{"id":"f15c6e6c-fdd7-4338-b4a2-4743b638e5f8","arxiv_id":"2506.16479","paper_version":1,"verdict":"CONDITIONAL","confidence":"MODERATE","novelty_score":7.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":6,"one_line_summary":"A phase-field fracture theory for viscoelastic elastomers couples deformation, strength, and fracture energy, and matches experiments on rubber bands, acrylic sheets, and rubber tearing tests.","lead":"The paper builds a mathematical model that predicts where and when cracks start and grow in stretchy plastics called elastomers under slow loading. It combines three measured material properties: how the material deforms over time, how strong it is, and how much energy is needed to tear it.","discovery_kind":"extension","skeptic_critique":{"model":"deepseek-v4-flash","headline":"Most load-bearing gap: the coefficient δ^ε in Eq. (26) is imported from elastic-brittle theory without derivation, yet it controls Griffith crack growth in the viscoelastic phase-field model.","rationale":"The reader's weakest assumption is the rate-independence of the strength surface. I think that assumption is less damaging than the paper's treatment of δ^ε. First, the experimental record cited in §2.2.1 (Smith's failure envelopes, Knauss's surface) gives direct support for a rate-independent failure locus in (S,F) space; different rates then appear because the stress-deformation loading path itself depends on rate. The δ^ε formula, in contrast, is a mathematical calibration on which all crack-growth predictions depend, and the paper provides only an analogy to the elastic case. Second, the central claim is that (29)–(31) predict propagation and nucleation from cracks, not just bulk nucleation; every such prediction passes through δ^ε and the equilibrium-energy Griffith condition. Without a derivation or a convergence study, the numerical matches in §5.2–§5.3—which are at one ε value and partly qualitative—cannot distinguish a correct theory from a tuned interpolation. I therefore flag this as the single most load-bearing concern. I agree with the reader's overall CONDITIONAL verdict, and no change to that verdict is needed; the missing piece is an independently checkable derivation or ε-convergence test. I credit the paper for its explicit acknowledgment in §6 of excluded mechanisms (fatigue, strain-induced crystallization, inertia), which appropriately bounds the “complete” claim.","tokens_in":43329,"tokens_out":10521,"duration_ms":117284,"concrete_test":"Perform a sharp-interface asymptotic check or an ε-convergence study: re-derive Eq. (26) from the limit ε → 0 of the phase-field equations (29)–(31) for a stationary crack in a finite-strain viscoelastic solid, or, computationally, rerun the §5.2 pure-shear test with ε = 0.25, 0.5, and 1.0 mm (keeping h = ε/5) at the six experimental stretch rates and compare the predicted onset stretch Λ_c with the sharp-interface value Λ_c = 3.63 from Eq. (7). If Λ_c shifts by more than the experimental scatter across ε or differs systematically from 3.63, the imported formula is not valid for viscoelastic materials; if it is ε-insensitive and matches, the concern is resolved.","verdict_should_be":"UNCHANGED","load_bearing_attack":"The paper's crack-growth predictions hinge on the coefficient δ^ε defined in Eq. (26). This coefficient appears in the phase-field equation (31) as the multiplier of εG_c in the divergence term and also in the driving force c_e through Eqs. (27)–(28); it is the piece that ensures nucleation from large pre-existing cracks and propagation follow the sharp Griffith condition (7). The paper does not derive Eq. (26) for viscoelastic materials. Section 3.5.4 simply states that because c_e and the coefficients (25) are analogous to those used by Kamarei et al. (2024) for purely elastic elastomers, “the same type of formula derived for δ^ε by these authors applies here.” That transfer is non-trivial: the invariants I_1, I_2 entering c_e through Eq. (21) include the non-equilibrium stress through ψ_NEq and C_v, so the elastic derivation does not automatically carry over. If Eq. (26) is not the correct asymptotic calibration for finite viscoelasticity, the predicted critical stretch and stress for crack growth will deviate from −∂W^Eq/∂Γ_0 = G_c, and the central claim of a complete predictive fracture theory fails at its most consequential point. The Section 5.2 comparison uses a single regularization length ε = 0.5 mm, and Section 5.3 is only qualitative; neither supplies the missing derivation nor an ε-convergence check.","agreement_with_reader":"disagree"},"referee_report":{"model":"deepseek-v4-flash","summary":"The paper develops a macroscopic phase-field theory of fracture nucleation and propagation in viscoelastic elastomers under quasistatic loading. It extends the Kumar-Francfort-Lopez-Pamies phase-field approach by coupling finite viscoelasticity (a two-potential model with internal variable C_v), a Drucker-Prager-type strength surface in stress-deformation space, and a Griffith criticality condition in which