{"id":"8a18e226-9f6e-4afd-b6d1-0b64ca81caf4","arxiv_id":"1908.01952","paper_version":1,"verdict":"CONDITIONAL","confidence":"MODERATE","novelty_score":6.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":0,"one_line_summary":"An asymptotic Wiener-Hopf factorization yields a first-order approximate scattered field for two staggered semi-infinite cracks on a square lattice, with far-field agreement to numerics that degrades as the stagger increases.","lead":"This paper derives an approximate formula for how a plane wave scatters off two parallel semi-infinite cracks on a square lattice when the crack tips are staggered. It uses an asymptotic Wiener-Hopf factorization and compares the far field with numerical simulations.","discovery_kind":"new_application","skeptic_critique":{"model":"deepseek-v4-flash","headline":"Final inverse-transform and far-field formulas (93)-(94), (113)-(114) appear to place the incident-wave pole at z_P^{-1} rather than z_P; because z_P=e^{ik cosTheta} by (31), the stated xi_P=k cosTheta has the wrong sign unless a different convention is intended.","rationale":"The reader's weakest assumption concerns the uncontrolled smallness of the perturbation matrix N_M and the numerically verified Liouville step; that concern is legitimate and supports the CONDITIONAL verdict. However, a more direct and more load-bearing problem appears earlier in the printed argument: the inverse-transform step from (91)-(92) to (93)-(94) changes the incident-wave pole from z_P to z_P^{-1}, and (109) fixes xi_P with the sign opposite to what (31)-(32) require. Because the final far-field amplitudes depend on K_-^{-1}(xi_P) and on the denominator (e^{i(xi_s-xi_P)}-1), this is not a cosmetic issue: the displayed formulas (113)-(114) are not the stationary-phase evaluation of the field defined by (91)-(92) unless an unstated convention is adopted. The paper's own caveats about degrading agreement with increasing M and about anomalous contour-integral points further indicate that the numerical evaluation needs scrutiny, but the pole inconsistency is independent of those issues. Since this is likely a correctable sign/convention error rather than a collapse of the whole method, the reader's CONDITIONAL verdict is retained; the paper should be revised to fix the pole convention and to rerun the far-field comparison with the corrected expression.","tokens_in":19143,"tokens_out":22738,"duration_ms":218343,"concrete_test":"Independently re-derive (93) from (91)-(92) by applying the inverse discrete Fourier transform u_{x,y}=(1/2 pi i) oint uF_y(z) z^{x-1} dz and tracking deltaD+(z z_P^{-1})=z/(z-z_P). If the re-derived integrand has denominator z-z_P and K_-^{-1}(z_P), then the printed (93)-(94) are inconsistent as written; equivalently, recompute the Figure 4 far-field comparison using xi_P=-k cosTheta instead of xi_P=+k cosTheta in (113)-(114) and check which choice matches the numerical curve.","verdict_should_be":"UNCHANGED","load_bearing_attack":"Equation (31) fixes z_P=e^{ik cosTheta}, and (32) defines the forcing through deltaD+(z z_P^{-1}), whose simple pole is at z=z_P. Consistently, the Wiener-Hopf solution step leading to (84)-(85) and the inverse transforms (91)-(92) contain K_-^{-1}(z_P) and denominator z-z_P. Yet the printed inversion integrals (93)-(94) contain K_-^{-1}(z_P^{-1}) and denominator z-z_P^{-1}. With z=e^{-i xi}, this amounts to placing the pole at xi_P=k cosTheta, exactly as written in (109), whereas (31)-(32) require xi_P=-k cosTheta. The stationary-phase amplitudes (113)-(114) inherit this: they are evaluated at the wrong point unless an unstated reciprocal convention is introduced. Thus the central displayed formulas are not the far-field asymptotics of the field defined by (91)-(92); this is an internal inconsistency, not merely a missing error bound. If it is a typographical error, it must be corrected and the numerics rerun with the corrected sign.","agreement_with_reader":"partial"},"referee_report":{"model":"deepseek-v4-flash","summary":"The paper formulates scattering of a time-harmonic plane wave on a square lattice by two parallel semi-infinite cracks with staggered edges as a 2x2 matrix Wiener-Hopf equation (36). Because the matrix kernel is not exactly factorable, the authors adapt an asymptotic factorization (61) based on the