{"id":"c0a4574e-cfba-4e26-8c23-d6e9b5a70bc2","arxiv_id":"2411.18283","paper_version":1,"verdict":"CONDITIONAL","confidence":"MODERATE","novelty_score":6.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":5,"one_line_summary":"A crosstalk aperture between two choked-flow channels exhibits self-sustained whistling whose frequency scales with the Strouhal number based on aperture length.","lead":"This paper shows that the small gap between neighboring combustion chambers in gas turbines can whistle loudly on its own, even without flames or acoustic reflections. The whistle's frequency follows a simple Strouhal scaling with gap length, which could help engineers design turbine interfaces that suppress these damaging vibrations.","discovery_kind":"new_application","skeptic_critique":{"model":"deepseek-v4-flash","headline":"The intrinsic-mode claim is not fully secured: the choked vane throat is a strong reflector, and its upstream reflection coefficient is never measured, so a local acoustic feedback path is not excluded.","rationale":"The reader's conditional verdict rests on the same weakest assumption I identify: the anechoic terminations are necessary but not sufficient because the choked vane throat is an internal reflector. The paper's own introduction cites [38] for choked-nozzle sound reflection, yet no upstream-side characterization of the vane-throat reflection is reported and no control experiment removes it. The experimental observations are convincing and the LES reproducing two of three cases is real support, but the mechanism label 'intrinsic' is stronger than 'self-sustained in a rig with anechoic horns.' A local feedback path through the choked vane must be excluded before accepting the no-feedback interpretation. The d=7.2 mm LES/experiment mismatch is a secondary issue and does not weaken the experimental central observation. Since this concern is the same as the reader's, and the current evidence does not resolve it, the conditional verdict should stand until the proposed test is run.","tokens_in":16726,"tokens_out":8316,"duration_ms":85627,"concrete_test":"Run paired LES for the d=7.2 mm misaligned rectangular-edge case: (i) the baseline configuration with choked vanes, and (ii) an otherwise identical domain in which the choked vane row is replaced by a non-reflecting NSCBC outlet placed immediately downstream of the aperture while the channel mass flow and upstream pressure are held fixed. If the self-sustained oscillation disappears or its frequency and amplitude change substantially in case (ii), the choked-vane reflection participates in the feedback loop and the intrinsic-instability claim is not established for the present geometry.","verdict_should_be":"UNCHANGED","load_bearing_attack":"Section 2.3 validates the anechoic character of the upstream and downstream horns above 600 Hz, and all whistling peaks lie above 600 Hz. However, the microphones used for the downstream characterization (x5–x8) are located downstream of the vanes; they characterize the outlet horn, not the reflection presented to upstream-traveling waves by the choked vane passage. The introduction explicitly notes that the sonic region at the vanes reflects sound [38]. Thus the inference in Section 3.1 that 'no longitudinal acoustic mode participate' rules out global duct modes between anechoic terminations, but it does not rule out a short local loop: shear-layer oscillation at the aperture radiates downstream, reflects at the choked vane throat, and returns to seed the shear layer. The aperture is immediately upstream of the vanes, so this path is geometrically plausible. The Strouhal scaling and the mass-flow independence are consistent with a hydrodynamic mode, but they do not discriminate between a purely intrinsic instability and an edge-tone-type feedback involving the vane throat. If the choked-vane reflection is essential, the central claim that the whistling results from an intrinsic aeroacoustic mode would need qualification.","agreement_with_reader":"agree"},"referee_report":{"model":"deepseek-v4-flash","summary":"This paper reports an experimental and numerical study of self-sustained whistling at the crosstalk aperture between two choked-flow channels that model the combustor-to-turbine interface of a can-annular gas turbine. The rig has nominally anechoic upstream and downstream terminations, and the aperture geometry is varied in terms of axial length, edge shape, and vane alignment. The authors observe intense narrow-band acoustic peaks for several configurations, with a peak frequency that decreases as the aperture length increases, and they interpret the phenomenon as an intrinsic aeroacoustic instability governed