{"id":"a3f95a7f-f399-40a7-b0c9-ce0237e69bf2","arxiv_id":"2607.06429","paper_version":1,"verdict":"CONDITIONAL","confidence":"MODERATE","novelty_score":5.0,"correctness_risk":"unknown","formal_verification":"none","parameter_count":0,"one_line_summary":"A nested-domain volume-surface integral equation for the Helmholtz equation with heterogeneous density is derived and verified against BEM and FEM-BEM benchmarks.","lead":"This paper derives a volume-surface integral equation (VSIE) for acoustic wave propagation through nested materials with density discontinuities, and benchmarks it against FEM-BEM and analytical solutions. It matters because integral equation methods can avoid artificial boundary conditions and reduce mesh requirements for unbounded-domain acoustic simulations.","discovery_kind":"unclear","skeptic_critique":{"model":"glm-5.2","headline":"The L²-decoupled discretization breaks the pressure-continuity assumption used to derive the surface terms; without evidence that the discrete solution recovers interface continuity, the convergence stall in Benchmark 4 is unexplained.","rationale":"The reader correctly identified the L²-decoupled discretization as the weakest link. The derivation of the continuous VSIE (Theorem 7) is sound: the chain of lemmas from the perturbed Helmholtz equation through integration by parts and jump relations follows standard operator identities, and the mass function construction at interfaces is verified by both interior and exterior traces giving the same expression. The formulation correctly reduces to the Lippmann-Schwinger equation when density is constant (Remark 10) and to a pure VIE when density is globally continuous (Remark 9), which provides sanity checks. The benchmark suite is a genuine verification effort with four test cases of increasing complexity. However, the L²-decoupled discretization (Eq. 42) is an ad hoc modeling choice whose convergence is not independently established. The derivation's integration by parts step (Lemma 3) explicitly uses pressure continuity to combine boundary terms, yet the discrete space does not enforce this continuity. The integral operators may recover it weakly, but this is unproven, and the convergence stall in Benchmark 4 — the only test case with both heterogeneity and a density discontinuity — provides empirical evidence that the recovery may be incomplete at high contrasts. The asymmetric mesh comparison (2× finer for FEM-BEM) is a secondary concern: it is a reasonable pragmatic choice given that FEM suffers from pollution effects, but it weakens the quantitative convergence claims. The absence of performance data, while limiting the 'large-scale' claim in the abstract, does not affect the correctness of the formulation. The CONDITIONAL verdict is appropriate: the formulation is a legitimate contribution, but the discretization needs either a convergence proof or additional evidence (such as the proposed jump measurement) that interface continuity is recovered before the convergence claims are fully established.","tokens_in":23316,"tokens_out":4687,"duration_ms":276058,"concrete_test":"On Benchmark 4, extract the collocation-point values of p_Ω₁, p_Ω₂, and p_Γ₂ nearest to Γ₂ at each mesh refinement level. Compute the discrete pressure jumps ‖p_Ω₁ − p_Γ₂‖₂ and ‖p_Ω₂ − p_Γ₂‖₂ (evaluated at matching nearest-neighbor points). If these jumps do not decrease at least as O(h) with mesh refinement h, the L²-decoupled discretization fails to recover interface continuity, and the convergence stall in Figure 10 is attributable to the discretization rather than to low-order quadrature or staircase effects. If the jumps do decrease appropriately, the concern does not land and the stall has another cause.","verdict_should_be":"UNCHANGED","load_bearing_attack":"The continuous VSIE (37) is derived under the assumption that p is continuous across interfaces. Specifically, in Lemma 3 (Eq. 23), the two boundary terms at Γ₂ from integration by parts over Ω₁ and Ω₂ are combined into a single term ∫_{Γ₂} G(α₂−α₁)p n̂₂ dy using the fact that the same p appears on both sides of Γ₂ (interface condition (6)). This yields a single surface unknown p_Γ₂ weighted by the density jump (α₂−α₁). The L²-decoupled discretization (Eq. 42) then treats p_Ω₁, p_Ω₂, p_Γ₁, p_Γ₂ as four independent L² functions. In the discrete system (43), the surface integral at Γ₂ couples only to p_Γ₂, while the volume integrals couple to p_Ω₁ and p_Ω₂. Nothing in the discrete formulation forces p_Ω₁|_{Γ₂} = p_Γ₂ or p_Ω₂|_{Γ₂} = p_Γ₂. The authors state this continuity is recovered 'via the integral operators in the VSIE,' but no error analysis or