small parameter epsilon = lambda^N sin(xi M/2), solve the Wiener-Hopf equation formally, and give integral representations of the scattered field in (93)-(94). The far field is then approximated by stationary phase, leading to (113)-(114), and these formulas are compared with direct numerical lattice simulations and with the low-frequency continuum solution of Abrahams and Wickham.","tokens_in":19463,"tokens_out":6961,"duration_ms":65429,"significance":"If correct, the paper would provide the first lattice analogue of the two staggered semi-infinite crack/plate diffraction problem and would demonstrate that asymptotic Wiener-Hopf factorization can be adapted to a circular contour for a discrete matrix kernel. The work has genuine strengths: the Wiener-Hopf setup is coherent, the zero-offset limit is tied to an exact solution in [50], the low-frequency comparison with the continuum Abrahams-Wickham solution is physically meaningful, and the numerical comparisons involve no fitted free parameters. However, the central displayed far-field formulas currently contain an internal sign/pole inconsistency, and several load-bearing steps (the first-order truncation, the Liouville argument, and the dropped pole contribution) lack analytic or quantitative control. These issues must be addressed before the claims can be accepted.","major_comments":[{"comment":"The first-order factorization G1,M(z) ≈ (I + N1M-(z))(I + N1M+(z)) is adopted without any error bound or convergence estimate. The parameter epsilon = lambda^N sin(xi M/2) is asserted to be small, and the paper itself notes that agreement with numerics degrades as M increases, but there is no quantitative criterion for when the truncated product is a valid replacement for G_M. Since (61) is the basis of the entire solution, this requires either a rigorous remainder estimate or a systematic numerical study of the truncation error.","section":"§3.1, Eqs. (57)-(61)"},{"comment":"The Liouville step J(z)=0 is justified only by numerical inspection of plots. The solution formulas (78)-(79) depend on the entire function J being identically zero; this requires analytic control of the growth/decay of the two sides of (77), not visual inspection of a few computed curves. Without such control, the derivation of (78)-(79) is incomplete.","section":"§3.2, Eq. (77) and Figs. 2-3"},{"comment":"There is an internal sign/pole inconsistency in the inversion and far-field formulas. Equation (31) sets zP = e^{ik cos Theta}, and equations (91)-(92) place the pole at z = zP. However, the printed inversion integrals (93)-(94) contain K_-^{-1}(zP^{-1}) and denominator z - zP^{-1}; with the mapping z = e^{-i xi} this places the pole at xi = k cos Theta, as stated in (109), whereas (31)-(32) require xi = -k cos Theta. The stationary-phase amplitudes (113)-(114) inherit the wrong evaluation point. Unless a reciprocal convention is introduced, the displayed far-field formulas are not the asymptotics of the field defined by (91)-(92). This must be corrected and the numerical comparisons rerun with the corrected sign.","section":"§4, Eqs. (93)-(94), (109), (113)-(114)"},{"comment":"The far-field asymptotics are stated 'modulo the contribution of pole', and the pole contribution is then dropped without an estimate. In scattering problems, the pole typically gives the reflected or plane-wave component, which can be of the same order as the saddle-point contribution. As written, (113)-(114) cannot be asserted as asymptotic equivalents unless the pole contribution is shown to be negligible in the relevant angular sector; otherwise it must be included in the comparison with numerical data.","section":"§4, after Eq. (112) and note following (114)"}],"minor_comments":[{"comment":"The phase in vi_2x is written as e^{ik(N sin Theta + M sin Theta)}, but the subsequent shifted Fourier transform (32) and the Wiener-Hopf forcing (37c) use M cos Theta; this should be reconciled.","section":"§1, Eq. (16)"},{"comment":"The sentence 'C+ and C+ are analytic' should presumably read 'C+ and C- are analytic'; as printed it is a typographical error.","section":"§3.2, Eq. (76)"},{"comment":"The Liouville argument refers to 'Examining (135), (136), and (59)' with equations from Appendix A that assume N is even, but the main solution is stated for general N; the scope of