by Strouhal scaling of the shear layer in the aperture. Compressible LES reproduces the whistling for two of the three simulated cases, and the paper uses the simulations to localize the sound source via the fluctuating Lamb vector. The claimed practical relevance is that such crosstalk whistling could be misattributed to thermoacoustic flame feedback in real engines.","tokens_in":17008,"tokens_out":4169,"duration_ms":46897,"significance":"If the intrinsic-instability interpretation is correct, the result is significant: it would demonstrate that a purely hydrodynamic shear-layer feedback at the aperture can produce high-amplitude self-sustained oscillations without a global acoustic mode, and it would establish a new mechanism relevant to can-annular combustor design. The experimental dataset is valuable: it systematically varies aperture length, edge geometry, and vane alignment, and it uses PDF shape as a stability indicator. The LES reproduces the main trends and provides a plausible source region, and the paper is honest about the one discrepant LES case. The main weakness is that the evidence does not fully exclude a local acoustic feedback path involving the choked vane throat, so the central claim needs additional support before the paper can be accepted as written.","major_comments":[{"comment":"The conclusion that the whistling is an intrinsic aeroacoustic instability is not fully secured. The anechoic character of the upstream and downstream horns, characterized above 600 Hz, excludes longitudinal modes spanning the entire rig, but it does not exclude a short local feedback loop: shear-layer oscillation at the aperture radiates downstream, reflects from the choked vane throat a few millimeters away, and returns to seed the shear layer. The microphones x5-x8 characterize the downstream horn from the downstream side of the vanes; they do not measure the upstream-looking reflection coefficient of the choked vane passage. The introduction itself cites [38] to state that choked nozzles reflect sound. To support the central claim, the authors should either measure or model the reflection presented to upstream-traveling waves by the vane passage, or provide a direct test such as an LES with the vane throat artificially made anechoic, showing that the oscillation persists without that reflection.","section":"Section 3.1 / Section 2.3"},{"comment":"The Strouhal numbers 0.62, 0.51, 0.55, and 0.41 are extracted from the measured peak frequencies at a single flow velocity and then used to draw the dashed scaling lines; calling this 'Strouhal scaling enables prediction' overstates the evidence, because the constants are calibrated to the same data. The data do clearly show that the whistling frequency decreases with aperture length, and a constant-St collapse is a reasonable summary, but it should be presented as a fit rather than a predictive test. Please report the fitted constants with uncertainties and, if predictive power is claimed, validate against an independent condition such as a different flow velocity or a different vane radius.","section":"Section 3.1, Eq. (2) and Fig. 3c"},{"comment":"The LES predicts self-sustained oscillations for d = 7.2 mm, while the experiment shows a stable Gaussian PDF for the same geometry; the paper attributes this to proximity of the instability threshold without quantitative support. This case is also one of the two used in Section 4.3 for the Lamb-vector source analysis. The discrepancy weakens the claim that the LES 'remarkably reproduce the whistling phenomenon.' Please provide a quantitative argument for the threshold-sensitivity interpretation, for example a linear growth-rate estimate, a sensitivity study to small geometric perturbations, or a stability map, so that the mismatch is not left as an ad hoc explanation.","section":"Section 4.2, d = 7.2 mm"}],"minor_comments":[{"comment":"The sentence 'In Fig. 2a, one can see a sharp peak for all crosstalk opening' appears to refer to Fig. 3a, the acoustic spectra, not Fig. 2a, the rig overview.","section":"Section 3.1, first paragraph"},{"comment":"The text states 'an alternance of upward (red) and downward (red) flow deflection'; the two colors should be different, likely 'downward (blue)', or the figure color scheme should be described accurately.","section":"Section 4.3, Fig. 8 discussion"},{"comment":"The isentropic choked-mass-flow equation is typeset with an ambiguous exponent; please use standard display math so that the exponent (γ+1)/(2(γ-1)) is clear.","section":"Eq. (1)"},{"comment":"The sentence 'an increase in the mass flow rate is accompanied with an of the mean pressure' is incomplete; it should read 'accompanied by an