{"id":"0d6408b9-490a-45f8-8adc-33246d61d9f6","arxiv_id":"1908.03050","paper_version":2,"verdict":"CONDITIONAL","confidence":"MODERATE","novelty_score":6.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":5,"one_line_summary":"A new energy-resolved laminarization probability shows spanwise wall oscillations nearly double the chance that strong perturbations decay in plane Couette flow, even though edge-state energy decreases.","lead":"This paper introduces a statistical protocol for measuring how often random perturbations of a given energy return a plane Couette flow to its laminar state, and uses it to test spanwise wall-oscillation control. The result matters because it shows a control that lowers the edge-state energy can still substantially suppress transition, so simple scalar indicators can mislead.","discovery_kind":"new_method","skeptic_critique":{"model":"deepseek-v4-flash","headline":"Control-effectiveness verdict rests on one hand-chosen perturbation ensemble; absent a robustness check across perturbation families, the headline claim that wall oscillations suppress transition remains conditional.","rationale":"The reader's weakest-assumption analysis identifies exactly the same load-bearing issue: the representativeness of the random perturbation ensemble. This is the correct locus of the central claim. The paper's novel protocol is internally coherent and the numerical results are presented with reasonable detail, but the conclusion that the control strategy suppresses transition is an average over one hand-chosen perturbation family. The authors are transparent about this in Section 4, calling the work a proof of concept and noting that the perturbation choice may not be ideal for a given application. That transparency does not remove the concern; it makes the conditional nature of the claim explicit. I considered other possible objections, including the ambiguous formulation of the controlled-case initial conditions relative to the oscillatory laminar state, the absence of error bars, and the reliance on fitted gamma curves for the 1.8-fold estimate. These are real but secondary: the first could be resolved by clarification, and the latter two affect the precision but not the qualitative ordering of the controlled and uncontrolled curves. The ensemble-dependence concern, by contrast, can invalidate the headline claim if a different physically motivated perturbation family changes the ordering. The reader's CONDITIONAL verdict already accounts for this, so no change to the verdict is needed. The concrete test proposed would settle whether the concern lands: if an alternative ensemble preserves the ordering, the concern is mitigated and the claim can be strengthened; if not, the claim should be explicitly restricted to the protocol's perturbation family.","tokens_in":11061,"tokens_out":11184,"duration_ms":128275,"concrete_test":"Repeat the Re = 500 controlled-versus-uncontrolled comparison using an alternative perturbation ensemble, e.g., independent random perturbations constructed around the minimal-seed / optimal transient-growth direction, normalized to the same 40 energy levels with N = 200 per level and identical laminarization criteria. If the controlled laminarization probability is not above the uncontrolled one at most energy levels, or if the difference is within overlapping binomial 95% confidence intervals, then the protocol's verdict is ensemble-dependent and the headline claim must be restricted to the original perturbation family.","verdict_should_be":"UNCHANGED","load_bearing_attack":"The central claim that spanwise wall oscillation suppresses transition is established only for the particular random-perturbation family defined in Section 2: uniform spectral coefficients, B drawn uniformly on [-2E/||Ulam||, 2E/||Ulam||] with equal numbers of positive and negative signs, and energies up to (2/3)Eturb. The controlled-case plateau (a = 0.286) is compared with the uncontrolled plateau (a = 0.0805) and the reported 1.8-fold mean increase is computed from fits to this ensemble. The authors concede in Section 4 that the perturbation form is a modeling choice and that the work is a proof of concept. If a physically relevant disturbance family, such as localized streamwise vortices or optimal transient-growth perturbations, produced the opposite ordering of the controlled and uncontrolled curves, then the conclusion 'transition is significantly suppressed' would fail, even though the scalar-indicator criticism (edge-state energy decreases while laminarization probability increases) would remain. The paper provides no sensitivity or invariance test for the perturbation family, so the external validity of the central claim is the least secure link.","agreement_with_reader":"agree"},"referee_report":{"model":"deepseek-v4-flash","summary":"The paper proposes a probabilistic protocol for assessing the robustness of the laminar state and the efficacy of transition-control