Pith. sign in

REVIEW 4 major objections 4 minor 55 references

Physics Augmented Machine Learning Discovery of Composition-Dependent Constitutive Laws for 3D Printed Digital Materials

T0 review · 4 major / 4 minor · reviewed 2026-08-06 · deepseek-v4-flash

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

Pith's one-line read A physics-augmented neural network learns composition-dependent stress and torque laws for 3D printed blends, predicting a held-out mixture accurately but not blends beyond the trained composition range

desk verdict Composition-aware pICNN+QLV model with a real held-out interpolation test; the main caveats are the overclaimed polyconvexity guarantee and the unverified large-strain incompressibility assumption. read the letter →

arxiv 2507.02991 v1 pith:CQJM23U7 submitted 2025-07-01 cs.LG physics.comp-ph

Steven Yang , Michal Levin , Govinda Anantha Padmanabha , Miriam Borshevsky , Ohad Cohen , D. Thomas Seidl , Reese E. Jones , Nikolaos Bouklas
show 1 more author
Noy Cohen
This is my paper · ORCID
classification cs.LGphysics.comp-ph
keywords Multi-material3DprintingPhysics-augmentedneuralnetworkPartiallyinputconvexConstitutivemodelingHyperelasticityViscoelasticityDigitalmaterialsL0sparsification
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

The paper sets out to show that a single composition-aware physics-augmented neural network can learn the mechanical response of 3D printed digital materials, a continuum of blends between a soft rubber-like resin and a stiff glassy resin, without fitting a separate constitutive model for each blend. The network learns a hyperelastic strain-energy function that is convex in the strain invariants and depends non-convexly on composition, then feeds that energy into a quasi-linear viscoelastic model whose relaxation coefficient also depends on composition. On experimental data from five blends tested in tension and torsion at multiple rates, the model reproduces the nonlinear, rate-dependent curves and, for a held-out intermediate blend (DM-40), predicts the response with $R^2 \geq 0.96$ in tension and $R^2 = 0.990$ in torsion. The broader goal is a scalable, automated route to constitutive-model discovery for multi-material printing, where the space of possible compositions is effectively continuous.

What carries the argument

The load-bearing mechanism is the partially input convex neural network (pICNN) coupled to a quasi-linear viscoelastic (QLV) model. The pICNN enforces polyconvexity of the strain-energy function with respect to strain invariants while allowing arbitrary non-convex dependence on composition, ensuring thermodynamic admissibility; the QLV layer with relaxation time $\tau = 10$ s fixed and composition-dependent coefficient $\gamma$ (output by an MLP) captures rate dependence; $L_0$ sparsification prunes the trained network to 17+4+11 active parameters, yielding the closed-form strain energy of Eq. 26 and $\gamma(c)$ of Eq. 27.

What would settle it

Measure volume change directly during uniaxial tension to large stretch ($\lambda > 1.2$) and during torsion on a printed DM-50 specimen; if the material dilates or compresses by more than a few percent, the incompressibility assumption ($J=1$) that defines all strain invariants and stresses in Sections 3.3 and 3.4 fails, and the learned energy function does not describe the true material.

Watch

Extended reading notes

Core claim

The central claim is that a partially input convex neural network (pICNN) coupled to a quasi-linear viscoelastic (QLV) kernel can serve as a unified, composition-aware constitutive law for Agilus/Digital-ABS digital materials. Strain invariants $\bar{I}_1$ and $\bar{I}_2$ enter the convex branch, the resin mixing ratio $c$ enters the non-convex branch, and a separate multilayer perceptron maps $c$ to the QLV relaxation coefficient $\gamma$; the resulting model is trained jointly on uniaxial tension at three stretch rates and torsion at 360 deg/min. After $L_0$ sparsification the network shrinks from 1,357+75 weights to 17+4+11 active parameters and yields closed-form expressions for the strain energy (Eq. 26) and $\gamma(c)$ (Eq. 27). The paper establishes that this model achieves high accuracy on training compositions ($R^2 > 0.98$ in 21 of 24 datasets), interpolates well to the held-out DM-40 composition ($R^2 \geq 0.964$ in tension, $R^2 = 0.990$ in torsion), and degrades sharply when extrapolating to the stiffer held-out DM-70 ($R^2$ as low as 0.251 in tension, negative in torsion).

Load-bearing premise

The printed digital materials are assumed to be isotropic and perfectly incompressible ($J=1$) over the full deformation range, based on a Poisson's ratio of about 0.47 measured only at small stretches ($\lambda < 1.05$).

