{"id":"b1ab490e-1002-4f85-b97f-fd4fb9c095d0","arxiv_id":"2505.18718","paper_version":1,"verdict":"CONDITIONAL","confidence":"MODERATE","novelty_score":6.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":8,"one_line_summary":"The paper proposes a revised log-plus-linear overlap model with a universal inner log constant κ_in ≈ 1/2.6 and finds two normal-stress models fit equally well closer to the wall.","lead":"This paper revises the model of the extended overlap region in wall-bounded turbulent flows by introducing separate log-law coefficients for the inner layer and the outer overlap region. It also re-tests two competing models for streamwise turbulence intensity in a near-wall region and finds both fit the data equally well.","discovery_kind":"extension","skeptic_critique":{"model":"deepseek-v4-flash","headline":"The revised Eq. 7 is fitted with three free parameters and a post hoc inner-layer boundary Yin; the claimed universal κ_in=2.6 is not robustly separated from the non-universal log+lin fit without an independent estimate of Yin.","rationale":"The reader's weakest assumption already identifies the universality of the inner pure-log layer as unde rived and post hoc; my concern sharpens this into a concrete parameter-sensitivity check on Eq. 7. The paper's proposed κ_in = 1/2.6 is the single most load-bearing number: if it is not robust to the admitted uncertainty in Yin (0.11–0.15), the central separation between universal inner and non-universal outer log coefficients collapses into a three-parameter curve fit. The normal-stress re-evaluation (Figs. 7–13) is also concerning because both models are fit in a region with 5–20% viscous stress, but it is presented as secondary evidence and does not bear the full weight of the headline claim. I agree with the reader that the paper is a plausible empirical revision, not a derivation, and that CONDITIONAL is the right verdict. The condition should be a demonstration that the claimed universal κ_in survives variation of the model's own undetermined boundary parameter Yin; absent that, the revised model has not been distinguished from a flexible fit.","tokens_in":10110,"tokens_out":1967,"duration_ms":15241,"concrete_test":"Re-analyze the Lee and Moser (2015) Reτ ≈ 5200 channel DNS and the Pirozzoli (2024) Reτ ≈ 12,055 pipe DNS using the paper's indicator-function method, and (1) compute κ_in from the plateau of Ξ over a range of y+ choices with Yin fixed at 0.12; (2) repeat with Yin varied over 0.10, 0.11, 0.13, 0.14, 0.15, re-fitting Eq. 7; (3) report κ_in and κ_o for each Yin. If κ_in stays within 0.38–0.39 for both flows across all Yin choices, the universality claim is supported. If κ_in drifts outside that band for either flow, the claim needs a separate derivation or a preselection criterion for Yin that is not currently in the paper. A secondary check: verify that the fitted κ_in is independent of the fitting lower bound y+ in the range 100–400.","verdict_should_be":"CONDITIONAL","load_bearing_attack":"The central claim is that the inner pure-log layer has a universal coefficient 1/κ_in = 2.6 for all wall-bounded flows, while the extended overlap log coefficient 1/κ_o is non-universal. The evidence, however, is the result of a fitting procedure whose region boundaries (Yin and Yout) are selected per flow and per pressure gradient, and whose parameter separation is not derived from matched asymptotics (the paper explicitly defers this: 'Additional matched asymptotic work is now under consideration, but is not covered here.'). Eq. 7 contains three parameters (κ_in, κ_o, S_o) plus the boundary Yin; figures 1–6 show fits for only a few high-Re cases. Because κ_in is estimated from a pure-log segment that is visually short (e.g., Y ≈ 0.03–0.12 in channel flow), the extracted value is sensitive to the chosen Yin, the selected y+ range, and the handling of the viscous and wake regions. The paper reports κ_in = 0.384 from the channel data but proposes 1/κ_in = 2.6 (κ_in ≈ 0.385) 'based on all measurements ... examined' without listing those cases, their Reynolds numbers, or the scatter. The reader is asked to accept universality of κ_in from a few high-Re DNS cases, while the non-universality of κ_o is inferred from the same data. A testable risk: for the same dataset, a different but equally plausible Yin choice (e.g., Yin = 0.10 or 0.14, within the range 0.11–0.15 the paper admits) will change both κ_in and κ_o; if κ_in moves outside 0.38–0.39, the universality claim loses its empirical anchor. Additional concerns in the normal-stress section (fitting a log law in a region where viscous stress is 5–20% of total stress) are secondary to the mean-velocity model, but they share the pattern of post hoc region selection.","agreement_with_reader":"agree"},"referee_report":{"model":"deepseek-v4-flash","summary":"This paper proposes a revision of the logarithmic-plus-linear (log+lin) extended-overlap model