{"id":"5a4e161a-7587-4364-949f-6cb59a8a2e66","arxiv_id":"1908.02027","paper_version":1,"verdict":"CONDITIONAL","confidence":"MODERATE","novelty_score":7.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":2,"one_line_summary":"Large streamwise or spanwise slip on channel walls introduces three-dimensional leading instabilities, and spanwise slip dramatically lowers the critical Reynolds number, while isotropic slip strongly stabilizes.","lead":"This paper shows that anisotropic velocity slip on channel walls, with large slip in either the streamwise or spanwise direction, triggers three-dimensional instabilities that earlier studies missed. It also finds that spanwise slip can lower the critical Reynolds number by an order of magnitude, while equal slip in both directions strongly stabilizes the flow.","discovery_kind":"extension","skeptic_critique":{"model":"deepseek-v4-flash","headline":"The η boundary condition in §II.E is the load-bearing unverified step; all reported 3-D critical Reynolds numbers depend on it, and no independent 3-D validation or convergence check is given.","rationale":"The reader's weakest_assumption is exactly the missing η boundary condition, and I agree. I considered whether this is merely a presentational gap rather than a substantive risk. It is substantive because the no-slip and 2-D validation cannot constrain the coupled 3-D boundary condition, and because the most dramatic claim (spanwise slip lowers Recr below 400) has no independent support. The paper is otherwise careful: the base flow is consistently normalized by fixed bulk speed, the no-slip non-modal benchmark matches Reddy & Henningson within tolerance, 2-D critical values match Ghosh et al. after unit conversion, and the physical discussion of lift-up versus Orr mechanisms is internally consistent. Those strengths do not, however, cover the 3-D modes. I therefore keep the reader's CONDITIONAL verdict; the requested derivation and an independent 3-D eigenvalue check would settle the issue. If the check fails, the central claim would have to be rejected; if it passes, the paper's novel instability results stand.","tokens_in":12911,"tokens_out":8707,"duration_ms":87750,"concrete_test":"Re-derive the wall-normal-vorticity boundary condition from (2), (9), and (10), and implement it in a primitive-variable Chebyshev-collocation eigenvalue solver (u,v,w,p) for the same linearized equations. Compute the most unstable eigenvalues for representative cases: (λ_z=0.2, Re=394, α≈0.6, β≈1.27) and (λ_x=0.2, Re=6280, α≈0.56, β≈1.06). If the primitive-variable solver does not reproduce a growing mode at these parameters, the reported critical Reynolds numbers are artifacts. As a second check, rerun the present velocity–vorticity code at N=80, 96, and 128 Chebyshev points in y for these cases; unstable eigenvalues that shift substantially or disappear indicate insufficient resolution or a spurious boundary treatment.","verdict_should_be":"UNCHANGED","load_bearing_attack":"The central claim—that sufficiently large streamwise or spanwise slip triggers 3-D leading instabilities—rests entirely on the velocity–vorticity formulation in §II.E. The paper solves for u_y (fourth order) and η (second order) and must impose the Navier slip condition (2) on u_x and u_z. Using (9)–(10), the slip conditions become a system of four boundary equations at y=±1 that couple η, η_y, u_yy, and u_y (with u_y=0 giving the remaining two conditions). This derivation and its final form are not shown. If these boundary equations are mis-derived or mis-implemented, the eigenvalues in Figs. 4, 5, 8, and 9 are not the eigenvalues of the original slip problem, and all Recr values, including Recr≈394 for λ_z=0.2, could be numerical artifacts. The only validations reported (Fig. 2) are for no-slip transient growth and for 2-D (β=0) streamwise-slip modes; neither exercises the 3-D coupled η boundary condition. No grid-convergence study or independent check of any 3-D eigenmode is provided, so the novel part of the paper is exactly the part that is unsupported.","agreement_with_reader":"agree"},"referee_report":{"model":"deepseek-v4-flash","summary":"The paper revisits the linear modal and non-modal stability of plane channel flow subject to Navier slip at the walls, treating streamwise and spanwise slip as two separate limiting cases of anisotropic slip and also considering isotropic slip. Using a velocity-vorticity formulation for modal stability and an adjoint-based time-stepper for transient growth, the authors report that sufficiently large streamwise slip (λx above about 0.008) or spanwise slip (λz above about 0.02) makes three-dimensional modes the leading unstable modes, in contrast to the two-dimensional leading modes of no-slip channel flow. They report that the critical Reynolds number is only mildly increased by streamwise slip (Recr about 5900 at λx=0.05 and 6280 at λx=0.2) but is drastically reduced