{"id":"d70db2e9-4947-43ab-a73d-50eecf924bb2","arxiv_id":"2412.13700","paper_version":2,"verdict":"CONDITIONAL","confidence":"MODERATE","novelty_score":6.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":3,"one_line_summary":"An explicit analytical driving force makes phase field fracture models reproduce any material strength surface that is linear in its material coefficients, illustrated on Mohr-Coulomb and Drucker-Prager surfaces.","lead":"This paper derives an explicit analytical formula for the crack nucleation driving force used in phase field models of brittle fracture, so the model reproduces a chosen material strength surface. The formula works for any strength surface that is linear in its material constants, and it is demonstrated on the Mohr-Coulomb and Drucker-Prager strength surfaces.","discovery_kind":"new_method","skeptic_critique":{"model":"deepseek-v4-flash","headline":"The cited δε calibration (Eq. 5) yields finite ε/δε as ε→0, so ̄ω_ε does not diverge and the claimed convergence F_ε→F in Eq. (20) fails under the calibration the paper invokes.","rationale":"The reader's weakest assumption flagged the unverified nucleation criterion and the δε calibration as inherited from prior work; I agree that the calibration is a concern, but the load-bearing issue is more specific and verifiable. The paper's consistency theorem, Eq. (20), requires ̄ω_ε → ∞, i.e., ε/δε → 0. The paper explicitly claims that the cited expression in Eq. (5) satisfies this requirement. A direct asymptotic analysis of Eq. (5) shows δε ~ 2fε/Gc, so ε/δε tends to the finite constant Gc/(2f). Hence ̄ω_ε tends to 3f/4, W̄^ε stays finite, and β̄^ε (through Eq. (26)) does not approach β. The phase-field strength surface would not converge to the material strength surface as ε→0 for the proposed calibration, directly contradicting the central claim. The algebraic construction itself remains valid for any δε with ε/δε→0, but the paper provides no such δε and its cited evidence points the other way. This does not require changing the conditional verdict, since the gap is precisely the kind of condition that must be resolved before the method can be considered validated, and it sharpens rather than replaces the reader's concern. A concrete check, computing the limit and re-running the figures with Eq. (5), would settle whether this concern lands.","tokens_in":13689,"tokens_out":6682,"duration_ms":57709,"concrete_test":"Evaluate the limit ε/δε using Eq. (5) for the graphite parameters in Eq. (42); if it is finite (≈Gc/(2f)), then recompute the phase-field strength surfaces in Figs. 1(a)–(b) for a decreasing sequence ε with δε from Eq. (5) and quantify the discrepancy with the Mohr-Coulomb surface away from the two anchor points; a nonzero limiting discrepancy falsifies the convergence part of the central claim for the cited calibration.","verdict_should_be":"UNCHANGED","load_bearing_attack":"The consistency of the driving force rests on the limit ̄ω_ε = (3/8)δεGc/ε → ∞, equivalently ε/δε→0, which is used to make W̄^ε vanish and β̄^ε→β in Eq. (20). The paper states that the cited phenomenological calibration, Eq. (5), δε = 2/(5+Gc/(fε)), satisfies this property. It does not: for small ε, δε ≈ 2fε/Gc, so ε/δε → Gc/(2f), a finite nonzero constant since f has dimensions of stress. Consequently ̄ω_ε → (3/4)f, which is finite, and W̄^ε in Eq. (15) does not vanish; moreover Δβ in Eq. (26) tends to a nonzero constant, so β̄^ε does not converge to β. The phase-field strength surface therefore does not reduce to the material strength surface in the limit ε→0 if the suggested δε is used. The convergence claim in Eq. (20) only holds under the extra assumption that δε is chosen so that ε/δε→0; no such δε preserving Griffith behavior is provided or shown to exist, and the cited calibration explicitly contradicts it. This is a load-bearing gap in the central claim, not merely a technical caveat.","agreement_with_reader":"partial"},"referee_report":{"model":"deepseek-v4-flash","summary":"The paper derives an explicit analytical construction of the crack-nucleation driving force c_e for modified phase-field brittle fracture models, for material strength surfaces of the form F(σ,β)=g(σ,β)-1 with g linear in the material coefficients β. The author sets c_e = -ω_ε g(σ,β+Δβ^ε) and chooses Δβ^ε so that the algebraic phase-field strength surface 2W^ε + F(σ,β^ε) = 0 coincides with the material strength surface at n prescribed stress states for every regularization length ε, while claiming that in the limit ε→0 the full material strength surface is recovered. Closed-form expressions for Δβ^ε and β^ε are given and applied to Mohr-Coulomb and Drucker-Prager strength surfaces, with plane-stress plots showing the finite-ε matching at the chosen calibration