, they can struggle with computational efficiency. Existing data-driven surrogate models partially address these challenges by predicting propagated damage or stress-strain behavior but fail to comprehensively capture the evolution of stress and damage throughout the entire deformation history, including crack initiation and propagation. This study proposes a novel auto-regressive composite U-Net deep learning model to simultaneously predict stress and damage fields during CFRC deformation. By leveraging the U-Net architecture's ability to capture spatial features and integrate macro- and micro-scale phenomena, the proposed model overcomes key limitations of prior approaches. The model achieves high accuracy in predicting evolution of stress and damage distribution within the microstructure of a CFRC under unidirectional strain, offering a speed-up of over 60 times compared to IGFEM.

Figures

Figures reproduced from arXiv: 2504.14143 by the authors.

Figure 1
Figure 1. (A) The schematic of a microstructural representation of the composite under uniaxial applied displacement with [PITH_FULL_IMAGE:figures/full_fig_p004_1.png] view at source ↗
Figure 2
Figure 2. Schematic representation of Composite-Net, consisting of three interconnected neural networks: Damage-Net (yel [PITH_FULL_IMAGE:figures/full_fig_p007_2.png] view at source ↗
Figure 3
Figure 3. Architecture of U-Net used in Damage-Net, UTS-Net and Necking-Net (Left). The simplified schematic of the U-Net [PITH_FULL_IMAGE:figures/full_fig_p008_3.png] view at source ↗
Figures from the paper (11 more)
Figure 4
Figure 4. Figure 4: Schematic representation of the Damage-Net for mapping microstructure to the final damage pattern. [PITH_FULL_IMAGE:figures/full_fig_p009_4.png]
Figure 5
Figure 5. Figure 5: Schematic representation of UTS-Net for predicting the von Mises stress and damage response from initial condition [PITH_FULL_IMAGE:figures/full_fig_p010_5.png]
Figure 6
Figure 6. Figure 6: Schematic representation of Necking-Net for predicting the von Mises stress and damage response from UTS to final [PITH_FULL_IMAGE:figures/full_fig_p011_6.png]
Figure 7
Figure 7. Figure 7: Post-processing pipeline for extracting the main crack path from predicted damage fields. (A) Ground truth damage [PITH_FULL_IMAGE:figures/full_fig_p013_7.png]
Figure 8
Figure 8. Figure 8: Histogram of RMSE values for the testing dataset using Damage-Net (A). Column B shows the ground truth damage [PITH_FULL_IMAGE:figures/full_fig_p014_8.png]
Figure 9
Figure 9. Figure 9: Rollout predictions and ground truth of stress (top) and damage (bottom) fields from UTS-Net. [PITH_FULL_IMAGE:figures/full_fig_p015_9.png]
Figure 10
Figure 10. Figure 10: Histogram of RMSE for the testing dataset using UTS-Net (A). The red star represents where the case in Figure [PITH_FULL_IMAGE:figures/full_fig_p016_10.png]
Figure 11
Figure 11. Figure 11: Rollout predictions and ground truth of stress (top) and damage (bottom) fields from Necking-Net. [PITH_FULL_IMAGE:figures/full_fig_p017_11.png]
Figure 12
Figure 12. Figure 12: Histogram of RMSE for the testing dataset using Necking-Net (A). The red star represents where the case in Figure [PITH_FULL_IMAGE:figures/full_fig_p018_12.png]
Figure 13
Figure 13. Figure 13: Rollout predictions of stress (top) and damage (bottom) fields for a representative test case using the Composite-Net. [PITH_FULL_IMAGE:figures/full_fig_p019_13.png]
Figure 14
Figure 14. Figure 14: Histogram of RMSE values for the testing dataset using the Composite-Net (A). Column B shows the ground truth [PITH_FULL_IMAGE:figures/full_fig_p020_14.png]

Discussion (0). Continue with ORCID to comment.

Reference graph

Works this paper leans on

30 extracted references · 15 canonical work pages

  1. [1]

    Prathap, Introduction, Springer Netherlands, Dordrecht, 1993, pp

    G. Prathap, Introduction, Springer Netherlands, Dordrecht, 1993, pp. 1–32. doi:10.1007/ 978-94-017-3319-9_1 . URL https://doi.org/10.1007/978-94-017-3319-9_1

    work page doi:10.1007/978-94-017-3319-9_1 1993

  2. [2]

    Zhang, J.-M

    K. Zhang, J.-M. Jin, P . H. Geubelle, An interface-enriched generalized fem for em analysis of com- posites with nonconformal meshes, in: 2016 IEEE/ACES International Conference on Wireless In- formation Technology and Systems (ICWITS) and Applied Computational Electromagnetics (ACES), 2016, pp. 1–2. doi:10.1109/ROPACES.2016.7465468

    work page arXiv 2016

  3. [3]