the numerical justification should be clarified.","section":"§3.2, near Eq. (77)"},{"comment":"The claimed agreement between the semi-analytical and numerical far fields is only visual; the paper would be substantially strengthened by a quantitative error metric (e.g., relative L2 or pointwise error as a function of M and omega).","section":"§4, Figs. 4-5"}],"recommendation":"major_revision","confidential_remarks":"The paper builds heavily on the authors' prior work [44, 46, 50], but the staggered-offset Wiener-Hopf kernel and the asymptotic factorization on the circle appear to be new. The main obstacle to acceptance is not novelty but internal consistency: the sign/pole error in (93)-(94) affects the central far-field formulas, and the Liouville step and the first-order truncation need analytic or quantitative support. If the sign error is a typo and the numerical comparisons can be rerun, the paper could become a useful contribution."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"Dear X,\n\nThis paper solves a new problem: scattering of a lattice plane wave by two staggered semi-infinite cracks on the square lattice, formulated as a 2x2 Wiener-Hopf equation and handled with the Mishuris-Rogosin asymptotic factorization adapted to the circle. The zero-offset limit checks against the Sharma-Maurya exact solution, and the low-frequency comparison with Abrahams-Wickham continuum results (Fig 5) is a useful sanity check. That part is legitimate and worth engaging with.\n\nThe problem is that the central far-field formulas are built on a sign error. Equation (31) defines z_P = e^{ik cosTheta}, and the forcing and solution up to (91)-(92) consistently put the incident pole at z = z_P. Then the printed inverse transforms (93)-(94) switch to z_P^{-1}, and (109) asserts \"by definition\" z_P = e^{-ik cosTheta}. Under the mapping z = e^{-i xi}, this places the pole at xi_P = +k cosTheta instead of -k cosTheta. The stationary-phase amplitudes (113)-(114) inherit that wrong point. Unless a reciprocal substitution in the contour integral is silently intended, the displayed far-field formulas are not the asymptotics of the field defined by (91)-(92). This is not a missing error bound; it is an internal inconsistency in the paper's own equations. Since Figures 4 compare evaluations of these same formulas against numerics, the apparent agreement is not persuasive evidence until the sign is corrected and the plots re-generated.\n\nThe other soft spots are the usual ones for this style of formal work: the first-order factorization (61) has no error estimate, the Liouville step is justified only by numerical plots, the pole contribution is dropped in the far field without a careful argument, and the contour integrals are computed numerically without error control. Those are limitations, not fatal flaws.\n\nIf the sign error is a typo, the fix is straightforward, but it propagates through the main results and the numerics must be redone with the correct pole location. I would send this to a referee: the problem is new, the Wiener-Hopf machinery is competently applied, and a careful referee could help the authors sort out the convention. I would not cite the far-field formulas in their current form.","headline":"New staggered-crack lattice scattering problem, but the far-field formulas are undermined by a sign error in the incident pole that must be fixed before the numerics can be trusted.","tokens_in":19895,"tokens_out":9075,"would_cite":false,"duration_ms":84708,"reading_group":"maybe","serious_thinker":"yes","would_accept_peer_review":true},"rs_alignment":null,"lean_confirmation":null,"pith_extraction":{"msc":["45E10","74J20"],"pacs":[],"model":"deepseek-v4-flash","headline":"The paper derives an approximate solution for two staggered semi-infinite cracks on a square lattice by first-order asymptotic Wiener-Hopf factorization, yielding scattered-field integrals and far-field formulas that match numerical…","keywords":["lattice waves","Wiener-Hopf factorization","staggered cracks","square lattice","scattering","asymptotic factorization","stationary phase method","anti-plane displacement"],"falsifier":"Compute the unit-circle norm of the difference between $G_M$ and the first-order product $(I+N_{1M-})(I+N_{1M+})$ for $M=1,2,3,\\ldots$ at fixed $N$ and frequency; if the