{"id":"7c01aec1-c631-410e-9064-3da3f98ed7ea","arxiv_id":"2411.12242","paper_version":1,"verdict":"CONDITIONAL","confidence":"MODERATE","novelty_score":5.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":4,"one_line_summary":"Increasing the amplitude of Gaussian wakes entering a T106A turbine cascade delays boundary-layer separation, shrinks separated-flow regions, and reduces skin friction drag by about 50% at the highest amplitude tested.","lead":"This computational study simulates the flow over a low-pressure turbine blade with incoming wake-like disturbances at nine different strengths, finding that stronger wakes delay boundary-layer separation and cut skin friction drag by about half. It matters for turbine design because reducing aerodynamic losses in low-pressure turbines directly improves jet engine fuel efficiency.","discovery_kind":"extension","skeptic_critique":{"model":"deepseek-v4-flash","headline":"Profile-loss claim contradicts the paper's own loss model: Eq. (13) makes loss proportional to trailing-edge momentum thickness, which the paper reports increases by 39.2% with awake, so the claimed loss benefit is unsupported and likely backwards.","rationale":"The reader's weakest assumption was the 2D/3D fidelity issue: vortex stretching and spanwise turbulent-spot dynamics are absent, so quantitative claims may not transfer to real turbine flows. That concern is legitimate, but it is about external validity; it does not undermine the paper's internal 2D argument. The present concern is stronger because it identifies an internal inconsistency: the paper's own loss model and its own reported momentum-thickness trend point in the opposite direction from the profile-loss benefit claimed in the abstract and conclusions. Even under the paper's 2D assumptions, the design-relevant conclusion is unsupported. This does not change the overall conditional disposition, because the core flow-physics observations (delayed separation, earlier reattachment, reduced separated-flow extent, and the CETE budget trends) could still be salvageable in a revised manuscript that either removes the profile-loss-benefit framing or supplies an actual ζp calculation, ideally together with a spanwise/3D check for the quantitative 50% skin-friction claim. The reader already noted that profile loss was never computed, but did not emphasize that the momentum-thickness proxy actually contradicts the claimed benefit; hence partial agreement.","tokens_in":19991,"tokens_out":7759,"duration_ms":89551,"concrete_test":"Compute ζp from Eq. (13) for awake = 0.1 and awake = 0.9 using the reported trailing-edge momentum thickness (Table 2), the displacement thickness, base pressure coefficient, trailing-edge thickness, pitch pb = 0.9306, and exit angle αex = -63.2°, or directly by integrating total pressure loss at the outflow plane. If ζp increases with awake, the profile-loss-improvement claim is contradicted and Section 4 must be revised; if ζp decreases, the paper must show explicitly which non-momentum terms in Eq. (13) overcome the 39.2% momentum-thickness increase.","verdict_should_be":"UNCHANGED","load_bearing_attack":"The central applied claim, that higher wake amplitude improves profile loss (Abstract and Section 4), is not supported by the paper's own loss proxy. Section 3.1 presents Eq. (13) for profile loss and states that the mixed-out loss term, proportional to trailing-edge momentum thickness, contributes about 90% of total loss. Immediately afterward, the paper reports that increasing awake from 0.1 to 0.9 increases maximum momentum thickness by 39.2% (Fig. 7, Table 2). Eq. (13) therefore predicts a roughly 30-40% increase in the dominant loss term, not a reduction. To invert this, one would need compensating changes in Cpb and δTE large enough to overturn the momentum-thickness increase, but the paper never computes ζp or provides such evidence. In fact, no direct profile-loss computation appears anywhere. Thus the abstract's 'suppress flow separation and improve profile loss' and the conclusion's 'potential benefit for reducing the profile loss' are unsupported extrapolations that run opposite to the paper's own displayed quantity. This is more damaging than the acknowledged 2D limitation because it fails even if all 2D assumptions are granted.","agreement_with_reader":"partial"},"referee_report":{"model":"deepseek-v4-flash","summary":"The paper presents two-dimensional implicit large-eddy simulations of the T106A low-pressure turbine cascade with periodically incoming Gaussian wakes, varying the wake amplitude awake from 0.1 to 0.9 in nine cases. The authors report that increasing wake amplitude delays suction-surface separation, moves reattachment upstream, reduces the separated-flow region, and lowers skin friction by about 50% at the highest amplitude. They also analyze vorticity spectra, space-time enstrophy maps, and TKE and