Editorial extensions

If this is right

  • A single trained model replaces per-composition constitutive fitting: once trained on a handful of blends, it supplies stress and torque predictions for any intermediate composition in the training range.
  • The $L_0$-sparsified result is a closed-form strain-energy function and $\gamma(c)$ mapping with only 17 convex, 4 non-convex, and 11 connection parameters, so the discovered law can be written down, inspected, and embedded in simulations.
  • The framework covers both uniaxial tension and torsion within one model, so multiaxial behavior of an interpolated composition can be predicted without new experiments.
  • Extrapolation in composition is not reliable: for held-out DM-70 the model's $R^2$ drops to 0.251 at the slowest tension rate and becomes negative in torsion, so the model's valid range is bounded by the trained composition interval.

Reading between the lines

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

  • A practical consequence the authors leave implicit: to apply this method to a new material family, the training set should bracket the intended composition range with endpoint formulations, because interpolation is reliable but extrapolation is not.
  • Because the torsion tests in this study stayed in the quasi-static regime, the relaxation coefficient $\gamma$ is constrained mainly by tension data; adding higher-rate or cyclic torsion would likely produce a different $\gamma$ and sharpen the viscoelastic branch.
  • The near-linearity of torque vs twist angle alongside strongly nonlinear tension suggests the second strain invariant $\bar{I}_2$ of the learned energy is weakly identified by the current experiments; biaxial or shear tests would pin down the $\bar{I}_2$ dependence.
  • The architecture is not specific to resin jetting; the same pICNN-QLV combination should transfer to any continuously tunable material family, provided the composition input is encoded monotonically with the property of interest.
Share X Bluesky LinkedIn Reddit HN

Editorial analysis

A structured set of objections, weighed in public.

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

Referee Report

4 major / 4 minor

Summary. The paper combines an experimental campaign on five PolyJet digital materials (A, DM-40, DM-50, DM-60, DM-70) under uniaxial tension and torsion at several rates with a physics-augmented neural network model. The model uses a partially input convex neural network (pICNN) to represent a composition-dependent hyperelastic strain energy and a quasi-linear viscoelastic (QLV) formulation with a composition-dependent relaxation coefficient to capture rate dependence. Training is performed on A, DM-50, and DM-60, with DM-40 held out for interpolation and DM-70 for extrapolation. The authors report high R^2 and low sMAPE on training data and on the interpolated DM-40, and they honestly document the failure to extrapolate to DM-70. L0 sparsification reduces the model to closed-form expressions for the energy and relaxation parameter.

Significance. If the physical assumptions behind the kinematic reductions are valid, this is a useful contribution: it demonstrates a scalable, composition-aware framework for constitutive model discovery, provides experimental data for an under-characterized class of printed materials, and includes a rare, clearly reported extrapolation failure that delimits the method's validity. The sparse closed-form expressions in Eqs. (26) and (27) are a concrete interpretability outcome, and the staged training and honest reporting of held-out performance are methodological strengths. The main reservations concern the strength of the 'polyconvexity' guarantee and the kinematic assumptions used to convert raw force/torque measurements into the invariants and stresses that define the training targets.