only the equilibrium free energy competes with the fracture energy. The governing equations are (29)-(31), and a numerical scheme based on Crouzeix-Raviart finite elements and explicit Runge-Kutta time stepping is presented. The theory is illustrated by simulations of uniaxial ring tests on polyurethane (Mueller 1968), pure-shear tests on VHB 4905 (Pharr et al. 2012), and trousers tests on SBR (Greensmith and Thomas 1955).","tokens_in":43697,"tokens_out":5962,"duration_ms":62297,"significance":"If the theory is correct, it would be a substantial contribution: it is, to my knowledge, the first phase-field framework for elastomers that simultaneously accounts for strength and viscoelasticity-related toughness in a way consistent with the equilibrium-energy Griffith condition, and it offers a complete numerical methodology. The manuscript is careful in assembling the experimental evidence, and the numerical implementation appears robust. The main caveats are that the central calibration coefficient delta^epsilon is imported from an elastic-brittle derivation without proof, and that parts of the validation reuse parameters fitted to the same experiments, making some comparisons descriptive rather than predictive. These issues do not invalidate the framework, but they need to be addressed before the paper can be judged as establishing a complete and predictive theory.","major_comments":[{"comment":"The coefficient delta^epsilon in Eq. (26) controls both the divergence term in the phase-field equation (31) and, through Eqs. (27)-(28), the magnitude of the driving force; it is the mechanism by which crack growth from large pre-existing cracks is forced to satisfy the Griffith condition (7). The paper does not derive this coefficient for the viscoelastic setting. Section 3.5.4 states that the prescription is analogous to that of Kamarei et al. (2024) and therefore the same formula applies. That inference is not immediate: the invariants I_1 and I_2 entering c_e through Eq. (21) contain the non-equilibrium stress through psi_NEq and C_v, so the asymptotic expansion that produced delta^epsilon in the elastic-brittle case must be re-examined for finite viscoelasticity. If Eq. (26) is not the correct asymptotic calibration, the predicted critical crack-growth stretch will deviate from -dW^Eq/dGamma_0 = G_c, and the central claim fails at its most consequential point. The manuscript should either supply a derivation or demonstrate by epsilon-convergence studies that the sharp-interface limit is recovered. Section 5.2 uses a single regularization length epsilon = 0.5 mm and Section 5.3 is only qualitative, so neither supplies that evidence.","section":"Section 3.5.4, Eq. (26)"},{"comment":"The VHB comparison in Section 5.2 does not independently test the crack-growth mechanism. The material constants in Table 4 and the fracture energy G_c = 634 N/m are taken from Shrimali and Lopez-Pamies (2023b), where they were used to explain the same Pharr et al. (2012) experiments; the strength constants in Table 5 are stipulated as being consistent with experiments. More importantly, Eq. (26) is constructed so that growth from large cracks obeys Eq. (7), and the predicted critical stretch Lambda_c = 3.63 is the value already imposed by Eq. (7). The agreement in Fig. 13 therefore confirms internal consistency and the numerical implementation, but it is not an independent validation that the theory predicts nucleation from pre-existing cracks. A test using independently measured strength and toughness, in a different geometry or loading history, would be needed.","section":"Section 5.2 and Eq. (7)"},{"comment":"For the Mueller ring tests, the viscoelastic constants in Table 2 are fitted to the slowest and fastest stress-stretch curves, and the strength constants in Table 3 are fitted to the critical stress-stretch pairs for all ten stretch rates. Consequently, Fig. 9 demonstrates that the framework can describe the data set used for calibration, but it does not provide predictive evidence in the sense claimed in the abstract and Section 6. The authors should either calibrate on a subset of rates and validate on the remaining ones, or temper the predictive claim for this example.","section":"Section 5.1, Tables 2 and 3"},{"comment":"The strength surface is assumed to be independent of loading history, so that it can be written as F(S,F) = 0. The manuscript cites experimental support to a first degree of approximation, but this is a central assumption for the nucleation mechanism, since c_e in Eqs. (27)-(28) contains no explicit time dependence. The available data are mostly uniaxial and equi-biaxial tests; no direct evidence is provided for arbitrary multiaxial histories or for non-monotonic loading paths. The authors should state this assumption more prominently and, if possible, test the predicted nucleation under a relaxation or non-monotonic history where rate effects would be distinguishable from the stress-deformation correlation.","section":"Section 2.2.1, Eq. (2)"}],"minor_comments":[{"comment":"There