increase of the mean pressure.'","section":"Appendix"},{"comment":"The statement that 'most of the acoustic energy travelling towards the terminations is absorbed' is qualitative; please report the numerical range of |R| above 600 Hz so the anechoic criterion is quantitative and reproducible.","section":"Section 2.3"}],"recommendation":"major_revision","confidential_remarks":"This is a solid experimental and numerical study of an industrially relevant phenomenon, and the dataset will be useful regardless of how the intrinsic-instability question is resolved. The main risk is over-interpretation of the anechoic terminations: the upstream and downstream horn measurements do not rule out a short acoustic feedback loop with the choked vane throat, which is geometrically very close to the aperture. I would be satisfied if the authors add a quantitative argument or a targeted measurement/simulation that isolates the role of the vane reflection, and if they reframe the Strouhal collapse as a fit rather than a prediction. No concerns about novelty disclosure; funding is acknowledged."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"Plainly: this is the first clean experimental look at aeroacoustic whistling in the crosstalk gap immediately upstream of choked turbine vanes, and the geometric trends are real. But the 'intrinsic aeroacoustic instability' headline is argued more strongly than the measurements support. The anechoic horns kill duct modes, but the paper never measures the reflection coefficient presented to upstream-traveling waves by the choked vane passage—and the introduction itself cites the sonic region as a strong reflector. With the aperture right at the vane throat, a short local feedback loop (shear-layer radiation downstream, reflection at the throat, seeding back at the aperture) is geometrically plausible and is not excluded by the Strouhal scaling or mass-flow independence. That is the soft spot in the central claim.\n\nWhat is genuinely good: the experiment is careful and modular; the PSDs show sharp peaks, the frequency decreases with aperture length, and the edge-shape and vane-alignment effects are robust. The Gaussian-to-bimodal PDF transition is a nice way to display a bifurcation, and the LES reproducing two of three cases with <10% frequency error is meaningful. The Lamb-vector analysis is a reasonable way to localize the sound source. The parametric coverage—seven aperture lengths, two edge shapes, two alignments—is a real contribution for design guidance.\n\nThe scaling lines are fitted, not predicted: four Strouhal numbers drawn through the same measured peaks. Calling that 'prediction' is a stretch, but the paper's main message does not depend on those exact values. The d=7.2 mm LES/experiment mismatch is confessed and blamed on threshold sensitivity; fair, but unquantified. Minor: no error bars on the peak frequencies.\n\nThe literature coverage looks solid, with the relevant can-annular and grazing-flow work cited; I do not see self-citation abuse.\n\nWho this is for: combustion and aeroacoustics researchers, especially people working on can-annular gas turbine instabilities. It deserves a serious referee. My recommendation: send it out, but ask the authors to either measure or model the choked-vane reflection and to soften the intrinsic claim if they cannot, and to clarify that the Strouhal constants are fitted from the data.","headline":"A careful first experimental map of crosstalk-aperture whistling, but the 'intrinsic instability' label outruns the evidence because the choked-vane reflection path is never closed.","tokens_in":17520,"tokens_out":2036,"would_cite":true,"duration_ms":18949,"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":"Crosstalk gaps between combustor cans can whistle on their own, with no acoustic feedback from the duct ends.","keywords":["can-annular combustors","crosstalk aperture","aeroacoustic instabilities","intrinsic instability","Strouhal scaling","shear-layer self-oscillation","choked turbine vanes","large eddy simulation"],"falsifier":"A decisive check would be to measure the upstream reflection coefficient at the exact whistling frequency for each geometry; if it is not small at those frequencies, then the claim that longitudinal modes play no role is unsupported. Alternatively, replacing the anechoic terminations with highly reflective walls and observing a frequency shift to a duct resonance would show that boundary feedback participates after all.","tokens_in":16548,"feed_emoji":"🎵","tokens_out":7244,"duration_ms":63225,"temperature":0.7,"pith_summary":"This paper reports that the narrow crosstalk opening between two neighbouring combustor channels can