major comments (4)
  1. [§3.1–3.1.2] The paper states that convexity of the learned strain energy with respect to (I1, I2) 'ensures the polyconvexity conditions required for hyperelastic constitutive modeling.' This is not sufficient. Polyconvexity of an isotropic incompressible energy requires additional structure beyond convexity in the two invariants, typically monotonicity conditions on the derivatives ∂Ψ/∂I1 and ∂Ψ/∂I2 (or an explicit representation in terms of F and cof F), and convexity in (I1, I2) alone does not imply rank-one convexity or ellipticity. Since the paper does not verify ellipticity, monotonicity, or a suitable invariant-based polyconvexity theorem, the guarantee is overstated. Please either prove the required conditions for the specific architecture or soften the claim to 'convex in I1 and I2' and discuss what stability property this actually provides.
  2. [§3.3–3.4, §2.1.1] The incompressibility and kinematic reductions are load-bearing because every training target is computed from them. Eq. (17) imposes transverse stretch λ^{-1/2} for the entire tensile deformation range up to rupture, justified only by ν ≥ 0.47 measured at λ < 1.05. The video extensometer recorded transverse strain; please report large-strain transverse stretch data, either to confirm J ≈ 1 over the training range or to use measured transverse stretches in training. Similarly, Eq. (19) assumes pure simple torsion with no radial expansion and no axial stretch. For an incompressible isotropic cylinder with a free lateral surface, this deformation is not generally an equilibrium solution at the twists used here; Poynting-type normal stresses and radial deformation can be significant at shear strains of order 0.5. If the actual deformation in the torsion tests was not verified, the invariants and shear stresses in Eqs. (20)–(21) may be systematically wrong, and the high reported R^2 values would be fits to a mis-specified kinematic target.
  3. [§2.2 vs §4.1/Fig. 6] The torsion data range is internally inconsistent. Section 2.2 states that torsion results are shown for φ ≤ 360° because at higher angles micro-cracks evolve and the rod buckles and deforms out-of-plane, but Section 4.1 and Figure 6 plot T L/Jp up to φ ≤ 720°, and Figure 2 shows a specimen at φ = 720°. Please clarify which angular range is used for training and how post-buckling or damaged data are excluded. This matters because the apparently linear torque response at large angles may include structural instability rather than intrinsic material response.
  4. [Eq. (14)] The QLV convolution is written as σ(t) = σe(t) + ∫_0^t D′(t−s) σe(t) ds, with σe(t) inside the integral. As written this is not a history-dependent convolution: σe(t) is constant with respect to the integration variable, so the integral reduces to a time-dependent scalar factor multiplying the current stress. The standard QLV form should have σe(s) under the integral. If the implementation uses the correct form, please correct the equation; if not, the model is not the QLV model claimed in the text. This is a central equation and needs to be fixed.
minor comments (4)
  1. [Table 3 and §5] The sentence 'Out of 24 training and test datasets for these compositions, 21 have an R2 above 0.98, and 22 have a sMAPE below 8%' is not supported by the rounded values in Table 3. For example, DM-60 has rows with R2 of 0.977, 0.956, and 0.925, DM-50 has rows of 0.980, and several sMAPE values exceed 8%. Please recompute these counts from unrounded metrics or adjust the sentence.
  2. [§2.1.1 and §4.1] It is unclear whether Poisson's ratio was measured in the present study or taken from reference [28]. Section 2.1.1 describes a calculation from the uniaxial tests, while Section 4.1 attributes the values to Levin and Cohen [28]. Please state this explicitly and report the measured ν values for each composition if they are new data.
  3. [Table 2] The mapping from % Digital ABS to the composition parameter c is not a simple linear scaling of the digital ABS percentages (e.g., 25.2% maps to 0.1755 while 27.2% maps to 1.0). Please provide the exact scaling formula or clarify how c is defined, since Eq. (26) and Eq. (27) depend on this input definition.
  4. [Eq. (11)] The first line of Eq. (11), 's = s(ζ − γ) + γ1', appears to contain a typo or undefined notation. Please correct or define the intermediate variable s̄ (or similar) used in the hard-sigmoid approximation.

Circularity Check

0 steps flagged · score 0.0 of 10

No significant circularity: the held-out DM-40 interpolation and the failed DM-70 extrapolation show the evaluation is a genuine benchmark, not a refit.

full rationale

The derivation chain is data-driven and independently benchmarked. The pICNN strain energy is trained on tension and torsion experiments for compositions A, DM-50, and DM-60, while DM-40 (c=0.1755) is held out entirely and DM-70 (c=2.0895) is held out as an extrapolation test; the reported DM-40 R2 and sMAPE values are predictions, not fitted constants, and the poor DM-70 results (R2 as low as 0.251 in tension, -0.118 in torsion) are the opposite of a forced outcome. The incompressibility and isotropy assumptions, including Eq. 17 and Eq. 19, define a kinematic modeling choice rather than a circular reduction: the training targets (axial stress and torque) come from independent force/displacement measurements, and no target is reinserted as an input or as a fitted prediction. The Poisson-ratio justification cites prior same-author work (Levin and Cohen [28]), but that is an independent experimental measurement, not a fitted parameter or a restatement of the learned energy, so it does not raise the circularity score. Self-citations to PANN methodology ([40], [41], [52]) are ordinary method references and are not load-bearing uniqueness claims; the central claim is validated against external held-out data.

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

The central claim rests on several domain assumptions: incompressibility and isotropy of the printed materials, adequacy of a single-relaxation-time QLV model, and the sufficiency of convexity in I1,I2 for polyconvexity. No new physical entities are introduced; the fitted quantities are network weights and chosen hyperparameters (τ, L0 penalties, composition scaling).

