{"id":"0831d945-7c61-4223-831b-dafb89aa4280","arxiv_id":"2505.16962","paper_version":1,"verdict":"CONDITIONAL","confidence":"MODERATE","novelty_score":6.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":3,"one_line_summary":"In strongly decelerating turbulent boundary layers, riblets achieve far larger drag reductions than on flat plates, and in some cases produce a net upstream force, due to intensified KH rollers in the grooves.","lead":"Direct numerical simulations show that riblets in decelerating turbulent boundary layers reduce drag far more than on flat plates, sometimes producing a net upstream force. The result suggests riblet films on airfoils and other adverse-pressure-gradient surfaces may be more effective than current design curves predict.","discovery_kind":"extension","skeptic_critique":{"model":"deepseek-v4-flash","headline":"Drag-reduction percentages in Eq. (5) may be an artifact of comparing at fixed x without quantitative Reynolds-number matching; the forward-force cases are exactly those excluded from the paper's 'most cases' Re match claim.","rationale":"The reader's weakest assumption matches the most load-bearing risk I find. The central claim is a quantitative statement about drag reduction, so the metric that produces those numbers must be sound. The paper itself flags the matching condition with 'in most cases' and gives no quantitative support; moreover, the extreme cases that yield the headline result are precisely the ones that may fall outside that caveat. This is not a generic convergence or error-bar concern but a direct threat to the definition of the measured quantity. The proposed test—computing Re_theta and H and then re-evaluating delta_tau_w at matched Re_theta—would settle whether the forward-force claim is an artifact. If the test fails, the central claim would need to be substantially revised; if it passes, the claim is supported. The paper's honest 'preliminary results' framing and its own statement that 'further analysis will be documented in the final paper' confirm that the current draft is not yet a fully secure quantitative result, which supports the conditional rather than immediate acceptance position.","tokens_in":14225,"tokens_out":12075,"duration_ms":113027,"concrete_test":"From the saved DNS fields, compute the momentum thickness theta(x) and shape factor H(x) for all eight cases (two smooth, six riblet). Plot the ratios theta_r/theta_s and H_r/H_s as functions of x, focusing on the stations where delta_tau_w is most negative (especially where delta_tau_w < -100%). If both ratios remain within 1–2% across the entire forward-force region, the fixed-x comparison is adequate and the central claim survives. If the ratios deviate significantly, recompute delta_tau_w by interpolating the smooth-wall tau_s at the location where the riblet case attains the same Re_theta (using the riblet-case theta and the common edge velocity), and check whether the forward-force regime and the 45–250% range are materially altered.","verdict_should_be":"UNCHANGED","load_bearing_attack":"The paper's evaluation of drag reduction hinges on Eq. (5), which compares riblet and smooth-wall shear stress at the same streamwise location. Section III.B justifies this by stating that Re_delta (delta*U_e/nu) remains nearly the same between the two walls 'in most cases,' but no quantitative evidence is provided. The relevant state parameter for boundary-layer comparison, Re_theta, is never reported. The caveat 'most cases' is not empty: Section III.A shows that for cases ⌫05;40, ⌫10;20, and ⌫10;40 the Clauser parameter tends to infinity, and the statement about negligible Re_delta variation is explicitly limited to 'all moderate-APG cases and the high-APG cases with small or medium riblets.' Case ⌫10;40, which produces the strongest forward force, is thereby excluded from the matched-Reynolds-number assertion. If the adverse pressure gradient drives the riblet and smooth boundary layers along different histories, a substantial part of the claimed 45–250% drag reduction, and in particular the negative-drag regime (delta_tau_w < -100%), could reflect unequal flow states rather than the riblets' intrinsic drag-modifying effect. The paper's self-described preliminary status does not remove the need for a quantitative matched-state check.","agreement_with_reader":"agree"},"referee_report":{"model":"deepseek-v4-flash","summary":"This manuscript reports direct numerical simulations of spatially developing turbulent boundary layers over sinusoidal riblets subjected to two adverse pressure gradients (maximum Clauser parameter roughly 5 and 10), with three riblet sizes selected to span the drag-reducing, Kelvin-Helmholtz-roller, and drag-increasing regimes at a ZPG reference plane. The central claims are that riblets reduce drag substantially more under APG than in ZPG flows, with reported reductions of roughly 45-250 percent; that for the largest riblets under the stronger APG the wall shear