pith. sign in

arxiv: 2604.10692 · v1 · submitted 2026-04-12 · 📡 eess.SY · cs.SY

i-Tac: Inverse Design of 3D-Printed Tactile Elastomers with Scalable and Tunable Optical and Mechanical Properties

Wen Fan , Dandan Zhang This is my paper

Pith reviewed 2026-05-10 15:47 UTC · model grok-4.3

classification 📡 eess.SY cs.SY
keywords inverse design3D printingelastomerstactile sensorsmixture designresponse surface modelsmulti-objective optimization
0
0 comments X

The pith

i-Tac inverse design finds printable resin mixtures that deliver user-specified transparency and hardness in a single fabrication step.

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

The paper introduces i-Tac, an inverse design pipeline that starts from desired optical and mechanical properties rather than from material recipes. It mixes three resins in a PolyJet printer, fits statistical response surface models to measured transparency and hardness values, and then runs multi-objective optimization to locate the exact composition window that meets any chosen targets. This replaces repeated trial-and-error preparation and testing with a direct calculation that yields a usable print file. A sympathetic reader would care because vision-based tactile sensors in robotics and prosthetics need different stiffness and light transmission depending on the sensing architecture, and the new route cuts the number of design cycles required to match those needs.

Core claim

i-Tac employs a mixture design methodology with three complementary resins to generate response surface models that map composition to transparency and hardness, thereby defining a scalable property space; a desirability-function-based multi-objective optimisation then identifies feasible composition regions and an optimal operating window, so that elastomers matching user-defined targets can be manufactured in a single iteration.

What carries the argument

Response surface models fitted to mixture-design experiments, combined with desirability-function multi-objective optimisation that inverts the conventional forward trial-and-error loop to output printable compositions directly from target properties.

If this is right

  • Elastomers can be produced with any combination of transparency and hardness inside the modelled space without iterative material adjustments.
  • Monolithic multi-material prints become feasible for both commercial and custom vision-based tactile sensor geometries.
  • The characterised property space expands continuously with the three-resin mixture range rather than being limited to discrete commercial grades.
  • Design time for new sensor architectures drops because the optimisation replaces repeated fabrication and measurement cycles.

Where Pith is reading between the lines

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

  • Adding further measured responses such as friction or tear strength to the same models would let the optimisation target complete sensor performance specifications at once.
  • Linking the optimisation output to finite-element simulations of contact deformation could create an end-to-end loop from application requirements to print-ready files.
  • The same mixture-design-plus-optimisation structure could be applied to other resin systems or additive processes where composition controls functional behaviour.

Load-bearing premise

The fitted response surface models accurately predict optical and mechanical properties for any untested composition inside the characterised region, and printing or curing steps introduce no large unmodeled deviations.

What would settle it

Print an elastomer at the composition returned by the optimisation for a chosen transparency and hardness target, then measure those two properties and find that the results lie outside the tolerance band around the targets.

Figures

Figures reproduced from arXiv: 2604.10692 by Dandan Zhang, Wen Fan.

