{"id":"20fa0522-a082-4322-9c4d-8b1c7d6bc8c1","arxiv_id":"2411.14616","paper_version":1,"verdict":"CONDITIONAL","confidence":"MODERATE","novelty_score":7.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":2,"one_line_summary":"In DNS, drag reduction occurs when streak-scale wall transpiration is in-phase with wall pressure; out-of-phase roller-scale transpiration generates spanwise rollers and drag increase.","lead":"This paper simulates a turbulent channel flow with wall suction and blowing to show that the phase between wall pressure and wall transpiration decides whether drag rises or falls. It proposes that streak and roller scale transpiration with specific pressure phases are the building blocks behind riblets, porous, and permeable surfaces.","discovery_kind":"extension","skeptic_critique":{"model":"deepseek-v4-flash","headline":"The p–v phase is diagnosed after opposition-control runs, not manipulated independently, so the central claim that Δθ controls the response remains correlational.","rationale":"The reader's weakest assumption was robustness of the phase dichotomy under slow-pressure variability; I agree that is a real caveat, but the most load-bearing issue is one level up: the phase relation is never independently imposed. In opposition control, ∠Â_d is the design variable; p is a response variable. The paper's strongest claim says Δθ is the controlling quantity, but the evidence is a correlation between drag response and a post-hoc phase measurement. This is not an accusation of error; it is a missing control experiment. The proposed pressure-feedback test is the natural experiment: it directly imposes the phase relation that §7.2 claims is transferable to porous/compliant surfaces. If that test works, the framework is strongly supported; if not, the paper should be read as a correlational study. Since the paper is otherwise careful and the evidence is promising, the appropriate verdict remains conditional rather than accept/reject. Hence UNCHANGED.","tokens_in":33562,"tokens_out":6678,"duration_ms":71736,"concrete_test":"Run DNS at the same Re_b=5600 with a pressure-feedback transpiration law v̂_κ = β e^{iφ} p̂_κ at the streak (κ_s) and roller (κ_r) wavenumber sets, with φ=0 (in-phase) and φ=π (out-of-phase), tuning β so the rms transpiration matches N25/P50 levels. If ξ<1 for φ=0 and ξ>1 for φ=π, with roller scales appearing for φ=π, the Δθ relation is causal; if the drag response does not follow the imposed phase, the central claim reduces to a correlation. This directly tests the porous-wall model of eq. (7.1) and removes the confounding ∠Â_d.","verdict_should_be":"UNCHANGED","load_bearing_attack":"The paper's central claim is that the phase difference Δθ_κ between wall pressure and transpiration is the controlling quantity for the three drag responses (abstract, §7.1). What is actually manipulated in every simulation is the controller phase ∠Â_d: eq. (2.5) sets ∠v_κ(y_w)=∠Â_d+∠v_κ(y_d)+π, and fig. 3 maps ξ against ∠Â_d. The p–v phase Δθ_κ (eq. 5.1) is computed post hoc from the resulting flow fields. The scale-restricted controllers in §3.3 and Appendix A vary ∠Â_d for selected wavenumbers, but they do not prescribe Δθ_κ; the phase is measured only after the fact. Thus the data cannot distinguish 'Δθ controls the response' from '∠Â_d controls the response, and Δθ is merely a covarying marker.' This distinction matters for §7.2: the porous-wall equivalence v=−βp' assumes that imposing an out-of-phase relation is what generates rollers, and that changing the phase would change the drag. The paper explicitly leaves this untested ('It would be interesting to verify...'). The admitted slow-pressure stochasticity for N25 (§5.2) is a secondary concern about robustness; the causal gap is more fundamental because even the two clean cases are consistent with a non-causal interpretation.","agreement_with_reader":"partial"},"referee_report":{"model":"deepseek-v4-flash","summary":"This paper performs DNS of a low-Reynolds-number turbulent channel flow (Re_b = 5600, Re_tau,0 = 180) with varying-phase opposition control, in which the complex controller gain A_d is chosen so that its phase angle A_d prescribes a streamwise shift of the wall transpiration. The authors analyze