free parameters (5)
  • Fixed QLV relaxation time τ = 10 s
    Chosen rather than fitted; γ and τ appear as a ratio in Eq. 15, so τ was fixed to 10 s to reduce optimization complexity.
  • pICNN sparsified weight coefficients (Eq. 26) = a1=6.84602, a2=6.86996, a3=6.88268, a4=6.93876, b1=2.37126, b2=2.1961, b3=2.56167, b4=1.9888, and other coefficients…
    Learned from experimental data through Adam optimization with L0 sparsification; these define the discovered hyperelastic strain energy.
  • MLP weight coefficients for γ (Eq. 27) = Exponents and coefficients in Eq. 27 (e.g., 0.423, 0.446, 0.552, 0.492, 0.054, 0.147, 0.254, 0.264; 1.381, 1.018…
    Learned from tension data across three strain rates; maps composition c to the QLV relaxation coefficient γ.
  • Composition scaling mapping = c = 0, 0.1755, 0.4669, 1.0, 2.0895 for A, DM-40, DM-50, DM-60, DM-70
    Chosen by hand so that c=0 corresponds to A and c=1 to DM-60; this scaling affects the non-convex path and extrapolation behavior.
  • L0 regularization coefficients = α_fc=5e-4, α_nc=1e-6, α_ncfc=1e-6
    Chosen hyperparameters ramped during Step 7; controls sparsity vs. data fit.
assumptions (5)
  • domain assumption Convexity of the strain energy with respect to I1 and I2 is sufficient to guarantee polyconvexity of the hyperelastic model.
    Stated in §1 and §3.1; the pICNN architecture only enforces convexity, not the monotonicity conditions needed for polyconvexity, so this is an unsupported assumption that may lead to non-physical energies.
  • domain assumption Agilus/Digital ABS digital materials are isotropic and nearly incompressible (ν≥0.47) throughout the testing range.
    Used in §3.1.1 and §3.3-3.4 to set J=1 and compute invariants from a single stretch or twist measurement; measured only at small stretches.
  • domain assumption A single-exponential QLV kernel with fixed τ=10 s captures the time-dependent response over two decades of strain rate.
    Adopted in §3.2; the rate data (Figs. 7-8) show a quasi-static plateau at low rates, but a single relaxation time may not capture a broad relaxation spectrum.
  • standard math The ICNN construction of Amos et al. [51] and the pICNN variant guarantee convexity with respect to the strain invariants in the passthrough architecture.
    Invoked in §3.1.2; with nonnegative output weights, convex nondecreasing activations, and passthrough additions, the output is convex in Ibar.
  • domain assumption The torsion test is equivalent to simple torsion of an incompressible rod, with no end effects or buckling over the data range used.
    Used in §3.4; the paper notes buckling beyond 360° yet plots data up to 720° (Figure 6), so the validity over the full range is unclear.

how reviews work

0 comments
Cite this review

Pith. "Pith review of Physics Augmented Machine Learning Discovery of Composition-Dependent Constitutive Laws for 3D Printed Digital Materials." pith.science (2026). https://pith.science/paper/CQJM23U7

@misc{pith2026250702991,
  author       = {Pith},
  title        = {Pith review of: Physics Augmented Machine Learning Discovery of Composition-Dependent Constitutive Laws for 3D Printed Digital Materials},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/CQJM23U7}},
  note         = {Machine review of arXiv:2507.02991}
}
abstract

Multi-material 3D printing, particularly through polymer jetting, enables the fabrication of digital materials by mixing distinct photopolymers at the micron scale within a single build to create a composite with tunable mechanical properties. This work presents an integrated experimental and computational investigation into the composition-dependent mechanical behavior of 3D printed digital materials. We experimentally characterize five formulations, combining soft and rigid UV-cured polymers under uniaxial tension and torsion across three strain and twist rates. The results reveal nonlinear and rate-dependent responses that strongly depend on composition. To model this behavior, we develop a physics-augmented neural network (PANN) that combines a partially input convex neural network (pICNN) for learning the composition-dependent hyperelastic strain energy function with a quasi-linear viscoelastic (QLV) formulation for time-dependent response. The pICNN ensures convexity with respect to strain invariants while allowing non-convex dependence on composition. To enhance interpretability, we apply $L_0$ sparsification. For the time-dependent response, we introduce a multilayer perceptron (MLP) to predict viscoelastic relaxation parameters from composition. The proposed model accurately captures the nonlinear, rate-dependent behavior of 3D printed digital materials in both uniaxial tension and torsion, achieving high predictive accuracy for interpolated material compositions. This approach provides a scalable framework for automated, composition-aware constitutive model discovery for multi-material 3D printing.

Figures

Figures reproduced from arXiv: 2507.02991 by the authors.