Figure 1
Figure 1. Figure 1: A: Physiological anatomy of human skin, in which the dermis consists of three distinct tissue components. B: Inverse design frame￾work for achieving target tunability in elastomer properties through (a) AC/TM/GM mixture design and (b) monolithic manufacturing. C: Scalable property space of AC/TM/GM mixture elastomers, covering a broad range from clear to opaque and from soft to rigid. Across different VBTS… view at source ↗
Figure 2
Figure 2. Figure 2: A: Comparison between forward design and inverse design paradigm in elastomer property tailoring. (a)/(b) Forward design requires prototyping in advance of parameter modification, though additive manufacturing can accelerate fabrication. (c) Inverse design of i-Tac helps additive manufacturing in rapid elastomer tailoring without iteration. B: Challenges to be addressed in elastomer property tailoring. abl… view at source ↗
Figure 3
Figure 3. Figure 3: i-Tac inverse design pipeline for elastomer property tailoring in VBTSs. A: Material formulation, construction of response surface models to map AC/TM/GM mixture designs to elastomer properties and establish a scalable property space. B: Property tailoring, identification of feasible operating windows for 3D printing from desired property targets using a desirability function, enabling target tunability. C… view at source ↗
Figure 4
Figure 4. Figure 4: A: Mixture design space of AC/TM/GM component, with centroid mixture (c1), 3-mixture (c2-c4), 2-mixture (a1-a5,t2-t4,g3-g5), forbidden mixture (t1,t5,g2), and special mixture (g1). B: Fifteen samples (a1-a5,t2-t4,g1,g3-g5,c1-c4) are fabricated for mixture experiment. B. Mixture Experiment 1) Selected Samples: From MDS, a total of 19 mixtures can be obtained through uniform sampling. Among these, c1 corresp… view at source ↗
Figure 5
Figure 5. Figure 5: A: Light transmission per wavelength of mixture design samples with 3mm thickness. B: Light transmission at 700nm with different sample thickness. C: Opacity density per unit thickness. D: Mechanical hardness of mixture design samples. Amix = − log10 Tmix + (1 − Tbias)  = copacity dthick (4) copacity = log10 Tmix + (1 − Tbias)  dthick (5) Finally, copacity characterises the effective ‘opacity density per… view at source ↗
Figure 6
Figure 6. Figure 6: A: Transparency metrics in optical mixture experiment. B: Hardness metrics in mechanical mixture experiment. C: Experiment results in terms of optical transparency and Shore 00 hardness demonstrate wide property scalability of mixture designs [PITH_FULL_IMAGE:figures/full_fig_p007_6.png] view at source ↗
Figure 7
Figure 7. Figure 7: A: Transparency ReSM for mixture designs. B: Hardness ReSM for mixture designs mixtures expand both optical and mechanical property scala￾bility of 3D-printed elastomers, also highlighting their strong potential for inverse design in target tunability. C. Response Surface Model (ReSM) 1) Quadratic Regression Formulation: Based on the ex￾perimental data summarised in Table III, a response surface model (ReS… view at source ↗
Figure 8
Figure 8. Figure 8: Feasible property space. A: FPS(AC), x1(Y1, Y2), effects more in higher hardness area. B: FPS(TM), x2(Y1, Y2), effects more in lower hardness area. C: FPS(GM), x3(Y1, Y2), effects more in lower clarity and ultra-soft area [PITH_FULL_IMAGE:figures/full_fig_p009_8.png] view at source ↗
Figure 9
Figure 9. Figure 9: Inverse design methodology for tailoring elastomer property. A: Desired elastomer property. B: Required material formulations and user preference. C: Multi-objective optimization between different property targets. D: Optimal inverse design. E: Optimal inverse design window for elastomer fabrication. F: Property prediction of inverse designed elastomer. G: Real evaluated property of elastomer product. TABL… view at source ↗
Figure 10
Figure 10. Figure 10: A: Design operating window W∆x with user-defined ±∆x%. B: Property operating window W∆Y with user-defined −∆Y %. C: Optimal operating window Woptimal, defines a set of i integer-percentage designs {(x1, x2, x3)0, ...,(x1, x2, x3)i} for subsequent printing. 1) Desirability Function: Given the property targets of both (TY1 , TY2 ), desirability function approach [17] is employed to achieve the target tunabi… view at source ↗
Figure 11
Figure 11. Figure 11: A: Influence of elastomer material properties on tactile sensing. (a) Hardness reflects the resistance of the sensor surface [PITH_FULL_IMAGE:figures/full_fig_p012_11.png] view at source ↗
Figure 12
Figure 12. Figure 12: Material property evaluation results. A: Uniaxial tensile testing results. (a) Representative stress-strain curve under [PITH_FULL_IMAGE:figures/full_fig_p012_12.png] view at source ↗
Figure 13
Figure 13. Figure 13: Case study of desirability optimization in terms of criteria guideline. A: NTB/NTB; B: NTB/LTB; C: LTB/STB; D: STB/STB [PITH_FULL_IMAGE:figures/full_fig_p014_13.png] view at source ↗
Figure 14
Figure 14. Figure 14: A: Mold-casting silicone (Ecoflex, DragonSkin, SORTA Clear). B: Optimal operating window tailoring from real silicone. C: Real transparency RY1 and predicted Yˆ1 (red edges) of inverse design samples. D: Real hardness RY2 and predicted Yˆ2 (red lines) of inverse design samples. E: Accumulated inverse design error compared between I1/I2/I3. F: Average inverse design error compared between Error1/Error2 [P… view at source ↗
Figure 9
Figure 9. Figure 9: As summarised in Table IX, Error 1, which represents the difference between the desired target TY and the ReSM prediction Yˆ , is generally small. Only Ecoflex 00-31 and Ecoflex 00-45 exhibit notable discrepancies, with Y1 Error 1 of 17% and Y2 Error 1 of 5-10%, respectively. By contrast, the predictions Yˆ for the other inverse designs are close to their corresponding TY , with most errors below 1%. This … view at source ↗
Figure 15
Figure 15. Figure 15: A: (a) ViTacTip-f/r with AC/TM elastomer, (b) ViTacTip-f/r with pure GM elastomer, (c) GBS with AC/TM elastomer, (d) DBS with AC/TM elastomer. B: (a) Comparison between 3D-print GBS and commercial GelSight; (b) Comparison between 3D-print and homemade ViTacTip-f. C: VBTS base mounted with 3D-print contact modules, (a) DIGIT base, (b) GelSight base, (c/d) ViTacTip base v1/v2. D: (a) Experiment setup of opt… view at source ↗
read the original abstract

