pith. machine review for the scientific record. sign in

arxiv: 2605.03186 · v1 · submitted 2026-05-04 · 💻 cs.CE

Recognition: unknown

Hybrid Machine Learning and Physical Modeling of Feedstock Deformation During Robotic 3D Printing of Continuous Fiber Thermoplastic Composites

Chady Ghnatios, Kazem Fayazbakhsh

Authors on Pith no claims yet

Pith reviewed 2026-05-08 02:11 UTC · model grok-4.3

classification 💻 cs.CE
keywords 3D printingcontinuous fiber compositesfeedstock deformationhybrid modelingneural ODEviscoelastic modelingrobotic depositioncomposite prepregs
0
0 comments X

The pith

A hybrid model of viscoelastic physics and neural ODEs predicts composite feedstock deformation during robotic 3D printing and generalizes above training temperatures.

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

The paper establishes a hybrid modeling approach to forecast how continuous fiber thermoplastic composite feedstock deforms as it is deposited by a robot. Experiments using DMA and DSC identify residual stress relief, drying, crystallization, and thermal stresses as the main drivers. These are captured by combining a Kelvin-Voigt physical model for viscoelastic response with a stabilized neural ODE for the remaining effects. The resulting model is then tested in actual robotic printing and reproduces observed behavior even when temperatures exceed the training range. A reader would care because accurate deformation prediction supports better path planning and fewer defects in manufacturing strong composite parts.

Core claim

The hybrid model, constructed from DMA and DSC experiments, uses Kelvin-Voigt viscoelastic modeling of the composite prepregs together with a stabilized neural ODE for drying and crystallization. When applied to robotic 3D printing, this model reproduces the prepreg deformation behavior far above the temperatures used in training, demonstrating robustness and generalization capability.

What carries the argument

The stabilized neural ODE coupled to the Kelvin-Voigt viscoelastic equations, which together represent the combined influence of residual stress relief, drying, crystallization, and thermal stresses on feedstock deformation.

If this is right

  • The model enables simulation of deposition paths to reduce manufacturing defects in robotic composite printing.
  • Validation against real printing data confirms the hybrid approach works outside laboratory conditions.
  • Generalization to higher temperatures supports use with different printing parameters without full retraining.
  • The framework can guide optimization of process variables such as speed and temperature to improve part quality.

Where Pith is reading between the lines

These are editorial extensions of the paper, not claims the author makes directly.

  • The same hybrid structure could be adapted to model deformation in other temperature-sensitive additive manufacturing processes.
  • Incorporating real-time sensor feedback into the neural ODE might further improve online path correction during printing.
  • Extending the model to include fiber orientation changes or inter-layer bonding effects would address additional defect sources.

Load-bearing premise

Residual stress relief, drying, crystallization, and thermal stresses are the dominant drivers of deformation, and the hybrid model fitted to lab data captures them well enough to predict behavior in real robotic printing.

What would settle it

A robotic printing trial at a temperature substantially higher than the training range in which the measured feedstock deformation deviates markedly from the model's prediction would falsify the generalization claim.

Figures

Figures reproduced from arXiv: 2605.03186 by Chady Ghnatios, Kazem Fayazbakhsh.