Figure 1
Figure 1. Tension experimental set-up: DM-60 sample marked with two longitudinal and two transverse marks, [PITH_FULL_IMAGE:figures/full_fig_p005_1.png] view at source ↗
Figure 2
Figure 2. Torsion experimental set-up: (a-c) A sample at twist angles of [PITH_FULL_IMAGE:figures/full_fig_p007_2.png] view at source ↗
Figure 3
Figure 3. Flow chart of the constitutive modeling framework. Strain invariants and material composition [PITH_FULL_IMAGE:figures/full_fig_p008_3.png] view at source ↗
Figures from the paper (9 more)
Figure 4
Figure 4. Figure 4: Architecture of the proposed partial input convex neural network (pICNN), where strain invariants serve [PITH_FULL_IMAGE:figures/full_fig_p011_4.png]
Figure 5
Figure 5. Figure 5: Experimental uniaxial tension data showing nominal stress versus stretch for various digital material [PITH_FULL_IMAGE:figures/full_fig_p017_5.png]
Figure 6
Figure 6. Figure 6: Experimental torsion data showing T L/Jp (MPa) versus rotation angle φ (deg) for various digital material (DM) compositions tested at twist rates of (a) φ˙ = 90 deg/min, (b) φ˙ = 180 deg/min, and (c) φ˙ = 360 deg/min [PITH_FULL_IMAGE:figures/full_fig_p018_6.png]
Figure 7
Figure 7. Figure 7: Select uniaxial tension plots for (a) A, (b) DM-50, and (c) DM-70 showing nominal stress versus stretch. [PITH_FULL_IMAGE:figures/full_fig_p019_7.png]
Figure 8
Figure 8. Figure 8: Select torsion tests for (a) A, (b) DM-50, and (c) DM-70 showing [PITH_FULL_IMAGE:figures/full_fig_p020_8.png]
Figure 9
Figure 9. Figure 9: Training and testing loss across model training steps 1-7. (a) The blue line shows the combined loss for [PITH_FULL_IMAGE:figures/full_fig_p025_9.png]
Figure 10
Figure 10. Figure 10: Evolution of the total number of model parameters in the pICNN during training. Shown are the [PITH_FULL_IMAGE:figures/full_fig_p027_10.png]
Figure 11
Figure 11. Figure 11: Comparison of model predictions (lines) and experimental data (points) for uniaxial tension. Axial [PITH_FULL_IMAGE:figures/full_fig_p029_11.png]
Figure 12
Figure 12. Figure 12: Comparison of model prediction (lines) and experimental data (points) for torsion experiments [PITH_FULL_IMAGE:figures/full_fig_p029_12.png]

Discussion (0). Continue with ORCID to comment.

Reference graph

Works this paper leans on

55 extracted references · 53 canonical work pages

  1. [1]

    Sharma, S., and Goel, S. A. (2019) 3D printing and its future in medical world.Journal of Medical Research and Innovation 3, e000141–e000141

    work page 2019

  2. [2]

    E.3D printing in Orthopaedic surgery; Elsevier, 2019; pp 1–15

    Jakus, A. E.3D printing in Orthopaedic surgery; Elsevier, 2019; pp 1–15

    work page 2019

  3. [3]

    G., Raut, D., and Shinde, D

    Gokhare, V . G., Raut, D., and Shinde, D. (2017) A review paper on 3D-printing aspects and various processes used in the 3D-printing.Int. J. Eng. Res. Technol 6, 953–958

    work page 2017

  4. [4]

    W., Pohl, R., Sun, C., Romer, G.-W., Huis int Veld, B., and Lohse, D

    Visser, C. W., Pohl, R., Sun, C., Romer, G.-W., Huis int Veld, B., and Lohse, D. (2015) Toward 3D printing of pure metals by laser-induced forward transfer.Advanced materials 27, 4087–4092

    work page 2015

  5. [5]

    A., Mykulowycz, N

    Gibson, M. A., Mykulowycz, N. M., Shim, J., Fontana, R., Schmitt, P., Roberts, A., Ketkaew, J., Shao, L., Chen, W., Bordeenithikasem, P., and others (2018) 3D printing metals like thermoplastics: Fused filament fabrication of metallic glasses.Materials Today 21, 697–702

    work page 2018

  6. [6]

    (1998) 3D printing with metals.Computing and Control Engineering Journal 9, 31–38

    Ribeiro, F. (1998) 3D printing with metals.Computing and Control Engineering Journal 9, 31–38

    work page 1998

  7. [7]

    (2019) 3D printing of ceramics: A review.Journal of the European Ceramic Society 39, 661–687

    Chen, Z., Li, Z., Li, J., Liu, C., Lao, C., Fu, Y ., Liu, C., Li, Y ., Wang, P., and He, Y . (2019) 3D printing of ceramics: A review.Journal of the European Ceramic Society 39, 661–687

    work page 2019

  8. [8]

    C., Rajoo, S., Noor, A

    Hwa, L. C., Rajoo, S., Noor, A. M., Ahmad, N., and Uday, M. (2017) Recent advances in 3D printing of porous ceramics: A review.Current Opinion in Solid State and Materials Science 21, 323–347

    work page 2017

Show all 55 references
  1. [9]

    (2017) 3D printing of polymer matrix composites: A review and prospective.Composites Part B: Engineering 110, 442–458

    Wang, X., Jiang, M., Zhou, Z., Gou, J., and Hui, D. (2017) 3D printing of polymer matrix composites: A review and prospective.Composites Part B: Engineering 110, 442–458. 34

    2017

  2. [10]

    G., and Lewis, J

    Compton, B. G., and Lewis, J. A. (2014) 3D-printing of lightweight cellular composites. Advanced materials 26, 5930–5935

    2014

  3. [11]

    (2018) A review of 3D printing technology for medical applications.Engineering 4, 729–742

