{"id":"f8b4edc1-4213-4353-b3a7-ed2143613703","arxiv_id":"2411.13324","paper_version":1,"verdict":"CONDITIONAL","confidence":"MODERATE","novelty_score":6.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":1,"one_line_summary":"A hybrid finite-volume method with a simplified one-variable double-flux scheme suppresses pressure oscillations at material interfaces while keeping correct shock capturing in thermally perfect multicomponent flows.","lead":"An adaptive CFD solver blends a non-conservative interface treatment with a conservative shock-capturing scheme to stop pressure oscillations at material interfaces. It targets multi-species, temperature-dependent gas mixtures, including reactions, and is verified on shock-bubble, detonation, and flame tests.","discovery_kind":"new_method","skeptic_critique":{"model":"deepseek-v4-flash","headline":"The hybrid claim rests on an unspecified, problem-dependent shock sensor threshold S_t (Eq. 40), with no sensitivity study or stated values; a mis-set threshold changes the weak solution or returns interface oscillations, so the reliability claim is not yet supported.","rationale":"The reader's weakest assumption identifies the same load-bearing concern: the unspecified, problem-dependent shock sensor threshold S_t. My reading of the manuscript confirms that Section 4.1 gives no threshold values or selection guidance, Algorithm 1's switching description is internally inconsistent, and all verification tests use fixed, unreported thresholds without sensitivity analysis. The paper's strongest claim is not merely that the method works for one tuned threshold, but that it is 'capable of handling multi-dimensional shock-interface problems reliably and efficiently' in general; that claim requires either a demonstrated robustness to S_t or a concrete default selection procedure. A single Sod run compared to the conservative scheme does not establish convergence to the correct weak solution, especially because the hybrid scheme is non-conservative at the material front and the paper itself notes the impacts of AMR coarse-fine operations on the double-flux scheme are not analyzed. The underlying numerical construction is reasonable and the verification suite is extensive, so the appropriate response is to require the missing sensitivity evidence rather than reject the method. The reader's conditional verdict already reflects this, so no verdict change is needed.","tokens_in":21204,"tokens_out":2925,"duration_ms":37848,"concrete_test":"Report the S_t values used for every benchmark, then rerun the 1-D Sod problem and the 2-D shock-bubble case with S_t spanning, for example, 1e-4, 3e-4, 1e-3, 3e-3, 1e-2, and 3e-2, at two resolutions (e.g., 200 and 800 cells in the principal direction). For each run measure: (i) L1 error of density and shock position relative to a high-resolution fully conservative reference; (ii) total mass and energy conservation loss; and (iii) the maximum pressure and velocity oscillation near the material interface. If shock position or interface profiles vary by more than the grid-convergence error across that S_t range, the 'same weak solution correctly' claim is not robust. Additionally, run a material-interface-crossing-shock test with a known or front-tracking reference to confirm threshold independence in the regime the abstract emphasizes.","verdict_should_be":"UNCHANGED","load_bearing_attack":"The central claim that the hybrid solver 'converges to the same weak solution correctly' and handles shock-interface problems reliably depends on the shock sensor threshold S_t in Eq. (40) selecting the fully conservative scheme everywhere that a non-conservative double-flux update would change the weak solution. Section 4.1 states that S_t is 'problem-related' and gives no value for any test case, and Section 4.3.1 reports only a single Sod comparison without a sensitivity study or grid-convergence demonstration. If S_t is too large, double-flux fluxes are applied in or near shocked cells, the energy update becomes non-conservative at the discontinuity, and the shock speed and strength can converge to the wrong weak solution. If S_t is too small, the pressure and velocity oscillations the method is designed to cure reappear. The risk is compounded by the cell-by-cell switching: at an interface between a double-flux cell and a conservative cell, the left and right fluxes are inconsistent, so local conservation error is not controlled by the smoothness property that justifies the sensor. Algorithm 1's description of the switching condition is also garbled, making the exact logic unreproducible from the text. None of the verification cases includes a threshold-sweep or a shock-interface test with an independent reference solution, so the correctness claim rests on an unexamined user choice.","agreement_with_reader":"agree"},"referee_report":{"model":"deepseek-v4-flash","summary":"This paper presents a second-order finite volume hybrid scheme for multicomponent compressible flows with thermally perfect species. The method extends the double-flux idea to require