are typographical errors in the reference list: 'Jounal of Applied Mechanics' in Chockalingam (2025) and 'Journal of the Mechancis and Physics of Solids' in Breedlove et al. (2024) should be corrected.","section":"References"},{"comment":"The h-correction formula for delta^epsilon would benefit from explicit parentheses; as typeset, '1 + 3/8 h/epsilon' is ambiguous and could be misread as 1 + 3/(8h/epsilon) rather than 1 + (3/8)(h/epsilon).","section":"Remark 10"},{"comment":"Table 5 lists a2 = 0 but omits b2, although equation (5) contains b2. The authors should state that b2 is unused when a2 = 0, or add the value for completeness.","section":"Table 5"},{"comment":"The trousers comparison is explicitly qualitative, but the text around Fig. 16(b) says the theory 'predicts accurately' the propagation described by Eq. (8). Since no quantitative comparison with the Greensmith and Thomas data is made, the wording should be softened to 'reproduces qualitatively' to avoid overclaiming.","section":"Section 5.3"}],"recommendation":"major_revision","confidential_remarks":"The main risk in this manuscript is the unproven transfer of the delta^epsilon formula from elastic-brittle fracture to finite viscoelasticity. I would ask the authors to provide either a derivation or a convincing epsilon-convergence study before publication. The validation claims should also be adjusted to distinguish descriptive fits from independent predictions. The paper is otherwise substantial and likely acceptable after these revisions."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"Colleague,\n\nHere is my read of 2506.16479. The genuinely new thing is the combination: a phase-field formulation for viscoelastic elastomers that includes a strength surface in stress-deformation space (not just stress space) and uses only the equilibrium stored energy in the Griffith competition. That is a real step beyond existing viscoelastic phase-field models, which put the entire stored elastic energy into the crack-driving term and have no strength-based nucleation. The framework itself—equations (29)-(31) with y, C_v, and z—is carefully assembled, and the numerical scheme (Crouzeix-Raviart plus explicit RK, with care for det C_v = 1) is a serious engineering contribution.\n\nThe three sample simulations are also instructive. The Mueller ring comparison shows the theory can reproduce the rate-dependent critical stress and stretch. The VHB pure-shear and SBR trousers simulations are consistent with the equilibrium-energy Griffith story. But the validation is not as strong as the word “predict” in the title suggests.\n\nFirst, many parameters enter from the same experiments being “predicted.” For Mueller, the viscoelastic constants are fit to the slow and fast curves, and the strength constants are fit to all ten rate data points. For VHB, the viscoelastic parameters and G_c are taken from a prior analysis and the strength constants are stipulated. For trousers, it is a canonical elastomer and the comparison is only qualitative.\n\nSecond, there is construction-level circularity. The driving force c_e is designed so that, by construction, uniform stress states nucleate exactly on the strength surface and large cracks grow at the equilibrium-energy Griffith condition. So the VHB onset at Lambda_c ≈ 3.63 is not an independent confirmation; it is the input (7) dressed in phase-field clothing.\n\nThird, and most important, the δ^ε coefficient in Eq. (26) is imported from the elastic-brittle paper by Kamarei et al. (2024) without a derivation for the viscoelastic case. That coefficient is the piece that makes crack growth follow −∂W^Eq/∂Γ_0 = G_c. Because the invariants in c_e now depend on the non-equilibrium stress through C_v, the elastic derivation does not automatically carry over. There is also no ε-convergence study. This is a load-bearing gap, not a cosmetic one. I would want the derivation or a careful asymptotic/numerical check before trusting the propagation predictions.\n\nThe paper is still worth a serious referee. The right outcome is not rejection but major revision: derive or justify δ^ε in finite viscoelasticity, add an ε-convergence check, and ideally include one blind prediction with all parameters fixed before the experiment is shown. I would not cite it yet as a validated model, but I would read it and probably discuss it.","headline":"A genuine first: a phase-field theory for viscoelastic elastomers that pairs a strength surface with the equilibrium-energy Griffith condition, but the load-bearing δ^ε coefficient is imported rather than derived and the validation is partly circular.","tokens_in":44235,"tokens_out":3205,"would_cite":false,"duration_ms":33124,"reading_group":"maybe","serious_thinker":"yes","would_accept_peer_review":true},"rs_alignment":null,"lean_confirmation":null,"pith_extraction":{"msc":["74R10","74D10","74S05"],"pacs":[],"model":"deepseek-v4-flash","headline":"One phase-field theory now predicts both nucleation and propagation of fracture in viscoelastic elastomers.","keywords":["viscoelastic