compressible-enstrophy budgets, and conclude that calmed regions induced by wake passing suppress separation and improve profile loss at low Reynolds numbers.","tokens_in":20070,"tokens_out":4010,"duration_ms":43310,"significance":"If the quantitative claims were properly supported, the paper would provide a useful parametric database for wake-induced transition on a realistic LPT profile, and the application of the CETE budget to a wake-perturbed turbine cascade would be a novel diagnostic contribution. The systematic nine-case amplitude sweep and the connection to classical turbulent-spot and calmed-region convection speeds are valuable. However, the main applied claim—improved profile loss—is contradicted by the paper's own loss proxy, and the headline skin-friction reduction is not defined precisely enough to be evaluated. The two-dimensional framework is acknowledged but limits transfer of the conclusions to real turbine flows.","major_comments":[{"comment":"The claim that higher wake amplitude 'improve[s] profile loss' is contradicted by the paper's own loss model. Equation (13) states that profile loss is dominated (about 90%) by the term proportional to trailing-edge momentum thickness θTE. Section 3.1 and Table 2 report that increasing awake from 0.1 to 0.9 increases the maximum momentum thickness by 39.2% (from 0.0369 to 0.0514). Without a compensating decrease in Cpb and δTE, Eq. (13) predicts a substantial increase, not a reduction, in ζp. No direct ζp computation is presented anywhere, so the abstract's 'improve profile loss' and Section 4's 'potential benefit for reducing the profile loss' are unsupported; as written, the reported data point in the opposite direction.","section":"Abstract; Sections 3.1 and 4; Eq. (13); Table 2"},{"comment":"The central quantitative claim of a 50% reduction in skin friction is not defined precisely. Figure 4 shows a strongly streamwise-varying local Cf, so the 50% figure is ambiguous: it could refer to a local value, a suction-surface integral, or a spatial average, and the manuscript does not state which. The claim also lacks uncertainty estimates and any grid-refinement or time-averaging convergence check, so the reader cannot determine whether the amplitude trend is numerical or physical.","section":"Section 3.1; Fig. 4"},{"comment":"The simulations solve the two-dimensional compressible Navier-Stokes equations, and Section 3.3 acknowledges that the vortex stretching term is absent in 2D. Because the paper's applied conclusions—calmed regions 'suppress flow separation and improve profile loss'—are framed for LPT blades at low Re, and because turbulent spots and calmed regions are intrinsically three-dimensional, the 2D results cannot by themselves support those applied claims. The manuscript should either clearly scope the conclusions as two-dimensional only, with a discussion of expected three-dimensional effects, or provide a three-dimensional verification case.","section":"Sections 2 and 3.3"}],"minor_comments":[{"comment":"The text refers to 'the spectrum in Fig. 9(c)' when describing the awake=0.3 case, but Fig. 9(c) is a time series and the spectrum is in Fig. 9(d); similar frame-number mismatches occur for the awake=0.5 and later cases.","section":"Section 3.2; Fig. 9"},{"comment":"The caption of Fig. 11 lists amplitudes awake=0.7 and 0.9, while the accompanying text describes awake=0.1, 0.3, and 0.5; the caption and text should be reconciled.","section":"Fig. 11 caption"},{"comment":"Section 2 states that 'five to six through-flows' are used to flush initial transients, while Section 3.1 says 'five through flows'; the numbers should be harmonized.","section":"Section 2 versus Section 3.1"}],"recommendation":"major_revision","confidential_remarks":"The profile-loss contradiction is the main obstacle: Eq. (13) and the reported momentum-thickness increase point to a loss increase, not the claimed benefit. If the authors remove or substantially reframe the profile-loss claim, define the skin-friction metric, and add convergence evidence, the parametric CFD and CETE diagnostic content could be salvageable. The two-dimensional limitation is significant but could be handled by scoping the conclusions explicitly."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"Dear colleague,\n\nQuick take: this is a workmanlike 2D CFD sweep, not a breakthrough. The separation-suppression trend with wake amplitude is internally consistent, but the central applied claim—higher wake amplitude improves profile loss—runs against the paper's own loss equation. The 50% skin-friction reduction is also not backed by convergence or uncertainty evidence.