stress reverses, producing a forward force; and that the mechanism is intensified Kelvin-Helmholtz rollers that generate a mean reverse flow within the grooves. The paper is explicitly a preliminary draft for a conference paper, with detailed statistical analysis deferred to a future 'full paper.'","tokens_in":14484,"tokens_out":6263,"duration_ms":53531,"significance":"If the result survives a matched-state comparison and a convergence check, it would be significant: it would show that ZPG-based riblet scaling laws (the l+ drag curve and the KH-roller thresholds) fail substantially in strong APGs, and it would identify a new forward-force regime with practical implications for drag-reduction applications. The paper's strengths are its clean parametric matrix, direct computation of wall stress by integration of the IBM force, stated grid resolution in wall units, and careful placement of the results against the established ZPG drag curve. The interpretation is, however, presently supported mainly by visualizations rather than by quantitative state matching or uncertainty quantification.","major_comments":[{"comment":"The drag-reduction metric compares wall shear stress at the same streamwise coordinate, justified by the statement that Re_delta is nearly equal between smooth and riblet cases 'in most cases.' No quantitative evidence is provided: Re_delta is not plotted or tabulated for the two walls, and Re_theta is not reported at all. Moreover, Section III.A itself restricts the claim: negligible increases in Re_delta are stated only for 'all moderate-APG cases and the high-APG cases with small or medium riblets,' which excludes exactly cases ⌫05;40, ⌫10;20, and ⌫10;40 - the cases that produce negative wall shear and enter the forward-force regime. If the APG drives the riblet and smooth boundary layers along different histories, a substantial part of the reported 45-250 percent reduction, and especially the δτ_w < -100 percent range, could reflect comparison of unlike flow states rather than the intrinsic drag modification by riblets. Please provide quantitative matching of boundary-layer state parameters (e.g., Re_theta(x), Re_delta(x), and shape factor H(x) for both walls) and either restrict the claims to matched cases or adjust the metric accordingly.","section":"III.B, Eq. (5) and III.A"},{"comment":"The paper states that statistics are collected after the flow reaches a statistically steady state, but it gives no averaging time, sample count, or convergence measure. For quantitative claims that include a mean negative wall shear stress whose magnitude is small relative to the smooth-wall baseline, statistical convergence is load-bearing: without it, the reported 45-250 percent range and the forward-force regime are not fully verified. Please add running means or confidence intervals for τ_w (or δτ_w) at representative streamwise stations, and report the averaging period in outer time units.","section":"II.C, III.B"},{"comment":"The causal statement that the enhanced drag reduction is 'a product of' Kelvin-Helmholtz rollers is supported only by instantaneous flow visualizations (Figs. 7 and 8). There is no spectral analysis, no roller convection velocity or passage frequency, and no conditional or phase-averaged link between roller passage and instantaneous wall shear. The manuscript itself defers such analysis to 'the full paper.' As written, the mechanism claim is plausible but not demonstrated; either provide quantitative evidence for the roller interpretation and its connection to the mean reverse groove flow, or soften the causal language to an explicitly stated hypothesis.","section":"III.D, III.E"}],"minor_comments":[{"comment":"Typographical errors should be corrected: Ref. [22] has 'rilets' instead of 'riblets', Ref. [24] has 'turbuelnt' instead of 'turbulent', and Section III.B contains 'fows' instead of 'flows'.","section":"References"},{"comment":"The contour plot has no colorbar, and the caption mentions that blue and red regions are saturated at different magnitudes; please add a colorbar or otherwise quantify the contour levels.","section":"Figure 3"},{"comment":"The notation for the Reynolds number, written as 'Re_delta = X*4/a', is difficult to parse; please define Δ* and U_e explicitly in one place and consistently distinguish Re_delta from Re_theta.","section":"II.C, III.A"},{"comment":"For δτ_w < -100 percent the surface produces thrust, not a reduction of drag; consider using wording such as 'drag reduction and thrust production' consistently rather than referring to the entire range as drag reduction.","section":"Abstract and Conclusion"},{"comment":"The fixed height-to-spacing ratio h/s = 3/π (if that is the intended value) is not motivated; please specify the rationale for this choice and how it relates to the sinusoid shape.","section":"II.B"}],"recommendation":"major_revision","confidential_remarks":"The manuscript reads as a preliminary conference paper: it repeatedly defers detailed statistics and mechanism analysis to a 'full paper.' The central claim is plausible and worth pursuing, but the fixed-x comparison without quantitative Reynolds-number matching and the absence of uncertainty quantification are load-bearing issues that must be resolved before this is acceptable as a journal paper. I would encourage the editor to request a revision that either supplies the missing quantitative support or explicitly narrows the claims to the cases for which matched-state comparison is demonstrated."