by spanwise slip (Recr about 394 at λz=0.2 and 336 at λz=0.25), while equal isotropic slip strongly stabilizes the flow. They also report that streamwise slip suppresses transient growth, spanwise slip enhances it and changes the optimal perturbation structure, and isotropic slip yields transient growth dominated by the streamwise-slip effect.","tokens_in":13160,"tokens_out":4989,"duration_ms":57113,"significance":"If the reported three-dimensional instabilities are correct, the paper provides a substantive correction to the earlier conclusion, attributed to Lauga & Cossu (2005) and Min & Kim (2005), that velocity slip suppresses linear instability: that conclusion was based on two-dimensional modes, and this paper shows that sufficiently large anisotropic slip can make oblique/three-dimensional modes unstable at lower Reynolds numbers, even below the no-slip critical value for spanwise slip. The non-modal results, especially the shift of the optimal perturbation from streamwise-invariant rolls to finite-small-α oblique structures under spanwise slip, are also physically interesting and falsifiable. The paper's strengths include two explicit benchmark validations (no-slip transient growth against Reddy & Henningson (1993), and two-dimensional streamwise-slip critical Reynolds numbers against Ghosh et al. (2014)), a systematically explored parameter range, and internally consistent smooth trends in the reported eigenvalue data.","major_comments":[{"comment":"The boundary condition for the wall-normal vorticity η is the load-bearing step for all three-dimensional modal results, but neither its derivation nor its final form is shown. The paper states that the η boundary condition 'can be derived using the slip boundary condition (2)' and that 'four boundary conditions coupling η and uy' result, yet the explicit conditions at y=±1 are absent. Since Eqs. (9)–(10) express ux and uz in terms of η and ∂uy/∂y, imposing Eq. (2) on ux and uz yields boundary relations involving η, ∂η/∂y, and second derivatives of uy; the reader cannot check the signs, the wall-normal orientation, or the handling of the (α²+β²)-¹ factor. If this derivation is incorrect, the eigenvalues in Figs. 4, 5, 8, and 9, and therefore every reported Recr value for three-dimensional modes, are not eigenvalues of the stated physical problem. The authors must present the full derivation and the final boundary conditions.","section":"§II.E, Eqs. (7)–(10)"},{"comment":"No validation or convergence check exercises the three-dimensional coupled η–uy boundary condition. The two reported validations are a no-slip transient-growth calculation using the primitive-variable time-stepper and two-dimensional (β=0) streamwise-slip critical Reynolds numbers; neither involves the coupling of η with uy at nonzero β that is central to the new three-dimensional instability results. Moreover, no grid-convergence study is reported for any of the eigenvalue computations shown in Figs. 4, 5, 8, or 9. I request an explicit N-convergence test for a representative unstable three-dimensional mode, e.g., the λz=0.2, Re=394, (α,β)=(0.6,1.27) case, together with an independent check against direct temporal integration of the primitive-variable linearized equations (3) for the same parameters.","section":"§III.A, Fig. 2; §III.B–C"},{"comment":"The claim that three-dimensional instabilities 'disappear' under isotropic slip is supported too thinly. The text reports that for λx=λz=0.05 the first instability occurs at Re≃1.85×10⁵ and that the most unstable mode is still two-dimensional, but it does not state the range of (α,β) searched, the eigenvalue algorithm's convergence at such high Reynolds numbers, or how the absence of unstable three-dimensional modes was established. Because this negative result is used to contrast anisotropic and isotropic slip, it needs the same level of documentation as the positive instability searches.","section":"§III.D"}],"minor_comments":[{"comment":"The sentence stating that subcritical transition 'can occur way below this Reynolds number at about Re=6601' is self-contradictory, since 6601 is above the no-slip critical value 5772; this appears to be a typo and should be corrected.","section":"§I"},{"comment":"There are numerous typographical errors that should be fixed in a revision, including 'Cartisian' (Cartesian), 'exerpiments' (experiments), 'boundanry' (boundary), 'stuctures' (structures), and 'steamwise' (streamwise).","section":"Throughout"},{"comment":"The caption should identify clearly which curve is the reference data from Ghosh et al. and which points are the present results; the current text says the reference values are plotted as a solid line but does not state the symbol/line style for 'our results'.","section":"Fig. 2(b) caption"},{"comment":"The claim that the adjoint system uses 'the same boundary condition (2)' would benefit from a one-line justification, since the adjoint