strategies. For plane Couette flow, the authors sample random initial perturbations at discrete kinetic-energy levels and estimate the laminarization probability Plam(E) as the fraction of trajectories that decay below an energy threshold. The uncontrolled results at Re = 400, 500, and 700 show a monotonically decreasing Plam(E) that saturates at a plateau value a, and the authors fit these curves with a reflected, saturated Gamma cumulative distribution function. They then apply the protocol to Re = 500 with spanwise wall oscillations (amplitude W = 0.3, frequency ω = 1/16), obtaining a plateau a = 0.286 versus a = 0.0805 uncontrolled, and report a mean relative increase in laminarization probability of 1.8. In contrast, the edge-state kinetic energy decreases by about 37% under control, which the authors interpret as demonstrating that scalar indicators such as edge-state or minimal-seed energy can be misleading for control assessment. The paper explicitly frames the work as a proof of concept and acknowledges that the perturbation family is a modeling choice.","tokens_in":11283,"tokens_out":6511,"duration_ms":73482,"significance":"If the result holds, the protocol offers a genuinely useful diagnostic for transition control: it replaces a single scalar criterion with an energy-resolved basin-stability measurement, and the controlled/uncontrolled comparison is made through direct numerical simulation rather than through the fitted curve. The paper is careful in its numerical methodology, uses standard edge-tracking and spectral DNS, and is explicit that the perturbation ensemble is a modeling choice. The main messages—that scalar proxies can disagree with the energy-resolved probability and that the protocol can be customized—are valuable and likely to influence subsequent work on transition control. The significance is tempered by the absence of statistical uncertainty quantification on the reported probabilities and by the fact that the headline control-effectiveness claim is demonstrated for one hand-chosen perturbation family only.","major_comments":[{"comment":"The manuscript reports Plam values and fitted parameters without any uncertainty quantification. Plam(E) is a binomial proportion estimated from N = 100 or 200 independent draws, so the standard error at the uncontrolled plateau (a = 0.0805, N = 200) is about 0.019 and at the controlled plateau (a = 0.286, N = 200) about 0.032; the raw plateau difference is therefore large relative to sampling error, but the central quantitative claim of a 1.8-fold mean relative increase is computed from fitted Gamma CDFs and carries no confidence interval. Please add binomial confidence intervals (or at least the raw counts) for Plam at each energy level, report uncertainties for the fitted parameters a, α, and β, and state whether the 1.8 figure is statistically significant in the presence of multiple-comparison and fitting uncertainty.","section":"Section 3, Fig. 4 and Table 1"},{"comment":"The random perturbation family—uniform spectral coefficients, B drawn uniformly from [-2E/||Ulam||, 2E/||Ulam||] with equal numbers of positive and negative signs, and energies capped at (2/3)Eturb—is acknowledged in the Discussion to be a modeling choice, yet the abstract and Section 3 state that 'transition is significantly suppressed' without that qualifier. Because a different physically relevant perturbation family (for example, localized streamwise vortices or optimally growing disturbances) could in principle give a different ordering of the controlled and uncontrolled Plam curves, the central claim is strictly an ensemble-dependent result. Please either add a robustness test across perturbation families or systematically qualify the conclusion in the abstract and throughout as being valid for the chosen perturbation ensemble.","section":"Section 2 (RP generation) and Section 4 (Discussion)"},{"comment":"In the controlled case the laminar base flow is time-dependent (spanwise wall oscillations), but the random perturbations are generated 'in the same way as for the uncontrolled case', i.e., using the steady profile Ulam = y ex, and the energy E is defined via Eq. (2.3) without specifying the reference state. The manuscript does not state whether u in the controlled case is the deviation from the oscillatory Stokes-layer solution or from Ulam. This distinction matters because the energy levels E(j), the thresholds Elam and Eturb, and the edge-state energy are all defined with respect to that decomposition. Please specify the reference base flow used in the controlled-case energy and perturbation generation, and if Ulam is retained, justify why that is the relevant state from which perturbations should be measured.","section":"Section 2, Eq. (2.3), and Section 3 (controlled case)"},{"comment":"The 'relative probability increase' of 1.8 is defined as (posc(E) - p(E))/p(E) averaged over 'the range of the considered energies', but the averaging measure is not