Figure 1
Figure 1. Figure 1: Robotic 3D printing setup used in this work and the feedstock. view at source ↗
Figure 2
Figure 2. Figure 2: Robotic 3D printing of a single tape. tape width of 6.35 mm. While the nominal layer thickness for this feedstock is 0.14 mm, a thicker layer of 0.200 mm was used for the first layer. This is the general practice in 3D printing to have a high-quality first layer and better bonding to the subsequent one and have a direct impact on the tape width. Furthermore, residual view at source ↗
Figure 3
Figure 3. Figure 3: DSC thermograms for LM PAEK-CCF pre-processed feedstock. view at source ↗
Figure 4
Figure 4. Figure 4: Prepreg tape inside the film tensile fixture inside the DMA testing view at source ↗
Figure 5
Figure 5. Figure 5: Strain VS temperature evolution in the performed DMA experiments. view at source ↗
Figure 6
Figure 6. Figure 6: Imposed air temperature T∞ in the DMA experiments as a function of time. The thermal simulation uses the Proper Generalized Decomposition (PGD) method to compute the solution of the differential equation provided in equation (1) [39, 40]. The thermal solution is therefore computed in a separate form using space-time and in-plane-out-of-plane decomposition: T(x, t) = Xn i=1 Fi(x, y) · Zi(z) · Gi(t) (4) The … view at source ↗
Figure 7
Figure 7. Figure 7: Solution of the thermal fields (in K) as a function of time for DMA-1. A view at source ↗
Figure 8
Figure 8. Figure 8: Solution of the thermal fields as a function of time for DMA-1, with view at source ↗
Figure 9
Figure 9. Figure 9: Solution of the mechanical deformation fields at the end of the first view at source ↗
Figure 10
Figure 10. Figure 10: Deformations’ magnitude (m) of the mechanical deformation fields at view at source ↗
Figure 11
Figure 11. Figure 11: Fitted strain in DMA-1 with a maximum temperature of view at source ↗
Figure 12
Figure 12. Figure 12: Solution of the mechanical deformation fields at the end of the first view at source ↗
Figure 13
Figure 13. Figure 13: Deformations’ magnitude (m) of the mechanical deformation fields at view at source ↗
Figure 14
Figure 14. Figure 14: Predicted strains in DMA-2 with a maximum heating temperature of view at source ↗
Figure 15
Figure 15. Figure 15: Neural networks are used to predict ϵ ini . The training of the neural network is performed over the first 80% of the time sequence data available from DMA-1. The remaining 20% are used to validate the results. The network prediction of the initial strain release is illustrated in view at source ↗
Figure 16
Figure 16. Figure 16: Results of the predicted ϵ ini in DMA-1 view at source ↗
Figure 17
Figure 17. Figure 17: Random permutation features importance method on the surrogate view at source ↗
Figure 18
Figure 18. Figure 18: Results of the ϵˆ ini 2 prediction, with the experimental inputs taken from DMA-2. tallization onsets start to germinate just above the crystallization temperature in similar thermoplastics [51, 52, 53]. It is also shown that and a measure maximum in the heat of crystallization, of 22J/g, immediately above the glass transition region is emitted in PEEK [51]. For our LM PAEK-CF material, the glass transiti… view at source ↗
Figure 19
Figure 19. Figure 19: Neural network used for the prediction of the crystallization effect on view at source ↗
Figure 20
Figure 20. Figure 20: Results for the prediction ϵ crys 2 during the first onset of crystallization, with the inputs (∆T, t, ϵs ) from DMA-2 when reaching a temperature T > Tg and remaining above Tg during the first heating cycle. Beyond the first onset of crystallization, the mechanical and thermal properties of the composite prepreg change. Therefore, the surrogates Gc and Hc are trained beyond the first onset of crystalliza… view at source ↗
Figure 21
Figure 21. Figure 21: Predictions of ϵ crys 2 for the complete experiment beyond Tg, with the inputs (∆T, t, ϵs ). T1 is the first heating stage, T2 the first cooling one, T3 the second heating stage and T4 the second cooling one. The relative errors e crys a of the predictions are shown to be 1.65% for the train set and 2.43% for the test set, computed using the following equation: e crys a = 100 × X N i=1 |ϵ crys i − ϵˆ crys… view at source ↗
Figure 22
Figure 22. Figure 22: Comparison of strain as a summation of the three built models with view at source ↗
Figure 23
Figure 23. Figure 23: Top view of the 3D printed tape with the simulation boundary con view at source ↗
Figure 24
Figure 24. Figure 24: Through-thickness thermal fields in the 3D printed tape at the instant view at source ↗
Figure 25
Figure 25. Figure 25: Thermal fields through the thickness of the 3D printed tape at an view at source ↗
Figure 26
Figure 26. Figure 26: Deformed shape of the 3D printed tape, with the outline showing the view at source ↗
Figure 27
Figure 27. Figure 27: Lateral deformation (in meters) of the 3D printed tape as predicted by view at source ↗
read the original abstract

Feedstock deformation during 3D printing of continuous fiber composites is a critical challenge in path planning and a main driver in the generation of manufacturing defects. The proposed work addressed the feedstock deformation during the deposition through several experimental and numerical pathways. The experimental setups and numerical simulations are used to identify the main driving phenomena in the deformation of feedstock through residual stress relief and drying, crystallization, and thermal stresses. A hybrid physics-based and data-driven modeling effort is performed, using Kelvin-Voigt viscoelastic modeling of the composite prepregs and a stabilized neural ODE for the modeling of drying and crystallization. The identified hybrid models from DMA and DSC experiments are used in robotic 3D printing to validate the deposition of a composite prepreg in real printing settings. The results show the ability of the model to reproduce the prepreg behavior far above the temperature used in the training, showcasing its robustness and generalization capability.