states.","tokens_in":14032,"tokens_out":15334,"duration_ms":138026,"significance":"If the convergence claim is valid, the paper provides a useful general recipe that replaces case-by-case derivations of the phase-field nucleation driving force, and it gives first explicit M-C results and new D-P calibration choices. The finite-ε matching construction is elegant: at the n chosen strength states the phase-field surface agrees with the material surface for all ε, and the algebra leading to Eqs. (29)-(30) is clean and internally consistent conditional on the stated scaling assumption ε/δ_ε→0. However, the central consistency statement depends on the regularization coefficient δ_ε, and the specific calibration Eq. (5) cited in the paper does not satisfy the required scaling. Because the paper neither supplies an alternative δ_ε nor verifies the phase-field PDEs in a boundary value problem, the central claim is not established in its current form.","major_comments":[{"comment":"Equation (20), namely F_ε→F as ε→0, is the load-bearing consistency statement. The text immediately after Eq. (20) asserts that the suggested calibration Eq. (5), δ_ε = 2/(5 + G_c/(fε)), satisfies the required property ε/δ_ε→0. This is false. For small ε, Eq. (5) gives δ_ε ≈ 2fε/G_c, so ε/δ_ε→G_c/(2f), a finite nonzero constant, and hence ω_ε→3f/4 rather than ∞. Consequently W^ε in Eq. (15) does not vanish in the limit, and Δβ^ε in Eq. (26) tends to the nonzero constant -(8/3)(σ*/f)J^{-1}W_vec, so β^ε does not converge to β. The phase-field strength surface therefore does not reduce to the material strength surface under the calibration cited by the author. The consistency claim can be restored only by providing a δ_ε with ε/δ_ε→0 that also preserves Griffith crack propagation, or by explicitly stating the result as conditional on such a δ_ε; the manuscript currently does neither.","section":"§2.3, Eqs. (5), (15), (20), (26)"},{"comment":"The paper verifies only the algebraic phase-field strength surface, not the phase-field boundary value problem. The criterion in Eq. (4) is asserted by reference to prior work and stability discussions, but the constructed c_e depends on the true stress σ(F,v), and no simulation of Eqs. (2)-(3) is provided to show that the proposed c_e actually drives nucleation according to Eq. (4) in nonuniform fields. If Eq. (4) is not guaranteed to describe nucleation in arbitrary boundary value problems, then statements that the formulation predicts crack nucleation go beyond what is demonstrated. This is partly acknowledged in the author's caveat that preservation of large-crack Griffith physics relies on numerical calibration of δ_ε, but the gap is load-bearing for the claim that the construction is consistent for the phase-field theory, not merely for the algebraic strength surface.","section":"§2.1–§2.3, Eq. (4)"}],"minor_comments":[{"comment":"W_bs is defined as W(α_bt ex⊗ey + α_bt ey⊗ex), which is a pure-shear stress state, but Case (b) is the biaxial tension state σ_s2 = α_bt(ex⊗ex + ey⊗ey); Eq. (44) is the correct biaxial-tension value. The argument in Eq. (43) should be corrected.","section":"§3.2, Eq. (43)"},{"comment":"The notation '#»W / ¯ω_ε' in Eq. (28) should be made consistent with the vector W^ε_i = W_i/¯ω_ε defined through Eqs. (14)-(15); as written it is easy to misread as a dimensionless stress vector divided by a stress-like quantity.","section":"§2.3, Eq. (28)"},{"comment":"The captions refer to a common legend, but the legend values of ¯ω_ε and the distinction between the exact material surface and the phase-field surfaces are not described in the text; please state the plotted values of ¯ω_ε and the line styles in the captions.","section":"Figures 1 and 2"},{"comment":"The word 'verified' overstates the evidence: the paper presents algebraic surface comparisons only, not PDE simulations or experimental validation. 'Illustrated' or 'checked algebraically' would be more accurate.","section":"Abstract and §3"}],"recommendation":"major_revision","confidential_remarks":"The core algebraic construction is sound conditional on a suitable regularization scaling, and the finite-ε matching idea is a useful contribution. The main obstacle is the inconsistency between the claimed convergence and the cited calibration Eq. (5); this needs to be fixed or the claims weakened before publication. No concerns about attribution or novelty beyond the author's own stated scope."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"Two things to know. First, the central algebraic result—eqs. (29)–(30) giving an explicit driving force for any strength surface linear in its material coefficients—is correct and genuinely new. The Mohr-Coulomb application is a first, and the construction reduces the problem to inverting a small matrix. Second, the paper's consistency claim, as stated, is not supported by the calibration it cites. That is a real problem, not a nitpick.