Elastomers are central to vision-based tactile sensors (VBTSs), where they transduce external contact into observable deformation. Different VBTS architectures, however, require distinct optical and mechanical properties, particularly transparency and hardness. Conventional elastomer design relies on a forward, trial-and-error optimisation process from material preparation to property evaluation, which is inefficient and offers limited property scalability and target tunability. In this work, we present i-Tac, an inverse design pipeline for tailoring 3D-printed tactile elastomers with target optical and mechanical properties. Inspired by the composite structure of the human dermis, i-Tac exploits multi-material PolyJet additive manufacturing with three complementary resins. A mixture design methodology is employed to characterise the printed elastomers and establish response surface models (ReSMs) that map material compositions to functional properties, thereby defining a scalable property space. Based on user-defined targets, a desirability-function-based multi-objective optimisation is then performed to identify feasible composition regions and derive an optimal operating window for fabrication. This enables elastomers with desired properties to be manufactured in a single iteration, thereby achieving efficient target tunability. Experimental results validate the proposed i-Tac framework in terms of both property scalability and inverse design performance, showing that i-Tac can effectively tailor elastomer transparency and hardness while reducing the iterative burden of conventional forward design. By fabricating physical sensor samples from both commercial and custom designs, the proposed framework further demonstrates the potential of inverse-designed, monolithically manufactured elastomers for customisable VBTS fabrication.

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 / 2 minor

Summary. The paper introduces i-Tac, an inverse-design pipeline for 3D-printed tactile elastomers that combines mixture-design experiments with response surface models (ReSMs) to map resin compositions to optical transmittance and hardness, followed by desirability-function multi-objective optimization to identify composition windows that meet user-specified targets. The central claim is that this enables fabrication of elastomers with desired properties in a single iteration, validated experimentally for both property scalability and inverse-design performance in vision-based tactile sensor applications.

Significance. If the ReSMs are shown to predict properties accurately at the optimized (untested) compositions, the work would provide a practical, scalable route to custom elastomers that reduces the trial-and-error burden of conventional forward design for VBTSs. The use of PolyJet multi-material printing to realize the optimized mixtures is a concrete strength, and the overall framing of inverse design for tunable optical-mechanical properties is timely for tactile sensing.

major comments (2)
  1. [Abstract] Abstract: The statement that 'experimental results validate... inverse design performance' is not supported by any reported quantitative metrics (e.g., R², RMSE, or leave-one-out error) for the ReSMs, the number of mixture compositions tested, or direct measured-versus-predicted comparisons on the final optimized samples. Without these, the single-iteration claim rests on unshown validation details.
  2. [Methods / Results] The pipeline fits ReSMs to a finite set of printed mixtures and then optimizes inside the feasible region; however, no cross-validation, hold-out set, or extrapolation-error analysis is described, so it remains possible that the reported performance reflects in-sample fit rather than reliable prediction at the desirability-optimized points.
minor comments (2)
  1. Notation for the desirability functions and the feasible-region constraints could be clarified with an explicit equation or pseudocode block.
  2. Figure captions should explicitly state the number of replicate prints and the measurement protocol for transmittance and hardness to aid reproducibility.

