{"id":"90497293-78f7-4bed-8b2e-08839f59fae7","arxiv_id":"2512.12614","paper_version":2,"verdict":"CONDITIONAL","confidence":"MODERATE","novelty_score":6.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":9,"one_line_summary":"A Lode-invariant energy-limiting constitutive model with separate tensile and compressive failure energies reproduces tension-compression asymmetry and predicts pure shear failure of agarose hydrogels from uniaxial calibration alone.","lead":"This paper builds a new model that lets soft materials \"decide\" whether to fail in tension or compression differently, using a single mathematical framework. Tests on agarose gels show it can predict an unseen deformation mode (pure shear) after calibration on tension and compression data alone.","discovery_kind":"new_method","skeptic_critique":{"model":"deepseek-v4-flash","headline":"Pure-shear validation pins β(K3) at only one intermediate point; the linear Lode-angle interpolation, Eq. (35), is untested and yet underlies the 3D failure-mapping claims.","rationale":"The reader's weakest-assumption identification is correct and is also the most load-bearing concern. The central contribution is the bi-failure construction with mode-dependent energy limiters, and the headline validation is the pure-shear prediction. That prediction is real, out-of-sample, and should be credited. But because β(K3=0)=1/2 is fixed by endpoints for any reasonable interpolation, the pure-shear match does not probe the shape of β elsewhere. The linear form is therefore an unvalidated assumption that becomes the sole basis for claims about arbitrary intermediate distortion modes, including the 'foundation for three-dimensional failure mapping.' The paper's own Section 7 explicitly concedes this. The energy-landscape 'stability' argument is also visual rather than formal, so it does not close the gap. The concern is not that the model is wrong, but that the evidence supports only K3 = 0, -π/6, and π/6. My verdict remains conditional, matching the reader's: accept the core result with the condition that β be justified or tested at additional Lode angles. I agree with the reader rather than raising a separate objection.","tokens_in":40823,"tokens_out":4513,"duration_ms":51650,"concrete_test":"Perform unequal biaxial tension tests on 2% w/v agarose with in-plane stretch ratios such as λ1:λ2 = 2:1 (e.g., λ1=2, λ2=1.5, λ3=1/3), which gives an intermediate Lode angle K3 ≈ -21°, between uniaxial compression (-30°) and pure shear (0°). Using parameters calibrated solely from uniaxial tension and compression, predict the biaxial stress-stretch response via Eq. (36) with the linear β. Compare predicted failure stress/stretch to experiment. A systematic deviation would falsify the linear interpolation; agreement would independently support it. As a control, also compute predictions with nonlinear β forms that preserve β(0)=1/2 to confirm the test's sensitivity to β away from K3=0.","verdict_should_be":"UNCHANGED","load_bearing_attack":"The strongest claim—that uniaxial calibration predicts pure shear—is credible and is a genuine out-of-sample test. However, the broader claim of a foundation for 3D failure mapping rests on the assumed linear weighting β(K3) = (K3 + π/6)/(π/3), Eq. (35). The only intermediate mode tested is pure shear, where K3 = 0 and β = 1/2. A match at this single point does not test linearity: any continuous β with β(-π/6)=0, β(π/6)=1, and β(0)=1/2 yields exactly the same pure-shear prediction. The paper itself acknowledges this in Section 7: the Lode-weighting function 'was assumed to vary linearly based on validation across the three primary distortional modes.' Thus the pure-shear experiment validates β only at K3 = 0, and no evidence distinguishes Eq. (35) from a nonlinear interpolation. The energy-landscape analysis in Section 6.4 is visual and does not formally establish thermodynamic admissibility or convexity; it is used as additional support for the interpolation but cannot resolve the β ambiguity. Consequently, the pure-shear prediction is supported, but the claim of a general distortion-mode-dependent failure framework across the full Lode space is not yet established.","agreement_with_reader":"agree"},"referee_report":{"model":"deepseek-v4-flash","summary":"The paper proposes a Lode-invariant-based hyperelastic softening model for soft materials that couples Volokh's energy-limiting framework with separate tensile and compressive energy limiters. Mode dependence is introduced through a linear Lode-angle weighting function β(K3), and the intact response is described by the Prasad–Kannan hyperelastic potential. The model is calibrated to combined uniaxial tension and compression data for agarose hydrogels at 1, 2, and 3% w/v, then used to predict pure shear response and, via power-law concentration scaling, the response at an intermediate 2.5% w/v concentration. The free-energy landscape over the (K2,K3) invariant space is presented as evidence of thermodynamic stability.","tokens_in":41230,"tokens_out":8927,"duration_ms":93181,"significance":"If the central claims hold, the contribution is significant: an eight-parameter, physically interpretable constitutive framework that maps failure energetics across distortion modes using only uniaxial tension and compression calibration. The pure-shear prediction is a genuine out-of-sample test, and the reported agreement with experiments is the strongest part