normalized residual does not shrink with the offset parameter, the first-order factorization fails. A direct check is to compare the far-field angle dependence from (113)-(114) with an independent lattice simulation at $M=3$, $N=4$, $\\omega=0.35$, where the paper's figures already show visible deviation.","tokens_in":18965,"feed_emoji":"🌊","tokens_out":10046,"duration_ms":101547,"temperature":0.7,"pith_summary":"An incident time-harmonic lattice wave hits two parallel, semi-infinite cracks on a square lattice whose tips are staggered by $M$ lattice spacings. The paper reduces the discrete scattering problem to a coupled $2\\times2$ Wiener-Hopf equation whose kernel cannot be factorized by standard methods, and it adapts an asymptotic factorization technique to the unit circle to get a first-order approximate factorization in the small-offset parameter. Using that factorization, the Wiener-Hopf technique yields an integral representation of the scattered displacement field and stationary-phase far-field formulas. The paper shows numerically that these formulas agree with a direct lattice simulation when the offset is small, that agreement worsens as $M$ grows, and that the low-frequency numerical solution reproduces the known continuum solution for two staggered plates.","feed_headline":"Two staggered lattice cracks solved to first order in offset","feed_subtitle":"Asymptotic Wiener-Hopf factorization gives far-field formulas matching lattice simulations at small offsets.","key_machinery":"The load-bearing object is the $2\\times2$ Wiener-Hopf kernel $K(z)=L(z)G_M(z)$, with $L=h/r$ the scalar factor familiar from the single semi-infinite crack and $G_M$ the symmetric matrix $\\begin{pmatrix}1 & z^{-M}\\lambda^N \\\\ z^M\\lambda^N & 1\\end{pmatrix}$ that carries the stagger. A similarity transformation by $R_M=\\mathrm{diag}(z^{-M/2},z^{M/2})$ recasts $G_M$ as a matrix $F$ with entries $1$ and $\\lambda^N$, which is exactly factorized through its eigenvalues $1\\pm\\lambda^N$. The remaining part is written as $I+N_M$ and split additively into factors analytic inside and outside the unit circle, giving the approximate product $G_M \\approx (I+N_{1M-})(I+N_{1M+})$. This approximate factorization is what permits the Liouville step that produces the displacement formulas.","core_discovery":"On the paper's own terms, the central claim is that the matrix kernel $K(z)=L(z)G_M(z)$ of the coupled Wiener-Hopf equations can be multiplicatively factorized to first order as $K_-K_+$, where $L$ is the scalar factor from the single-crack problem, $G_M$ is the symmetric matrix encoding the stagger, and the perturbation enters through a matrix $N_M$ whose size is controlled by $\\epsilon = \\lambda^N \\sin(\\xi M/2)$. Once this factorization is made, the standard Wiener-Hopf argument gives explicit expressions for the half-range transforms, hence the contour integrals (93)-(94) for the displacement and the stationary-phase far-field estimates (113)-(114). The paper presents these formulas as the approximate solution of the two-staggered-crack lattice problem for small offset, and it reports graphical agreement with numerical lattice solutions in that regime.","pith_inferences":["The residual $G_M-(I+N_{1M-})(I+N_{1M+})$ could be measured numerically on the unit circle; if it scales like $|\\epsilon|$ over a wider parameter range, the first-order factorization may be promoted to a rigorous asymptotic expansion by iterating the same correction.","The Chebyshev-polynomial product forms for the scalar factors suggest that for even $N$ the contour integrals defining the factors can be evaluated in closed form, removing the numerical contour integration the paper identifies as a practical difficulty.","A similar stagger matrix arises for cracks on triangular or honeycomb lattices, so the same $R_M F R_M^{-1}$ decomposition may yield first-order factorization for those geometries as well.","An analytic estimate of the growth of $K_-(z)K_-^{-1}(z_P)$ near $z=0$ would replace the numerical inspection used to justify the Liouville step and would tell whether the approximation remains uniform in the observation angle."],"forward_implications":["For small tip offset, the scattered field anywhere outside the crack region can be evaluated from one-dimensional contour integrals rather than by