Simulated Author's Rebuttal

2 responses · 0 unresolved

We thank the referee for the constructive feedback and recommendation for major revision. We address each major comment below and will update the manuscript to strengthen the quantitative validation of the ReSMs and inverse-design claims.

read point-by-point responses

  1. Referee: [Abstract] Abstract: The statement that 'experimental results validate... inverse design performance' is not supported by any reported quantitative metrics (e.g., R², RMSE, or leave-one-out error) for the ReSMs, the number of mixture compositions tested, or direct measured-versus-predicted comparisons on the final optimized samples. Without these, the single-iteration claim rests on unshown validation details.

    Authors: We agree that the abstract would be strengthened by explicit quantitative support. The original manuscript reports experimental validation of the i-Tac pipeline but does not highlight R², RMSE, leave-one-out error, the exact number of tested mixtures, or direct measured-versus-predicted values for the optimized samples. We will revise the abstract to include these metrics and add a results table or figure showing measured-versus-predicted comparisons on the final optimized compositions to better substantiate the single-iteration performance. revision: yes

  2. Referee: [Methods / Results] The pipeline fits ReSMs to a finite set of printed mixtures and then optimizes inside the feasible region; however, no cross-validation, hold-out set, or extrapolation-error analysis is described, so it remains possible that the reported performance reflects in-sample fit rather than reliable prediction at the desirability-optimized points.

    Authors: We thank the referee for this observation. While the mixture-design experiments provide a structured sampling of the composition space, the original submission does not describe cross-validation, hold-out testing, or explicit extrapolation-error analysis at the optimized points. We will add leave-one-out cross-validation results for the ReSMs and a targeted prediction-error analysis at the multi-objective optimized compositions in the revised Methods and Results sections to demonstrate that the reported performance reflects reliable prediction rather than in-sample fit alone. revision: yes

Circularity Check

0 steps flagged

No circularity: standard empirical modeling and experimental validation

full rationale

The derivation proceeds by printing a finite set of mixture compositions, fitting response surface models to the resulting measured optical and mechanical properties, then applying desirability-based multi-objective optimization over the model to select new compositions. The single-iteration claim is then tested by actually printing and measuring those optimized compositions. This sequence contains no self-definitional reduction, no renaming of fitted values as independent predictions, and no load-bearing self-citation; the final performance numbers rest on fresh experimental data outside the original design points.

Axiom & Free-Parameter Ledger

1 free parameters · 2 axioms · 0 invented entities

The approach depends on empirical fitting of response surface models to a finite set of printed samples and on the assumption that the resulting surfaces remain valid for optimization; no new physical entities are postulated.

free parameters (1)
  • Response surface model coefficients
    Coefficients of the ReSMs are fitted to experimental measurements of printed mixtures; these fitted values define the property predictions used by the optimizer.
axioms (2)
  • domain assumption Elastomer optical and mechanical properties vary smoothly and continuously with resin composition ratios.
    Required for the response surface models to produce usable predictions across the composition space.
  • domain assumption The desirability function correctly encodes the relative importance and acceptable ranges of transparency and hardness for the target application.
    Central to the multi-objective optimization step that selects the final composition window.

pith-pipeline@v0.9.0 · 5586 in / 1442 out tokens · 32187 ms · 2026-05-10T15:47:49.941101+00:00 · methodology

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

34 extracted references · 34 canonical work pages

  1. [1]

    Hardware technology of vision-based tactile sensor: A review,

    S. Zhang, Z. Chen, Y . Gao, W. Wan, J. Shan, H. Xue, F. Sun, Y . Yang, and B. Fang, “Hardware technology of vision-based tactile sensor: A review,”IEEE Sensors Journal, vol. 22, no. 22, pp. 21 410–21 427, 2022

    work page 2022

  2. [2]

    Vitactip: Design and verification of a novel biomimetic physical vision-tactile fusion sensor,

    W. Fan, H. Li, W. Si, S. Luo, N. Lepora, and D. Zhang, “Vitactip: Design and verification of a novel biomimetic physical vision-tactile fusion sensor,” in2024 IEEE International Conference on Robotics and Automation (ICRA). IEEE, 2024, pp. 1056–1062

    work page 2024

  3. [3]