specified: uniform in E, uniform over the 40 discrete energy levels, or something else? The statistic is also computed from the fitted Gamma CDFs rather than directly from the measured Plam values. Please define the average precisely, report the corresponding raw-data statistic, and provide its sampling uncertainty.","section":"Section 3, relative probability increase"}],"minor_comments":[{"comment":"The sentence beginning 'that of reducing the Reynolds number...' is a fragment and appears to be a vestige of an earlier draft; it should be completed or removed.","section":"Section 3, after 'laminarize'"},{"comment":"There is a typo: 'Reynods' should be 'Reynolds'.","section":"Section 4, first paragraph"},{"comment":"There is a typo: 'uncontroled' should be 'uncontrolled'.","section":"Section 3, controlled case"},{"comment":"The integral notation '∫_Ω u · u ∂Ω' appears to be a typo for 'dΩ' (or 'dV').","section":"Eq. (2.3)"},{"comment":"The notation γ(α, βE) for the lower incomplete gamma function is nonstandard; please define it explicitly as the regularized lower incomplete gamma function γ(α, βE)/Γ(α), so that p(E) is clearly a CDF.","section":"Section 3, fitting function"}],"recommendation":"major_revision","confidential_remarks":"The paper is a solid proof-of-concept and the direct DNS comparison is well executed, but the lack of statistical error bars and the ensemble dependence of the headline claim need to be addressed before publication. The perturbation-ensemble issue is acknowledged by the authors themselves, so the revision path is clear: either add a sensitivity test across perturbation families or qualify the central claim throughout. The base-flow ambiguity in the controlled case should be clarified because it is essential for interpreting the controlled energy levels. The paper fits the scope of the journal and, with these revisions, would be acceptable."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"The short version: this is a solid proof-of-concept for a more informative way to assess transition control than scalar indicators. The protocol—sampling random perturbations by energy and computing laminarization probability—works as advertised, and the demonstration that wall oscillations raise Plam while lowering edge-state energy is a genuinely useful caution about relying on edge-state energy.\n\nThe DNS is standard and competently done (Channelflow), the edge tracking is careful, and the perturbation generation is described in enough detail to reproduce. The authors also check the transient-lifetime issue properly, and the controlled/uncontrolled comparison is a direct numerical measurement, so the core observation is not an artifact of the gamma CDF fits.\n\nMain soft spot is the perturbation ensemble. Everything rests on one random family: uniform spectral coefficients, uniform B with equal signs, energies up to (2/3)Eturb. The authors explicitly concede this is a modeling choice and call the work a proof of concept, so the strong wording 'transition is significantly suppressed' should be read as conditional on that ensemble. They do not test robustness across perturbation families; a referee could reasonably ask for one additional ensemble to see whether the ordering holds. That is a strengthening request, not a fatal flaw.\n\nLesser issues: no error bars on Plam despite binomial sampling with N=100/200, which are trivial to compute; the thresholds Elam=Eturb/100 and tturb=400 are plausible and tturb is justified, but sensitivity to them is untested. The reported 1.8 mean increase is computed from the fitted curves, not the raw probabilities, so it is partly a property of the fit—the plateau comparison (0.286 vs 0.0805) is the more honest headline.\n\nCitations look appropriate; the link to basin stability is properly credited.\n\nBottom line: this deserves serious peer review. I would send it out, and my main requests would be error bars and at least one perturbation-family robustness check. The paper is worth reading for anyone working on transition control or basin-of-attraction methods.","headline":"Useful proof-of-concept for energy-resolved laminarization probabilities in control assessment; core claim holds, but missing error bars and a single perturbation ensemble make it conditional.","tokens_in":11803,"tokens_out":3008,"would_cite":false,"duration_ms":31544,"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":"Laminarization probabilities, not scalar energies, reveal when wall-oscillation control suppresses transition in plane Couette flow.","keywords":["plane Couette flow","transition to turbulence","laminarization probability","basin of attraction","edge state energy","minimal seed","spanwise wall oscillations","flow control"],"falsifier":"Recompute the laminarization-probability curves using a different, physically motivated perturbation family—for example, perturbations localised in the wall-normal direction or optimised for transient energy growth—and