whistle loudly on its own, with no flame and no acoustic feedback from the duct ends. Using a laboratory rig whose upstream and downstream terminations are anechoic above 600 Hz, the authors observe self-sustained oscillations whose frequency falls with aperture length and collapses onto a constant Strouhal number. They interpret this as an intrinsic aeroacoustic instability: the shear layer spanning the aperture oscillates and radiates plane waves upstream, without a longitudinal acoustic mode locking the frequency. Compressible large-eddy simulations reproduce the whistling and locate the sound source in the fluctuating Lamb vector inside the aperture. The result matters for can-annular gas turbines because such gaps sit between every pair of cans, and self-oscillation there can be as damaging as thermoacoustic vibration.","feed_headline":"Combustor crosstalk gaps whistle with no acoustic feedback","feed_subtitle":"If true, it explains damaging vibrations in gas turbines without any flame coupling.","key_machinery":"The load-bearing object is the crosstalk aperture itself: a small gap between two square channels, bounded upstream by the splitting-plate trailing edge and downstream by choked turbine vanes, with low-Mach grazing flow on both sides. Its behaviour is governed by a Strouhal number based on the aperture length, $St = f(d + r_v/2)/U$, where $d$ is the gap distance, $r_v$ the vane leading-edge radius, and $U$ the channel velocity; the collapse of all whistling peaks onto one constant Strouhal number is the evidence that a convective shear-layer instability sets the frequency. The other essential elements are the anechoic terminations, verified by multi-microphone measurement of reflection coefficients, which exclude longitudinal acoustic feedback, and the fluctuating Lamb vector $\\omega \\times u$, whose projection on the acoustic velocity identifies the region of sound production in the aperture.","core_discovery":"On the paper's own terms, the central discovery is that intense whistling at the crosstalk aperture is an intrinsic aeroacoustic instability, not an acoustically coupled duct mode. Because the inlet and outlet terminations absorb sound above 600 Hz, the observed peaks at 2–5 kHz cannot be maintained by reflection from the rig boundaries; instead, the hydrodynamic shear layer across the aperture self-oscillates and radiates plane acoustic waves upstream in both channels. The oscillation frequency obeys $St = f(d + r_v/2)/U$ with $St \\approx 0.62$ for rectangular and $0.51$ for round upstream edges in the misaligned vane configuration, and the probability density of the band-passed pressure changes from Gaussian to bimodal as the aperture grows, signalling a transition to a self-oscillating limit cycle. The same instability is reproduced in compressible large-eddy simulations, where the source region is identified through the fluctuating Lamb vector.","pith_inferences":["If the instability is truly intrinsic, then any two neighbouring cans whose outlet gap meets the same Strouhal and geometry conditions should whistle even with perfectly non-reflecting boundaries, so engine tests with reflective boundaries may under- or over-estimate the amplitude depending on phase.","A testable extension is to vary the vane leading-edge radius $r_v$ and check that the whistling frequency shifts through the same Strouhal relation, which would confirm that the relevant length scale is $d + r_v/2$ rather than $d$ alone.","The same mechanism might apply to other two-sided grazing-flow apertures with a downstream choked passage, such as bleed-air systems or inter-stage gaps, wherever a thin shear layer spans a narrow opening.","The simulations suggest that turbulent perturbations visibly modulate the amplitude of the self-oscillation, so a stochastic model of the unstable shear-layer mode could predict the observed pressure statistics without resolving the full turbulence."],"forward_implications":["Whistling frequency at a given geometry can be predicted from the Strouhal number once the channel velocity is known.","The onset of whistling can be suppressed or delayed by changing the shape of the aperture's upstream edge, rectangular versus round, and by aligning the splitting plate with a vane.","In a real can-annular combustor, acoustic reflections from burners and flames would add to the intrinsic source, potentially producing a stronger aero-thermo-acoustic limit cycle than the pure aeroacoustic one.","The Gaussian-to-bimodal probability-density transition gives a robust experimental marker for detecting the instability threshold.","Because the whistling is intrinsic, it can occur in non-reactive flow and may be misattributed