\n\nThe derivation is clean: linearity of g in beta, the phase-field strength surface (eq. 19), and the finite-epsilon matching condition (eq. 21) determine Delta beta uniquely. The plots for M-C and D-P do show the claimed behavior at finite epsilon. The effective-toughness interpretation (eq. 31) is a nice observation.\n\nThe soft spot is the epsilon -> 0 limit. Eq. (20) requires epsilon/delta_epsilon -> 0 (or omega_epsilon -> infinity). The paper states that the phenomenological calibration in eq. (5), delta_epsilon = 2/(5 + G_c/(f epsilon)), satisfies this. It does not. For small epsilon, delta_epsilon ~ 2 f epsilon / G_c, so epsilon/delta_epsilon -> G_c/(2f) and omega_epsilon -> 3f/4. Consequently W^epsilon does not vanish, Delta beta tends to a nonzero constant, and the phase-field strength surface does not reduce to the material surface. The consistency proof is conditional on choosing a different delta_epsilon, but none is provided, and the only cited calibration contradicts the condition. The paper could fix this by correcting the claim and either finding a calibration that satisfies the limit or weakening the consistency statement. As written, the central convergence theorem is unproven for the practical calibration.\n\nAlso worth noting: the bulk nucleation criterion (eq. 4) is inherited from prior work and not derived here, and the validation is algebraic—no boundary-value simulations. For a methods paper that is acceptable, but it means the 'successful prediction' claims rest on the prior literature.\n\nBottom line: the algebraic tool is useful and likely correct under the right scaling, and the Mohr-Coulomb extension matters for concrete and rock modeling. But the paper needs a fix before it can be trusted as a consistency proof. I'd send it to a serious referee, with the delta_epsilon issue front and center.","headline":"Clean algebraic generalization with a genuine convergence gap: the cited delta_epsilon calibration does not satisfy the scaling the consistency proof requires.","tokens_in":14473,"tokens_out":4328,"would_cite":true,"duration_ms":36222,"reading_group":"maybe","serious_thinker":"yes","would_accept_peer_review":true},"rs_alignment":null,"lean_confirmation":null,"pith_extraction":{"msc":["74R10"],"pacs":[],"model":"deepseek-v4-flash","headline":"This paper derives an explicit analytical driving force that makes modified phase-field fracture models reproduce any material strength surface linear in its material parameters, exactly at $n$ chosen stress states for finite…","keywords":["phase-field fracture","crack nucleation driving force","brittle fracture","fracture nucleation","Mohr-Coulomb strength surface","Drucker-Prager strength surface","effective toughness","strength surface"],"falsifier":"Run a phase-field simulation of eqs. (2)--(3) with the proposed $c_e$ under a spatially uniform stress increment and record the stress at which $v$ first drops below $1$; if that stress does not pass through the $n$ chosen material strength states for finite $\\varepsilon$, or does not approach the full strength surface as $\\varepsilon\\to 0$, the central claim is falsified. For the Mohr-Coulomb Case (a) calibration in Section 3.1, this means checking numerically that uniaxial tension and uniaxial compression nucleate exactly at $\\sigma_{ts}$ and $\\sigma_{cs}$ for a finite $\\varepsilon$.","tokens_in":13520,"feed_emoji":"💥","tokens_out":16061,"duration_ms":119295,"temperature":0.7,"pith_summary":"Phase-field fracture models describe how existing cracks grow, but they need an additional, carefully chosen driving force to predict when a crack first nucleates in pristine material. This paper gives a closed-form recipe for that driving force, valid for any strength surface whose failure condition is linear in its material parameters: the model matches the true strength surface exactly at $n$ chosen stress states for every regularization length $\\varepsilon$, and matches the whole surface as $\\varepsilon\\to 0$. The recipe is worked out explicitly for Mohr-Coulomb and Drucker-Prager strength surfaces, with two different anchor-state choices for each, so it covers tensile, compressive, shear, and biaxial calibrations. If correct, it removes the need to re-derive the driving force for each new strength criterion and gives phase-field models a direct route to predicting nucleation in pressure-sensitive brittle materials such as concrete, rock, and ceramics.","feed_headline":"Formula aligns phase-field crack nucleation with strength surfaces","feed_subtitle":"It works for Mohr-Coulomb, Drucker-Prager, and any linear strength surface without re-derivation.","key_machinery":"The load-bearing