for the mean velocity profile in wall-bounded turbulent flows. The key modification is to separate the logarithmic coefficient into an inner pure-log coefficient κ_in, claimed to be universal (1/κ_in = 2.6, or κ_in ≈ 0.385), and an outer overlap coefficient κ_o, claimed to depend on the flow geometry and pressure gradient. The revision is motivated by indicator-function analysis of DNS and experimental profiles for channel, pipe, and zero-pressure-gradient boundary-layer flows. The paper also re-evaluates two competing models for the streamwise normal stress (a logarithmic model, Eq. (1), and a defect-power model, Eq. (2)) in a region closer to the wall than the traditional overlap region, reporting that both models fit the data equally well there. The conclusions emphasize the need for higher Reynolds numbers and, for the mean-velocity model, defer a matched-asymptotic derivation to future work.","tokens_in":10582,"tokens_out":8330,"duration_ms":59582,"significance":"The paper offers a clearly testable and falsifiable prediction: if a universal inner pure-log coefficient 1/κ_in = 2.6 exists, it can be checked against high-Reτ data and would change how the von Kármán 'constant' is interpreted in wall-bounded turbulence. A strength of the work is its use of multiple high-quality DNS and experimental datasets and its explicit acknowledgment of the limitations of current Reynolds numbers and of the need for matched asymptotic analysis. However, the central evidence is based on fits with several free parameters and on visual reading of indicator functions; the paper does not provide a derivation of Eq. (7), a sensitivity analysis in the region boundary Yin, or quantitative measures of fit quality for the normal-stress comparisons. These gaps currently prevent the claims from being fully load-bearing.","major_comments":[{"comment":"The revised log+lin model is proposed without a derivation, and the paper explicitly states that 'Additional matched asymptotic work is now under consideration, but is not covered here.' Because the central claims—the universality of κ_in and the non-universality of κ_o—rest entirely on this proposed functional form, the manuscript should either supply the matched-asymptotic justification or clearly frame the model as a conjecture and quantify how robustly the fitted parameters support it. As written, the reader is asked to accept a model on the basis of fits to a small number of high-Reτ profiles.","section":"Section III, Eq. (7)"},{"comment":"The pure-log region boundary Yin is chosen post hoc: Fig. 6 uses Yin = 0.12, and the text states that Yin 'may be in the range from 0.11 up to 0.15, and flow type or pressure gradient dependent.' Since κ_in is extracted from the segment below Yin and κ_o from the segment above Yin, a variation of Yin from 0.11 to 0.15 can shift both fitted coefficients. The paper reports no sensitivity analysis over this admitted range, nor does it list the individual data sets, Reynolds numbers, and scatter behind the proposed value 1/κ_in = 2.6. Without such evidence, the separation into a universal κ_in and a non-universal κ_o is not convincingly separated from the fitting procedure.","section":"Section III, Fig. 6 and discussion of Yin"},{"comment":"The claim that the logarithmic and defect-power models 'agree equally well' with the near-wall normal-stress data is not quantitatively supported. Each model is fitted with two free parameters (A1, B1 for Eq. (1); α1, β1 for Eq. (2)), but no residuals, goodness-of-fit statistics, or parameter uncertainties are reported; Figs. 12 and 13 show best-fit values without error bars. Since both functional forms are flexible two-parameter fits to the same data, the apparent agreement is not a strong test and should be quantified before drawing conclusions about the 'perplexing' behavior. In addition, the fitting region is defined by the 20% viscous-stress threshold, which is itself chosen without a robustness check.","section":"Section IV, Figs. 8–13"}],"minor_comments":[{"comment":"Equation (7) has a notation inconsistency: the second equality writes κ^{-1} without the subscript o, although the first equality uses κ_o^{-1}.","section":"Section III, Eq. (7)"},{"comment":"The text states 'Figure 7 shows the normal stress...' but the Figure 7 caption (and the surrounding discussion) describes the ratio of viscous to total stress; the normal-stress fits appear in Figs. 8 and 9. Please correct the figure references.","section":"Section IV, text near Fig. 7"},{"comment":"The abstract and Section III use both κ_in and k_in for the same quantity; please standardize the notation.","section":"Abstract and Section III"},{"comment":"Figure captions 9 and 10 have typos in the citations: 'Pirozzoli (2024 [12]' and 'Samie et al. (2018 [9]' are missing closing parentheses.","section":"Figure captions 9 and 10"},{"comment":"Section V says the fits were made 'outside of the region