of the paper. The use of Hencky-Lode invariants is well motivated, and the absence of internal damage variables is a practical advantage. The power-law interpolation across concentrations is also a useful practical tool. However, the broader claim that the framework provides a foundation for three-dimensional failure mapping rests on an assumed linear interpolation in K3 that is only validated at a single intermediate point, and the thermodynamic-consistency claim is supported by visual inspection rather than formal analysis. These issues are load-bearing for the paper's most general claims.","major_comments":[{"comment":"The linear Lode-weighting function β(K3)=(K3+π/6)/(π/3) is assumed, not derived or independently tested. The pure-shear prediction fixes β only at K3=0, where β=1/2; any continuous function with β(−π/6)=0, β(π/6)=1, and β(0)=1/2 gives exactly the same pure-shear prediction as Eq. (49). The manuscript's own Section 7 recognizes this: the Lode-weighting function 'was assumed to vary linearly based on validation across the three primary distortional modes.' Since every intermediate-mode prediction in the paper, including the stated 3D failure-mapping foundation, passes through this β, the cross-mode predictive claim is currently limited to pure shear. Additional multiaxial experiments at other K3 values, or a physics-based derivation for β, are needed before the 3D failure-mapping claim can be considered established.","section":"§7, Eq. (35)"},{"comment":"The printed pure-shear stress formula contains a factor-of-two error. Equation (48) gives T(ps)=√2 (∂W/∂K2)(∂ψprop/∂W). Since P=T/λ, Eq. (49a) should read P(ps)=√2/λ (∂W/∂K2)(∂ψprop/∂W), not 1/(√2 λ) times that quantity. As written, Eq. (49a) cannot reproduce the pure-shear curves in Fig. 8 unless the implementation used a different factor. This needs correction and verification against the code used for Fig. 8. Separately, the statement that γ2=1/K2 ∂W/∂K3 vanishes at K3=0 is not generally true for the Prasad–Kannan potential, since G'(0) is generally nonzero. For pure shear, T1 happens to be independent of γ2 because the 11 and 33 components of N2 coincide, but the derivation as written is not correct.","section":"§4.2.2, Eq. (49a)"},{"comment":"Thermodynamic admissibility is asserted from visual inspection of the free-energy surfaces. The text states that 'no regions of singularity or discontinuous curvature are observed' and that 'continuous and convex topology' validates internal consistency. For a softening elastic model, convexity of the energy is not the relevant criterion near failure; loss of strong ellipticity is expected in the softening regime. The paper does not provide formal conditions for boundedness of ∂ψ/∂K2 and ∂ψ/∂K3, positivity of the dissipation, or the range of stretches over which the model remains well posed. A formal analysis, or at least a delineation of the conditions under which the energy remains rank-one convex in the intact regime, is required to support the thermodynamic-consistency claim.","section":"§6.4, Eq. (36)"},{"comment":"The power-law concentration scaling is applied to all eight parameters, including m+. However, Table 3 shows no systematic concentration trend for m+: the averages are 139.22, 96.54, and 210.20 for 1, 2, and 3% w/v, with within-concentration replicate scatter as large as m+=5.99 vs. 287.34 at the same concentration. Fitting a power law to three averages, with no reported R² or confidence intervals, and then interpolating m+=154.47 at 2.5% w/v is not a reliable procedure. The 2.5% w/v validation is a legitimate independent test, but for m+ it is a model-selection artifact rather than a meaningful power-law prediction. The power-law claims should be restricted to parameters with clear monotone concentration dependence, and goodness-of-fit/uncertainty should be reported.","section":"§6.5, Table 3 and Table 5"}],"minor_comments":[{"comment":"The relative-error metric uses max{0.1max(P), |P_i|} in the denominator, which mixes a data-dependent scale with the pointwise value. This makes the reported 'average residual error' in Table 3 nonstandard and difficult to compare across datasets. Consider also reporting normalized RMSE or R² values.","section":"§5.3, Eq. (51)"},{"comment":"The paper repeatedly states that 'thermodynamic admissibility' is enforced during calibration and that the model is 'thermodynamically consistent,' but no formal admissibility constraint is visible in the optimization, except parameter positivity. Please clarify what admissibility condition is enforced in the fitting procedure.","section":"§6.1, Table 3"},{"comment":"Uniaxial tests are reported with n=4 per concentration, while pure shear experiments appear to use n=3. State explicitly why the replicate counts differ, and whether the pure-shear prediction uses average parameters from all four uniaxial replicates or a subset.","section":"§5.2.3 and Fig. 8"},{"comment":"The data availability statement says 'Data will be made available on request.' For a constitutive-validation paper of this type, deposition of the reduced stress–stretch datasets and fitting/prediction scripts would substantially increase reproducibility.","section":"Data accessibility"}],"recommendation":"major_revision","confidential_remarks":"The out-of-sample pure-shear prediction is the paper's main strength and appears credible; the concentration interpolation is also a useful test. The two issues that need the most attention are the factor error in Eq. (49a) and the unsupported linearity of β(K3) relative to the 3D failure-mapping claim. The thermodynamic-consistency argument also needs to be made quantitative. These are fixable within the scope of the manuscript, so I recommend major revision rather than rejection."