only one frozen auxiliary variable, \\hat\\gamma, and blends the extended double-flux update with a conservative HLLC/MUSCL scheme using a normalized pressure-curvature shock sensor, with characteristic decomposition used near shocks and primitive-variable interpolation elsewhere. The solver is embedded in block-structured AMR using the AMROC framework. Verification and demonstration cases include 1D and 2D entropy-wave convergence, advection of an H2 bubble, the 1D Sod problem, 2D inert shock-bubble interaction, DDT in a smooth channel, and a non-premixed planar RDE.","tokens_in":21442,"tokens_out":9783,"duration_ms":103721,"significance":"If the central claims hold, the paper offers a practical and relatively inexpensive cure for interface pressure oscillations in multicomponent reacting-flow simulations. The algebraic derivation of pressure equilibrium in Section 3.1 is clean, the method uses only one auxiliary variable, smooth entropy-wave tests show second-order convergence, the CPU overhead is reported to be below 10%, and the benchmark suite is substantial, including viscous and reactive cases. The main risk is that the correctness of the hybrid weak solution depends on a user-specified shock-sensor threshold for which no values or sensitivity analysis are given, and on an AMR nesting assumption that is not verified; these issues need to be addressed before the reliability claim is fully supported.","major_comments":[{"comment":"The central reliability claim depends on the shock sensor threshold S_t, but the paper provides no values, no sensitivity study, and no automatic selection criterion for any test case. Section 4.1 states that S_t is \"problem-related,\" and Section 4.3.1 reports only a single Sod comparison at one resolution. Since a too-large threshold applies non-conservative double-flux updates in or near shocks and can change the weak solution, while a too-small threshold restores interface oscillations, the claim that the hybrid solver \"converges to the same weak solution correctly\" is not yet demonstrated. Algorithm 1 is also not reproducible as written: Step 2 contains the incomplete condition \"> S_T^k using shock-capturing scheme, > S_T^k with extended double-flux scheme,\" and Step 3 then branches on S_k < S_T^k; the exact switching logic and the values used must be stated clearly.","section":"Section 4.1, Eq. (40), Algorithm 1"},{"comment":"The hybrid update is non-conservative at the material interface and at the interface between double-flux and conservative cells, so convergence to the correct weak solution needs a quantitative demonstration. The Sod test compares one resolution of the hybrid scheme with one resolution of the conservative scheme; no mesh-refinement study, no grid-convergence of the shock speed and strength, and no independent reference solution are provided. In addition, the paper does not quantify the conservation error produced by the cell-by-cell switching, where inconsistent left and right fluxes meet at a double-flux/conservative edge. Please add a grid-convergence study for a shock-interface problem and report the local conservation error at switching boundaries.","section":"Section 4.3.1"},{"comment":"The AMR claim is load-bearing but rests on an unverified premise. The paper states that the effects of coarse-fine prolongation and restriction on the non-conservative double-flux scheme have not been taken into account, and that the material interface is assumed to remain nested within the finest refinement level. For the AMR-based DDT simulation in Section 5.2, no evidence is given that this nesting condition is satisfied at all times, and no analysis of conservation errors at coarse-fine boundaries is provided. Please either verify the nesting condition for the presented AMR runs or analyze and control the conservation error introduced at coarse-fine interfaces.","section":"Section 4.2"}],"minor_comments":[{"comment":"The sign of the pressure term in Eq. (31) is incorrect: from Eqs. (14) and (30) one obtains E = h_f0 + p/(\\hat\\gamma-1) + (u^2+v^2)/2, not minus. Equation (32) and Algorithm 1, Step 10 use the plus sign, so this appears to be a typo, but it should be corrected because Eq. (31) is central to the derivation.","section":"Section 3.1.1, Eq. (31)"},{"comment":"As written, e_n = |\\psi_n - \\psi_0|/\\psi_n with \\psi_0 = 0 and \\psi_n the relative total-change measure evaluates to 1 identically; please restate the intended conservation-loss measure, probably e_n = |\\sum_j(U_j^n - U_j^0)| / \\sum_j U_j^n.","section":"Section 3.2.2, Eq. (39)"},{"comment":"The text says the temperature spikes are compared \"with the same calculation time\" at t = t1, but the figure caption reports t1 = 0.1 ms for the double-flux result and t2 = 0.008 ms for the conservative result; please clarify this apparent inconsistency.","section":"Section 5.1, Figure 12"},{"comment":"Many typos and inconsistent notation remain, including \"Correponding,\" \"auxillary,\" \"Langrangian,\" \"charatertistic,\" and \"the the original\" in the Highlights, as well as S_T^k versus