convergence proof is provided for this claim. If the coupling is insufficient — particularly at high density contrasts where (α₂−α₁) is large — the surface terms that carry the jump physics are evaluated with a p_Γ₂ that may not be consistent with the volumetric pressure near the interface. This would explain both the localized artifacts near Γ₂ in Figures 5 and 9 and the convergence stall in Benchmark 4 (Figure 10), where the density jump is 1000→2000 kg/m³. The paper cites [35] for the L² extension of the integral operator, but that result concerns the continuous operator on L², not the convergence of a decoupled piecewise-constant collocation scheme that does not enforce the trace identity used in the derivation.","agreement_with_reader":"agree"},"referee_report":{"model":"glm-5.2","summary":"This manuscript derives a volume-surface integral equation (VSIE) for the Helmholtz equation in the presence of two nested domains with heterogeneous material parameters and high-contrast density jumps at interfaces. The derivation proceeds through a sequence of standard steps (convolution theorem, integration by parts, jump relations) to shift all derivatives off the unknown pressure field, yielding a formulation amenable to L² discretization with piecewise-constant basis functions. The formulation is verified against analytical solutions, BEM, and FEM-BEM on four benchmarks of increasing complexity. The derivation is parameter-free, and the code is openly available.","tokens_in":23580,"tokens_out":1379,"duration_ms":2033902,"significance":"The extension of VSIE to nested domains with density discontinuities is a meaningful contribution to computational acoustics, addressing a gap between the standard Lippmann-Schwinger equation (constant density) and coupled FEM-BEM approaches. The parameter-free derivation (α and β defined from physical parameters via Eqs. 11-12) and the reproducible code (GitHub repository) are notable strengths. The benchmark suite, while modest in scale, is well-designed and provides credible verification. The L²-decoupled discretization allowing P0 basis functions is practically attractive.","major_comments":[{"comment":"§2.3.2, Eqs. (42)-(43): The L²-decoupled discretization treats p_Ω1, p_Ω2, p_Γ1, p_Γ2 as four independent L² functions, with no explicit enforcement of pressure continuity across interfaces. The continuous VSIE (37) is derived under the assumption that p is continuous across Γ₂ (used in Lemma 3, Eq. 23, to combine two boundary terms into a single integral weighted by (α₂−α₁)). The authors state that continuity is recovered 'via the integral operators in the VSIE,' but no convergence analysis or error estimate is provided for this claim. The convergence stall observed in Benchmark 4 (Figure 10, density jump 1000→2000 kg/m³) and the localized artifacts near Γ₂ in Figures 5 and 9 are consistent with the possibility that this implicit enforcement is insufficient at high contrasts. The authors should either (a) provide a reference or argument justifying that the discrete system enforces p_Ω1|","section":null},{"comment":"§2.2.6, Lemma 6, Eqs. (30)-(31): The jump relations are applied to obtain the Dirichlet traces at Γ₁ and Γ₂. For the trace at Γ₂ (Eq. 31), the jump relation (33) is applied to the double-layer operator on Γ₂, yielding the coefficient ½(ρ₀/ρ₂ + ρ₀/ρ₁). However, the adjoint double-layer volume integral (the ∇ₓ·∫ G(∇_yα)p dy term) also has a singularity when x approaches Γ₂ from either side. The manuscript does not discuss whether this volume integral operator has a jump across Γ₂. If it does, the trace equation (31) may be missing a jump contribution from this term. Please clarify whether the adjoint double-layer volume operator is continuous across Γ₂, and if so, why.","section":null},{"comment":"§3.5, Figure 10: The convergence study for Benchmark 4 shows the L² error leveling off after approximately 12 elements per wavelength, with the L∞ error remaining notably higher. The text attributes this to 'localized errors near the high-contrast interface' but does not analyze the cause. Given that this is the only benchmark exercising the full nested VSIE with both heterogeneous density and a density jump, the convergence stall is load-bearing for the paper's central claim that the formulation is correct and verified. Please provide a more specific diagnosis: is the stall due to the L²-decoupled discretization (see comment above), the midpoint quadrature for weakly-singular integrals, the staircase approximation, or another factor? A single refinement study isolating one of these factors would strengthen the claim