object is the shifted parameter vector $\\beta^\\varepsilon = \\beta + \\Delta\\beta^\\varepsilon$ that appears inside the strength function $g$ when the driving force is assembled. Because $g$ is linear in $\\beta$, the phase-field strength surface splits into two additive pieces: the strain-energy term $2\\bar{W}^\\varepsilon(\\sigma)$, which vanishes as $\\omega_\\varepsilon\\to\\infty$, and the shifted material strength function $F(\\sigma,\\beta^\\varepsilon)$. The shift $\\Delta\\beta^\\varepsilon$ is obtained by solving an $n\\times n$ linear system at $n$ distinct strength states $\\sigma_{si}$, with 'distinct' meaning the matrix $\\partial F_i/\\partial \\beta$ has linearly independent rows; this is what makes the finite-$\\varepsilon$ match exact at those states. The same linear-system structure also expresses the original parameters $\\beta$ from the chosen strength states. The construction's main structural consequence is the effective toughness $\\hat{G}_c^\\varepsilon = -\\delta_\\varepsilon F(\\sigma,\\beta^\\varepsilon) G_c$, which replaces $G_c$ in the nucleation equation and thereby encodes strength through a state-dependent fracture energy.","core_discovery":"The central claim is that a material strength surface written as $F(\\sigma,\\beta) = g(\\sigma,\\beta) - 1 = 0$, with $g$ linear in the dimensionless parameters $\\beta$, can be built into the modified phase-field theory through the explicit driving force $c_e = -\\frac{3}{8}\\frac{\\delta_\\varepsilon G_c}{\\varepsilon} g(\\sigma,\\beta + \\Delta\\beta^\\varepsilon)$. Here $\\Delta\\beta^\\varepsilon = -\\frac{16}{3}\\frac{\\varepsilon}{\\delta_\\varepsilon}\\frac{\\sigma_*}{G_c}\\left(\\frac{\\partial F}{\\partial \\beta}\\right)^{-1} W$, evaluated at $n$ distinct strength states $\\sigma_{si}$: $W$ collects the strain energies $W(\\sigma_{si})$ and $\\partial F/\\partial \\beta$ collects the derivatives of $F$ at those states. Substituting this $c_e$ into the phase-field strength surface $2W - c_e - \\omega_\\varepsilon = 0$, with $\\omega_\\varepsilon = \\frac{3}{8}\\frac{\\delta_\\varepsilon G_c}{\\varepsilon}$, makes the predicted locus pass through the $n$ chosen strength states for every $\\varepsilon$ and reduces to the material strength surface as $\\varepsilon\\to 0$, provided $\\delta_\\varepsilon$ is calibrated so that $\\omega_\\varepsilon\\to\\infty$. The paper also shows that the nucleation equation can be rewritten with a stress-dependent effective toughness $\\hat{G}_c^\\varepsilon = -\\delta_\\varepsilon F(\\sigma,\\beta^\\varepsilon) G_c$ taking the place of $G_c$; this effective toughness is zero at strength-based initiation and reaches $\\delta_\\varepsilon G_c$ in the fully cracked state. Application to Mohr-Coulomb and Drucker-Prager surfaces, with two different choices of exact-match states for each, verifies the claim.","pith_inferences":["The $n$-state matching property suggests a calibration strategy not stated in the paper: choose the exact-match states to cover the dominant loading directions of a specific simulation, so the finite-$\\varepsilon$ error is pushed away from the stress states that matter most.","Because the paper verifies only the algebraic strength surface, a natural next test is a full boundary-value simulation with the proposed $c_e$ under non-uniform stress; if the predicted nucleation stress deviates from the matched strength states, the algebraic criterion would need a dynamical justification.","The effective-toughness reading suggests that $\\delta_\\varepsilon$ could be calibrated from a traction-separation law rather than from a single pure-shear or edge-notch test, which would give the parameter a physical interpretation independent of the regularization.","For strength surfaces nonlinear in $\\beta$, the paper notes its formula is only a first-order approximation and cannot be shown consistent; extending the closed-form shift to such surfaces would require a separate argument, for example by re-solving the matching system at each state."],"forward_implications":["Given any strength surface of the form $F = g(\\sigma,\\beta) - 1 = 0$ with $g$ linear in $\\beta$, the driving force is assembled directly from eqs. (29)--(30), so no case-by-case derivation is required.","For a finite regularization length, the phase-field strength surface passes exactly through the $n$ chosen strength states, letting a modeler anchor the prediction to the failure modes most relevant to the problem.","As $\\varepsilon \\to 0$, the phase-field strength surface converges to the full material strength surface, restoring the exact strength criterion in the sharp-interface limit.","The nucleation equation acquires a stress-dependent effective toughness that vanishes at strength-based initiation and rises to $\\delta_\\varepsilon G_c$ in the fully cracked state, which connects the modified theory to cohesive-zone thinking.","The explicit Mohr-Coulomb formulas, new for this theory, extend the approach to concrete, rock, and other pressure-sensitive brittle materials, and the Drucker-Prager cases cover tensile-shear and tensile-biaxial calibrations."],"supporting_citations":[{"why":"It introduces the modified phase-field equations with a crack-nucleation driving force and defines what it means for that force to be consistent, which is the concept this paper generalizes.","marker":"[6]"},{"why":"It establishes the phase-transition formulation whose nucleation term the present construction inherits.","marker":"[5]"},{"why":"It provides the two-step variational form of the modified theory whose governing equations, eqs. (2)--(3), are the starting point of the derivation.","marker":"[18]"},{"why":"It supplies the explicit functional form of the calibration coefficient in eq. (5) that the consistency argument relies on to make the normalized energy diverge as the regularization length tends to zero.","marker":"[15]"},{"why":"It shows that classical variational phase-field models cannot predict fracture nucleation, which motivates the addition of the driving force.","marker":"[4]"},{"why":"It exemplifies the correction-factor version of the driving force and prior Drucker-Prager applications that this paper recasts in explicit form.","marker":"[11]"},{"why":"It supports the stability and localization discussion behind the phase-field strength surface in eq. (4).","marker":"[21]"}],"fun_headline_variants":["Explicit crack-driving force matches Mohr-Coulomb and Drucker-Prager","Crack nucleation driving force now explicit for any linear strength surface","Phase-field fracture: explicit driving force nails Mohr-Coulomb and Drucker-Prager","Explicit phase-field driving force matches strength surfaces exactly","New analytical formula for crack nucleation in Mohr-Coulomb and Drucker-Prager"],"cache_read_input_tokens":3200,"weakest_assumption_plain":"The argument assumes that crack nucleation under uniform monotonic stress is governed by the algebraic phase-field strength surface $2W - c_e - \\omega_\\varepsilon = 0$, and that $\\delta_\\varepsilon$ can be calibrated so that $\\omega_\\varepsilon \\to \\infty$ as $\\varepsilon \\to 0$ while Griffith crack propagation is preserved; the paper neither derives this criterion from the phase-field partial differential equations nor simulates a boundary value problem.","fun_headline_variants_meta":{"raw":{"variants":["Explicit crack-driving force matches Mohr-Coulomb and Drucker-Prager","Crack nucleation driving force now explicit for any linear strength surface","Phase-field fracture: explicit driving force nails Mohr-Coulomb and Drucker-Prager","Explicit phase-field driving force matches strength surfaces exactly","New analytical formula for crack nucleation in Mohr-Coulomb and Drucker-Prager"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.000906,"raw_usage":{"total_tokens":4027,"prompt_tokens":1209,"completion_tokens":2818,"prompt_tokens_details":{"cached_tokens":384},"prompt_cache_hit_tokens":384,"prompt_cache_miss_tokens":825,"completion_tokens_details":{"reasoning_tokens":2719}},"tokens_in":825,"tokens_out":2818,"duration_ms":19286,"temperature":1.0,"reasoning_tokens":2719,"cache_read_input_tokens":384,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-11T12:53:03.977847+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"Run a phase-field simulation of eqs. (2)--(3) with the proposed $c_e$ under a spatially uniform stress increment and record the stress at which $v$ first drops below $1$; if that stress does not pass through the $n$ chosen material strength states for finite $\\varepsilon$, or does not approach the full strength surface as $\\varepsilon\\to 0$, the central claim is falsified. For the Mohr-Coulomb Case (a) calibration in Section 3.1, this means checking numerically that uniaxial tension and uniaxial compression nucleate exactly at $\\sigma_{ts}$ and $\\sigma_{cs}$ for a finite $\\varepsilon$.","supporting_citations":[{"cited_title":"Kumar, B","cited_arxiv_id":null,"evidence_quote":"It introduces the modified phase-field equations with a crack-nucleation driving force and defines what it means for that force to be consistent, which is the concept this paper generalizes."},{"cited_title":"Kumar, G","cited_arxiv_id":null,"evidence_quote":"It establishes the phase-transition formulation whose nucleation term the present construction inherits."},{"cited_title":"Lopez-Pamies, J","cited_arxiv_id":null,"evidence_quote":"It shows that classical variational phase-field models cannot predict fracture nucleation, which motivates the addition of the driving force."}],"review_version":1}