Editorial analysis

A structured set of objections, weighed in public.

Desk editor's note, referee report, simulated authors' rebuttal, and a circularity audit. Tearing a paper down is the easy half of reading it; the pith above is the substance, this is the friction.

Referee Report

2 major / 1 minor

Summary. The paper claims that a hybrid model combining Kelvin-Voigt viscoelasticity with a stabilized neural ODE (identified from DMA and DSC experiments on residual stress relief, drying, crystallization, and thermal stresses) can predict feedstock deformation during robotic 3D printing of continuous fiber thermoplastic composites, with validation showing reproduction of prepreg behavior at temperatures well above the training range, demonstrating robustness and generalization.

Significance. If the generalization claim holds with quantitative support, the work could provide a practical tool for defect reduction and path planning in composite additive manufacturing by bridging small-scale material characterization with process-scale predictions.

major comments (2)
  1. [Abstract] Abstract: the central claim of reproducing prepreg behavior 'far above the temperature used in the training' and demonstrating 'robustness and generalization capability' is presented without any quantitative error metrics, statistical measures, temperature deltas, or ablation studies against a pure physics baseline, leaving the evidence for extrapolation under robotic printing conditions (shear rates, time scales, nozzle constraints) weakly supported.
  2. [Validation] Validation section (implied by abstract description of 'used in robotic 3D printing to validate'): the manuscript states that DMA/DSC-derived parameters are applied directly to real printing trials but supplies no details on how residual stress relief, drying, crystallization, and thermal stresses were confirmed as dominant under high-speed deposition versus potential flow-induced or pressure-driven effects, nor any comparison of predicted vs. measured deformation.
minor comments (1)
  1. [Modeling] Notation for the stabilized neural ODE and its coupling to the Kelvin-Voigt model should be clarified with explicit equations and stability constraints to aid reproducibility.

Simulated Author's Rebuttal

2 responses · 0 unresolved

We thank the referee for their constructive and detailed feedback on our manuscript. We address each major comment below and have prepared revisions to strengthen the presentation of our results.

read point-by-point responses

  1. Referee: [Abstract] Abstract: the central claim of reproducing prepreg behavior 'far above the temperature used in the training' and demonstrating 'robustness and generalization capability' is presented without any quantitative error metrics, statistical measures, temperature deltas, or ablation studies against a pure physics baseline, leaving the evidence for extrapolation under robotic printing conditions (shear rates, time scales, nozzle constraints) weakly supported.

    Authors: We agree that the abstract would benefit from quantitative support for the generalization claim. We will revise the abstract to include a concise summary of the key error metrics, temperature ranges, and any baseline comparisons reported in the validation experiments. revision: yes

  2. Referee: [Validation] Validation section (implied by abstract description of 'used in robotic 3D printing to validate'): the manuscript states that DMA/DSC-derived parameters are applied directly to real printing trials but supplies no details on how residual stress relief, drying, crystallization, and thermal stresses were confirmed as dominant under high-speed deposition versus potential flow-induced or pressure-driven effects, nor any comparison of predicted vs. measured deformation.

    Authors: We acknowledge that the validation section requires additional detail to clarify the dominance of the modeled phenomena and to present direct comparisons. We will expand this section to explain the basis for identifying residual stress relief, drying, crystallization, and thermal stresses as primary drivers under the deposition conditions, and to include quantitative predicted-versus-measured deformation results from the robotic printing trials. revision: yes

Circularity Check

0 steps flagged

No significant circularity: hybrid model parameters identified from independent DMA/DSC data and validated on separate printing trials

full rationale

The paper identifies Kelvin-Voigt viscoelastic parameters and neural ODE terms for drying/crystallization directly from DMA and DSC experimental datasets. These fitted components are then applied without re-fitting to robotic 3D printing deposition trials as an external validation step. No equation in the provided abstract or reader's summary reduces a claimed prediction to the fitting inputs by algebraic identity, and no load-bearing uniqueness theorem or ansatz is imported via self-citation. The generalization claim (reproduction above training temperature) rests on the physical separation between calibration experiments and printing validation rather than on any self-referential construction.

Axiom & Free-Parameter Ledger

0 free parameters · 0 axioms · 0 invented entities

Abstract provides insufficient detail to enumerate specific free parameters, axioms, or invented entities; no equations or fitting procedures are shown.

pith-pipeline@v0.9.0 · 5463 in / 1085 out tokens · 44006 ms · 2026-05-08T02:11:21.790556+00:00 · methodology

Review history (2 revisions) →

discussion (0)

Sign in with ORCID, Apple, or X to comment. Anyone can read and Pith papers without signing in.