check whether the controlled curve still lies above the uncontrolled one at Re = 500; if the ordering reverses, the protocol's verdict is not robust.","tokens_in":10853,"feed_emoji":"🌊","tokens_out":8469,"duration_ms":72023,"temperature":0.7,"pith_summary":"This paper introduces a probabilistic protocol for judging whether a flow-control strategy actually protects the laminar state. Instead of relying on scalar measures such as the energy of the edge state or the minimal seed, the authors sample random perturbations at fixed energies near the laminar flow, time-integrate them, and record the fraction that laminarise. Applied to plane Couette flow at Reynolds number 500, the protocol shows that spanwise wall oscillations increase the average laminarization probability by a factor of about 1.8 even though the edge-state energy drops by roughly 37%. The paper concludes that energy-resolved transition probabilities, not scalar indicators, are the reliable way to assess control, and that the protocol generalises to any nonlinear system with finite-amplitude instability.","feed_headline":"Wall oscillations nearly double laminarization odds in Couette flow","feed_subtitle":"Scalar indicators like edge-state energy fall even though control succeeds, so energy-resolved sampling is needed.","key_machinery":"The carrying object is the laminarization-probability curve $P_{\\mathrm{lam}}(E)$, estimated by Monte Carlo sampling over a random perturbation ensemble at each of 40 energy levels up to $(2/3)E_{\\mathrm{turb}}$. Each random perturbation is built as $u = A u_\\perp + B U_{\\mathrm{lam}}$, with $u_\\perp$ an incompressible field orthogonal to the laminar flow, spectral coefficients drawn uniformly, $B$ drawn uniformly from $[-2E/\\|U_{\\mathrm{lam}}\\|, 2E/\\|U_{\\mathrm{lam}}\\|]$, equal numbers of positive and negative $B$, and $A$ fixed by the energy constraint. The perturbations are integrated forward until their energy drops below $E_{\\mathrm{lam}} = E_{\\mathrm{turb}}/100$ (laminarization) or exceeds $E_{\\mathrm{turb}}$ (transition); the empirical fraction at each energy is the probability. The fit $p(E) = 1 - (1-a)\\gamma(\\alpha,\\beta E)$ with the lower incomplete gamma function compresses each curve into three parameters, making controlled and uncontrolled cases quantitatively comparable.","core_discovery":"The central claim is that the robustness of the laminar state, and therefore the success of a control strategy, should be quantified by the probability that random finite-amplitude perturbations of a given energy return to laminar flow. In uncontrolled plane Couette flow, the laminarization probability $P_{\\mathrm{lam}}(E)$ decreases from about 1 at small perturbation energies and saturates at a plateau $a$, with fitted values $a = 0.244$, $0.0805$, and $0.0484$ at Re = 400, 500, and 700. With spanwise wall oscillations of amplitude $W = 0.3$ and frequency $\\omega = 1/16$ at Re = 500, the plateau rises to $0.286$ and the average relative increase of $P_{\\mathrm{lam}}$ across the sampled energy range is 1.8, while the edge-state energy falls from roughly $1.82\\times10^{-2}$ to $1.15\\times10^{-2}$. Because the scalar indicators move in the opposite direction from the basin expansion, the paper argues that they cannot reliably assess control and that the full probability curve is required.","pith_inferences":["If the protocol is right, a control study that reports only edge-state energy could reject a strategy that actually expands the laminar basin, since the edge-state energy here falls by about 37% while the laminarization probability rises.","The sensitivity of the curves to the perturbation family suggests a robustness test: repeat the sampling with perturbations shaped like experimental noise or optimal transient-growth disturbances, and treat agreement between families as evidence for the control verdict.","The non-zero plateau at large energies reflects the fractal interleaving of laminar and turbulent basins, so the protocol measures basin geometry as well as basin volume; a natural extension is to compute an energy-resolved basin-stability map for other multistable systems."],"forward_implications":["Control assessment in subcritical shear flows should be based on energy-resolved laminarization probabilities rather than on edge-state or minimal-seed energies alone.","At Re = 500, spanwise wall oscillations with amplitude 0.3 and frequency 1/16 nearly double the average laminarization probability, with the largest gains for large-energy perturbations.","The controlled flow at Re = 500 behaves like the uncontrolled flow at Re = 400, except that small-energy perturbations are unaffected by the control.","The fitting form $p(E) = 1 - (1-a)\\gamma(\\alpha,\\beta E)$ provides a compact way to compare the effect of a control strategy at a given Reynolds number.","The same sampling protocol can be applied to any system with finite-amplitude instability, including other shear