significantly.","section":null}],"minor_comments":[{"comment":"§2.2.6, Theorem 6: The text references 'the VSIE (37)' in the statement of Lemma 6, but Theorem 7 (which states Eq. 37) has not yet been stated at that point in the manuscript. The forward reference should be clarified.","section":null},{"comment":"§2.2.4, Eq. (28): The surface integral over Γ₂ in Eq. (28) uses (α₁−α₂) with normal n̂₁, while Eq. (24) uses (α₂−α₁) with n̂₂. The sign convention is consistent (since n̂₁ = −n̂₂ at Γ₂), but a brief note would help the reader verify.","section":null},{"comment":"§3.1: The statement 'we used mass-matrix preconditioning for the BEM' could benefit from a citation or brief description, as the BEM formulation is referenced to [47] but the preconditioning choice is not discussed there.","section":null},{"comment":"§3.3, Figure 5: The text mentions 'slight differences at the material interface' in the VSIE solution. It would help to quantify these differences (e.g., as a percentage of the field amplitude) to contextualize the visual comparison.","section":null},{"comment":"§4: The phrase 'ambiguous factor of two in mesh resolution between VSIE and FEM-BEM' is unclear. Please rephrase to clarify whether this refers to the 2:1 ratio used in the benchmarks or an inherent ambiguity in the comparison methodology.","section":null},{"comment":"References [9] and [38] are dated 2026; please verify these are not preprints with updated publication status.","section":null}],"recommendation":"major_revision","confidential_remarks":"The reader's concern about the L²-decoupled discretization is well-founded and is the most important issue to address. The convergence stall in Benchmark 4 is a concrete symptom that needs diagnosis. However, I do not view this as a fatal flaw: the continuous formulation appears correct, and the discretization issue is likely addressable with either a convergence argument or a modified discrete space. The paper is a solid contribution if the authors can close this gap. The benchmark suite, while not exhaustive, is sufficient for verification at the level claimed."},"author_rebuttal":{"model":"glm-5.2","summary":"We thank the referee for a careful and constructive report. The three major comments all identify legitimate gaps in the manuscript's justification and analysis. We address each below: (1) the L²-decoupled discretization's implicit enforcement of pressure continuity lacks a rigorous justification, and we will add a discussion and a numerical diagnostic; (2) the adjoint double-layer volume operator's continuity across Γ₂ needs explicit clarification, which we will provide; (3) the convergence stall in Benchmark 4 requires a more specific diagnosis, which we will supply via an isolated refinement study. All three points lead to manuscript revisions.","responses":[{"response":"The referee correctly identifies a gap between the continuous formulation (which assumes pressure continuity across Γ₂) and the discrete formulation (which treats p_Ω1, p_Ω2, p_Γ1, p_Γ2 as independent L² functions with no explicit continuity constraint). We agree that the manuscript's current statement that continuity is recovered 'via the integral operators in the VSIE' is insufficiently justified. We will revise the manuscript to address this in two ways. First, we will add a discussion of the mechanism: the double-layer boundary integral operators on Γ₁ and Γ₂ couple the surface unknowns p_Γ1, p_Γ2 to the volumetric unknowns p_Ω1, p_Ω2 through the off-diagonal blocks K_{Ωi,Γj} and K_{Γi,Ωj} in the linear system (43). At the continuous level, the jump relations (Lemma 6) ensure that the interior and exterior traces of the VSIE yield the same equation on each interface, which is the mechanism by which continuity is implicitly enforced. At the discrete level, this coupling transfers information across the interface but does not exactly enforce pointwise continuity. Second, we will add a numerical diagnostic to Benchmark 4 reporting the discrete jump |p_Ω1 − p_Γ2| and |p_Ω2 − p_Γ2| at interface elements as a function of mesh refinement, to quantify how well continuity is recovered in practice. We acknowledge that we cannot currently provide a rigorous error estimate for the implicit continuity enforcement at high contrasts, and we will state this limitation transparently. revision_made = 'partial'","revision_made":"partial","referee_comment":"§2.3.2, Eqs. (42)-(43): The L²-decoupled discretization treats p_Ω1, p_Ω2, p_Γ1, p_Γ2 as four independent L² functions, with no explicit enforcement of pressure continuity across interfaces. The continuous VSIE (37) is derived under the assumption that p is continuous across Γ₂ (used