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"Bottom line: this is a genuinely surprising result from a credible DNS setup. Strong adverse pressure gradients appear to turn riblets into far better drag reducers than ZPG predicts — 45–250% reduction — and for the largest riblets under the strongest APG the mean wall shear reverses, giving thrust. That is new: prior APG studies stopped at ~15% reduction. The smooth-wall baseline and the direct computation of wall stress from the IBM force are the right approach, the grid resolution is stated (Δy+ ≤ 0.6 in all cases), and the authors are honest that this is a preliminary conference-draft report.\n\nThe soft spots, in order. First, the drag-reduction metric compares riblet and smooth wall at fixed x, justified by a claim that Re_δ stays nearly the same 'in most cases'. No quantitative comparison is given, and the thrust-generating case (V10;40) is exactly the one excluded from that statement — its Clauser parameter goes to infinity. Until they show matched boundary-layer state evolution, part of the 45–250% could be a flow-state mismatch rather than riblet physics. Second, no error bars or convergence measures are reported; a 250% mean from a finite sampling window needs a confidence interval. Third, the KH-roller mechanism is inferred from instantaneous contours and visual coherence, not from quantitative roller diagnostics. The authors explicitly defer that analysis to the 'full paper,' which is acceptable for a conference draft but means the central mechanism claim is unverified.\n\nThe citation pattern looks right, and the claimed novelty relative to the weak-APG riblet literature is credible. The circularity burden is low because drag reduction is computed directly from DNS wall stress against a smooth-wall baseline.\n\nWho this is for: anyone working on riblets in pressure-gradient flows, and RANS modelers who carry ZPG drag curves into airfoil and blade applications. It deserves a serious referee. Send it to peer review; the revision must include the matched-state check, uncertainty quantification, and quantitative KH diagnostics. For the AIAA conference, I'd accept conditionally at most.","headline":"Strong APGs can make riblets dramatically more effective — even producing mean thrust — but the 45–250% numbers rest on an unverified fixed-x comparison and mechanism evidence that is still qualitative.","tokens_in":15011,"tokens_out":2658,"would_cite":false,"duration_ms":21313,"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":"Riblets reduce drag by 45-250 percent in attached decelerating turbulent boundary layers.","keywords":["turbulent boundary layer","adverse pressure gradient","riblets","drag reduction","Kelvin-Helmholtz rollers","direct numerical simulation","skin friction","forward force"],"falsifier":"Compute a matched-history variant in which the smooth and riblet boundary layers are forced to share the same streamwise displacement-thickness or momentum-thickness development, then measure the wall-shear difference; if the 45-250 percent reduction and negative wall shear do not survive the matching, the claim that riblets alone produce the forward force would be refuted.","tokens_in":13989,"feed_emoji":"🌊","tokens_out":7054,"duration_ms":59734,"temperature":0.7,"pith_summary":"The paper seeks to establish that streamwise riblets reduce skin-friction drag far more in attached, decelerating turbulent boundary layers than in zero-pressure-gradient flows. In direct numerical simulations of two adverse-pressure-gradient strengths and three riblet sizes, drag reduction reaches 45 to 250 percent, and for the largest riblets under the stronger gradient the wall shear stress reverses, producing a net upstream force. The authors attribute this to Kelvin-Helmholtz roller vortices near the riblet crests, which grow in size, strength, and frequency during deceleration and create a mean reverse flow inside the grooves. The results imply that standard viscous-scaled riblet metrics developed for zero-pressure-gradient flows will not predict drag modification when the pressure gradient is non-negligible.","feed_headline":"Riblets cut drag up to 250 percent in decelerating flow","feed_subtitle":"Deceleration strengthens Kelvin-Helmholtz rollers, letting riblets beat their zero-pressure-gradient drag performance.","key_machinery":"The argument is carried by direct numerical simulations of a spatially developing boundary