    Crystaltac: Vision-based tactile sensor family fabricated via rapid monolithic manufacturing,

    W. Fan, H. Li, and D. Zhang, “Crystaltac: Vision-based tactile sensor family fabricated via rapid monolithic manufacturing,”Cyborg and Bionic Systems, vol. 6, p. 0231, 2025

    work page 2025

  4. [4]

    9dtact: A compact vision-based tactile sensor for accurate 3d shape reconstruction and generalizable 6d force estimation,

    C. Lin, H. Zhang, J. Xu, L. Wu, and H. Xu, “9dtact: A compact vision-based tactile sensor for accurate 3d shape reconstruction and generalizable 6d force estimation,”IEEE Robotics and Automation Letters, vol. 9, no. 2, pp. 923–930, 2023

    work page 2023

  5. [5]

    Taceva: A performance evaluation framework for vision-based tactile sensors,

    Q. Cong, S. Oh, W. Fan, S. Luo, K. Althoefer, and D. Zhang, “Taceva: A performance evaluation framework for vision-based tactile sensors,” Advanced Intelligent Systems, p. e202501179, 2025

    work page 2025

  6. [6]

    Tirgel: A visuo-tactile sensor with total internal reflection mechanism for external observation and contact detection,

    S. Zhang, Y . Sun, J. Shan, Z. Chen, F. Sun, Y . Yang, and B. Fang, “Tirgel: A visuo-tactile sensor with total internal reflection mechanism for external observation and contact detection,”IEEE Robotics and Automation Letters, vol. 8, no. 10, pp. 6307–6314, 2023

    work page 2023

  7. [7]

    Gelsight: High-resolution robot tactile sensors for estimating geometry and force,

    W. Yuan, S. Dong, and E. H. Adelson, “Gelsight: High-resolution robot tactile sensors for estimating geometry and force,”Sensors, vol. 17, no. 12, p. 2762, 2017

    work page 2017

  8. [8]

    Design and benchmarking of a multimodality sensor for robotic manipulation with gan-based cross-modality interpretation,

    D. Zhang, W. Fan, J. Lin, H. Li, Q. Cong, W. Liu, N. F. Lepora, and S. Luo, “Design and benchmarking of a multimodality sensor for robotic manipulation with gan-based cross-modality interpretation,” IEEE Transactions on Robotics, vol. 41, pp. 1278–1295, 2025

    work page 2025

  9. [9]

    Geltip: A finger-shaped optical tactile sensor for robotic manipulation,

    D. F. Gomes, Z. Lin, and S. Luo, “Geltip: A finger-shaped optical tactile sensor for robotic manipulation,” in2020 IEEE/RSJ international conference on intelligent robots and systems (IROS). IEEE, 2020, pp. 9903–9909

    work page 2020

  10. [10]

    Using fiber optic bundles to miniaturize vision-based tactile sensors,

    J. Di, Z. Dugonjic, W. Fu, T. Wu, R. Mercado, K. Sawyer, V . R. Most, G. Kammerer, S. Speidel, R. E. Fanet al., “Using fiber optic bundles to miniaturize vision-based tactile sensors,”IEEE Transactions on Robotics, vol. 41, pp. 62–81, 2024

    work page 2024

  11. [11]

    Mechanical behaviour, contact pose estimation, and finite element analysis of vision based tactile sensors fabricated by molding and direct ink writing: A comparative study,

    A. Solayman, M. Halwani, I. M. Zaid, R. B. Ambade, R. Chang, A. Ayyad, F. AlMaskari, Y . Zweiri, and Y . Abdul Samad, “Mechanical behaviour, contact pose estimation, and finite element analysis of vision based tactile sensors fabricated by molding and direct ink writing: A comparative study,”Advanced Engineering Materials, vol. 26, no. 22, p. 2400630, 2024

    work page 2024

  12. [12]

    Design and evaluation of a rapid monolithic manufacturing technique for a novel vision-based tactile sensor: C-sight,