    Yan, Q., Dong, H., Su, J., Han, J., Song, B., Wei, Q., and Shi, Y . (2018) A review of 3D printing technology for medical applications.Engineering 4, 729–742

    2018

  4. [12]

    A., Finne-Wistrand, A., and Gasser, T

    Liu, H., Ahlinder, A., Yassin, M. A., Finne-Wistrand, A., and Gasser, T. C. (2020) Computational and experimental characterization of 3D-printed PCL structures toward the design of soft biological tissue scaffolds.Materials & Design 188, 108488

    2020

  5. [13]

    L., Peng, C., Pille, P., Leary, M., and Tran, P

    Tee, Y . L., Peng, C., Pille, P., Leary, M., and Tran, P. (2020) PolyJet 3D printing of composite materials: experimental and modelling approach.Jom 72, 1105–1117

    2020

  6. [14]

    Materials Horizons 6, 394–404

    Lei, D., Yang, Y ., Liu, Z., Chen, S., Song, B., Shen, A., Yang, B., Li, S., Yuan, Z., Qi, Q., and others (2019) A general strategy of 3D printing thermosets for diverse applications. Materials Horizons 6, 394–404

    2019

  7. [15]

    (2020) 3D printing technologies: techniques, materials, and post- processing.Current Opinion in Chemical Engineering 28, 134–143

    Karakurt, I., and Lin, L. (2020) 3D printing technologies: techniques, materials, and post- processing.Current Opinion in Chemical Engineering 28, 134–143

    2020

  8. [16]

    Patpatiya, P., Chaudhary, K., Shastri, A., and Sharma, S. (2022) A review on polyjet 3D printing of polymers and multi-material structures.Proceedings of the Institution of Mechanical Engineers, Part C: Journal of Mechanical Engineering Science 236, 7899– 7926

    2022

  9. [17]

    (2024) Design principles for 3D-printed thermally activated shape-morphing structures.International Journal of Mechanical Sciences 262, 108716

    Hanuhov, T., and Cohen, N. (2024) Design principles for 3D-printed thermally activated shape-morphing structures.International Journal of Mechanical Sciences 262, 108716

    2024

  10. [18]

    J., and O’Sullivan, L

    Shannon, A., O’Connell, A., O’Sullivan, A., Byrne, M., Clifford, S., O’Sullivan, K. J., and O’Sullivan, L. (2020) A radiopaque nanoparticle-based ink using PolyJet 3D printing for medical applications.3D Printing and Additive Manufacturing 7, 259–268

    2020

  11. [19]

    (2017) Reconstruction of complex 35 mandibular defects using integrated dental custom-made titanium implants.British Journal of Oral and Maxillofacial Surgery 55, 425–427

    Rachmiel, A., Shilo, D., Blanc, O., and Emodi, O. (2017) Reconstruction of complex 35 mandibular defects using integrated dental custom-made titanium implants.British Journal of Oral and Maxillofacial Surgery 55, 425–427

    2017

  12. [20]

    Barone, S., Casinelli, M., Frascaria, M., Paoli, A., and Razionale, A. V . (2016) Interactive design of dental implant placements through CAD-CAM technologies: from 3D imaging to additive manufacturing.International Journal on Interactive Design and Manufacturing 10, 105–117

    2016

  13. [21]

    (2023) Personalized 3D printed eye gear for microscopic surgeons amidst and beyond COVID-19.Bioengineering 10, 1129

    Singh, R., Singh, R., and Suri, A. (2023) Personalized 3D printed eye gear for microscopic surgeons amidst and beyond COVID-19.Bioengineering 10, 1129

    2023

  14. [22]

    J., Wu, J., Hamel, C

    Kuang, X., Roach, D. J., Wu, J., Hamel, C. M., Ding, Z., Wang, T., Dunn, M. L., and Qi, H. J. (2019) Advances in 4D printing: materials and applications.Advanced Functional Materials 29, 1805290

    2019

  15. [23]

    (2018) 4D printing: history and recent progress.Chinese Journal of Polymer Science 36, 563–575

    Wu, J.-J., Huang, L.-M., Zhao, Q., and Xie, T. (2018) 4D printing: history and recent progress.Chinese Journal of Polymer Science 36, 563–575

    2018

  16. [24]

    Raviv, D., Zhao, W., McKnelly, C., Papadopoulou, A., Kadambi, A., Shi, B., Hirsch, S., Dikovsky, D., Zyracki, M., Olguin, C., and others (2014) Active printed materials for complex self-evolving deformations.Scientific reports 4, 1–8

    2014

  17. [25]

    I., Vladimirsky, D., Klein, G., and Rudykh, S

    Slesarenko, V ., Engelkemier, S., Galich, P. I., Vladimirsky, D., Klein, G., and Rudykh, S. (2018) Strategies to control performance of 3d-printed, cable-driven soft polymer actuators: From simple architectures to gripper prototype.Polymers 10, 846