in Lemma 3, Eq. 23, to combine two boundary terms into a single integral weighted by (α₂−α₁)). The authors state that continuity is recovered 'via the integral operators in the VSIE,' but no convergence analysis or error estimate is provided for this claim. The convergence stall observed in Benchmark 4 (Figure 10, density jump 1000→2000 kg/m³) and the localized artifacts near Γ₂ in Figures 5 and 9 are consistent with the possibility that this implicit enforcement is insufficient at high contrasts. The authors should either (a) provide a reference or argument justifying that the discrete system enforces p_Ω1|"},{"response":"We thank the referee for this careful observation. The adjoint double-layer volume operator T[q](x) = ∇ₓ · ∫_{Ω₁∪Ω₂} G_{k₀}(x,y) (∇_y α(y)) q(y) dy is indeed singular when x approaches Γ₂, and the manuscript should discuss this explicitly. The key point is that the singularity structure of this operator differs from that of the double-layer boundary integral operator. The volume integral involves ∇_y α(y), which is a bounded function supported in the interior of Ω₁ and Ω₂ (away from the interfaces, since α is C¹ inside each subdomain). The gradient ∇ₓ acts on the Green's function, producing a kernel with a 1/|x−y|²-type singularity integrated over a volume. When x approaches Γ₂, the integration domain excludes a neighborhood of x itself (since the integral is over the open subdomains), and the volume integral's limit exists and is the same from both sides of Γ₂. This is because the volume integral operator with a weakly singular kernel is continuous across surfaces that are not part of the integration domain's boundary in a way that would produce a jump. More precisely, the adjoint double-layer volume operator maps L²(Ω) to H^{1/2}(Ω) continuously (see, e.g., Steinbach [19, Section 3.1] and Costabel [35]), and its trace is well-defined and continuous across interior interfaces. This contrasts with the double-layer boundary integral operator, which has a jump because the integration surface coincides with the trace surface. We will add a remark to the manuscript clarifying this distinction and citing the relevant functional-analytic results. revision_made = 'yes'","revision_made":"yes","referee_comment":"§2.2.6, Lemma 6, Eqs. (30)-(31): The jump relations are applied to obtain the Dirichlet traces at Γ₁ and Γ₂. For the trace at Γ₂ (Eq. 31), the jump relation (33) is applied to the double-layer operator on Γ₂, yielding the coefficient ½(ρ₀/ρ₂ + ρ₀/ρ₁). However, the adjoint double-layer volume integral (the ∇ₓ·∫ G(∇_yα)p dy term) also has a singularity when x approaches Γ₂ from either side. The manuscript does not discuss whether this volume integral operator has a jump across Γ₂. If it does, the trace equation (31) may be missing a jump contribution from this term. Please clarify whether the adjoint double-layer volume operator is continuous across Γ₂, and if so, why."},{"response":"The referee is correct that the convergence stall in Benchmark 4 is insufficiently analyzed and that it is load-bearing for the paper's claims. We will conduct an isolated refinement study to diagnose the cause. Based on our computational experience and the referee's suggestion, we suspect the stall is primarily due to the midpoint quadrature for the weakly-singular self-interaction integrals on the surface elements at Γ₂, combined with the L²-decoupled discretization's imperfect enforcement of continuity at high contrast. To isolate the factor, we will run two additional studies: (1) a refinement study on Benchmark 4 where the weakly-singular integrals near Γ₂ are evaluated with a higher-order quadrature rule instead of the analytical sphere approximation, holding the mesh fixed; and (2) a comparison of the discrete pressure jump |p_Ω1 − p_Γ2| at Γ₂ across refinement levels to assess whether the implicit continuity enforcement degrades at high contrast. We will report these results in the revised manuscript and provide a specific diagnosis. We acknowledge that if the stall persists under these interventions, it may indicate a fundamental limitation of the P0 L²-decoupled discretization at high density contrasts, which we will state honestly. revision_made = 'yes'","revision_made":"yes","referee_comment":"§3.5, Figure 10: The convergence study for Benchmark 4 shows the L² error leveling off after approximately 12 elements per wavelength, with the L∞ error remaining notably higher. The text attributes this to 'localized errors near the high-contrast interface' but does not analyze the cause. Given that this is the only benchmark exercising the full nested VSIE with both heterogeneous density and a density jump, the convergence stall is load-bearing for the paper's central claim that the formulation is correct and verified. Please provide a more specific diagnosis: is the stall due to the L²-decoupled discretization (see comment above), the midpoint quadrature for weakly-singular integrals, the staircase approximation, or another factor? A single refinement study isolating one of these factors would strengthen the claim significantly."