the wall-pressure field and its phase difference Delta_theta_kappa = angle(p_hat_kappa(y_w)) - angle(v_hat_kappa(y_w)) (Eq. 5.1) using circular statistics. They decompose the pressure into fast, slow, and Stokes components and show that transpiration makes the Stokes pressure leading-order. They identify two scale families, streak scales (exemplified by kappa_s = (0.5,11)) and roller scales (exemplified by kappa_r = (6.5,0)). For controller N25 (angle A_d = -pi/4, 21% drag reduction), streak-scale transpiration is in-phase with the wall pressure and suppresses the near-wall cycle; for P50 (angle A_d = +pi/2, drag increase), roller-scale transpiration is out-of-phase and energizes spanwise rollers; for N75 (-3pi/4), the phase data have large circular variance. The paper proposes conditions for robust pressure-transpiration phase relations based on the Green's function domain of dependence and temporal frequency sparsity, and draws analogies to porous, permeable, and riblet surfaces.","tokens_in":33773,"tokens_out":6703,"duration_ms":65871,"significance":"If the claimed phase dichotomy holds, the wall-pressure/transpiration phase difference would be a useful wall-based parameter unifying active opposition control and passive tailored surfaces, and it is experimentally testable. The study is careful in several respects: the DNS and pressure Poisson solvers are validated against Lee and Moser (2015); circular statistics are used appropriately for phase data; scale-restricted controllers in Section 3.3 and Appendix A support the attribution of drag changes to specific scale families; and the paper states its own limitations, such as the large circular variance in the N75 case and the stochasticity of the slow pressure. It also makes falsifiable predictions for riblets and porous walls. The principal caveat is that Delta_theta is diagnosed post hoc and never independently imposed, so the central claim that Delta_theta controls the response is correlational; moreover, the drag-reducing N25 case is not fully explained by the proposed mechanism.","major_comments":[{"comment":"The paper states that the phase difference Delta_theta_kappa 'controls' or 'parametrizes' the flow response, but in every simulation only the controller phase angle A_d is prescribed; Delta_theta_kappa is measured after the simulation from the resulting pressure and transpiration fields. Figure 3 maps the drag ratio against angle A_d, not against Delta_theta. The data are therefore equally consistent with angle A_d being the causal parameter and Delta_theta being a covarying marker. This distinction matters because Section 7.2's porous-wall equivalence assumes that imposing an out-of-phase relation is what generates rollers. I request either rewriting the central claims in explicitly correlational or falsifiable language, or adding a test in which Delta_theta is imposed independently of angle A_d (for example, a pressure-driven boundary condition with an adjustable complex impedance) and the response is shown to track Delta_theta.","section":"Abstract and §7.1, with Eqs. (2.3)–(2.5) and (5.1)"},{"comment":"The paper explicitly states: 'Given the significant variation in the direction of the slow pressure, it remains somewhat unclear how controller N25 establishes a robust overall phase relationship.' This matters because N25 is the only drag-reducing case and is the linchpin of the in-phase branch. At kappa_s the slow and Stokes pressures dominate (Fig. 7e), yet the proposed mechanisms for a robust phase relation do not apply there: the fast pressure is the smallest component at that scale, and Fig. 12a shows broadband temporal content rather than the sparsity invoked for the Stokes pressure. The robust total-pressure phase for N25 is therefore not explained by the paper's framework. Please provide a quantitative explanation, for example a time-resolved vector addition of the pressure components or a conditional analysis, or explicitly restrict the claimed mechanism to the cases it actually covers.","section":"§5.2 and Figs. 7 and 10"},{"comment":"The asserted dichotomy that streak scales attenuate the near-wall cycle when in-phase and amplify it 'otherwise' rests on one clean amplifying streak-scale case, N75, whose mean