    W. Fan, H. Li, Y . Xing, and D. Zhang, “Design and evaluation of a rapid monolithic manufacturing technique for a novel vision-based tactile sensor: C-sight,”Sensors, vol. 24, no. 14, p. 4603, 2024

    work page 2024

  13. [13]

    Magictac: A novel high-resolution 3d multi-layer grid-based tactile sensor,

    W. Fan, H. Li, and D. Zhang, “Magictac: A novel high-resolution 3d multi-layer grid-based tactile sensor,” in2024 IEEE International Conference on Robotics and Automation (ICRA). IEEE, 2024, pp. 388–394

    work page 2024

  14. [14]

    Willemet,The Biomechanics of the Tactile Perception of Friction

    L. Willemet,The Biomechanics of the Tactile Perception of Friction. Springer, 2022

    work page 2022

  15. [15]

    A retrospective view of mixture experiments,

    J. A. Cornell, “A retrospective view of mixture experiments,”Quality engineering, vol. 23, no. 4, pp. 315–331, 2011

    work page 2011

  16. [16]

    Response surface methodology,

    A. I. Khuri and S. Mukhopadhyay, “Response surface methodology,” Wiley interdisciplinary reviews: Computational statistics, vol. 2, no. 2, pp. 128–149, 2010

    work page 2010

  17. [17]

    Desirability function approach: A review and performance evaluation in adverse conditions,

    N. R. Costa, J. Lourenc ¸o, and Z. L. Pereira, “Desirability function approach: A review and performance evaluation in adverse conditions,” Chemometrics and Intelligent Laboratory Systems, vol. 107, no. 2, pp. 234–244, 2011

    work page 2011

  18. [18]

    Deltact: A vision- based tactile sensor using a dense color pattern,

    G. Zhang, Y . Du, H. Yu, and M. Y . Wang, “Deltact: A vision- based tactile sensor using a dense color pattern,”IEEE Robotics and Automation Letters, vol. 7, no. 4, pp. 10 778–10 785, 2022

    work page 2022

  19. [19]

    Digit: A novel design for a low-cost compact high-resolution tactile sensor with application to in-hand manipulation,

    M. Lambeta, P.-W. Chou, S. Tian, B. Yang, B. Maloon, V . R. Most, D. Stroud, R. Santos, A. Byagowi, G. Kammereret al., “Digit: A novel design for a low-cost compact high-resolution tactile sensor with application to in-hand manipulation,”IEEE Robotics and Automation Letters, vol. 5, no. 3, pp. 3838–3845, 2020

    work page 2020

  20. [20]

    Ai-driven inverse design of materials: Past, present, and future,

    X.-Q. Han, X.-D. Wang, M.-Y . Xu, Z. Feng, B.-W. Yao, P.-J. Guo, Z.-F. Gao, and Z.-Y . Lu, “Ai-driven inverse design of materials: Past, present, and future,”Chinese Physics Letters, vol. 42, no. 2, p. 027403, 2025

    work page 2025

  21. [21]

    Progress and opportunities for machine learning in materials and processes of additive manufacturing,

    W. L. Ng, G. L. Goh, G. D. Goh, J. S. J. Ten, and W. Y . Yeong, “Progress and opportunities for machine learning in materials and processes of additive manufacturing,”Advanced Materials, vol. 36, no. 34, p. 2310006, 2024

    work page 2024

  22. [22]

    Rapid inverse design of metamaterials based on prescribed mechanical behavior through machine learning,

    C. S. Ha, D. Yao, Z. Xu, C. Liu, H. Liu, D. Elkins, M. Kile, V . Deshpande, Z. Kong, M. Bauchyet al., “Rapid inverse design of metamaterials based on prescribed mechanical behavior through machine learning,”Nature Communications, vol. 14, no. 1, p. 5765, 2023

    work page 2023

  23. [23]

    Tailoring stress–strain curves of flexible snapping mechanical metamaterial for on-demand mechanical responses via data-driven in- verse design,

    Z. Chai, Z. Zong, H. Yong, X. Ke, J. Zhu, H. Ding, C. F. Guo, and Z. Wu, “Tailoring stress–strain curves of flexible snapping mechanical metamaterial for on-demand mechanical responses via data-driven in- verse design,”Advanced Materials, vol. 36, no. 33, p. 2404369, 2024

    work page 2024

  24. [24]