Robin condition is not automatic for every non-self-adjoint formulation; a short derivation or citation would remove ambiguity.","section":"§II.B"}],"recommendation":"major_revision","confidential_remarks":"The core physical claims are interesting and, if verified, would warrant publication in a fluid-dynamics journal. I see no circularity: slip lengths are prescribed inputs and the benchmark validations are external. The main gate is the missing η boundary-condition derivation and the lack of any independent three-dimensional validation; these are fixable within the manuscript's scope, so I recommend major revision rather than rejection. The authors should also be asked to provide the isotropic-slip search details, as the negative three-dimensional result is currently asserted rather than demonstrated."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"Dear colleague,\n\nRead this. It revisits slip channel flow stability and reports that large anisotropic slip triggers 3-D leading instabilities: spanwise slip can push Recr down to ~394, streamwise slip only modestly changes it, and isotropic slip strongly stabilizes. If correct, this corrects the earlier conclusion that slip suppresses instability.\n\nWhat's good: the parameter sweep is new and the contrast between anisotropic and isotropic slip is sharp. The no-slip transient growth matches Reddy & Henningson, and the 2-D streamwise-slip critical Re matches Ghosh et al. So the two validation benchmarks are solid. The base-flow modification for streamwise slip and the invariances for spanwise slip are clearly explained.\n\nSoft spots. The main one is in §II.E. The boundary condition for η is derived from the Navier slip condition, but the derivation and final expressions are not shown. The text only says it couples η and u_y. All the new 3-D results—Figs. 4, 5, 8, 9 and the Recr values—depend on that coupled boundary condition. The validations in Fig. 2 don't exercise it: no-slip transient growth and β=0 modes are exactly the cases where the η coupling is trivial or absent. There is no independent check of any 3-D eigenmode and no grid-convergence study. Given that the paper's whole novel claim depends on that one unshown derivation, this is a genuine gap, not a style complaint. I don't see evidence the result is wrong—the trends look smooth and the physics is plausible—but the paper as written doesn't let the reader verify the load-bearing step.\n\nMinor: the claim that streamwise slip slightly destabilizes in 0.07<λx<0.11 is based on Recr slightly below 5772; close to a benchmark, so it should be checked with the same scrutiny.\n\nVerdict: this deserves a serious referee. The missing derivation and the lack of 3-D validation are fixable, and if the authors provide them the paper is a solid step forward. I'd send it to review rather than desk reject, and I'd ask specifically for the η boundary conditions and an independent eigenvalue check (or at least a convergence study) before acceptance.\n\nCheers.","headline":"The 3-D instability results are interesting and likely right, but the paper omits the derivation of the one boundary condition everything new depends on, so it needs a serious referee and an independent check.","tokens_in":13642,"tokens_out":2413,"would_cite":false,"duration_ms":24923,"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":"With sufficiently large slip, both streamwise and spanwise wall slip switch the leading instability of channel flow to three-dimensional modes; spanwise slip can reduce the critical Reynolds number to about 336, while streamwise slip…","keywords":["slip channel flow","linear stability","Navier slip boundary condition","three-dimensional instability","critical Reynolds number","non-modal transient growth","anisotropic slip","velocity-vorticity formulation"],"falsifier":"Run an independent primitive-variable linear stability calculation (solving the linearized Navier–Stokes equations directly with the Navier slip condition, without the velocity–vorticity $\\eta$ boundary condition) for $\\lambda_z=0.2$ at $Re=400$: if no mode with streamwise wavenumber $\\alpha\\simeq0.6$ and spanwise wavenumber $\\beta\\simeq1.27$ is unstable, the paper's $\\eta$ boundary condition is incorrect; if the mode is reproduced, the claim is confirmed.","tokens_in":12721,"feed_emoji":"🌊","tokens_out":10448,"duration_ms":87714,"temperature":0.7,"pith_summary":"This paper revisits the linear and non-modal stability of pressure-driven channel flow with tangential velocity slip at the walls, extending earlier analyses that focused on small slip and on spanwise-invariant two-dimensional modes. It claims that once the slip length is large enough, both streamwise slip ($\\lambda_x$) and spanwise slip ($\\lambda_z$) make three-dimensional oblique modes the leading instabilities, and the two directions act oppositely: streamwise slip only modestly raises the critical Reynolds number (about 5900 at $\\lambda_x=0.05$ and 6280 at $\\lambda_x=0.2$, with a small slip range where it falls below the no-slip value 5772), whereas spanwise slip lowers it dramatically (about 394 at $\\lambda_z=0.2$ and 336 at $\\lambda_z=0.25$). Streamwise slip suppresses non-modal transient growth, while spanwise slip enlarges it and shifts the optimal perturbation from streamwise rolls to tilted, long-wavelength structures. If correct, these results mean that anisotropic slip—rather than slip itself—can trigger earlier instability and stronger transient growth, which matters for flows over superhydrophobic surfaces where effective slip lengths can be large.","feed_headline":"Spanwise slip drops channel-flow instability threshold to Re≈336","feed_subtitle":"Anisotropic wall slip can trigger instabilities far below the no-slip onset at Re=5772.","key_machinery":"The central object that carries the argument is the velocity–vorticity form of the linearized incompressible Navier–Stokes equations, in which the wall-normal velocity $u_y$ satisfies a fourth-order Orr–Sommerfeld-type equation and the wall-normal vorticity $\\eta=\\partial u_x/\\partial z-\\partial u_z/\\partial x$ satisfies a second-order equation, coupled at the walls through the Navier slip condition. This formulation lets the authors build a spectral Fourier–Chebyshev eigenvalue problem and scan the four-parameter space $(Re,\\alpha,\\beta,\\lambda)$ to locate the first unstable mode. For non-modal growth, they use an adjoint-based time-stepping method in primitive variables, iterated with a Krylov subspace solver, which gives the maximum transient growth $G(t)$ and the optimal perturbation. The key conceptual distinction is two-dimensional ($\\beta=0$) versus three-dimensional modes, because the earlier conclusion that slip stabilizes the flow came from considering only the two-dimensional modes.","core_discovery":"The central discovery is that the linear stability of slip channel flow is controlled by the direction of slip. Using the velocity–vorticity formulation of the linearized Navier–Stokes equations with the Navier slip boundary condition at both walls, the paper computes eigenvalues over the streamwise and spanwise wavenumbers $(\\alpha,\\beta)$ for a wider range of slip lengths than earlier work. For streamwise slip ($\\lambda_x$), two-dimensional modes ($\\beta=0$) do become more stable as $\\lambda_x$ grows, but above $\\lambda_x\\simeq 0.008$ three-dimensional modes are the most dangerous, so the critical Reynolds number stays near the no-slip value $Re_{cr}\\simeq 5772$, reaching about 6280 at $\\lambda_x=0.2$ and even dipping below 5772 for $0.07\\lesssim\\lambda_x\\lesssim0.11$. For spanwise slip ($\\lambda_z$), which does not alter the parabolic base flow, three-dimensional instabilities appear once $\\lambda_z$ exceeds about 0.02 and the critical Reynolds number falls sharply, to about 394 at $\\lambda_z=0.2$ and 336 at $\\lambda_z=0.25$. Equal slip in both directions removes these three-dimensional instabilities and raises the critical Reynolds number to about $1.85\\times10^5$ at $\\lambda=0.05$. The paper also reports that spanwise slip enlarges the non-modal transient growth and moves the optimal perturbations to small finite streamwise wavenumbers, producing tilted streak structures instead of streamwise rolls.","pith_inferences":["If the critical Reynolds number for spanwise slip is as low as roughly 336, then in millimeter-scale channels with effective slip lengths on the order of 0.2 times the half-gap, linear instability should be observable at flow speeds far below the usual transition threshold; this is a testable experimental prediction with superhydrophobic surfaces that slip only in the spanwise direction.","The finding that spanwise slip destabilizes while leaving the base flow unchanged suggests the instability is produced by the modified wall condition acting on fluctuations rather than by reduced shear; an energy-budget or resolvent analysis of the three-dimensional eigenmodes could identify the production term responsible.","The same tilted three-dimensional modes likely connect the single-phase slip-channel results to oblique-wave instabilities reported in earlier two-fluid slippery channel studies; the paper's mode shapes could be compared directly with those earlier eigenfunctions to test whether the mechanism is shared."],"forward_implications":["Three-dimensional modes, not spanwise-invariant Tollmien–Schlichting waves, set the stability threshold for slip lengths above $\\lambda_x\\simeq0.008$ or $\\lambda_z\\simeq0.02$; any stability analysis that restricts to two-dimensional modes will overestimate the stabilizing effect of slip.","Spanwise slip can make a channel linearly unstable at Reynolds numbers an order of magnitude below the no-slip threshold—$Re_{cr}\\simeq394$ at $\\lambda_z=0.2$—so flows previously assumed linearly stable in this range