Reference graph

Works this paper leans on

57 extracted references · 35 canonical work pages

  1. [1]

    Polymers (Basel)17(12), 1601 (2025)

    Zheng, H., Zhu, S., Chen, L., Wang, L., Zhang, H., Wang, P., Sun, K., Wang, H., Liu, C.: 3D printing continuous fiber reinforced polymers: A review of material selection, process, and mechanics-function integration for targeted applications. Polymers (Basel)17(12), 1601 (2025)

    2025

  2. [2]

    Polymer Composites46(14), 12619–12649 (2025)

    Zhang, Y., Zheng, W., Wang, Y., Ma, K., Feng, X., Ji, Q., Guo, W., Lu, B.: A review of 3d printing continuous carbon fiber reinforced thermoplastic polymers: Materials, processes, performance enhancement, and failure analysis. Polymer Composites46(14), 12619–12649 (2025). doi:10.1002/pc.29895. https://4spepublications.onlinelibrary.wiley.com/doi/pdf/10.10...

    work page doi:10.1002/pc.29895 2025

  3. [3]

    Composite Structures232, 111476 (2020)

    Kabir, S.M.F., Mathur, K., Seyam, A.-F.M.: A critical review on 3d printed continuous fiber-reinforced composites: History, mechanism, materials and properties. Composite Structures232, 111476 (2020). doi:10.1016/j.compstruct.2019.111476

    work page doi:10.1016/j.compstruct.2019.111476 2020

  4. [4]

    Composites Science and Technology267, 111209 (2025)

    Ouyang, S., Li, D., Zhu, W., Fu, L., Zhang, Z., Wang, N., Zhi, Q.: Process modeling and deformation prediction of 3d printed continuous fiber-reinforced composites based on in-situ micro-scale measuring. Composites Science and Technology267, 111209 (2025). doi:10.1016/j.compscitech.2025.111209

    work page doi:10.1016/j.compscitech.2025.111209 2025

  5. [5]

    Additive Manufacturing55, 102796 (2022)

    Ghnatios, C., Fayazbakhsh, K.: Warping estimation of continuous fiber-reinforced composites made by robotic 3d printing. Additive Manufacturing55, 102796 (2022). doi:10.1016/j.addma.2022.102796

    work page doi:10.1016/j.addma.2022.102796 2022

  6. [6]

    Additive Manufacturing102, 104749 (2025)

    Diaz, M., Baur, J.W.: Warpage of thin additively manufactured continuous fiber thermoset composites. Additive Manufacturing102, 104749 (2025). doi:10.1016/j.addma.2025.104749

    work page doi:10.1016/j.addma.2025.104749 2025

  7. [7]

    International Journal of Advanced Manufacturing Technology104(5-8), 1661–1676 (2019)

    El Moumen, A., Tarfaoui, M., Lafdi, K.: Modelling of the temperature and residual stress fields during 3D printing of polymer composites. International Journal of Advanced Manufacturing Technology104(5-8), 1661–1676 (2019)

    2019

  8. [8]

    Composites Part A: Applied Science and Manufacturing188, 108519 (2025)

    Fereidouni, M., Hoa, S.V.: Transverse squeeze flow of thermoplastic composite tape during in-situ consolidation via automated fiber placement. Composites Part A: Applied Science and Manufacturing188, 108519 (2025). doi:10.1016/j.compositesa.2024.108519

    work page doi:10.1016/j.compositesa.2024.108519 2025

  9. [9]

    Miri, S., Rana, J., Fayazbakhsh, K., Ghnatios, C.: Numerical and experimental study of the consolidation of continuous carbon fiber thermoplastics made by robotic 3D printing. Int. J. Mater. Form.18(1) (2025)

    2025

  10. [10]

    Advanced Manufacturing: Polymer & Composites Science9(1), 2171940 (2023)

    Jiang, B., Chen, Y., Ye, L., Chang, L., Dong, H.: Residual stress and warpage of additively manufactured scf/pla composite parts. Advanced Manufacturing: Polymer & Composites Science9(1), 2171940 (2023). doi:10.1080/20550340.2023.2171940

    work page doi:10.1080/20550340.2023.2171940 2023

  11. [11]

    Journal of Manufacturing Processes70, 389–399 (2021)