dominated by viscous effects; that is, the ratio of viscous over total stresses greater than 20%'; this is inconsistent with the abstract, which describes the fitting region as where viscous stresses are less than 20% of total. Please clarify which side of the threshold is meant.","section":"Section V, Conclusions"},{"comment":"The sentence 'This is consistent with the result 0.38 < κ_in < 0.39 found for ZPG boundary layers by Monkewitz and Nagib (2023), ...' attributes to those references a quantity (κ_in) that, to my knowledge, those papers did not define; please clarify whether those values were originally reported as the coefficient of the pure-log segment or as the overlap-region κ.","section":"Section III, last paragraph"}],"recommendation":"major_revision","confidential_remarks":"The manuscript is essentially a research letter built from re-analysis of existing datasets and from the author's own prior works. The central new quantity, κ_in ≈ 0.385, is asserted as universal on the basis of fits not fully documented in the paper. I would encourage the editor to ask for a supplementary table of all cases used (flow type, Reτ, y+ range, Yin, Yout, fitted κ_in, κ_o, So) and for a sensitivity analysis with respect to Yin. If that material is provided, the paper could be publishable; without it, the claim is not verifiable."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"Quick take: this is a plausible empirical revision of the overlap model, worth taking seriously, but the headline claim of a universal κ_in is not yet backed by enough independent evidence or a derivation.\n\nWhat is actually new: the split between an inner pure-log layer with its own coefficient κ_in and the extended overlap log+lin coefficient κ_o, Equation 7. This is a genuine extension of Monkewitz and Nagib (2023) and Baxerras et al. (2024). The observation that the pure-log region is short and does not coincide with a constant shear-stress layer in the DNS is a useful caution. The near-wall evaluation of the normal stress models (region just past the peak, viscous stress <20%) is also a new way to test the two competing trends, and it honestly reports that both models fit equally well – that is a fair thing to put on record.\n\nThe soft spots are real. The paper proposes Equation 7 without derivation and explicitly defers the matched asymptotic work. The value κ_in = 0.384 from one channel DNS case and the proposed universal 1/κ_in = 2.6 are based on fits whose region boundaries (Y_in, Y_out) are chosen per flow and per pressure gradient. The stress-test note is right: if Y_in can be 0.11 to 0.15, and the pure-log segment is visually short, the extracted κ_in is sensitive to that choice. No error bars, no list of all cases examined, no data or code. So the universality claim is an empirical conjecture, not a robust result. The normal stress fits share the same post hoc region selection pattern. These are addressable issues, not fatal ones, but they do limit what the paper can currently support.\n\nI take the paper at face value: it is candid about its own limitations, and the Reynolds number warning in the conclusions is sensible. It is a subfield paper, for people working on overlap layers, log laws, and normal stress scaling. It deserves serious refereeing: the questions are important, the data are high-quality DNS and experiments from well-known sources, and the author is clearly an expert in this debate. I would send it to review, with the expectation that the referee asks for a more transparent fitting procedure and a sensitivity analysis on Y_in. I would not cite the universal κ_in claim directly without checking the underlying fits, but I would cite the split model as a proposed revision.","headline":"A plausible empirical revision of the overlap model with a genuine new split between κ_in and κ_o, but the universal κ_in claim is not yet robustly supported and needs a transparent fitting procedure.","tokens_in":11082,"tokens_out":2056,"would_cite":true,"duration_ms":18145,"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":"Wall-bounded turbulent flows carry a universal pure-log inner layer with $1/\\kappa_{in}=2.6$, beyond which the log coefficient $\\kappa_o$ of the extended overlap becomes flow-dependent.","keywords":["wall-bounded turbulence","logarithmic plus linear law","von Kármán constant","inner layer","overlap region","streamwise normal stress","defect-power model","indicator function"],"falsifier":"Compute the indicator function $\\Xi = y^+ \\mathrm{d}U^+/\\mathrm{d}y^+$ from a well-resolved DNS of an adverse-pressure-gradient boundary layer at $Re_\\tau > 10{,}000$. If within $y^+$ beyond the viscous region up to $Y \\approx 0.12$ the indicator function fails to reach a plateau at $2.6$, or if that plateau value changes with pressure gradient, the claimed universality of the inner pure-log layer is refuted.","tokens_in":9928,"feed_emoji":"📐","tokens_out":8151,"duration_ms":65573,"temperature":0.7,"pith_summary":"This