to flame-acoustic coupling in engines."],"supporting_citations":[{"why":"Supplies the linear-stability framework for the aeroacoustic response of crosstalk apertures, which motivates the present experimental and numerical investigation.","marker":"[27]"},{"why":"Provides the vortex-sound theory, via the Lamb vector and its projection on the acoustic velocity, used here to locate the sound-production region in the aperture.","marker":"[28]"},{"why":"Gives the canonical experimental baseline for grazing-flow aperture acoustics that the present choked-vane geometry extends.","marker":"[36]"},{"why":"Documents sound reflection and production by choked nozzle flows, used to justify the distinct acoustic boundary conditions around the choked vanes.","marker":"[38]"},{"why":"Provides the multi-microphone method used to characterise the anechoic terminations and to reconstruct the acoustic pressure fields in the channels.","marker":"[39]"},{"why":"Supplies the Aeolian-tone precedent for flow-instability tones that scale with Strouhal number without acoustic feedback from boundaries.","marker":"[40]"},{"why":"Establishes the probability-density criterion for distinguishing linearly stable oscillators from self-oscillating ones, applied here to detect the whistling threshold.","marker":"[45]"},{"why":"Introduces and reviews the concept of intrinsic thermoacoustic instabilities, the analogue that the paper extends to the aeroacoustic case.","marker":"[42]"}],"fun_headline_variants":["Crosstalk gap whistles from shear-layer self-oscillation","No acoustic feedback needed: gap whistles on its own","Intrinsic gap instability sets gas-turbine crosstalk whistling","Self-oscillating shear layer makes combustor gaps whistle","Crosstalk aperture whistles without duct modes"],"cache_read_input_tokens":3200,"weakest_assumption_plain":"The load-bearing premise is that the terminations are truly anechoic at all whistling frequencies and that the choked vane passage itself provides no hidden acoustic reflection path back to the aperture.","fun_headline_variants_meta":{"raw":{"variants":["Crosstalk gap whistles from shear-layer self-oscillation","No acoustic feedback needed: gap whistles on its own","Intrinsic gap instability sets gas-turbine crosstalk whistling","Self-oscillating shear layer makes combustor gaps whistle","Crosstalk aperture whistles without duct modes"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.000171,"raw_usage":{"total_tokens":1315,"prompt_tokens":1031,"completion_tokens":284,"prompt_tokens_details":{"cached_tokens":384},"prompt_cache_hit_tokens":384,"prompt_cache_miss_tokens":647,"completion_tokens_details":{"reasoning_tokens":199}},"tokens_in":647,"tokens_out":284,"duration_ms":3107,"temperature":1.0,"reasoning_tokens":199,"cache_read_input_tokens":384,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-12T11:20:26.276151+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"A decisive check would be to measure the upstream reflection coefficient at the exact whistling frequency for each geometry; if it is not small at those frequencies, then the claim that longitudinal modes play no role is unsupported. Alternatively, replacing the anechoic terminations with highly reflective walls and observing a frequency shift to a duct resonance would show that boundary feedback participates after all.","supporting_citations":[{"cited_title":"Pedergnana, N","cited_arxiv_id":null,"evidence_quote":"Supplies the linear-stability framework for the aeroacoustic response of crosstalk apertures, which motivates the present experimental and numerical investigation."},{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"Provides the vortex-sound theory, via the Lamb vector and its projection on the acoustic velocity, used here to locate the sound-production region in the aperture."},{"cited_title":"Kooijman, A","cited_arxiv_id":null,"evidence_quote":"Gives the canonical experimental baseline for grazing-flow aperture acoustics that the present choked-vane geometry extends."},{"cited_title":"Weilenmann, N","cited_arxiv_id":null,"evidence_quote":"Documents sound reflection and production by choked nozzle flows, used to justify the distinct acoustic boundary conditions around the choked vanes."},{"cited_title":"Schuermans, W","cited_arxiv_id":null,"evidence_quote":"Provides the multi-microphone method used to characterise the anechoic terminations and to reconstruct the acoustic pressure fields in the channels."},{"cited_title":"Etkin, G","cited_arxiv_id":null,"evidence_quote":"Supplies the Aeolian-tone precedent for flow-instability tones that scale with Strouhal number without acoustic feedback from boundaries."}],"review_version":1}