elastomers","phase-field fracture","fracture nucleation","strength surface","Griffith fracture energy","finite viscoelasticity","quasistatic loading","internal variables"],"falsifier":"Run pure-shear crack-nucleation tests on the same elastomer over stretch rates spanning at least six decades; if the critical stretch at which the pre-existing crack starts to grow changes by substantially more than experimental scatter, the theory's claim that only the equilibrium stored energy enters the Griffith condition is wrong.","tokens_in":43062,"feed_emoji":"💥","tokens_out":6343,"duration_ms":63473,"temperature":0.7,"pith_summary":"This paper seeks to establish a complete macroscopic phase-field theory that predicts where and when cracks nucleate and how they propagate in viscoelastic elastomers under slow, quasistatic loads. It argues from experiments that any such theory needs three ingredients: finite viscoelasticity, a strength surface defined in stress-deformation space, and a fracture energy. The central claim is that nucleation is governed by this strength surface while propagation and growth from large pre-existing cracks follow a Griffith condition in which only the equilibrium part of the stored elastic energy competes with the fracture energy. Simulations of three classic experiments—rubber-band uniaxial tension, pure-shear crack nucleation, and trousers tearing—match published data, which the authors present as evidence that the framework is complete. The theory reduces to solving two nonlinear PDEs coupled with a nonlinear ODE for the deformation field, a tensorial internal variable, and the phase field.","feed_headline":"Phase-field theory predicts when and where rubber cracks","feed_subtitle":"Couples viscoelasticity, a strength surface, and equilibrium-energy fracture to match classic experiments.","key_machinery":"The core object is the phase field z(X,t), an order parameter taking z=1 in intact material and z=0 in fractured material, coupled to an internal variable C_v that stores the viscous part of the deformation. The load-bearing identity is the reduced Griffith criticality condition -∂W^Eq/∂Γ0 = G_c, which says that crack growth is driven only by the equilibrium stored energy W^Eq competing with the fracture energy G_c. The strength of the elastomer enters through a Drucker-Prager-type strength surface F(S,F)=0, whose violation under uniform stress activates a driving force c_e inside the phase-field equation. These ingredients together generate the governing system (29)-(31): balance of linear momentum, the internal-variable evolution equation, and the phase-field equation with inequalities expressing irreversibility and the bounds 0≤z≤1.","core_discovery":"The paper claims that fracture in viscoelastic elastomers can be described by a single phase-field model in which the deformation field y(X,t), a tensorial internal variable C_v(X,t) tracking viscous relaxation, and a phase field z(X,t) evolve together under quasistatic loading. The key physical assertion is that the Griffith energy competition for crack growth involves only the equilibrium portion of the stored elastic energy—the energy that would remain if the material were held at fixed deformation until all viscous relaxation finished—and not the total stored or dissipated energy. Nucleation under spatially uniform stress is instead assigned to a strength surface F(S,F)=0 in stress-deformation space, which enters the phase-field equation through a constitutive driving force c_e. The full system, equations (29)-(31), is shown to reproduce experimental measurements of nucleation in the bulk, nucleation from large pre-existing cracks, and steady crack propagation, supporting the claim that the theory is complete for quasistatic loading.","pith_inferences":["A natural next test would be to use the theory to predict the critical flaw size at which nucleation switches from strength-dominated to Griffith-dominated behavior, and to compare that prediction with the measured fracto-cohesive length scale in materials such as VHB 4905.","Because the strength surface is assumed history independent, the theory is likely to break down at very high loading rates or under cyclic loading; a plausible extension would make the strength surface depend on a rate-sensitive internal variable rather than on time explicitly.","The same phase-field structure could be adapted to fatigue by introducing a second internal variable that progressively degrades either the strength surface or the fracture energy with accumulated cycles, although the paper explicitly leaves non-monotonic loading for future work.","If the equilibrium-energy Griffith condition holds generally, then measurements of the critical tearing energy at different rates can be used to infer the equilibrium stored energy of an elastomer, offering a new macroscopic route to test constitutive viscoelastic models."],"forward_implications":["If the theory is correct, a single set of field equations can predict both crack nucleation and crack growth in viscoelastic