    2018

  18. [26]

    G., Preti, M

    Shi, G., Palombi, A., Lim, Z., Astolfi, A., Burani, A., Campagnini, S., Loizzo, F. G., Preti, M. L., Vargas, A. M., Peperoni, E., and others (2020) Fluidic haptic interface for mechano-tactile feedback.IEEE transactions on haptics 13, 204–210

    2020

  19. [27]

    (2019) Fast-response, stiffness-tunable soft actuator by hybrid multimaterial 3D printing.Advanced Functional Materials 29, 1806698

    Zhang, Y .-F., Zhang, N., Hingorani, H., Ding, N., Wang, D., Yuan, C., Zhang, B., Gu, G., and Ge, Q. (2019) Fast-response, stiffness-tunable soft actuator by hybrid multimaterial 3D printing.Advanced Functional Materials 29, 1806698. 36

    2019

  20. [28]

    (2023) Swelling under Constraints: Exploiting 3D-Printing to Optimize the Performance of Gel-Based Devices.Advanced Materials Technologies 8, 2202136

    Levin, M., and Cohen, N. (2023) Swelling under Constraints: Exploiting 3D-Printing to Optimize the Performance of Gel-Based Devices.Advanced Materials Technologies 8, 2202136

    2023

  21. [29]

    F., and Ghajari, M

    Abayazid, F. F., and Ghajari, M. (2020) Material characterisation of additively manufactured elastomers at different strain rates and build orientations.Additive Manufacturing 33, 101160

    2020

  22. [30]

    (2018) Towards mechanical characterization of soft digital materials for multimaterial 3D-printing.International Journal of Engineering Science 123, 62–72

    Slesarenko, V ., and Rudykh, S. (2018) Towards mechanical characterization of soft digital materials for multimaterial 3D-printing.International Journal of Engineering Science 123, 62–72

    2018

  23. [31]

    M., and Costa, C

    V olpato, N., Solis, D. M., and Costa, C. A. (2016) An analysis of Digital ABS as a rapid tooling material for polymer injection moulding.International Journal of Materials and Product Technology 52, 3–16

    2016

  24. [32]

    (2009) Nonlinear Viscoelastic Solids - A Review.Mathematics and Mechanics of Solids 14, 300–366

    Wineman, A. (2009) Nonlinear Viscoelastic Solids - A Review.Mathematics and Mechanics of Solids 14, 300–366

    2009

  25. [33]

    Aziz, S., and Spinks, G. M. (2020) Torsional artificial muscles.Materials Horizons 7, 667–693

    2020

  26. [34]

    (2017) Effect of intrinsic twist and orthotropy on extension–twist–inflation coupling in compressible circular tubes.Journal of Elasticity 128, 175–201

    Singh, R., Kumar, S., and Kumar, A. (2017) Effect of intrinsic twist and orthotropy on extension–twist–inflation coupling in compressible circular tubes.Journal of Elasticity 128, 175–201

    2017

  27. [35]

    (2021) Inversion and perversion in twist incompatible isotropic tubes.Extreme Mechanics Letters 46, 101303

    Emuna, N., and Cohen, N. (2021) Inversion and perversion in twist incompatible isotropic tubes.Extreme Mechanics Letters 46, 101303

    2021

  28. [36]

    (2021) A new type of soft pneumatic torsional actuator with helical chambers for flexible machines.Journal of Mechanisms and Robotics 13

    Xiao, W., Hu, D., Chen, W., Yang, G., and Han, X. (2021) A new type of soft pneumatic torsional actuator with helical chambers for flexible machines.Journal of Mechanisms and Robotics 13. 37

    2021

  29. [37]

    (2021) Inflation-induced twist in geometrically incompatible isotropic tubes.Journal of Applied Mechanics 88

    Emuna, N., and Cohen, N. (2021) Inflation-induced twist in geometrically incompatible isotropic tubes.Journal of Applied Mechanics 88

    2021

  30. [38]

    (2022) Electrically-induced twist in geometrically incompatible dielectric elastomer tubes.International Journal of Solids and Structures111707

    Bazaev, K., and Cohen, N. (2022) Electrically-induced twist in geometrically incompatible dielectric elastomer tubes.International Journal of Solids and Structures111707

    2022

  31. [39]

    Kolesnikov, A. M. (2022) Finite deformations of a nonlinearly elastic electrosensitive tube reinforced by two fiber families.Continuum Mechanics and Thermodynamics 34, 1237– 1255

    2022

  32. [40]

    N., Jones, R

    Fuhg, J. N., Jones, R. E., and Bouklas, N. (2024) Extreme sparsification of physics- augmented neural networks for interpretable model discovery in mechanics.Computer Methods in Applied Mechanics and Engineering 426, 116973