flows and non-fluid multistable systems."],"supporting_citations":[{"why":"It supplies the edge-tracking algorithm used to measure the edge-state energies that the scalar-indicator comparison is based on.","marker":"Skufca et al. 2006"},{"why":"It provides the reference plane-Couette edge state used to identify the uncontrolled edge trajectory.","marker":"Schneider et al. 2008"},{"why":"It provides the wall-oscillation control strategy, its parameter values, and the earlier claim that it raises the minimal-seed energy.","marker":"Rabin et al. 2014"},{"why":"It defines minimal seeds and their energy decay with Reynolds number, the scalar indicator the protocol finds inconclusive.","marker":"Duguet et al. 2013"},{"why":"It establishes that spanwise wall oscillations suppress turbulence, the physical effect whose efficiency the protocol reassesses.","marker":"Jung et al. 1992"},{"why":"It documents drag reduction by spanwise wall oscillations and motivates the control configuration tested here.","marker":"Quadrio & Ricco 2004"},{"why":"It explains the interleaved, fractal edge geometry that produces the non-vanishing laminarization plateau at large energies.","marker":"Chantry & Schneider 2014"},{"why":"It describes upper-edge-of-chaos events, justifying the waiting-time rule that distinguishes transient overshoot from transition.","marker":"Budanur et al. 2020"},{"why":"It fixes the domain size and numerical resolution used in all simulations.","marker":"Pershin et al. 2019"}],"fun_headline_variants":["Probability curves beat edge energy for judging flow control","Laminarization odds rise despite edge-energy drop","Sampling phase space reveals control effectiveness scalars miss","Probabilistic protocol rates control beyond traditional scalars","For Couette flow, probability sampling wins over scalar metrics"],"cache_read_input_tokens":3200,"weakest_assumption_plain":"The conclusion that wall oscillations suppress transition rests on the assumption that the randomly generated perturbation family used in the sampling is representative of the perturbations that would actually trigger transition in a real flow; the authors note this is a modelling choice.","fun_headline_variants_meta":{"raw":{"variants":["Probability curves beat edge energy for judging flow control","Laminarization odds rise despite edge-energy drop","Sampling phase space reveals control effectiveness scalars miss","Probabilistic protocol rates control beyond traditional scalars","For Couette flow, probability sampling wins over scalar metrics"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.000821,"raw_usage":{"total_tokens":3596,"prompt_tokens":948,"completion_tokens":2648,"prompt_tokens_details":{"cached_tokens":384},"prompt_cache_hit_tokens":384,"prompt_cache_miss_tokens":564,"completion_tokens_details":{"reasoning_tokens":2572}},"tokens_in":564,"tokens_out":2648,"duration_ms":20664,"temperature":1.0,"reasoning_tokens":2572,"cache_read_input_tokens":384,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-14T14:25:13.398050+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"Recompute the laminarization-probability curves using a different, physically motivated perturbation family—for example, perturbations localised in the wall-normal direction or optimised for transient energy growth—and check whether the controlled curve still lies above the uncontrolled one at Re = 500; if the ordering reverses, the protocol's verdict is not robust.","supporting_citations":[{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"It provides the reference plane-Couette edge state used to identify the uncontrolled edge trajectory."},{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"It provides the wall-oscillation control strategy, its parameter values, and the earlier claim that it raises the minimal-seed energy."},{"cited_title":", Monokrousos, A","cited_arxiv_id":null,"evidence_quote":"It defines minimal seeds and their energy decay with Reynolds number, the scalar indicator the protocol finds inconclusive."},{"cited_title":", Mangiavacchi, N","cited_arxiv_id":null,"evidence_quote":"It establishes that spanwise wall oscillations suppress turbulence, the physical effect whose efficiency the protocol reassesses."},{"cited_title":"& Ricco, P","cited_arxiv_id":null,"evidence_quote":"It documents drag reduction by spanwise wall oscillations and motivates the control configuration tested here."},{"cited_title":"& Schneider, T","cited_arxiv_id":null,"evidence_quote":"It explains the interleaved, fractal edge geometry that produces the non-vanishing laminarization plateau at large energies."},{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"It describes upper-edge-of-chaos events, justifying the waiting-time rule that distinguishes transient overshoot from transition."},{"cited_title":", Beaume, C","cited_arxiv_id":null,"evidence_quote":"It fixes the domain size and numerical resolution used in all simulations."}],"review_version":1}