    Samy, A.A., Golbang, A., Harkin-Jones, E., Archer, E., Tormey, D., McIlhagger, A.: Finite element analysis of residual stress and warpage in a 3d printed semi-crystalline polymer: Effect of ambient temperature and nozzle speed. Journal of Manufacturing Processes70, 389–399 (2021). doi:10.1016/j.jmapro.2021.08.054

    work page doi:10.1016/j.jmapro.2021.08.054 2021

  12. [12]

    Pourali, M., Adisa, A., Salunke, S., Peterson, A.M.: Crystallization modeling of two semi-crystalline polyamides during material extrusion additive manufacturing. Sci. Rep.14(1), 26297 (2024)

    2024

  13. [13]

    Journal of Materials Processing Technology316, 117961 (2023)

    Malagutti, L., Charlon, S., Mazzanti, V., Mollica, F.: Effects of printed bead volume on thermal history, polymer degree of crystallinity and mechanical properties in large scale additive manufacturing. Journal of Materials Processing Technology316, 117961 (2023). doi:10.1016/j.jmatprotec.2023.117961

    work page doi:10.1016/j.jmatprotec.2023.117961 2023

  14. [14]

    Advances in Industrial and Manufacturing Engineering4, 100085 (2022)

    Lee, A., Wynn, M., Quigley, L., Salviato, M., Zobeiry, N.: Effect of temperature history during additive manufacturing on crystalline morphology of peek. Advances in Industrial and Manufacturing Engineering4, 100085 (2022). doi:10.1016/j.aime.2022.100085

    work page doi:10.1016/j.aime.2022.100085 2022

  15. [15]

    part i: A theory that accounts for phase change, shrinkage, warpage and residual stress

    Sreejith, P., Kannan, K., Rajagopal, K.R.: A thermodynamic framework for the additive manufacturing of crystallizing polymers. part i: A theory that accounts for phase change, shrinkage, warpage and residual stress. International Journal of Engineering Science183, 103789 (2023). doi:10.1016/j.ijengsci.2022.103789

    work page doi:10.1016/j.ijengsci.2022.103789 2023

  16. [16]

    Composites Part A: Applied Science and Manufacturing200, 109310 (2026)

    Wang, Y., Xin, Z., Yuan, J., Duan, Y., Xiao, H., Yang, F., Zhang, D., Li, F.: Through-thickness crystallinity gradient controls warpage reduction in cf/paek via in-situ consolidation automated fiber placement. Composites Part A: Applied Science and Manufacturing200, 109310 (2026). doi:10.1016/j.compositesa.2025.109310

    work page doi:10.1016/j.compositesa.2025.109310 2026

  17. [17]

    Additive Manufacturing38, 101775 (2021)

    Zhang, H., Chen, J., Yang, D.: Fibre misalignment and breakage in 3d printing of continuous carbon fibre reinforced thermoplastic composites. Additive Manufacturing38, 101775 (2021). doi:10.1016/j.addma.2020.101775

    work page doi:10.1016/j.addma.2020.101775 2021

  18. [18]

    Composite Structures376, 119819 (2026)

    Kumar, S., Yoo, D.-H., Song, J.-S., Kim, H.-S.: A critical review of the past, present, and future of 3d printing for continuous and short fiber composites. Composite Structures376, 119819 (2026). doi:10.1016/j.compstruct.2025.119819

    work page doi:10.1016/j.compstruct.2025.119819 2026

  19. [19]

    Virtual and Physical Prototyping19(1), 2412192 (2024)

    Lu, L., Yuan, Y., Xie, Y., Yuan, S., Song, J., Luo, H., Li, Y., Zhu, J., Zhang, W.: Autonomous intelligent additive manufacturing of continuous fiber-reinforced composites: data-enhanced knowledgebase and multi-sensor fusion. Virtual and Physical Prototyping19(1), 2412192 (2024). doi:10.1080/17452759.2024.2412192

    work page doi:10.1080/17452759.2024.2412192 2024

  20. [20]

    Afanasev, A., Höfer, P., Holtmannspötter, J., Zimmer, F., Ehrlich, I.: Development of a continuous fiber-reinforced 3D printing process with a 6-axis robot arm: Process design and equipment. Int. J. Adv. Manuf. Technol. (2026)

    2026

  21. [21]

    Composites Part A: Applied Science and Manufacturing173, 107654 (2023)

    Yadav, N., Schledjewski, R.: Review of in-process defect monitoring for automated tape laying. Composites Part A: Applied Science and Manufacturing173, 107654 (2023). doi:10.1016/j.compositesa.2023.107654

    work page doi:10.1016/j.compositesa.2023.107654 2023

  22. [22]