layer in which a hyperbolic-tangent freestream deceleration imposes a strong, growing adverse pressure gradient, with sinusoidal riblets enforced by an immersed boundary method. The central quantity is the drag curve, a plot of percentage wall-shear change against the viscous-scaled groove size $\\ell_g^+ = \\ell_g u_\\tau/\\nu$, where $\\ell_g$ is the square root of the groove cross-sectional area; in zero-pressure-gradient flows this curve collapses riblet performance across geometries. The physical mechanism proposed is the Kelvin-Helmholtz roller: spanwise-coherent vortices that roll up from the shear layer at the riblet crest, whose lower halves induce local reverse flow. The paper argues that the adverse pressure gradient strengthens this shear layer, so the rollers grow and persist even as $\\ell_g^+$ falls below the zero-pressure-gradient threshold, and their time-averaged passage produces a sustained mean reverse flow in the grooves.","core_discovery":"The central discovery is that in attached, decelerating turbulent boundary layers, riblets reduce drag far beyond the zero-pressure-gradient benchmark: the drag reduction ranges from 45 to 250 percent, and for the largest riblets under the stronger adverse pressure gradient the wall shear stress reverses sign, so the riblet surface produces a forward force. The paper argues that this is caused by Kelvin-Helmholtz roller vortices forming near the riblet crest. The adverse pressure gradient augments the rollers' size, strength, and frequency, and because the lower halves of the rollers move upstream, their time-averaged passage creates a mean reverse flow inside the grooves. Even though the flow at the riblet crest remains attached, this reverse flow cancels the positive shear from the crest, which is why drag reduction can exceed 100 percent.","pith_inferences":["If the adverse-pressure-gradient-sustained shear layer is the controlling mechanism, a more robust design rule for non-equilibrium flows would use local pressure-gradient or shear-layer parameters rather than $\\ell_g^+$ alone; varying the freestream deceleration shape while holding riblet geometry fixed would test this directly.","The forward-force regime suggests riblets might be placed selectively on the decelerating portions of wings or nacelles, but real geometries add sweep and spanwise pressure gradients, so the mechanism would need to survive three-dimensionality.","The mean reverse flow inside the grooves means a riblet-covered wall in an adverse pressure gradient behaves somewhat like a partially separated surface, which could affect noise and heat transfer, not just drag, in downstream applications."],"forward_implications":["Existing zero-pressure-gradient drag-prediction metrics, based only on viscous-scaled riblet size, systematically underpredict drag reduction once an adverse pressure gradient is strong enough; a pressure-gradient-dependent correction is needed.","A riblet surface can produce a net upstream force while the outer boundary layer remains attached, opening a passive-thrust regime for decelerating flows over airfoils, diffusers, and other expanding geometries.","Kelvin-Helmholtz rollers, usually a sign of riblet drag penalty in zero-pressure-gradient flows, become drag reducers in strong adverse pressure gradients once they are intense enough to sustain a mean reverse flow inside the grooves.","Drag reduction in adverse-pressure-gradient riblet flows grows with riblet size and with pressure-gradient strength, opposite to what the zero-pressure-gradient drag curve predicts as $\\ell_g^+$ decreases."],"supporting_citations":[{"why":"Establishes the $\\ell_g^+$-based drag curve for zero-pressure-gradient riblet flows that the adverse-pressure-gradient results are compared against.","marker":"[13]"},{"why":"Defines the viscous regime and the onset of Kelvin-Helmholtz rollers in zero-pressure-gradient riblet flows, the thresholds the paper tests under adverse pressure gradients.","marker":"[14]"},{"why":"Provides the double-averaging decomposition used to separate mean, dispersive, and turbulent fields over riblets.","marker":"[15]"},{"why":"Early experiment showing riblet drag reduction can improve under longitudinal pressure gradients, motivating the comparison with zero-pressure-gradient predictions.","marker":"[21]"},{"why":"Weak-adverse-pressure-gradient experiment reporting modestly increased riblet drag reduction, the prior baseline that the present 45-250 percent range exceeds.","marker":"[22]"},{"why":"Numerical study of riblets in mild adverse pressure gradients whose roughly 15 percent reductions are the direct computational predecessor extended here.","marker":"[25]"},{"why":"Direct numerical simulations of riblet shapes in minimal-span channels supplying the Kelvin-Helmholtz roller visual signatures and regime behavior used to identify the rollers.","marker":"[33]"},{"why":"Shows how the immersed-boundary force integrated over the wall yields the total