phase difference is approximately +pi/2 rather than pi and whose circular variance is acknowledged to be large. The phase difference spectra in Fig. 13(a) also vary substantially from scale to scale. Without additional phase shifts (for example angle A_d = -pi/2, -pi/8, +pi/4) analyzed with the same Delta_xi_t,kappa and Delta_theta_kappa metrics, the smooth transition from suppression to amplification remains speculative. I recommend either adding such cases or softening the abstract and Section 7.1 to state the observed correlation for the three cases rather than a general phase-controlled dichotomy.","section":"§6.3, Figs. 10 and 15"}],"minor_comments":[{"comment":"The histograms are informative, but the paper would benefit from reporting quantitative circular variance or bootstrap confidence intervals for the circular means. This would sharpen the claim that N25 and P50 are statistically strong while N75 is ambiguous.","section":"§5.1 and Fig. 10"},{"comment":"The explanation that the Green's function contributes an additional phase of pi because it is negative is correct, but the sign conventions are easy to confuse with the +pi appearing in the control law, Eq. (2.5). A brief explicit statement distinguishing the two sources of pi would improve readability.","section":"§5.3, Eq. (5.6)"},{"comment":"The phrase 'parametrized by the wall pressure' could be misread as implying the wall pressure alone is the parameter; the actual parameter is the phase difference between wall pressure and transpiration defined in Eq. (5.1). Please clarify.","section":"Abstract and §7.1"},{"comment":"The text contains a typo: 'conduced' should be 'conducted'.","section":"§7.1"},{"comment":"There are several encoding artifacts and minor typos, for example 'G ´omez-de Segura' in the body text and 'futher' near Eq. (2.12). A careful proofreading pass is recommended.","section":"Throughout"}],"recommendation":"major_revision","confidential_remarks":"This is a solid DNS study with careful analysis and appropriate caveats. The main risk is over-interpretation of correlational phase data as causal. The authors already acknowledge some of the weaknesses (N75 variance, N25 slow-pressure stochasticity), which makes the issues fixable through rephrasing and targeted additional analysis. I support considering the revised version for JFM."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"You should read this one. It is the first full nonlinear DNS evidence that the phase between wall pressure and transpiration separates the three canonical drag responses: in-phase streak scales suppress the near-wall cycle, out-of-phase roller scales energize spanwise rollers and increase drag. The two clean cases (N25 and P50) make the point convincingly, and the analysis is careful throughout—validated DNS and pressure Poisson solvers, proper circular statistics, and a genuinely useful decomposition into fast, slow, and Stokes pressure. The Green's function argument for why fast pressure locks to the transpiration phase, and why slow pressure usually does not, is the most instructive part of the paper. The Stokes pressure result—that transpiration makes it leading-order even at high Reynolds number—deserves attention on its own.\n\nThe soft spots are real but mostly minor. The biggest one, which the stress-test note correctly identifies, is that the controller manipulates the phase shift ∠A_d relative to the sensor signal; the p–v phase Δθ is measured after the fact. So the data cannot strictly distinguish \"Δθ controls the response\" from \"∠A_d controls the response and Δθ is a covarying marker.\" The paper partially closes this gap with the scale-restricted controllers and the mechanistic domain-of-dependence analysis, and it is honest about the remaining gap (\"It would be interesting to verify...\"). But the abstract and summary do claim more than the experimental design supports. That is worth flagging in review, not rejecting over.\n\nOther soft spots: one Reynolds number, three in-depth phase shifts, and the N75 case has large circular variance and is appropriately caveated. The admitted difficulty in explaining how N25 establishes a robust total-pressure phase when the slow-pressure direction wanders (§5.2) is a secondary concern, not a fatal one. No public code or data, which makes reproduction harder but is common for DNS papers. The self-citation pattern is legitimate—the prior resolvent and DNS results are independently obtained and the current paper stands on its own.