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"Short take: this is a genuinely new constitutive construction, and the pure-shear prediction is a real out-of-sample success. The paper deserves refereeing, but it overclaims at the edges.\n\nWhat's new: the bi-failure limiter with distinct Φ±, m±, interpolated by Lode angle β(K3) over the Prasad–Kannan potential, is not in the prior literature. The derivation is careful, and the pure-shear response predicted from uniaxial-only calibration tracks the agarose data at all three concentrations. That is a legitimate demonstration, not a restatement of earlier work. The uniaxial fits are solid, with errors mostly below 10%.\n\nSoft spots, in order of seriousness. First, the linear Lode weighting β(K3) in Eq. (35) is assumed. The pure-shear test only exercises K3 = 0, so it pins the interpolation to a single intermediate point. Any continuous curve through the three endpoints reproduces that prediction. The paper acknowledges this in Section 7, but the abstract's broader claim of a foundation for 3D failure mapping is not yet earned. Second, 'thermodynamic consistency' is asserted from visual inspection of the energy surfaces. A formal convexity or stability check is missing, and the claim should be downgraded until it is supplied. Third, the 2.5% w/v prediction relies on power-law interpolation, but m+ shows no systematic concentration trend; the interpolated m+ is essentially a free value forced through a power-law fit. That prediction is correspondingly less persuasive than the pure-shear test. Fourth, data are 'available on request' and no code is shipped, so independent verification is not possible.\n\nWho should read it: anyone modeling soft-tissue or gel failure who wants a compact mode-dependent energy-limiter formulation. It is a solid, useful paper, not a field-changer. Send it to peer review, but the authors should either justify the linear β with a second intermediate deformation mode or soften the 3D-mapping claim, add a formal stability check, and address the m+ interpolation. That is a reasonable revision path.","headline":"Useful new construction with a credible pure-shear out-of-sample test, but the linear Lode interpolation and thermodynamic consistency claims are softer than the abstract implies.","tokens_in":41685,"tokens_out":2553,"would_cite":true,"duration_ms":28601,"reading_group":"maybe","serious_thinker":"yes","would_accept_peer_review":true},"rs_alignment":null,"lean_confirmation":null,"pith_extraction":{"msc":["74B20","74R20"],"pacs":["62.20.F-","62.20.mm"],"model":"deepseek-v4-flash","headline":"A single Lode-weighted energy potential captures tension–compression-asymmetric softening and failure in soft materials, with uniaxial-calibrated parameters predicting pure shear without refitting.","keywords":["hyperelasticity","energy limiters","tension–compression asymmetry","Lode invariants","softening and failure","agarose hydrogels","constitutive modeling","soft materials"],"falsifier":"Measure failure under a distortion mode away from pure shear — unequal biaxial tension, or combined compression–shear — and compare the predicted failure stress and stretch, computed from uniaxial-calibrated parameters through β(K3), against experiment. A discrepancy that grows as the loading moves off the pure-shear midpoint would falsify the linear interpolation; repeating the comparison on a non-hydrogel soft material would test whether the construction generalizes beyond agarose.","tokens_in":40720,"feed_emoji":"🧪","tokens_out":10804,"duration_ms":101592,"temperature":0.7,"pith_summary":"This paper sets out to show that tension–compression-asymmetric softening and failure in soft materials can be captured by one strain-energy potential that never invokes internal damage variables: it gives tensile and compressive distortion their own energy ceilings and blends them with the Lode angle, the coordinate that labels the deformation mode. Fitted only to combined uniaxial tension and compression data for agarose hydrogels (1, 2, and 3% w/v), the model is claimed to predict the pure-shear response—stiffness, softening, and rupture—with no additional fitting, a cross-mode transfer most existing continuum damage laws do not offer. If the claim holds, complete failure characterization reduces to one tension test and one compression test per material, expressed through eight physically interpretable parameters: four elastic, four failure-related. The framework also yields power-law concentration scaling, letting an untested gel composition (2.5% w/v) be predicted rather than measured. A sympathetic reader would care because mode-aware failure is exactly the missing piece for predicting rupture in gels, tissues, and elastomers under multiaxial loading.","feed_headline":"Two uniaxial tests predict how soft gels fail in shear","feed_subtitle":"A Lode-angle blend of tension and compression energy limits forecasts pure-shear softening and rupture.","key_machinery":"Key machinery: the Lode angle K3 of the logarithmic strain, which uniquely labels the distortion mode (−π/6 uniaxial compression, 0 