elastomers under arbitrary quasistatic loads, with no need to prescribe a crack path in advance.","The apparent rate dependence of tearing energy emerges naturally from viscous dissipation during loading, while the critical stretch at which a large crack starts to grow in pure shear is predicted to be essentially rate independent.","Fracture nucleation from small flaws, notches, or other nonuniform stress states is predicted to be a mediation between the strength surface and the equilibrium-energy Griffith condition, so the theory interpolates between the two classical limiting criteria.","All material inputs—viscoelastic parameters, uniaxial and hydrostatic strength, and fracture energy—can in principle be measured from standard experiments such as uniaxial tension, poker-chip tests, pure-shear tests, and trousers tests.","The numerical scheme based on non-conforming finite elements in space and explicit Runge-Kutta time stepping provides a practical route to simulate finite-deformation, near-incompressible elastomer fracture at large strains."],"supporting_citations":[{"why":"Supplies the base phase-field theory for elastic brittle materials that this paper extends to viscoelasticity.","marker":"Kumar, Francfort and Lopez-Pamies (2018a)"},{"why":"Provides the reduced Griffith criticality condition in which only the equilibrium stored energy competes with the fracture energy.","marker":"Shrimali and Lopez-Pamies (2023b)"},{"why":"Supplies the two-potential finite viscoelastic constitutive model used for the bulk behavior.","marker":"Kumar and Lopez-Pamies (2016)"},{"why":"Introduces the strength-surface concept for fracture nucleation under uniform stress in the elastic setting.","marker":"Kumar and Lopez-Pamies (2020)"},{"why":"Provides the blueprint for constructing the phase-field driving force from a strength surface.","marker":"Kumar et al. (2020)"},{"why":"Supplies the explicit formulas for the regularization-length-dependent coefficients used in the driving force.","marker":"Kamarei et al. (2024)"},{"why":"Provides the polyurethane rubber-band experimental data used to validate bulk nucleation under uniform uniaxial tension.","marker":"Mueller (1968)"},{"why":"Provides the VHB 4905 pure-shear experimental data used to validate nucleation from a large pre-existing crack.","marker":"Pharr et al. (2012)"},{"why":"Provides the trousers-test experimental data used to validate steady crack propagation.","marker":"Greensmith and Thomas (1955)"}],"fun_headline_variants":["Phase-field theory unifies nucleation and growth of cracks in rubber","Complete phase-field description for fracture in viscoelastic elastomers","Rubber fracture: one phase-field model from nucleation to propagation","Phase-field theory captures viscoelastic fracture from start to steady crack","Elastomer cracking: phase-field with viscoelasticity, strength, and energy"],"cache_read_input_tokens":3200,"weakest_assumption_plain":"The theory assumes that the strength surface of a viscoelastic elastomer can be written as a single hypersurface in stress-deformation space with no explicit dependence on loading history or rate.","fun_headline_variants_meta":{"raw":{"variants":["Phase-field theory unifies nucleation and growth of cracks in rubber","Complete phase-field description for fracture in viscoelastic elastomers","Rubber fracture: one phase-field model from nucleation to propagation","Phase-field theory captures viscoelastic fracture from start to steady crack","Elastomer cracking: phase-field with viscoelasticity, strength, and energy"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.00062,"raw_usage":{"total_tokens":2936,"prompt_tokens":1064,"completion_tokens":1872,"prompt_tokens_details":{"cached_tokens":384},"prompt_cache_hit_tokens":384,"prompt_cache_miss_tokens":680,"completion_tokens_details":{"reasoning_tokens":1780}},"tokens_in":680,"tokens_out":1872,"duration_ms":12541,"temperature":1.0,"reasoning_tokens":1780,"cache_read_input_tokens":384,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-15T19:25:22.056929+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"Run pure-shear crack-nucleation tests on the same elastomer over stretch rates spanning at least six decades; if the critical stretch at which the pre-existing crack starts to grow changes by substantially more than experimental scatter, the theory's claim that only the equilibrium stored energy enters the Griffith condition is wrong.","supporting_citations":[{"cited_title":", author Lopez-Pamies, O","cited_arxiv_id":null,"evidence_quote":"Supplies the two-potential finite viscoelastic constitutive model used for the bulk behavior."},{"cited_title":", year 1968","cited_arxiv_id":null,"evidence_quote":"Provides the polyurethane rubber-band experimental data used to validate bulk nucleation under uniform uniaxial tension."},{"cited_title":", author Sun, J.S","cited_arxiv_id":null,"evidence_quote":"Provides the VHB 4905 pure-shear experimental data used to validate nucleation from a large pre-existing crack."}],"review_version":2}