\n\nWho is this for? Anyone working on opposition control, porous/permeable surfaces, riblets, or wall-based drag reduction. It deserves a proper peer review at JFM or a comparable venue; the causal gap and the small parameter coverage should be addressed, but the core observation is solid and new.","headline":"First solid nonlinear-DNS evidence that wall pressure–transpiration phase tracks drag-reducing vs. drag-increasing responses, though the phase is diagnosed rather than imposed, so the causal claim stays correlational.","tokens_in":34327,"tokens_out":1583,"would_cite":true,"duration_ms":18892,"reading_group":"yes","serious_thinker":"yes","would_accept_peer_review":true},"rs_alignment":null,"lean_confirmation":null,"pith_extraction":{"msc":[],"pacs":[],"model":"deepseek-v4-flash","headline":"Wall-pressure phase decides whether transpiration cuts or boosts drag.","keywords":["turbulent channel flow","drag reduction","wall transpiration","wall pressure","phase difference","varying-phase opposition control","spanwise rollers","tailored surfaces"],"falsifier":"Run a DNS or experiment with streak-scale transpiration at a higher Reynolds number, e.g. $Re_\\tau\\approx 1000$, where the slow pressure is relatively stronger, and measure the circular mean of $\\Delta\\theta_\\kappa$ at the energetic streak scales: if drag is reduced while the mean phase difference is not near zero, the central mapping is falsified. Alternatively, measure the wall-pressure/transpiration phase on a drag-reducing riblet surface at its effective virtual wall: in-phase streak scales should appear if the claimed dynamical equivalence with tailored surfaces holds.","tokens_in":33319,"feed_emoji":"🌊","tokens_out":8246,"duration_ms":70403,"temperature":0.7,"pith_summary":"The paper tries to establish that the effect of wall transpiration on a turbulent channel flow is governed by one wall quantity: the phase difference between the transpiration and the wall pressure. Using direct numerical simulations with a controller that shifts the streamwise phase of the transpiration, it shows that slow, streamwise-elongated “streak” transpiration suppresses the near-wall cycle and reduces drag when it is in phase with the wall pressure, and amplifies it otherwise; short, wide “roller” transpiration generates spanwise rollers and drag increase when it is out of phase. The phase relation is itself a dynamical outcome of the coupled pressure–transpiration system, and the paper identifies conditions under which the relation is robust. If correct, the result unifies active and passive flow control: riblets, porous and permeable walls can be understood as transpiration boundary conditions whose drag effect is set by the same two scale families and phase relations.","feed_headline":"Wall pressure phase decides whether transpiration cuts or boosts drag","feed_subtitle":"The phase between wall pressure and transpiration governs drag reduction, amplification, and spanwise rollers.","key_machinery":"The machinery is the phase difference between wall pressure and transpiration at each Fourier mode, $\\Delta\\theta_\\kappa(t)=\\angle\\hat p_\\kappa(y_w,t)-\\angle\\hat v_\\kappa(y_w,t)$, together with the decomposition of the pressure into fast (linear source), slow (nonlinear source), and Stokes (boundary-condition) components. The varying-phase opposition control law $\\hat v_\\kappa(y_w)=-\\hat A_d\\hat v_\\kappa(y_d)$ with complex gain $\\hat A_d$ makes the phase shift $\\angle\\hat A_d$ an effective streamwise shift of the transpiration relative to the sensor signal, so one control parameter sweeps the phase range. The FIK decomposition of the friction coefficient into a weighted integral of Reynolds stresses turns drag change into a sum over wavenumbers, and circular statistics (the circular mean) reduce noisy phase histograms at each scale to a single number.","core_discovery":"The central claim is that the wall-pressure/transpiration phase difference $\\Delta\\theta_\\kappa = \\angle\\hat p_\\kappa(y_w)-\\angle\\hat v_\\kappa(y_w)$ at a spatial scale $\\kappa=(k_x,k_z)$ selects among the three canonical responses in a low-$Re_\\tau$ turbulent channel flow. At streak scales, which are associated with the near-wall cycle, $\\Delta\\theta_\\kappa\\approx 0$ suppresses the cycle and reduces drag, while other phases amplify it. At roller scales, $\\Delta\\theta_\\kappa\\approx\\pi$ coincides with the emergence of spanwise rollers and a large drag increase. The paper further shows that these phase relations are not imposed but emerge from the coupled dynamics: the fast pressure picks up a robust phase because its Green’s-function weight sits in the wall-normal layer where the controller constrains $\\hat v_\\kappa$; the slow pressure’s nonlinear source term is typically too decorrelated to set a robust phase; and the Stokes pressure sets a robust phase only when the temporal frequency content of the transpiration is approximately sparse, as happens when an amplified eigenmode dominates.","pith_inferences":["The phase-difference criterion could be tested as a wall-only sensor metric in experiments: measure the circular mean of $\\Delta\\theta_\\kappa$ on a riblet or porous surface at its effective virtual wall, and check whether drag-reducing configurations show in-phase streak scales.","Because the slow pressure is stochastic and its direction varies, the robust in-phase relation in the drag-reducing case may be carried mainly by the Stokes and fast components; at higher $Re_\\tau$ or with different controller gains the total-pressure phase could decorrelate from $\\angle\\hat A_d$, so the mapping may need a statistical rather than deterministic statement.","The same logic suggests a concrete design target for meta-material surfaces: the surface response to wall pressure should be scale-dependent, with zero phase at streak scales and $\\pi$ phase at roller scales, which is testable beyond the paper’s simulations."],"forward_implications":["Drag reduction by transpiration requires streak-scale actuation that is in phase with the wall pressure; small phase shifts do this, while larger positive or negative shifts amplify the near-wall cycle and increase drag.","Roller-scale transpiration with an out-of-phase pressure–transpiration relation generates spanwise rollers and drag increase, and this is dynamically equivalent to riblets past the viscous regime and to porous or permeable walls.","Wall pressure can serve as a wall-based proxy for the background flow state in control design, but pressure data from uncontrolled canonical flows will not transfer because transpiration makes the Stokes pressure leading-order.","Passive pressure-driven tailored surfaces will struggle to reduce drag unless they impose a scale-dependent response and an in-phase relation between pressure and transpiration, for example through complex-valued permeability or resonator-type surfaces."],"supporting_citations":[{"why":"Establishes the varying-phase opposition control formulation and shows that the complex gain changes the attainable drag reduction.","marker":"Toedtli et al. 2019a"},{"why":"Documents that positive phase shifts generate spanwise rollers and an amplified eigenvalue, the drag-increase mechanism compared throughout the present study.","marker":"Toedtli et al. 2020"},{"why":"Provides the scale-restricted controller results on which the streak/roller scale-family decomposition rests.","marker":"Toedtli 2021"},{"why":"Proposes the hypothesis that in-phase wall-normal velocity and pressure are needed for drag reduction, which this study tests in the full nonlinear flow.","marker":"Xu et al. (2003)"},{"why":"Supplies the fast/slow/Stokes pressure decomposition and Green’s function solution used to explain how each pressure component establishes its phase.","marker":"Kim (1989)"},{"why":"Gives the Reynolds-stress decomposition of skin friction underlying the scale-by-scale drag metric.","marker":"Fukagata et al. (2002)"},{"why":"Identifies spanwise rollers as the cause of riblet viscous-regime breakdown, the tailored-surface analogue for roller scales.","marker":"García-Mayoral & Jiménez 2011"},{"why":"Models permeable substrates with a transpiration coefficient and shows the linear regime