    Pfaff, M

    T . Pfaff, M. Fortunato, A. Sanchez-Gonzalez, P . W . Battaglia, Learning mesh-based simulation with graph networks, CoRR abs/2010.03409 (2020). arXiv:2010.03409. URL https://arxiv.org/abs/2010.03409

    arXiv 2020

  4. [4]

    X. Fan, D. Akhare, J.-X. Wang, Neural differentiable modeling with diffusion-based super-resolution for two-dimensional spatiotemporal turbulence, Computer Methods in Applied Mechanics and En- gineering 433 (2025) 117478. doi:https://doi.org/10.1016/j.cma.2024.117478. URL https://www.sciencedirect.com/science/article/pii/S0045782524007321

    arXiv 2025

  5. [5]

    X.-Y. Liu, M. H. Parikh, X. Fan, P . Du, Q. Wang, Y.-F . Chen, J.-X. Wang, Confild-inlet: Synthetic turbu- lence inflow using generative latent diffusion models with neural fields (2024).arXiv:2411.14378. URL https://arxiv.org/abs/2411.14378

    arXiv 2024

  6. [6]

    Q. Wang, P . Ren, H. Zhou, X.-Y. Liu, Z. Deng, Y. Zhang, R. Chengze, H. Liu, Z. Wang, J.-X. Wang, J.-R. Wen, H. Sun, Y. Liu, P 2c2net: Pde-preserved coarse correction network for efficient prediction of spatiotemporal dynamics (2024). arXiv:2411.00040. URL https://arxiv.org/abs/2411.00040

    arXiv 2024

  7. [7]

    P . Du, M. H. Parikh, X. Fan, X.-Y. Liu, J.-X. Wang, Conditional neural field latent diffusion model for generating spatiotemporal turbulence, Nature Communications 15 (1) (2024) 10416

    work page 2024

  8. [8]

    R. Li, J. Zhou, J.-X. Wang, T . Luo, Physics-informed bayesian neural networks for solving phonon boltzmann transport equation in forward and inverse problems with sparse and noisy data, ASME Journal of Heat and Mass Transfer (2024) 1–33 arXiv:https://asmedigitalcollection.asme. 31 org/heattransfer/article-pdf/doi/10.1115/1.4067163/7406060/ht-24-1263.pdf ...

    work page doi:10.1115/1.4067163/7406060/ht-24-1263.pdf 2024

Show all 30 references
  1. [9]

    Akhare, T

    D. Akhare, T . Luo, J.-X. Wang, Diffhybrid-uq: Uncertainty quantification for differentiable hybrid neural modeling (2023). arXiv:2401.00161. URL https://arxiv.org/abs/2401.00161

    2023 arXiv

  2. [10]

    S. Kim, T . Luo, E. Lee, I.-S. Suh, Distributed quantum approximate optimization algorithm on inte- grated high-performance computing and quantum computing systems for large-scale optimization (2024). arXiv:2407.20212. URL https://arxiv.org/abs/2407.20212

    2024 arXiv

  3. [11]

    B. P . Croom, M. Berkson, R. K. Mueller, M. Presley, S. Storck, Deep learning prediction of stress fields in additively manufactured metals with intricate defect networks, Mechanics of Materials 165 (2022) 104191. doi:https://doi.org/10.1016/j.mechmat.2021.104191. URL https://...

    2022

  4. [12]

    Akhare, Z

    D. Akhare, Z. Chen, R. Gulotty, T . Luo, J.-X. Wang, Probabilistic physics-integrated neural differen- tiable modeling for isothermal chemical vapor infiltration process, npj Computational Materials 10 (1) (2024) 120

    2024

  5. [13]

    Akhare, T

    D. Akhare, T . Luo, J.-X. Wang, Physics-integrated neural differentiable (pindiff) model for compos- ites manufacturing, Computer Methods in Applied Mechanics and Engineering 406 (2023) 115902. doi:https://doi.org/10.1016/j.cma.2023.115902. URL https://www.sciencedirect.com/sc...

    2023

  6. [14]

    S. M. Estalaki, C. S. Lough, R. G. Landers, E. C. Kinzel, T . Luo, Predicting defects in laser powder bed fusion using in-situ thermal imaging data and machine learning, Additive Manufacturing 58 (2022) 103008. doi:https://doi.org/10.1016/j.addma.2022.103008. URL https://www.s...

    2022

  7. [15]

    J. G. Michopoulos, A. Bhaduri, F . Chinesta, E. Cueto, D. Liu, S. K. Ravi, J.-X. Wang, Spe- cial issue: Scientific machine learning for manufacturing processes and material sys- tems, Journal of Computing and Information Science in Engineering 24 (11) (2024) 110301. 32 arXiv:h...