wall stress, the quantity behind the drag-reduction numbers.","marker":"[37]"},{"why":"Establishes the empirical $\\ell_g^+$ range for Kelvin-Helmholtz roller occurrence in zero-pressure-gradient flows; the paper shows this threshold fails under strong adverse pressure gradients.","marker":"[48]"}],"fun_headline_variants":["Riblets achieve negative drag under adverse pressure gradients","Decelerating flow helps riblets cut drag by up to 250%","Riblets produce forward force in decelerating boundary layers","KH rollers boost riblet drag reduction beyond 100 percent","Riblets turn drag into thrust in decelerating flows"],"cache_read_input_tokens":3200,"weakest_assumption_plain":"The load-bearing premise is that the smooth-wall and riblet cases can be compared at the same streamwise station because their momentum-thickness Reynolds numbers are nearly equal; if the deceleration alters the two boundary layers' histories differently, part of the measured drag gap could reflect comparing different flow states rather than the riblets' effect.","fun_headline_variants_meta":{"raw":{"variants":["Riblets achieve negative drag under adverse pressure gradients","Decelerating flow helps riblets cut drag by up to 250%","Riblets produce forward force in decelerating boundary layers","KH rollers boost riblet drag reduction beyond 100 percent","Riblets turn drag into thrust in decelerating flows"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.000963,"raw_usage":{"total_tokens":4068,"prompt_tokens":881,"completion_tokens":3187,"prompt_tokens_details":{"cached_tokens":384},"prompt_cache_hit_tokens":384,"prompt_cache_miss_tokens":497,"completion_tokens_details":{"reasoning_tokens":3101}},"tokens_in":497,"tokens_out":3187,"duration_ms":18905,"temperature":1.0,"reasoning_tokens":3101,"cache_read_input_tokens":384,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-07T14:52:14.180732+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"Compute a matched-history variant in which the smooth and riblet boundary layers are forced to share the same streamwise displacement-thickness or momentum-thickness development, then measure the wall-shear difference; if the 45-250 percent reduction and negative wall shear do not survive the matching, the claim that riblets alone produce the forward force would be refuted.","supporting_citations":[{"cited_title":"Drag reduction by riblets,","cited_arxiv_id":null,"evidence_quote":"Establishes the $\\ell_g^+$-based drag curve for zero-pressure-gradient riblet flows that the adverse-pressure-gradient results are compared against."},{"cited_title":"Hydrodynamic stability and breakdown of the viscous regime over riblets,","cited_arxiv_id":null,"evidence_quote":"Defines the viscous regime and the onset of Kelvin-Helmholtz rollers in zero-pressure-gradient riblet flows, the thresholds the paper tests under adverse pressure gradients."},{"cited_title":"Dispersive stresses in turbulent flow over riblets,","cited_arxiv_id":null,"evidence_quote":"Provides the double-averaging decomposition used to separate mean, dispersive, and turbulent fields over riblets."},{"cited_title":"Effects of longitudinal pressure gradients on turbulent drag reduction with riblets,","cited_arxiv_id":null,"evidence_quote":"Early experiment showing riblet drag reduction can improve under longitudinal pressure gradients, motivating the comparison with zero-pressure-gradient predictions."},{"cited_title":"The reduction of skin friction by rilets under the influence of an adverse pressure gradient,","cited_arxiv_id":null,"evidence_quote":"Weak-adverse-pressure-gradient experiment reporting modestly increased riblet drag reduction, the prior baseline that the present 45-250 percent range exceeds."},{"cited_title":"Riblet drag reduction in mild adverse pressure gradients: A numerical investigation,","cited_arxiv_id":null,"evidence_quote":"Numerical study of riblets in mild adverse pressure gradients whose roughly 15 percent reductions are the direct computational predecessor extended here."},{"cited_title":"Direct numerical simulations of turbulent flow over various riblet shapes in minimal-span channels,","cited_arxiv_id":null,"evidence_quote":"Direct numerical simulations of riblet shapes in minimal-span channels supplying the Kelvin-Helmholtz roller visual signatures and regime behavior used to identify the rollers."},{"cited_title":"Numerical simulations of sink-flow boundary layers over rough surfaces,","cited_arxiv_id":null,"evidence_quote":"Shows how the immersed-boundary force integrated over the wall yields the total wall stress, the quantity behind the drag-reduction numbers."},{"cited_title":"Influence of riblet shapes on the occurrence of Kelvin-Helmholtz rollers,","cited_arxiv_id":null,"evidence_quote":"Establishes the empirical $\\ell_g^+$ range for Kelvin-Helmholtz roller occurrence in zero-pressure-gradient flows; the paper shows this threshold fails under strong adverse pressure gradients."}],"review_version":1}