and roller-scale breakdown central to the out-of-phase analysis.","marker":"Gómez-de Segura & García-Mayoral 2019"},{"why":"Shows riblets suppress or amplify the near-wall cycle depending on spacing, the analogue for streak scales.","marker":"Chavarin & Luhar 2020"},{"why":"Provides the canonical channel DNS data used to validate the flow solver and the pressure Poisson solver.","marker":"Lee & Moser 2015"}],"fun_headline_variants":["Wall pressure phase dictates if transpiration slashes or spikes drag","Phase between wall pressure and transpiration steers turbulent drag","In-phase transpiration weakens turbulence, out-of-phase amplifies it","Phase of wall pressure toggles drag reduction vs. amplification","Wall-pressure phase selects between drag reduction and spanwise rollers"],"cache_read_input_tokens":3200,"weakest_assumption_plain":"The load-bearing premise is that the in-phase/out-of-phase dichotomy for the total wall pressure survives the slow pressure component; the paper itself notes it is unclear how the drag-reducing controller establishes a robust overall phase relation given the large directional variation of the slow pressure, so if slow pressure dominates in other regimes or at higher Reynolds number the mapping could break down.","fun_headline_variants_meta":{"raw":{"variants":["Wall pressure phase dictates if transpiration slashes or spikes drag","Phase between wall pressure and transpiration steers turbulent drag","In-phase transpiration weakens turbulence, out-of-phase amplifies it","Phase of wall pressure toggles drag reduction vs. amplification","Wall-pressure phase selects between drag reduction and spanwise rollers"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.000695,"raw_usage":{"total_tokens":3196,"prompt_tokens":1051,"completion_tokens":2145,"prompt_tokens_details":{"cached_tokens":384},"prompt_cache_hit_tokens":384,"prompt_cache_miss_tokens":667,"completion_tokens_details":{"reasoning_tokens":2069}},"tokens_in":667,"tokens_out":2145,"duration_ms":13074,"temperature":1.0,"reasoning_tokens":2069,"cache_read_input_tokens":384,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-12T15:06:28.192048+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"Run a DNS or experiment with streak-scale transpiration at a higher Reynolds number, e.g. $Re_\\tau\\approx 1000$, where the slow pressure is relatively stronger, and measure the circular mean of $\\Delta\\theta_\\kappa$ at the energetic streak scales: if drag is reduced while the mean phase difference is not near zero, the central mapping is falsified. Alternatively, measure the wall-pressure/transpiration phase on a drag-reducing riblet surface at its effective virtual wall: in-phase streak scales should appear if the claimed dynamical equivalence with tailored surfaces holds.","supporting_citations":[{"cited_title":"International Journal of Heat and Fluid Flow 85 , 108651","cited_arxiv_id":null,"evidence_quote":"Documents that positive phase shifts generate spanwise rollers and an amplified eigenvalue, the drag-increase mechanism compared throughout the present study."},{"cited_title":"PhD thesis","cited_arxiv_id":null,"evidence_quote":"Provides the scale-restricted controller results on which the streak/roller scale-family decomposition rests."},{"cited_title":"Journal of Fluid Mechanics 205 , 421--451","cited_arxiv_id":null,"evidence_quote":"Supplies the fast/slow/Stokes pressure decomposition and Green’s function solution used to explain how each pressure component establishes its phase."},{"cited_title":"Journal of Fluid Mechanics 678 , 317--347","cited_arxiv_id":null,"evidence_quote":"Identifies spanwise rollers as the cause of riblet viscous-regime breakdown, the tailored-surface analogue for roller scales."},{"cited_title":"& García-Mayoral, R","cited_arxiv_id":null,"evidence_quote":"Models permeable substrates with a transpiration coefficient and shows the linear regime and roller-scale breakdown central to the out-of-phase analysis."},{"cited_title":"AIAA Journal 58 (2), 589--599 , arXiv:arXiv: https://doi.org/10.2514/1.J058205","cited_arxiv_id":null,"evidence_quote":"Shows riblets suppress or amplify the near-wall cycle depending on spacing, the analogue for streak scales."}],"review_version":1}