    Ghnatios, C., Rai, K.E., Hascoet, N., Pires, P.-A., Duval, J.-L., Lambarri, J., Hascoet, J.-Y., Chinesta, F.: Reduced order modeling of selective laser melting: from calibration to parametric part distortion. Int. J. Mater. Form.14(5), 973–986 (2021)

    2021

  23. [23]

    Journal of the Mechanics and Physics of Solids184, 105542 (2024)

    Ghnatios, C., Cazacu, O., Revil-Baudard, B., Chinesta, F.: A new methodology for anisotropic yield surface description using model order reduction techniques and invariant neural network. Journal of the Mechanics and Physics of Solids184, 105542 (2024). doi:10.1016/j.jmps.2024.105542 Ghnatios and Fayazbakhsh Page 29 of 30

    work page doi:10.1016/j.jmps.2024.105542 2024

  24. [24]

    Ghnatios, C., Barasinski, A.: A nonparametric probabilistic method to enhance PGD solutions with data-driven approach, application to the automated tape placement process. Adv. Model. Simul. Eng. Sci.8(1) (2021)

    2021

  25. [25]

    Composites Part B: Engineering309, 113114 (2026)

    Tuo, H., Wang, L., Zhang, J., Zhang, X., Dong, S., Li, X.: The defect quantification in additive manufacturing composites using fused infrared thermography and artificial intelligence. Composites Part B: Engineering309, 113114 (2026). doi:10.1016/j.compositesb.2025.113114

    work page doi:10.1016/j.compositesb.2025.113114 2026

  26. [26]

    Machine Learning with Applications18, 100600 (2024)

    Daghigh, V., Daghigh, H., Lacy, T.E., Naraghi, M.: Review of machine learning applications for defect detection in composite materials. Machine Learning with Applications18, 100600 (2024). doi:10.1016/j.mlwa.2024.100600

    work page doi:10.1016/j.mlwa.2024.100600 2024

  27. [27]

    Polymers17(18) (2025)

    Malashin, I., Martysyuk, D., Tynchenko, V., Gantimurov, A., Nelyub, V., Borodulin, A.: Data-driven optimization of discontinuous and continuous fiber composite processes using machine learning: A review. Polymers17(18) (2025). doi:10.3390/polym17182557

    work page doi:10.3390/polym17182557 2025

  28. [28]

    Data-Centric Engineering6, 23 (2025)

    Ebel, H., van Delden, J., LÃijddecke, T., Borse, A., Gulakala, R., Stoffel, M., Yadav, M., Stender, M., Schindler, L., de Payrebrune, K.M., et al.: Data publishing in mechanics and dynamics: challenges, guidelines, and examples from engineering design. Data-Centric Engineering6, 23 (2025). doi:10.1017/dce.2025.13

    work page doi:10.1017/dce.2025.13 2025

  29. [29]

    Journal of sound and vibration452, 112–131 (2019)

    Clot, A., Meggitt, J.W.R., Langley, R.S., Elliott, A.S., Moorhouse, A.T.: Development of a hybrid fe-sea-experimental model. Journal of sound and vibration452, 112–131 (2019)

    2019

  30. [30]

    Comptes rendus mécaniqueIn Press(2021)

    Ghnatios, C.: A hybrid modeling combining the proper generalized decomposition (pgd) approach to data-driven model learners, with application to non-linear biphasic materials. Comptes rendus mécaniqueIn Press(2021)

    2021

  31. [31]

    Composites Part B: Engineering265, 110958 (2023)

    Ferdousi, S., Advincula, R., Sokolov, A.P., Choi, W., Jiang, Y.: Investigation of 3d printed lightweight hybrid composites via theoretical modeling and machine learning. Composites Part B: Engineering265, 110958 (2023). doi:10.1016/j.compositesb.2023.110958

    work page doi:10.1016/j.compositesb.2023.110958 2023

  32. [32]

    In: Conference on Neural Information Processing Systems (NeurIPS 2018), vol

    Chen, R.T.Q., Rubanova, Y., Bettencourt, J., Duvenaud, D.: Neural ordinary differential equations. In: Conference on Neural Information Processing Systems (NeurIPS 2018), vol. 32. Montreal, Canada, pp. 1–18 (2018)

    2018

  33. [33]