pure shear, +π/6 uniaxial tension). On it rests a bi-failure potential: two energy-limiting branches, each capping the intact energy W at a pseudo-failure energy Φ± with softening sharpness m±, are blended by a linear weight β(K3) = (K3 + π/6)/(π/3), so pure shear (β = 1/2) inherits the arithmetic mean of tensile and compressive stress-reduction factors. The intact response comes from a Lode-invariant hyperelastic potential with a mode-dependent modulus G(K3), giving eight total parameters fitted from combined uniaxial data; no internal damage variables appear.","core_discovery":"The central claim is that soft-material failure can be made an explicit function of the distortion mode through a bi-failure construction: the strain-energy density splits into a mode-dependent failure energy and a mode-dependent recoverable elastic energy, each a Lode-angle-weighted blend of separate tensile and compressive branches. The Lode angle K3 of the logarithmic strain labels the mode (−π/6 compression, 0 pure shear, +π/6 tension); a linear weight β(K3) interpolates between branches. Each branch caps the intact energy W at a pseudo-failure energy Φ± with softening sharpness m±, yielding the stress-reduction factor exp[−(W/Φ±)^m±]; at pure shear the prediction is the arithmetic mean","pith_inferences":["The linear Lode-angle weight β(K3) is the link that is assumed rather than measured: the paper checks it only at the pure-shear midpoint (K3 = 0) and only for agarose. A test under a second intermediate mode — unequal biaxial stretch or combined compression–shear — would reveal whether linear mode interpolation survives or needs a nonlinear form.","The intact potential used as the elastic baseline was originally built for brain tissue, so the same bi-failure construction is transferable to other soft materials; the open question is whether the linear mode rule is generic or specific to agarose-like networks.","The formulation is presented for monotonic loading with reversibility (failure energy can elastically recover on unloading); extending to tearing, cyclic, or post-failure protocols would require activating the irreversible switch the paper introduces but sets to zero.","If the interpolation rule is later fitted to true multiaxial data instead of assumed, the same structure becomes a practical tool for mapping full three-dimensional failure envelopes in (K2, K3) space — the paper's own stated next step."],"forward_implications":["One tension test plus one compression test per material may suffice to fix failure behavior across the distortion-mode range, with pure shear following automatically from the Lode-angle midpoint.","The pure-shear prediction is explicit and testable: the stress-reduction factor there equals the arithmetic mean of the tensile and compressive reduction factors.","Failure energies in agarose are roughly an order of magnitude larger in compression than in tension; the model encodes that ratio as a continuous function of the Lode angle rather than as two disconnected fits.","Because fitted parameters scale as power laws in gel concentration, response at intermediate compositions can be interpolated — demonstrated at 2.5% w/v for both uniaxial and pure-shear loading.","The framework yields a single free-energy surface over the (K2, K3) invariant space, which the paper identifies as the foundation for constructing full three-dimensional failure maps of soft materials."],"fun_headline_variants":["Two uniaxial tests predict shear failure in soft gels","Lode-angle weighting predicts gel shear failure from uniaxial tests","Bi-failure model makes pure-shear predictions from tension/compression data","Shear softening in gels predicted from uniaxial calibration"],"cache_read_input_tokens":2304,"weakest_assumption_plain":"The load-bearing premise is that softening and failure vary linearly with the Lode angle between the compression and tension extremes: the weighting function β(K3) is assumed, not derived, and is checked only at the pure-shear midpoint and only for agarose hydrogels.","fun_headline_variants_meta":{"raw":{"variants":["Two uniaxial tests predict shear failure in soft gels","Lode-angle weighting predicts gel shear failure from uniaxial tests","Bi-failure model makes pure-shear predictions from tension/compression data","Shear softening in gels predicted from uniaxial calibration"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.000412,"raw_usage":{"total_tokens":2011,"prompt_tokens":825,"completion_tokens":1186,"prompt_tokens_details":{"cached_tokens":256},"prompt_cache_hit_tokens":256,"prompt_cache_miss_tokens":569,"completion_tokens_details":{"reasoning_tokens":1116}},"tokens_in":569,"tokens_out":1186,"duration_ms":11847,"temperature":1.0,"reasoning_tokens":1116,"cache_read_input_tokens":256,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-03T16:34:29.469941+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"Measure failure under a distortion mode away from pure shear — unequal biaxial tension, or combined compression–shear — and compare the predicted failure stress and stretch, computed from uniaxial-calibrated parameters through β(K3), against experiment. A discrepancy that grows as the loading moves off the pure-shear midpoint would falsify the linear interpolation; repeating the comparison on a non-hydrogel soft material would test whether the construction generalizes beyond agarose.","supporting_citations":[],"review_version":1}