solving the full two-dimensional lattice problem.","The factorization separates the geometry: the single-crack scalar factor $L_\\pm$ and the zero-offset matrix factors are universal, while the crack separation $N$ enters through $\\lambda^N$ and explicit Chebyshev product factors.","In the aligned-tip limit $M=0$ the perturbation matrix vanishes and the formulas reduce to the exact zero-offset solution, giving a built-in consistency check.","The same factorization can be applied to a wave incident from the waveguide between the cracks; only the right-hand side of the Wiener-Hopf equation changes.","Because agreement with numerics degrades as $M$ increases, the derived formulas are reliable only for small-to-moderate stagger relative to the crack separation."],"supporting_citations":[{"why":"supplies the asymptotic method of factorization for matrix functions whose standard factorization fails; the paper adapts it to the unit circle.","marker":"[37]"},{"why":"gives the unit-circle adaptation of the asymptotic factorization used in the approximate product (61).","marker":"[31]"},{"why":"provides the single semi-infinite crack Wiener-Hopf setup, explicit scalar factors $L_\\pm$, and stationary-phase far-field analysis extended here to two cracks.","marker":"[44]"},{"why":"contains the exact continuum solution for two staggered semi-infinite plates used as the low-frequency benchmark in Figure 5.","marker":"[4, 5]"},{"why":"provides the exact zero-offset solution used to verify the special case $M=0$ of the matrix formulation.","marker":"[50]"}],"fun_headline_variants":["Staggered lattice cracks solved to first order","First-order Wiener-Hopf for staggered crack pair on lattice","Asymptotic factorization tackles staggered lattice cracks","Staggered crack tips on lattice: first-order far field","Two staggered cracks: first-order lattice wave solution"],"cache_read_input_tokens":3200,"weakest_assumption_plain":"The load-bearing premise is that the offset between the two crack tips is small enough that keeping only the first-order correction to the zero-offset kernel is accurate; the paper supplies no error bound and notes that agreement with numerics gets worse as the offset increases.","fun_headline_variants_meta":{"raw":{"variants":["Staggered lattice cracks solved to first order","First-order Wiener-Hopf for staggered crack pair on lattice","Asymptotic factorization tackles staggered lattice cracks","Staggered crack tips on lattice: first-order far field","Two staggered cracks: first-order lattice wave solution"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.000911,"raw_usage":{"total_tokens":3913,"prompt_tokens":941,"completion_tokens":2972,"prompt_tokens_details":{"cached_tokens":384},"prompt_cache_hit_tokens":384,"prompt_cache_miss_tokens":557,"completion_tokens_details":{"reasoning_tokens":2897}},"tokens_in":557,"tokens_out":2972,"duration_ms":20573,"temperature":1.0,"reasoning_tokens":2897,"cache_read_input_tokens":384,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-14T14:58:19.615343+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"Compute the unit-circle norm of the difference between $G_M$ and the first-order product $(I+N_{1M-})(I+N_{1M+})$ for $M=1,2,3,\\ldots$ at fixed $N$ and frequency; if the normalized residual does not shrink with the offset parameter, the first-order factorization fails. A direct check is to compare the far-field angle dependence from (113)-(114) with an independent lattice simulation at $M=3$, $N=4$, $\\omega=0.35$, where the paper's figures already show visible deviation.","supporting_citations":[{"cited_title":"An asymptotic method of factorization of a class of matrix functions","cited_arxiv_id":null,"evidence_quote":"supplies the asymptotic method of factorization for matrix functions whose standard factorization fails; the paper adapts it to the unit circle."},{"cited_title":"On some problems involving multiple scattering due to edges","cited_arxiv_id":null,"evidence_quote":"gives the unit-circle adaptation of the asymptotic factorization used in the approximate product (61)."},{"cited_title":"Discrete scattering by a pair of parallel defects","cited_arxiv_id":null,"evidence_quote":"provides the exact zero-offset solution used to verify the special case $M=0$ of the matrix formulation."}],"review_version":1}