\n\nWhat's new: it is the first systematic sweep of Gaussian wake amplitude (0.1 to 0.9) on the T106A cascade, and it applies the author's compressible enstrophy transport equation to this configuration. The spot leading-edge speed (83.34% of free stream) and calmed-region speed (25.86%) are compared with Schubauer-Klebanoff and Gostelow, which is the right kind of external check. The numerics are DRP compact schemes with prior validation; nine cases at 14,000 core hours each is a real computational effort.\n\nSoft spots, in rough order of severity.\n\nFirst, the profile-loss conclusion is not just unproven, it looks backwards on the paper's own terms. Eq. (13) says the mixed-out loss term, proportional to trailing-edge momentum thickness, contributes about 90% of total loss. Immediately after, the paper reports that momentum thickness at the trailing edge increases by 39.2% from awake=0.1 to 0.9 (Fig. 7, Table 2). No ζp is computed, and no compensating change in Cpb or δ_TE is shown. So the abstract's 'improve profile loss' is unsupported by the displayed quantity. This is the main reason I can't accept the paper as is.\n\nSecond, the 50% skin-friction reduction is reported without saying where Cf is integrated, without error bars, and without grid- or time-convergence checks. As stated, it is a number, not a result.\n\nThird, the 2D assumption is acknowledged—vortex stretching is absent—and that is honest, but it means turbulent spots and calmed regions are being represented by 2D dynamics. The qualitative trends might survive in 3D, but the quantitative claims need verification.\n\nFourth, no code or data is shipped; 'available on request' limits reproducibility.\n\nThe paper's internal trends—delayed separation, earlier reattachment, smaller net separated region, lower TKE production—are coherent and consistent across figures and tables. The CETE terms are diagnostic, not fitted, and the wake parameters come from the literature, so circularity is not an issue.\n\nWho should read it: people working on LPT wake-induced transition, especially those interested in 2D parametric sensitivity and enstrophy budgets. It deserves a serious referee, but with a major-revision request: compute ζp directly, define the Cf reduction properly, add convergence checks, and either justify the 2D premise or soften the applied claims.\n\nRecommendation: engage with it in review, but do not let the profile-loss claim through without direct computation.","headline":"A workmanlike 2D CFD sweep of wake amplitude on a T106A cascade, internally consistent on separation trends but carrying an applied profile-loss claim that its own momentum-thickness data appear to contradict.","tokens_in":20776,"tokens_out":3073,"would_cite":false,"duration_ms":32120,"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":"Increasing the amplitude of periodic Gaussian wakes entering a T106A low-pressure turbine delays suction-surface separation and cuts skin friction by about 50%.","keywords":["wake-induced transition","low-pressure turbine","T106A cascade","Gaussian wake amplitude","separation suppression","skin friction","compressible enstrophy","calmed region"],"falsifier":"Repeat the same nine Gaussian-wake cases in a spanwise-periodic three-dimensional simulation, or in a linear cascade experiment with pulsed wakes, and compare the time-averaged skin-friction coefficient along the suction surface: if the roughly 50% drop between $a_{wake}=0.1$ and $a_{wake}=0.9$, or the 23.3% shortening of the trailing-edge separation bubble, is not reproduced, the paper's central claim is falsified.","tokens_in":2034,"feed_emoji":"🌀","tokens_out":6666,"duration_ms":162554,"temperature":0.7,"pith_summary":"The paper sets out to show that the strength of periodic Gaussian wakes entering a T106A low-pressure turbine cascade controls how the suction-surface boundary layer transitions. In a series of two-dimensional compressible simulations, raising the wake amplitude from 0.1 to 0.9 delays separation, moves transition and reattachment upstream, and shrinks the net separated-flow region, cutting the time-averaged skin-friction coefficient by roughly 50 percent. The periodic wakes produce puffs, streaks, turbulent spots, and 'calmed regions' behind the spots, and those calmed regions are presented as the mechanism that suppresses separation at the low Reynolds numbers typical of low-pressure turbines, where profile loss is the largest efficiency penalty. The paper explains the rotational-energy side of the process through a compressible enstrophy transport equation, in which viscous-stress and baroclinic terms dominate. If the claim holds, wake amplitude becomes a practical lever for reducing profile loss on high-lift blading and, through it, engine fuel burn.","feed_headline":"Wake amplitude cuts T106A drag by 50%","feed_subtitle":"Larger periodic wakes delay separation and shrink the separated region on the suction surface, lowering profile loss.","key_machinery":"The control parameter is the amplitude $a_{wake}$ of a periodic Gaussian wake superimposed on the inlet streamwise velocity, $u(w)=a_{wake}\\exp[-\\beta(\\mathrm{mod}(y/y_{max}+2t/t_{wake}-1,2))^2]$, with $t_{wake}=0.35$ and $\\beta=19$; nine amplitudes from 0.1 to 0.9 are compared. The numerical engine is an implicit large-eddy simulation of the two-dimensional compressible Navier-Stokes equations using dispersion-relation-preserving compact schemes and an optimized Runge-Kutta time integrator. The diagnostic that carries the energy argument is the compressible enstrophy transport equation (CETE), which splits the rate of change of enstrophy, a measure of rotational energy, into vortex stretching, compressibility, baroclinic, bulk-viscosity, and viscous terms; in two dimensions the stretching term is absent, so the remaining balance shows what the paper identifies as the dominant roles of baroclinicity and viscous stress. This decomposition, together with a turbulent kinetic energy production budget, is what connects the wake amplitude to separation suppression and profile loss.","core_discovery":"On the suction surface of the T106A blade, increasing the nondimensional Gaussian wake amplitude $a_{wake}$ from 0.1 to 0.9 changes the time-averaged boundary layer in a consistent direction: the leading-edge separation bubble disappears for amplitudes above 0.4, the trailing-edge separation bubble is delayed and shortened by 23.3 percent in streamwise extent, reattachment moves upstream, and the time-averaged skin-friction coefficient falls by roughly 50 percent. At the same time the trailing-edge momentum thickness grows by 39.2 percent and the maximum unsteady separation-bubble half-height grows by 37 percent, so the wake does not simply thin the boundary layer. Enstrophy space-time maps show the wake-induced transition sequence: longitudinal puffs (compact turbulent patches) stretch into streaks, break into turbulent spots, and leave calmed regions behind them; the spot leading edge convects at about 83 percent of the free-stream speed and the calmed region at about 26 percent, both close to classic measurements. Turbulent kinetic energy production decreases with amplitude and its peak moves toward the wall, and the compressible enstrophy budget is dominated first by the viscous-stress term and second by the baroclinic term, with the baroclinic share increasing at higher amplitudes. The paper reads these trends as evidence that stronger wakes suppress separated flow and improve profile loss at the low Reynolds numbers relevant to low-pressure turbines.","pith_inferences":["The 50% drag reduction is a two-dimensional result; because the paper itself notes that vortex stretching is absent in 2D and turbulent spots are intrinsically three-dimensional, I would not transfer the factor to engine blading until a spanwise-periodic 3D run at the same Reynolds number reproduces the trend.","If wake amplitude is viewed as a control knob, the opposing trends in skin friction and momentum thickness imply a design trade-off: there may be an intermediate amplitude that minimizes total profile loss rather than the largest one.","The CETE results hint at a reduced-order route: separation suppression at low Reynolds numbers could be predicted from a baroclinic-versus-viscous enstrophy balance without resolving the full broadband turbulence spectrum.","Because the paper finds that wake amplitude mirrors the effects of Mach number and free-stream turbulence but more strongly, a combined scaling in amplitude, Mach number, and turbulence intensity might collapse separation-onset data on the T106A surface into one curve."],"forward_implications":["If the 2D result represents the real flow, increasing upstream wake amplitude is a viable separation-control input: the suction surface sees about half the skin friction at $a_{wake}=0.9$ compared with $a_{wake}=0.1$.","Calmed regions behind turbulent spots are the physical mechanism that periodically suppresses the separation bubble, so wake-passing design can be optimized for low-Reynolds-number operation rather than treated only as a disturbance source.","The measured spot and calmed-region convection speeds (about 83% and 26% of the free-stream speed) give quantitative anchors for transition models in turbomachinery codes.","Because trailing-edge momentum thickness increases by 39.2% while skin friction drops by 50%, loss estimates must track boundary-layer growth and drag separately; a wake that reduces drag can still thicken the downstream boundary layer.","The CETE term ordering (viscous first, baroclinic second) is roughly independent of amplitude, suggesting that a single enstrophy-based diagnostic can characterize wake-induced transition across a range of wake strengths."],"supporting_citations":[{"why":"Supplies the baseline DNS of a separating T106A cascade with incoming wakes and the flow-configuration comparison used here.","marker":"[15]"},{"why":"Provides the experimental T106A benchmark data that fix the computational domain, Reynolds number, and validation