paper revises the account of the mean velocity profile in wall-bounded turbulent flows. It argues that the previously unified extended logarithmic overlap region actually consists of a thin pure-logarithmic inner layer, whose slope is universal at $1/\\kappa_{in}=2.6$, followed by a logarithmic-plus-linear overlap whose logarithmic coefficient $\\kappa_o$ depends on the flow geometry and pressure gradient. The paper also re-evaluates the two leading models for the streamwise normal stress and finds that an inviscid wall-scaled-eddies (logarithmic) model and a viscous bounded-dissipation (defect-power) model fit the data equally well just outside the viscous-dominated wall layer, with fitted parameters that differ from the accepted overlap-region values. The conclusions sharpen the Reynolds-number requirements for testing asymptotic wall-turbulence models and separate questions about the near-wall inner layer from questions about the outermost overlap and wake.","feed_headline":"2.6 is the universal slope of wall turbulence's inner log layer","feed_subtitle":"Beyond that thin layer, the log-plus-linear overlap slope depends on pressure gradient.","key_machinery":"The load-bearing object for the mean-velocity analysis is the indicator function $\\Xi = y^+ \\mathrm{d}U^+/\\mathrm{d}y^+ = Y\\,\\mathrm{d}U^+/\\mathrm{d}Y$, whose plateau in a pure-log layer equals $1/\\kappa$. The revised model writes $\\Xi_{OL} = 1/\\kappa_o + S_o(Y - Y_{in})$, inserting the inner-layer boundary $Y_{in}\\approx 0.11$ to $0.15$ and assigning the universal value $1/\\kappa_{in}=2.6$ to the pure-log segment. For the normal-stress comparison, the key device is the ratio of viscous to total stress, which fixes the fitting region just outside the inner peak ($y^+\\gtrsim 20$, viscous stress below 20% of total) and supplies the four fit parameters $A_1$, $B_1$, $\\alpha_1$, and $\\beta_1$ from DNS and experimental profiles.","core_discovery":"The central claim is that observed mean velocity profiles in channels, pipes, and zero-pressure-gradient boundary layers are best represented by a revised logarithmic-plus-linear law in which the single von Kármán coefficient is replaced by two: a universal inner-layer coefficient $\\kappa_{in}\\approx 1/2.6$ for a pure-log segment lying just above the viscous wall region at sufficiently high $Re_\\tau$ (at least 10,000), and a non-universal coefficient $\\kappa_o$ for the logarithmic term in the extended overlap region, whose value is set by the pressure gradient imposed by the flow geometry. A second claim is that the constant-shear-stress layer does not coincide with a constant indicator function, so the classical derivation of a logarithmic law from a constant-stress region is refuted by the DNS data examined here. For the streamwise normal stress, the paper claims that the inviscid logarithmic model and the viscous defect-power model both agree with experiments and DNS in a region close to the wall where viscous stress is a minority of the total stress, and that the four fitted parameters $A_1$, $B_1$, $\\alpha_1$, and $\\beta_1$ vary consistently across $200 < Re_\\tau < 100{,}000$.","pith_inferences":["The fixed value $1/\\kappa_{in}=2.6$ could be used as a calibration anchor when extracting overlap-region parameters from measurements, reducing the fitted degrees of freedom by one in boundary-layer studies.","The 20% viscous-stress threshold provides a Reynolds-number-independent recipe for choosing the lower fitting bound for normal-stress models, a rule that might transfer to other near-wall quantities such as passive scalar fluctuations.","If $\\kappa_o$ varies with pressure gradient, wall models in large-eddy simulation that assume a single constant log-law slope may need the two-layer split, especially in separated or strongly accelerated flows.","Applying the revised equation to adverse-pressure-gradient boundary layers at $Re_\\tau > 10{,}000$ would test whether $Y_{in}$ shifts systematically with pressure-gradient strength, a prediction the paper leaves open."],"forward_implications":["If the inner pure-log layer is universal at $1/\\kappa_{in}=2.6$, then the non-universality of the von Kármán coefficient in the literature is confined to the extended overlap coefficient $\\kappa_o$, not to the inner layer.","The absence of a constant indicator function over the constant-shear-stress layer rules out the constant-stress derivation of the log law as a valid basis for the overlap region.","Because both normal-stress models fit equally well in the near-wall outer edge of the inner layer, fits there cannot discriminate between the inviscid logarithmic and viscous defect-power descriptions, and their parameters must not be compared to overlap-region values.","Genuine tests of the assumed pure-log inner layer require friction Reynolds numbers of at least 10,000, so many existing