    Deep residual learning for image recognition

    He, K., Zhang, X., Ren, S., Sun, J.: Deep residual learning for image recognition. In: 2016 IEEE Conference on Computer Vision and Pattern Recognition (CVPR), pp. 770–778 (2016). doi:10.1109/CVPR.2016.90

    work page doi:10.1109/cvpr.2016.90 2016

  34. [34]

    Hochreiter and J

    Hochreiter, S., Schmidhuber, J.: Long short-term memory. Neural Computation9(8), 1735–1780 (1997). doi:10.1162/neco.1997.9.8.1735. https://direct.mit.edu/neco/article-pdf/9/8/1735/813796/neco.1997.9.8.1735.pdf

    work page doi:10.1162/neco.1997.9.8.1735 1997

  35. [35]

    CMES - Computer Modeling in Engineering and Sciences143(3), 3033–3064 (2025)

    Shawly, T., Alsheikhy, A.A.: A neural ode-enhanced deep learning framework for accurate and real-time epilepsy detection. CMES - Computer Modeling in Engineering and Sciences143(3), 3033–3064 (2025). doi:10.32604/cmes.2025.065264

    work page doi:10.32604/cmes.2025.065264 2025

  36. [36]

    Energies16(15) (2023)

    Ghnatios, C., Kestelyn, X., Denis, G., Champaney, V., Chinesta, F.: Learning data-driven stable corrections of dynamical systemsâĂŤapplication to the simulation of the top-oil temperature evolution of a power transformer. Energies16(15) (2023). doi:10.3390/en16155790

    work page doi:10.3390/en16155790 2023

  37. [37]

    Archives of Computational Methods in Engineering17(4), 327–350 (2010)

    Chinesta, F., Ammar, A., Cueto, E.: Recent advances in the use of the proper generalized decomposition for solving multidimensional models. Archives of Computational Methods in Engineering17(4), 327–350 (2010)

    2010

  38. [38]

    TEIJIN Datasheet T_Prepreg; TPUD PAEK HTS45 P12 12K. (2022)

    2022

  39. [39]

    Advanced modeling and simulation in engineering sciencessubmitted(2021)

    Ghnatios, C., Barasinski, A.: A nonparametric probabilistic method to enhance pgd solutions with data-driven approach, application to the automated tape placement process. Advanced modeling and simulation in engineering sciencessubmitted(2021)

    2021

  40. [40]

    Computer Methods in Applied Mechanics and Engineering213-216, 29–41 (2012)

    Ghnatios, C., Masson, F., Huerta, A., Leygue, A., Cueto, E., Chinesta, F.: Proper generalized decomposition based dynamic data-driven control of thermal processes. Computer Methods in Applied Mechanics and Engineering213-216, 29–41 (2012). doi:10.1016/j.cma.2011.11.018

    work page doi:10.1016/j.cma.2011.11.018 2012

  41. [41]

    In: VII International Conference on Computational Methods for Coupled Problems in Science and Engineering, Coupled Problems 2017, Rhodes Island, Greece (2017)

    Ghnatios, C., Barasinski, A., Villegas, I., Palardy, G., Chinesta, F.: Computational vamedecum of the coupled mechanical/thermal behavior of composite materials during ultrasonic welding. In: VII International Conference on Computational Methods for Coupled Problems in Science and Engineering, Coupled Problems 2017, Rhodes Island, Greece (2017)

    2017

  42. [42]

    Taylor and Francis group, ??? (1999)

    Jones, R.: Mechanics of Composite Materials, 2nd edn. Taylor and Francis group, ??? (1999)

    1999

  43. [43]

    Journal of Composite Materials3(4), 732–734 (1969)

    Halpin, J.C.: Stiffness and expansion estimates for oriented short fiber composites. Journal of Composite Materials3(4), 732–734 (1969). doi:10.1177/002199836900300419

    work page doi:10.1177/002199836900300419 1969

  44. [44]

    In: Proceedings of the American Society for Composites âĂŞ 36th Technical Conference (2021)

    Kulkarni, S.S., Khan, K.A., Umer, R.: Quasi-linear viscoelastic modelling of uncured prepregs under compaction. In: Proceedings of the American Society for Composites âĂŞ 36th Technical Conference (2021)

    2021

  45. [45]

    Materials12(18), 2853 (2019)

    Xiong, H., Hamila, N., Boisse, P.: Consolidation modeling during thermoforming of thermoplastic composite prepregs. Materials12(18), 2853 (2019)

    2019

  46. [46]

    Polymers16(4), 558 (2024)