    2024

  33. [41]

    A., Safta, C., Bouklas, N., and Jones, R

    Padmanabha, G. A., Safta, C., Bouklas, N., and Jones, R. E. (2024) Condensed Stein Variational Gradient Descent for Uncertainty Quantification of Neural Networks.arXiv preprint arXiv:2412.16462

    2024 arXiv

  34. [42]

    K., Roth, F

    Klein, D. K., Roth, F. J., Valizadeh, I., and Weeger, O. (2023) Parametrized polyconvex hyperelasticity with physics-augmented neural networks.Data-Centric Engineering 4, Publisher: Cambridge University Press (CUP)

    2023

  35. [43]

    Tac, V ., Sahli Costabal, F., and Tepole, A. B. (2022) Data-driven tissue mechanics with polyconvex neural ordinary differential equations.Computer Methods in Applied Mechanics and Engineering 398, 115248, Publisher: Elsevier BV

    2022

  36. [44]

    (2024) A Physics-Guided Machine Learning Model for Predicting Viscoelasticity of Solids at Large Deformation.Polymers 16, 3222, Publisher: MDPI AG

    Qin, B., and Zhong, Z. (2024) A Physics-Guided Machine Learning Model for Predicting Viscoelasticity of Solids at Large Deformation.Polymers 16, 3222, Publisher: MDPI AG

    2024

  37. [45]

    A., Brummund, J., Sun, W., and KÃ Cstner, M

    Rosenkranz, M., Kalina, K. A., Brummund, J., Sun, W., and KÃ Cstner, M. (2024) Viscoelasticty with physics-augmented neural networks: model formulation and training methods without prescribed internal variables.Computational Mechanics 74, 1279–1301, Publisher: Springer Science...

    2024

  38. [47]

    N., Van Wees, L., Obstalecki, M., Shade, P., Bouklas, N., and Kasemer, M

    Fuhg, J. N., Van Wees, L., Obstalecki, M., Shade, P., Bouklas, N., and Kasemer, M. (2022) Machine-learning convex and texture-dependent macroscopic yield from crystal plasticity simulations.Materialia 23, 101446, Publisher: Elsevier BV

    2022

  39. [48]

    Jailin, C., Benady, A., Legroux, R., and Baranger, E. (2024) Experimental Learning of a Hyperelastic Behavior with a Physics-Augmented Neural Network.Experimental Mechanics 64, 1465–1481, Publisher: Springer Science and Business Media LLC

    2024

  40. [49]

    A., Gebhart, P., Brummund, J., Linden, L., Sun, W., and KÃ Cstner, M

    Kalina, K. A., Gebhart, P., Brummund, J., Linden, L., Sun, W., and KÃ Cstner, M. (2024) Neural network-based multiscale modeling of finite strain magneto-elasticity with relaxed convexity criteria.Computer Methods in Applied Mechanics and Engineering 421, 116739, Publisher: El...

    2024

  41. [50]

    Louizos, C., Welling, M., and Kingma, D. P. (2017) Learning sparse neural networks throughL_0regularization.arXiv preprint arXiv:1712.01312

    2017 arXiv

  42. [51]

    Amos, B., Xu, L., and Kolter, J. Z. Input convex neural networks. International Conference on Machine Learning. 2017; pp 146–155

    2017

  43. [52]

    A., Kalina, K

    Jadoon, A. A., Kalina, K. A., Rausch, M. K., Jones, R., and Fuhg, J. N. (2024) Inverse design of anisotropic microstructures using physics-augmented neural networks.arXiv preprint arXiv:2412.13370

    2024 arXiv

  44. [53]

    K., Kalina, K

    Linden, L., Klein, D. K., Kalina, K. A., Brummund, J., Weeger, O., and Kästner, M. (2023) Neural networks meet hyperelasticity: A guide to enforcing physics.Journal of the Mechanics and Physics of Solids 179, 105363

    2023

  45. [54]

    D., and Parnell, W

    De Pascalis, R., Abrahams, I. D., and Parnell, W. J. (2014) On nonlinear viscoelastic 39 deformations: a reappraisal of Fung’s quasi-linear viscoelastic model.Proceedings of the Royal Society A: Mathematical, Physical and Engineering Sciences 470, 20140058

    2014

  46. [55]

    O., and Saccomandi, G

    Horgan, C. O., and Saccomandi, G. (1999) Simple torsion of isotropic, hyperelastic, incompressible materials with limiting chain extensibility.Journal of elasticity 56, 159– 170

    1999

  47. [56]

    P., and Ba, J

    Kingma, D. P., and Ba, J. (2014) Adam: A method for stochastic optimization.arXiv preprint arXiv:1412.6980

    2014 arXiv

Pith tools

Reviewed August 6, 2026 · model on record in the stance chip above.