S_t^i for the threshold. A careful proofread is needed.","section":"Throughout"},{"comment":"Approach B is claimed to guarantee pressure and velocity equilibrium by \"a derivation analogous to Eq. (32),\" but the derivation is not shown; since \\hat C_p in Eq. (34) includes the enthalpy of formation whereas Eq. (30) does not, please state the analogous relations explicitly or note any difference.","section":"Section 3.1.2"},{"comment":"The statement that the pressure-curvature sensor is \"resolution independent\" is plausible for smooth pressure but is not demonstrated under AMR refinement; a short numerical demonstration or discussion would help.","section":"Section 4.1"}],"recommendation":"major_revision","confidential_remarks":"This is a promising applied paper, but the missing threshold values and the unverified AMR nesting condition are real reproducibility risks for a computational physics journal. I would encourage the editor to request the actual S_t values used in Sections 4.3.1, 4.3.2, 5.2, and 5.3, together with a sensitivity test on the Sod problem, before acceptance. The manuscript also lacks a data/code availability statement; given the empirical nature of the validation, this is worth requesting."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"Quick take: this is a solid, incremental contribution to compressible multicomponent flow solvers. The genuinely new piece is the one-variable reformulation of the double-flux method (Approaches A and B) and the hybrid conservative/double-flux switching with characteristic-variable interpolation at shocks. I think the central derivations hold up.\n\nThe paper does a lot right. Section 3.1 is clean: freezing an averaged gamma is enough to preserve pressure and velocity equilibrium at a material interface, and the two approaches are compared carefully. The convergence studies in 3.2.1 show second-order accuracy, and the H2-bubble advection test confirms the interface behavior. The Sod and shock-bubble tests support the hybrid idea, and the DDT and RDE cases are substantial extra validation that the method works with viscosity, chemistry, and AMR.\n\nThe soft spots are real but not fatal. The shock sensor threshold S_t (Eq. 40) is problem-dependent and no values are reported for any test case. The correctness claim for the hybrid scheme depends on the conservative flux being active exactly at shocks; a threshold too high can produce the wrong weak solution, too low brings back interface oscillations. A threshold sweep or at least reported values would close this. The paper states the threshold 'can be resolved separately', which is hand-wavy.\n\nSecond, the interaction of AMR with the non-conservative scheme is explicitly unanalyzed; the paper says the interface is kept inside the finest level. That's an honest limitation, but it means the 'reliable multi-dimensional' claim is qualified.\n\nThird, Algorithm 1's step 2 looks garbled in the manuscript, making the exact switching logic hard to reproduce. Also, the 'only method' claim for thermally perfect species with temperature-dependent Cp overstates things, since Johnson & Kercher and Lv & Ihme already handle that in DG frameworks.\n\nOverall, the math is sound and the method is useful. I'd send it to review and ask for threshold values, a sensitivity study, and a fix to Algorithm 1. The right reader is someone doing practical combustion CFD with multicomponent flows; they'll get a working method with a clear caveat.","headline":"A useful incremental double-flux variant with clean derivations and broad validation, but the unspecified shock sensor threshold and AMR caveat need addressing.","tokens_in":21981,"tokens_out":3044,"would_cite":true,"duration_ms":32274,"reading_group":"maybe","serious_thinker":"yes","would_accept_peer_review":true},"rs_alignment":null,"lean_confirmation":null,"pith_extraction":{"msc":[],"pacs":[],"model":"deepseek-v4-flash","headline":"A hybrid double-flux and conservative scheme suppresses interface pressure oscillations while preserving correct shock capturing for thermally perfect multicomponent flows.","keywords":["double-flux algorithm","hybrid scheme","thermally perfect gas","multicomponent flow","material interface","pressure oscillation","adaptive mesh refinement","reacting flow"],"falsifier":"On the one-dimensional Sod problem using thermally perfect nitrogen, set the shock sensor threshold so that the double-flux scheme is active across the shock front; if the computed shock speed or post-shock pressure differs measurably from the fully conservative reference solution (or from the exact weak solution), the claim that the hybrid solver converges to the correct weak solution is falsified.","tokens_in":20968,"feed_emoji":"🔥","tokens_out":5918,"duration_ms":51342,"temperature":0.7,"pith_summary":"This paper develops a finite-volume solver for compressible multicomponent flows with temperature-dependent specific heats, combining two strategies: an extended double-flux method that suppresses spurious pressure and velocity oscillations at material interfaces, and a