can be unstable.","Streamwise slip does not strongly stabilize the flow: $Re_{cr}$ stays around 5900–6300 for $\\lambda_x$ up to 0.2, with a small window of $\\lambda_x$ where the flow is slightly less stable than no-slip flow.","Isotropic slip with equal streamwise and spanwise lengths suppresses the three-dimensional instabilities and greatly raises the critical Reynolds number, so the destabilizing effect is specifically due to anisotropy rather than to slip itself.","Spanwise slip changes the shape of the most amplified transient perturbations, replacing streamwise rolls with tilted long-wavelength structures, while streamwise slip reduces transient growth without changing the dominant structure."],"supporting_citations":[{"why":"Earlier modal stability analysis of slip channel flow limited to two-dimensional modes; the paper extends this by showing three-dimensional modes set the threshold.","marker":"21"},{"why":"Earlier study of streamwise and spanwise slip effects on transition and non-normal growth; the paper revisits it with larger slip and reports different optimal structures.","marker":"22"},{"why":"Earlier linear stability analysis of two-fluid flow in a slippery channel; the converted critical Reynolds numbers for two-dimensional modes validate the eigenvalue solver.","marker":"24"},{"why":"Supplies the no-slip transient-growth benchmark values used to validate the adjoint-based non-modal method.","marker":"3"},{"why":"Describes the adjoint-based direct optimal-growth time-stepper used for the non-modal calculations.","marker":"29"},{"why":"Provides the projection scheme used in the time-stepping discretization of the linearized and adjoint equations.","marker":"31"}],"fun_headline_variants":["Spanwise slip drops channel-flow instability onset to Re 336","Wall slip direction decides channel flow stability threshold","Anisotropic slip triggers 3D instabilities at Re 336","Equal slip in both directions stabilizes channel flow","Spanwise slip: channel flow instability at Re 336 vs 5772 no-slip"],"cache_read_input_tokens":3200,"weakest_assumption_plain":"The three-dimensional results rest on the paper's unshown derivation of the boundary condition that couples the wall-normal vorticity $\\eta$ to derivatives of the wall-normal velocity $u_y$ at the channel walls; if that boundary condition or its numerical implementation is wrong, the reported three-dimensional eigenvalues and critical Reynolds numbers would be invalid.","fun_headline_variants_meta":{"raw":{"variants":["Spanwise slip drops channel-flow instability onset to Re 336","Wall slip direction decides channel flow stability threshold","Anisotropic slip triggers 3D instabilities at Re 336","Equal slip in both directions stabilizes channel flow","Spanwise slip: channel flow instability at Re 336 vs 5772 no-slip"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.001127,"raw_usage":{"total_tokens":4769,"prompt_tokens":1113,"completion_tokens":3656,"prompt_tokens_details":{"cached_tokens":384},"prompt_cache_hit_tokens":384,"prompt_cache_miss_tokens":729,"completion_tokens_details":{"reasoning_tokens":3569}},"tokens_in":729,"tokens_out":3656,"duration_ms":66324,"temperature":1.0,"reasoning_tokens":3569,"cache_read_input_tokens":384,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-14T14:56:01.084085+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"Run an independent primitive-variable linear stability calculation (solving the linearized Navier–Stokes equations directly with the Navier slip condition, without the velocity–vorticity $\\eta$ boundary condition) for $\\lambda_z=0.2$ at $Re=400$: if no mode with streamwise wavenumber $\\alpha\\simeq0.6$ and spanwise wavenumber $\\beta\\simeq1.27$ is unstable, the paper's $\\eta$ boundary condition is incorrect; if the mode is reproduced, the claim is confirmed.","supporting_citations":[{"cited_title":"Kumar, S","cited_arxiv_id":null,"evidence_quote":"Earlier modal stability analysis of slip channel flow limited to two-dimensional modes; the paper extends this by showing three-dimensional modes set the threshold."},{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"Earlier study of streamwise and spanwise slip effects on transition and non-normal growth; the paper revisits it with larger slip and reports different optimal structures."},{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"Earlier linear stability analysis of two-fluid flow in a slippery channel; the converted critical Reynolds numbers for two-dimensional modes validate the eigenvalue solver."},{"cited_title":"Ghosh, R","cited_arxiv_id":null,"evidence_quote":"Describes the adjoint-based direct optimal-growth time-stepper used for the non-modal calculations."},{"cited_title":"Chattopadhyay, R","cited_arxiv_id":null,"evidence_quote":"Provides the projection scheme used in the time-stepping discretization of the linearized and adjoint equations."}],"review_version":1}