    Machine learning of material proper- ties: Predictive and interpretable multilinear models,

    A. E. Allen and A. Tkatchenko, “Machine learning of material proper- ties: Predictive and interpretable multilinear models,”Science advances, vol. 8, no. 18, p. eabm7185, 2022

    work page 2022

  25. [25]

    Extreme vertices design of mixture experiments,

    R. McLean and V . Anderson, “Extreme vertices design of mixture experiments,”Technometrics, vol. 8, no. 3, pp. 447–454, 1966

    work page 1966

  26. [26]

    Use of extreme vertices method for analysis of how proportional composition affects component interactions and product distribution during hydrothermal treatment,

    Y . Gao, B. Yu, S. Xin, T. Mi, Y . Chen, Q. Yuan, L. Yang, and P. Li, “Use of extreme vertices method for analysis of how proportional composition affects component interactions and product distribution during hydrothermal treatment,”BioResources, vol. 11, no. 2, pp. 4899– 4920, 2016

    work page 2016

  27. [27]

    A hybrid multi-objective optimization of functional ink composition for aerosol jet 3d printing via mixture design and response surface methodology,

    H. Zhang, Z. Liu, S. Yin, and H. Xu, “A hybrid multi-objective optimization of functional ink composition for aerosol jet 3d printing via mixture design and response surface methodology,”Scientific Reports, vol. 13, no. 1, p. 2513, 2023

    work page 2023

  28. [28]

    The beer-lambert law,

    D. F. Swinehart, “The beer-lambert law,”Journal of chemical education, vol. 39, no. 7, p. 333, 1962

    work page 1962

  29. [29]

    The fitting of scheff ´e-type models for estimating solubil- ities of multisolvent systems,

    J. A. Cornell, “The fitting of scheff ´e-type models for estimating solubil- ities of multisolvent systems,”Journal of Biopharmaceutical Statistics, vol. 1, no. 2, pp. 303–329, 1991

    work page 1991

  30. [30]

    Topology and material optimization in ultra-soft magneto- active structures: Making advantage of residual anisotropies,

    C. Perez-Garcia, R. Ortigosa, J. Mart ´ınez-Frutos, and D. Garcia- Gonzalez, “Topology and material optimization in ultra-soft magneto- active structures: Making advantage of residual anisotropies,”Advanced Materials, p. e18489, 2025. APPENDIX A. Material Property Evaluation Details

    work page 2025

  31. [31]

    In tactile sensing applications, an elastomer with appropriate compli- ance can respond effectively to contact stimuli through mea- surable surface deformation

    Hardness:For elastomers, softness is a key material attribute that distinguishes them from rigid solids. In tactile sensing applications, an elastomer with appropriate compli- ance can respond effectively to contact stimuli through mea- surable surface deformation. Hardness is commonly used as a quantitative metric (the inverse of softness), describing th...

    work page

  32. [32]

    In tactile sensing applications, near-incompressibility (v= 0.5) implies that geometric deformation under external loading occurs with limited volume variation

    Poisson’s Ratio:Poisson’s ratiovis a fundamental mechanical parameter describing the relationship between transverse strainε trans and axial strainε axial under uniaxial loading. In tactile sensing applications, near-incompressibility (v= 0.5) implies that geometric deformation under external loading occurs with limited volume variation. v=− dεtrans dεaxi...

    work page

  33. [33]

    Its primary objective is to obtain stress-strain curves for analysing tensile properties such as elastic modulus and tensile hysteresis

    Uniaxial Tensile Experiment:Uniaxial tensile testing is widely used to characterise the mechanical behaviour of elastomers under unidirectional loading. Its primary objective is to obtain stress-strain curves for analysing tensile properties such as elastic modulus and tensile hysteresis. ASTM D412 is a well-established standard for evaluating the tensile...

    work page

  34. [34]

    Consequently, the elastomer layer of the tactile sensor pri- marily undergoes compressive deformation during operation

    Uniaxial Compressive Experiment:In practical tactile sensing scenarios, the dominant loading mode is compression, as normal force is required to establish effective contact. Consequently, the elastomer layer of the tactile sensor pri- marily undergoes compressive deformation during operation. Understanding the compressive behaviour of candidate mate- rial...

    work page