}],"tokens_in":23283,"tokens_out":1752,"duration_ms":441221,"standing_objections":[]},"desk_editor":{"model":"glm-5.2","letter":"The main thing to know: this paper derives a VSIE formulation for two nested domains with discontinuous density across interfaces (Theorem 7, Eq. 37) and backs it up with a four-case benchmark suite against analytical solutions, BEM, and FEM-BEM. The derivation is clean — each step (convolution theorem, integration by parts, jump relations) is standard and correctly applied. The formulation is parameter-free: α and β are defined from physical material parameters, no constants are fitted. The specific combination of volume and surface integrals for nested domains with density jumps is genuinely new relative to the cited prior work on single-domain VSIE and the Lippmann-Schwinger equation. Credit is earned here.","headline":"The derivation is sound and the benchmarks are genuine; the L²-decoupling concern is real but probably not fatal.","tokens_in":24188,"tokens_out":212,"would_cite":false,"duration_ms":104223,"reading_group":"maybe","serious_thinker":"yes","would_accept_peer_review":true},"rs_alignment":null,"lean_confirmation":null,"pith_extraction":{"msc":[],"pacs":[],"model":"glm-5.2","headline":"Integral equations handle nested acoustic materials with sharp density jumps","keywords":[],"falsifier":"A benchmark scenario with very high density contrast (e.g., ratio exceeding 10:1) across an interface where the VSIE solution exhibits pressure discontinuities that do not diminish under mesh refinement, or where the relative error against FEM-BEM fails to converge below a practical threshold, would indicate that the L²-decoupled discretization cannot adequately capture the interface physics.","tokens_in":23492,"feed_emoji":"🔊","tokens_out":1064,"duration_ms":277529,"temperature":0.7,"pith_summary":"This paper derives a volume-surface integral equation (VSIE) for acoustic wave propagation through two nested domains with heterogeneous materials and discontinuous density at their interfaces. The central object is a single integral equation (Theorem 7, Eq. 37) that combines a single-layer volume integral over the interior domains, an adjoint double-layer volume integral involving gradients of the material contrast, and double-layer boundary integrals at both material interfaces. A mass function m(x) encodes the local density contrast, averaging the ratio of exterior to interior density at each point, including at interfaces where the density jumps. The key algebraic achievement is shifting all derivatives off the unknown pressure field and onto known material parameters and the Green's function, which permits discretization with piecewise-constant basis functions in L². The paper then benchmarks this formulation against analytical solutions, boundary element methods, and coupled finite-element–boundary-element methods across four test cases of increasing complexity, reporting mesh convergence and relative errors below 5% for most scenarios.","feed_headline":"Integral equations tame nested acoustic materials with sharp density jumps","feed_subtitle":"A single volume-surface equation replaces FEM-BEM coupling for waves crossing layered media, verified against four benchmarks.","key_machinery":"The mass function m(x) (Eq. 36), which encodes the local density contrast ρ₀/ρ(x) in the interior and its average at interfaces; the perturbed Helmholtz equation (Eq. 13) that localizes heterogeneity to the right-hand side; integration by parts (Lemma 3) that generates surface integrals at material interfaces; the exterior Green's function identity (Eq. 27) that eliminates the Laplacian; and the L²-decoupled discretization (Eq. 42) that treats volume and surface pressures as four independent functions.","core_discovery":"The paper proves that the heterogeneous Helmholtz transmission problem for two nested domains with discontinuous density can be solved by a single VSIE (Eq. 37) whose integrand contains no derivatives of the pressure. This is achieved by defining a mass function m(x) that averages the density contrast at interfaces, combined with integration-by-parts and the exterior Helmholtz equation to eliminate second-order