target.","marker":"[25]"},{"why":"Supplies the unsteady-wake parameters (spacing constant, passing period) used in the Gaussian inflow model.","marker":"[11]"},{"why":"Prior T106A cascade study used to validate the solver and to compare the effect of changing Mach number against wake amplitude.","marker":"[26]"},{"why":"Derives the compressible enstrophy transport equation that the paper applies to the budget of rotational energy.","marker":"[24]"},{"why":"Describes puffs, streaks, and turbulent spots in wake-induced transition, the structural features read from the enstrophy maps.","marker":"[23]"},{"why":"Provides the classic turbulent-spot convection-speed measurement that the paper matches to its enstrophy-derived spot speed.","marker":"[43]"},{"why":"Provides the classic calmed-region convection-speed measurement that the paper matches to its 26% value.","marker":"[45]"},{"why":"Introduces the negative-jet effect used to explain how wakes carry turbulent fluid to the boundary-layer edge.","marker":"[12]"}],"fun_headline_variants":["Wake amplitude halves T106A skin friction drag","Larger wakes delay T106A separation, cut drag 50%","Stronger wakes suppress T106A separation, reduce drag","Wake amplitude growth cuts T106A drag and delays separation","Higher wake amplitude cuts T106A drag by half"],"cache_read_input_tokens":22784,"weakest_assumption_plain":"The central claim rests on the two-dimensional simulations being a faithful stand-in for a real three-dimensional blade flow; the paper itself notes that vortex stretching is absent in 2D and offers no three-dimensional or grid-convergence check, so if spanwise motions change the separation and reattachment response, the 50-percent drag reduction need not carry over to the engine.","fun_headline_variants_meta":{"raw":{"variants":["Wake amplitude halves T106A skin friction drag","Larger wakes delay T106A separation, cut drag 50%","Stronger wakes suppress T106A separation, reduce drag","Wake amplitude growth cuts T106A drag and delays separation","Higher wake amplitude cuts T106A drag by half"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.000659,"raw_usage":{"total_tokens":3069,"prompt_tokens":1058,"completion_tokens":2011,"prompt_tokens_details":{"cached_tokens":384},"prompt_cache_hit_tokens":384,"prompt_cache_miss_tokens":674,"completion_tokens_details":{"reasoning_tokens":1930}},"tokens_in":674,"tokens_out":2011,"duration_ms":16472,"temperature":1.0,"reasoning_tokens":1930,"cache_read_input_tokens":384,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-12T17:45:31.157303+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"Repeat the same nine Gaussian-wake cases in a spanwise-periodic three-dimensional simulation, or in a linear cascade experiment with pulsed wakes, and compare the time-averaged skin-friction coefficient along the suction surface: if the roughly 50% drop between $a_{wake}=0.1$ and $a_{wake}=0.9$, or the 23.3% shortening of the trailing-edge separation bubble, is not reproduced, the paper's central claim is falsified.","supporting_citations":[{"cited_title":"Wissink, DNS of separating, low Reynolds number flow in a turbine cascade with incoming wakes, International Journal of Heat and Fluid Flow 24 (4) (2003) 626–635","cited_arxiv_id":null,"evidence_quote":"Supplies the baseline DNS of a separating T106A cascade with incoming wakes and the flow-configuration comparison used here."},{"cited_title":"Stadtm¨ uller, L","cited_arxiv_id":null,"evidence_quote":"Provides the experimental T106A benchmark data that fix the computational domain, Reynolds number, and validation target."},{"cited_title":"Karaca, A","cited_arxiv_id":null,"evidence_quote":"Supplies the unsteady-wake parameters (spacing constant, passing period) used in the Gaussian inflow model."},{"cited_title":"Sengupta, P","cited_arxiv_id":null,"evidence_quote":"Prior T106A cascade study used to validate the solver and to compare the effect of changing Mach number against wake amplitude."},{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"Derives the compressible enstrophy transport equation that the paper applies to the budget of rotational energy."},{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"Describes puffs, streaks, and turbulent spots in wake-induced transition, the structural features read from the enstrophy maps."},{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"Provides the classic turbulent-spot convection-speed measurement that the paper matches to its enstrophy-derived spot speed."},{"cited_title":"Gostelow, G","cited_arxiv_id":null,"evidence_quote":"Provides the classic calmed-region convection-speed measurement that the paper matches to its 26% value."},{"cited_title":"Addison, H","cited_arxiv_id":null,"evidence_quote":"Introduces the negative-jet effect used to explain how wakes carry turbulent fluid to the boundary-layer edge."}],"review_version":1}