standard conservative scheme for shock capturing. The two are blended by a shock sensor so that the non-conservative scheme is used only where pressure is smooth. The paper's claim is that this hybrid converges to the correct weak solution in shock-interface problems, keeps conservation loss at the material front small, and remains stable and efficient for viscous and chemically reacting flows, including a rotating detonation engine configuration. A practical refinement is that the double-flux method is simplified to require only one frozen auxiliary variable, the averaged specific-heat ratio $\\hat{\\gamma}$, instead of two.","feed_headline":"Hybrid fluxes stop gas-interface pressure spikes, keep shocks","feed_subtitle":"Double-flux scheme needs only one frozen variable and blends with conservative fluxes for reacting multicomponent flows.","key_machinery":"The load-bearing mechanism is the hybridized flux and interpolation strategy built around an extended double-flux scheme. At each cell, the normalized pressure curvature $S_i = |\\bar{p}_{i-1}-2\\bar{p}_i+\\bar{p}_{i+1}|/(\\bar{p}_{i-1}+2\\bar{p}_i+\\bar{p}_{i+1})$ acts as a shock sensor: if $S_i$ stays below a user-set threshold, the cell is updated with the extended double-flux scheme, which freezes the averaged specific-heat ratio $\\hat{\\gamma}$ during the two RK stages and then corrects the total energy; if the sensor fires, the update switches to a fully conservative scheme with characteristic-decomposition MUSCL interpolation. The extended double-flux reformulation expresses total energy as $E = h^f_0 - p/(\\hat{\\gamma}-1) + (u^2+v^2)/2$ (Approach A) or $E = p/(\\hat{\\gamma}-1) + (u^2+v^2)/2$ (Approach B), so that freezing $\\hat{\\gamma}$ cancels the terms that would otherwise produce $\\delta p \\neq 0$ at a material front.","core_discovery":"The paper's central claim is that a hybridized finite-volume scheme, blending an extended double-flux algorithm with a traditional conservative scheme via a pressure-curvature shock sensor, solves multicomponent shock-interface problems for thermally perfect species without the pressure-oscillation artifacts of fully conservative schemes and without the incorrect-shock artifacts of non-conservative ones. The extended double-flux method freezes a single averaged specific-heat ratio $\\hat{\\gamma}$ during the cell update; this single auxiliary variable is enough to guarantee pressure and velocity equilibrium at a material front, and it reduces conservation loss in the energy compared with the original two-variable double-flux model. The hybrid solver is verified on entropy-wave advection, an inert H$_2$ bubble, the Sod problem, a two-dimensional viscous shock-bubble interaction, a reactive H$_2$ bubble, a deflagration-to-detonation transition, and a non-premixed planar rotating detonation engine; in the Sod test it \"converges to the same weak solution correctly as compared to the fully conservative one,\" and in the detonation cases it avoids artificial partial extinction that the conservative scheme exhibits.","pith_inferences":["The shock-sensor threshold $S_t$ is left user-specified with no guidance or default value; a natural next step is an automatic estimator based on local smoothness that adapts $S_t$ per cell, which would remove the main reliability caveat.","The hybrid strategy should extend to higher-order base schemes such as WENO or discontinuous Galerkin as long as the double-flux energy correction is applied consistently per cell; the paper only demonstrates second-order MUSCL, but the mechanism is not tied to MUSCL.","If the method generalizes to three dimensions and other reaction mechanisms, it offers a practical path for large-scale detonation and shock-interface simulations with detailed chemistry, where fully conservative schemes produce spurious mixing that can change the combustion outcome."],"forward_implications":["The hybrid solver reproduces the correct weak solution in the Sod problem, meaning the blending does not disrupt shock speed or post-shock states while removing the pressure spike at the material front.","Only one auxiliary variable $\\hat{\\gamma}$ is frozen during the double-flux update, which simplifies implementation relative to the original double-flux method and reduces energy conservation loss; in the inert H2 bubble test, mass and momentum loss stay at machine accuracy and total energy loss is 0.002913 over the run.","Because the shock sensor is resolution-independent, the hybrid scheme combines with block-structured adaptive mesh refinement; the DDT and RDE cases show it works with 3 to 4 refinement levels and reduces wall-clock time relative to a uniform fine grid.","For reactive cases, the hybrid scheme avoids spurious oscillation-induced mixing errors: in the non-premixed planar RDE, the conservative scheme drives partial detonation extinction through locally high equivalence ratios while the hybrid scheme maintains a self-sustained detonation.","Both extended double-flux approaches (A and B) give virtually identical results, with Approach A slightly better for