    2024 doi

  8. [16]

    Maurizi, C

    M. Maurizi, C. Gao, F . Berto, Predicting stress, strain and deformation fields in materials and structures with graph neural networks, Scientific Reports 12 (12 2022). doi:10.1038/ s41598-022-26424-3

    2022

  9. [17]

    Yacouti, M

    M. Yacouti, M. Shakiba, Integrated convolutional and graph neural networks for predicting me- chanical fields in composite microstructures, Composites Part A: Applied Science and Manufactur- ing 190 (2025) 108618. doi:10.1016/j.compositesa.2024.108618. URL https://www.scienced...

    2025

  10. [18]

    M. S. Khorrami, J. R. Mianroodi, N. H. Siboni, P . Goyal, B. Svendsen, P . Benner, D. Raabe, An artifi- cial neural network for surrogate modeling of stress fields in viscoplastic polycrystalline materials (2022). arXiv:2208.13490. URL https://arxiv.org/abs/2208.13490

    2022 arXiv

  11. [19]

    Yacouti, M

    M. Yacouti, M. Shakiba, Performance evaluation of deep learning approaches for predicting mechanical fields in composites, Eng. with Comput. 40 (5) (2024) 3073–3086. doi:10.1007/ s00366-024-01966-4 . URL https://doi.org/10.1007/s00366-024-01966-4

    2024 doi

  12. [20]

    Sepasdar, A

    R. Sepasdar, A. Karpatne, M. Shakiba, A data-driven approach to full-field nonlinear stress distri- bution and failure pattern prediction in composites using deep learning, Computer Methods in Applied Mechanics and Engineering 397 (2022) 115126. doi:https://doi.org/10.1016/j.c...

    2022

  13. [21]

    H. Yan, H. Yu, S. Zhu, Y. Yin, L. Guo, Machine learning based framework for rapid forecasting of the crack propagation, Engineering Fracture Mechanics 307 (2024) 110278. doi:https://doi.org/ 10.1016/j.engfracmech.2024.110278. URL https://www.sciencedirect.com/science/article/p...

    2024

  14. [23]

    Ronneberger, P

    O. Ronneberger, P . Fischer, T . Brox, U-net: Convolutional networks for biomedical image segmenta- tion, CoRR abs/1505.04597 (2015). arXiv:1505.04597. URL http://arxiv.org/abs/1505.04597

    2015 arXiv

  15. [24]

    Sepasdar, M

    R. Sepasdar, M. Shakiba, Micromechanical study of multiple transverse cracking in cross-ply fiber- reinforced composite laminates, Composite Structures 281 (2022) 114986.doi:https://doi.org/ 10.1016/j.compstruct.2021.114986. URL https://www.sciencedirect.com/science/article/pi...

    2022

  16. [25]

    Q. Liu, Z. Qiao, Y. Lv, Pyvt: A python-based open-source software for visualization and graphic anal- ysis of fluid dynamics datasets, Aerospace Science and Technology 117 (2021) 106961. doi:https: //doi.org/10.1016/j.ast.2021.106961. URL https://www.sciencedirect.com/science/...

    2021

  17. [26]

    Ganaie, M

    M. Ganaie, M. Hu, A. Malik, M. Tanveer, P . Suganthan, Ensemble deep learning: A review, Engi- neering Applications of Artificial Intelligence 115 (2022) 105151.doi:10.1016/j.engappai.2022. 105151. URL http://dx.doi.org/10.1016/j.engappai.2022.105151

    2022

  18. [27]

    H. Xu, W . Fan, A. C. Taylor, D. Zhang, L. Ruan, R. Shi, Crack-net: Prediction of crack propagation in composites (2023). arXiv:2309.13626. URL https://arxiv.org/abs/2309.13626

    2023 arXiv

  19. [28]

    W . C. CUI, 5 - fatigue cracking in aged structures, in: J. Paik, R. Melchers (Eds.), Condition Assess- ment of Aged Structures, Woodhead Publishing Series in Civil and Structural Engineering, Wood- head Publishing, 2008, pp. 107–148. doi:https://doi.org/10.1533/9781845695217....

    2008 doi

  20. [29]

    Ortiz, A

    M. Ortiz, A. Pandolfi, Finite-deformation irreversible cohesive elements for three-dimensional crack-propagation analysis, International Journal for Numerical Methods in Engineering 44 (9) (1999) 1267–1282. 34

    1999

  21. [30]

    Sepasdar, M

    R. Sepasdar, M. Shakiba, Overcoming the convergence difficulty of cohesive zone models through a Newton-Raphson modification technique, Engineering Fracture Mechanics (2020) 107046

    2020

  22. [31]

    Y. Gao, A. Bower, A simple technique for avoiding convergence problems in finite element simula- tions of crack nucleation and growth on cohesive interfaces, Modelling and Simulation in Materials Science and Engineering 12 (3) (2004) 453. 35

    2004

Pith tools

Reviewed August 16, 2026 · model on record in the stance chip above.