    Wang, G., et al.: Preparation and mechanical properties of flexible prepreg resin with high strength and low creep. Polymers16(4), 558 (2024)

    2024

  47. [47]

    Master’s thesis, Virginia Tech (2018)

    Chan, K.J.: Investigation of processing conditions and viscoelastic properties on frictional sliding behavior of unidirectional carbon fiber epoxy prepreg. Master’s thesis, Virginia Tech (2018)

    2018

  48. [48]

    In: Proceedings of the American Society for Composites âĂŞ 38th Technical Conference (2023)

    Wei, Q., Sun, Y., Zhang, D., Enos, R.: An experimentally validated hyper-viscoelastic model for the bending responses of pre-impregnated tapes under processing conditions. In: Proceedings of the American Society for Composites âĂŞ 38th Technical Conference (2023)

    2023

  49. [49]

    In: 2022 7th International Advanced Research Workshop on Transformers (ARWtr), pp

    Kestelyn, X., Denis, G., Champaney, V., Hascoet, N., Ghnatios, C., Chinesta, F.: Towards a hybrid twin for infrastructure asset management: Investigation on power transformer asset maintenance management. In: 2022 7th International Advanced Research Workshop on Transformers (ARWtr), pp. 109–114 (2022). doi:10.23919/ARWtr54586.2022.9959914

    work page doi:10.23919/arwtr54586.2022.9959914 2022

  50. [50]

    Journal of Computational Physics474, 111838 (2023)

    Linot, A.J., Burby, J.W., Tang, Q., Balaprakash, P., Graham, M.D., Maulik, R.: Stabilized neural ordinary differential equations for long-time forecasting of dynamical systems. Journal of Computational Physics474, 111838 (2023). doi:10.1016/j.jcp.2022.111838

    work page doi:10.1016/j.jcp.2022.111838 2023

  51. [51]

    Polymer26(6), 905–912 (1985)

    Kemmish, D.J., Hay, J.N.: The effect of physical ageing on the properties of amorphous peek. Polymer26(6), 905–912 (1985). doi:10.1016/0032-3861(85)90136-3

    work page doi:10.1016/0032-3861(85)90136-3 1985

  52. [52]

    Polymer43(3), 731–735 Ghnatios and Fayazbakhsh Page 30 of 30 (2002)

    Atkinson, J.R., Hay, J.N., Jenkins, M.J.: Enthalpic relaxation in semi-crystalline peek. Polymer43(3), 731–735 Ghnatios and Fayazbakhsh Page 30 of 30 (2002). doi:10.1016/S0032-3861(01)00668-1. Mattice Special Issue

    work page doi:10.1016/s0032-3861(01)00668-1 2002

  53. [53]

    Biomaterials28(32), 4845–4869 (2007)

    Kurtz, S.M., Devine, J.N.: Peek biomaterials in trauma, orthopedic, and spinal implants. Biomaterials28(32), 4845–4869 (2007). doi:10.1016/j.biomaterials.2007.07.013

    work page doi:10.1016/j.biomaterials.2007.07.013 2007

  54. [54]

    Theses, Ecole Centrale de Nantes (ECN) (October 2012).https://theses.hal.science/tel-00867281

    Ghnatios, C.: Simulation avancée des problèmes thermiques rencontrés lors de la mise en forme des composites. Theses, Ecole Centrale de Nantes (ECN) (October 2012).https://theses.hal.science/tel-00867281

    2012

  55. [55]

    John Wiley & Sons, Nashville, TN (2003)

    Donea, J., Huerta, A.: Finite Element Methods for Flow Problems. John Wiley & Sons, Nashville, TN (2003)

    2003

  56. [56]

    Computer Methods in Applied Mechanics and Engineering346, 769–787 (2019)

    Ghnatios, C., Hachem, E.: A stabilized mixed formulation using the proper generalized decomposition for fluid problems. Computer Methods in Applied Mechanics and Engineering346, 769–787 (2019). doi:10.1016/j.cma.2018.09.030

    work page doi:10.1016/j.cma.2018.09.030 2019

  57. [57]

    Polymer Engineering & Science64(8), 3802–3816 (2024)

    Ledesma, R., Wohl, C., Grimsley, B.: Crystallization kinetics analysis and modeling of aerospace paek materials. Polymer Engineering & Science64(8), 3802–3816 (2024). doi:10.1002/pen.26813. https://4spepublications.onlinelibrary.wiley.com/doi/pdf/10.1002/pen.26813

    work page doi:10.1002/pen.26813 2024