momentum conservation."],"supporting_citations":[{"why":"Original adaptive double-flux shock-capturing algorithm for unsteady reactive flows; this paper extends it and uses its test configuration for the reactive H2 bubble.","marker":"[15]"},{"why":"Introduces the original double-flux model for thermally perfect gases with two auxiliary variables, which the present method simplifies to one.","marker":"[18]"},{"why":"Quasi-conservative approach to prevent pressure oscillations in multicomponent flow; the baseline non-conservative strategy this hybrid builds on.","marker":"[3]"},{"why":"High-order WENO extension of the double-flux method; provides comparison for conservation loss and confirms interface oscillation suppression.","marker":"[16]"},{"why":"Discontinuous Galerkin extension of the double-flux method for chemically reacting Navier-Stokes equations; a recent alternative framework for the same problem.","marker":"[17]"},{"why":"Comparison of algorithms for multicomponent reacting flows; supplies the inert H2 bubble advection test case used for verification.","marker":"[21]"},{"why":"HLLC Riemann solver, whose inherent velocity-equilibrium preservation at material interfaces is crucial for the double-flux construction.","marker":"[38]"},{"why":"Front-tracking/ghost-fluid method providing reference trajectories for the shock-bubble interaction validation.","marker":"[47]"},{"why":"Source of the normalized pressure-curvature shock sensor used for hybridization between double-flux and conservative schemes.","marker":"[27]"}],"fun_headline_variants":["Hybrid double-flux scheme kills gas-interface pressure spikes","Adaptive flux blending stops shock-interface pressure errors","Single frozen gamma tames pressure oscillation at interfaces","Efficient multi-species solver avoids pressure spikes and wrong shocks","Blend double-flux with conservative fluxes for accurate shock-interface runs"],"cache_read_input_tokens":3200,"weakest_assumption_plain":"The central reliability claim rests on the user-chosen shock sensor threshold $S_t$: if the threshold is set too high, the non-conservative double-flux scheme is applied near shocks and may converge to the wrong weak solution, and if set too low, interface oscillations are not cured; the paper gives no guidance or default values for $S_t$.","fun_headline_variants_meta":{"raw":{"variants":["Hybrid double-flux scheme kills gas-interface pressure spikes","Adaptive flux blending stops shock-interface pressure errors","Single frozen gamma tames pressure oscillation at interfaces","Efficient multi-species solver avoids pressure spikes and wrong shocks","Blend double-flux with conservative fluxes for accurate shock-interface runs"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.000183,"raw_usage":{"total_tokens":1281,"prompt_tokens":878,"completion_tokens":403,"prompt_tokens_details":{"cached_tokens":384},"prompt_cache_hit_tokens":384,"prompt_cache_miss_tokens":494,"completion_tokens_details":{"reasoning_tokens":320}},"tokens_in":494,"tokens_out":403,"duration_ms":5133,"temperature":1.0,"reasoning_tokens":320,"cache_read_input_tokens":384,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-12T16:33:09.957722+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"On the one-dimensional Sod problem using thermally perfect nitrogen, set the shock sensor threshold so that the double-flux scheme is active across the shock front; if the computed shock speed or post-shock pressure differs measurably from the fully conservative reference solution (or from the exact weak solution), the claim that the hybrid solver converges to the correct weak solution is falsified.","supporting_citations":[{"cited_title":"Billet, R","cited_arxiv_id":null,"evidence_quote":"Original adaptive double-flux shock-capturing algorithm for unsteady reactive flows; this paper extends it and uses its test configuration for the reactive H2 bubble."},{"cited_title":"Abgrall, How to prevent pressure oscillations in multicomponent flow calculations: a quasi conservative approach, Journal of Computational Physics 125 (1) (1996) 150–160","cited_arxiv_id":null,"evidence_quote":"Quasi-conservative approach to prevent pressure oscillations in multicomponent flow; the baseline non-conservative strategy this hybrid builds on."},{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"High-order WENO extension of the double-flux method; provides comparison for conservation loss and confirms interface oscillation suppression."},{"cited_title":"Bando, M","cited_arxiv_id":null,"evidence_quote":"Comparison of algorithms for multicomponent reacting flows; supplies the inert H2 bubble advection test case used for verification."},{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"HLLC Riemann solver, whose inherent velocity-equilibrium preservation at material interfaces is crucial for the double-flux construction."},{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"Source of the normalized pressure-curvature shock sensor used for hybridization between double-flux and conservative schemes."}],"review_version":1}