derivatives. The resulting formulation supports L² discretization with piecewise-constant basis functions, decoupling the four geometric components (two volumes, two surfaces) into independent L² spaces where interface continuity is recovered implicitly through the integral operators.","pith_inferences":["If the L²-decoupled approach proves insufficient at very high density contrasts (e.g., air-water interfaces with ratios near 1000:1), a hybrid discretization that enforces H¹ conformity only at the highest-contrast interface while keeping L² freedom elsewhere could recover accuracy without sacrificing the simplicity of piecewise-constant basis functions.","The mass function's averaging structure at interfaces—taking the mean of density ratios from both sides—resembles a Robin-type condition and may connect to established transmission conditions in layered media theory, though the paper does not draw this parallel explicitly.","The localized numerical artifacts near density-jump interfaces visible in Figures 5 and 9, combined with the convergence stall in Benchmark 4, are consistent with the hypothesis that the weak enforcement of pressure continuity becomes less effective as the contrast increases and the mesh remains coarse relative to the interface physics."],"forward_implications":["The VSIE formulation can be extended to three or more nested domains by following the same design principles, broadening the class of geometries addressable without finite-element meshes in the unbounded exterior.","Because the formulation supports fast matrix arithmetic via hierarchical compression and the discrete system is dense but structured, large-scale biomedical acoustic simulations (e.g., transcranial ultrasound with CT-derived skull properties) could become computationally feasible.","The L²-decoupled discretization strategy—where physical interface continuity is enforced weakly through integral operators rather than through the function space—may be applicable to other transmission problems where enforcing H¹ conformity is computationally expensive.","The convergence slowdown observed at high-contrast interfaces (Benchmark 4) suggests that adaptive mesh refinement or higher-order quadrature near density jumps would be natural next steps for improving accuracy."],"fun_headline_variants":["Nested VSIE formulation removes derivatives from layered acoustic integrals","Single integral equation solves Helmholtz transmission across nested density jumps","Mass-averaged VSIE decouples volumes and surfaces for layered acoustic media","No FEM-BEM coupling needed: one VSIE handles nested heterogeneous acoustics","L²-discretized VSIE recovers interface continuity implicitly for layered waves"],"cache_read_input_tokens":0,"weakest_assumption_plain":"The L²-decoupled discretization does not explicitly enforce pressure continuity across material interfaces in the discrete function space; instead, it relies on the integral operators to recover continuity implicitly. If this implicit enforcement is insufficient at high contrasts or on coarse meshes, the surface integral terms that carry the density-jump physics may produce localized numerical artifacts and convergence may stall.","fun_headline_variants_meta":{"raw":{"variants":["Nested VSIE formulation removes derivatives from layered acoustic integrals","Single integral equation solves Helmholtz transmission across nested density jumps","Mass-averaged VSIE decouples volumes and surfaces for layered acoustic media","No FEM-BEM coupling needed: one VSIE handles nested heterogeneous acoustics","L²-discretized VSIE recovers interface continuity implicitly for layered waves","Nested volume-surface equation targets high-contrast acoustic transmission","Derivative-free VSIE integrand verified for discontinuous density interfaces"]},"model":"glm-5.2","effort":"high","cost_usd":0.0,"raw_usage":{"total_tokens":589,"prompt_tokens":448,"completion_tokens":141,"prompt_tokens_details":null},"tokens_in":448,"tokens_out":141,"duration_ms":11331,"temperature":1.0,"reasoning_tokens":31,"cache_read_input_tokens":0,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-07-08T05:56:47.782442+00:00","model_set":{"reader":"glm-5.2"},"falsifier":"A benchmark scenario with very high density contrast (e.g., ratio exceeding 10:1) across an interface where the VSIE solution exhibits pressure discontinuities that do not diminish under mesh refinement, or where the relative error against FEM-BEM fails to converge below a practical threshold, would indicate that the L²-decoupled discretization cannot adequately capture the interface physics.","supporting_citations":[],"review_version":1}