channel, pipe, and boundary-layer datasets fall short of the asymptotic regime."],"supporting_citations":[{"why":"Introduces the logarithmic-plus-linear overlap model and its indicator-function analysis that the revised inner-layer model extends.","marker":"[2]"},{"why":"Provides evidence that even ZPG boundary layers lack a pure-log extended overlap, motivating the split between inner and extended layers.","marker":"[3]"},{"why":"Supplies the evaluation method and the overlap-region comparison of the logarithmic and defect-power normal-stress models.","marker":"[4]"},{"why":"Documents non-universality of the von Kármán coefficient, the basis for a flow-dependent $\\kappa_o$.","marker":"[5]"},{"why":"Fully resolved ZPG boundary-layer hot-wire data used for fitting both mean-velocity and normal-stress models.","marker":"[9]"},{"why":"DNS of pipe flow that supplies the pipe-profile fits and the near-wall dissipation-limit value 0.28.","marker":"[12]"},{"why":"NSTAP pipe-flow measurements at high Reynolds numbers used as the experimental pipe dataset.","marker":"[16]"},{"why":"DNS of channel flow up to $Re_\\tau \\approx 5200$ used for indicator functions, constant-stress-layer comparisons, and normal-stress fits.","marker":"[19]"}],"fun_headline_variants":["Universal inner log slope 1/2.6 for wall turbulence","Wall flow log law: inner slope fixed, outer varies with pressure","Reconsidered log law: inner slope universal, outer pressure-dependent","Inner layer log slope 1/2.6, outer slope set by pressure"],"cache_read_input_tokens":3200,"weakest_assumption_plain":"The load-bearing premise is that a short pure-logarithmic inner layer exists above the viscous wall region at friction Reynolds numbers of roughly 10,000 or more and has the same universal slope $1/\\kappa_{in}=2.6$ in every wall-bounded flow; the paper states this as an assumption rather than a derived result.","fun_headline_variants_meta":{"raw":{"variants":["Universal inner log slope 1/2.6 for wall turbulence","Wall flow log law: inner slope fixed, outer varies with pressure","Reconsidered log law: inner slope universal, outer pressure-dependent","Inner layer log slope 1/2.6, outer slope set by pressure"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.000772,"raw_usage":{"total_tokens":3509,"prompt_tokens":1129,"completion_tokens":2380,"prompt_tokens_details":{"cached_tokens":384},"prompt_cache_hit_tokens":384,"prompt_cache_miss_tokens":745,"completion_tokens_details":{"reasoning_tokens":2303}},"tokens_in":745,"tokens_out":2380,"duration_ms":16825,"temperature":1.0,"reasoning_tokens":2303,"cache_read_input_tokens":384,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-07T14:26:15.637627+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"Compute the indicator function $\\Xi = y^+ \\mathrm{d}U^+/\\mathrm{d}y^+$ from a well-resolved DNS of an adverse-pressure-gradient boundary layer at $Re_\\tau > 10{,}000$. If within $y^+$ beyond the viscous region up to $Y \\approx 0.12$ the indicator function fails to reach a plateau at $2.6$, or if that plateau value changes with pressure gradient, the claimed universality of the inner pure-log layer is refuted.","supporting_citations":[{"cited_title":"Monkewitz and H","cited_arxiv_id":null,"evidence_quote":"Introduces the logarithmic-plus-linear overlap model and its indicator-function analysis that the revised inner-layer model extends."},{"cited_title":"Baxerras, R","cited_arxiv_id":null,"evidence_quote":"Provides evidence that even ZPG boundary layers lack a pure-log extended overlap, motivating the split between inner and extended layers."},{"cited_title":"A method for evaluating relations of turbulent normal-stresses by experimental data over a wide range of Reynolds numbers","cited_arxiv_id":"2410.23669","evidence_quote":"Supplies the evaluation method and the overlap-region comparison of the logarithmic and defect-power normal-stress models."},{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"Documents non-universality of the von Kármán coefficient, the basis for a flow-dependent $\\kappa_o$."},{"cited_title":"Samie, I","cited_arxiv_id":null,"evidence_quote":"Fully resolved ZPG boundary-layer hot-wire data used for fitting both mean-velocity and normal-stress models."},{"cited_title":"Pirozzoli, On the streamwise velocity variance in the near-wall region of turbulent flows, Journal of Fluid Mechanics 989, A5 (2024)","cited_arxiv_id":null,"evidence_quote":"DNS of pipe flow that supplies the pipe-profile fits and the near-wall dissipation-limit value 0.28."},{"cited_title":"Hultmark, M","cited_arxiv_id":null,"evidence_quote":"NSTAP pipe-flow measurements at high Reynolds numbers used as the experimental pipe dataset."},{"cited_title":"Lee and R","cited_arxiv_id":null,"evidence_quote":"DNS of channel flow up to $Re_\\tau \\approx 5200$ used for indicator functions, constant-stress-layer comparisons, and normal-stress fits."}],"review_version":1}