{"id":"60043f7b-832f-4e7c-bae5-aaa946a344a7","arxiv_id":"2411.10052","paper_version":3,"verdict":"CONDITIONAL","confidence":"MODERATE","novelty_score":6.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":2,"one_line_summary":"The 3D Prandtl equations with the special structure v=Ku admit unique stable solutions on arbitrarily long time intervals when the initial data are small monotone perturbations of a shear profile.","lead":"This paper proves long-time existence, uniqueness, and stability of solutions to the three-dimensional Prandtl boundary layer equations when the tangential velocities satisfy a special proportional structure (v = K u). It extends the two-dimensional long-time theory to 3D under a monotonicity assumption and small perturbations of a shear flow.","discovery_kind":"extension","skeptic_critique":{"model":"deepseek-v4-flash","headline":"Assumption H forces K≡const: nonconstant global smooth solutions of ∂xK+K∂yK=0 do not exist, so the '3D special structure' reduces to 2D Prandtl and the claimed 3D extension is not supported.","rationale":"The reader identified the structural assumption v=Ku as load-bearing, and that is the right area, but the sharpest objection is stronger than 'not physically justified': under the paper's own global regularity hypothesis, the only admissible K is constant. The proof of this is short and unconditional: differentiating the stationary Burgers equation shows that any nonzero ∂yK produces finite-time blow-up of the gradient in one of the two x-directions, contradicting K∈W^{m+1,∞}(R²). With K constant, the system becomes a parameterized 2D Prandtl equation, so the theorem's advertised '3D' content collapses to a coordinate change of the 2D result. I emphasize that I am not claiming the energy estimates are wrong; for K≡c the estimates may well be correct and the theorem true in that restricted class. But the central claim as stated—long-time well-posedness for 3D Prandtl with a special structure—is not supported, because the structure has no nonconstant global smooth representatives. The paper's local predecessors could legitimately use nonconstant K because they only needed local-in-x existence; the global Sobolev setting of Theorem 1.1 removes that possibility. A conditional verdict is appropriate: the authors should either exhibit a nonconstant admissible K or explicitly reframe the theorem as a 2D result with a parameter and adjust the title, abstract, and novelty claims accordingly.","tokens_in":51368,"tokens_out":10417,"duration_ms":122882,"concrete_test":"Check whether any nonconstant K∈W^{m+1,∞}(R²) solves ∂xK+K∂yK=0 by computing the characteristic equation for q=∂yK: q(x)=q0/(1+q0(x−x0)); global boundedness in both x directions forces q0=0, hence K constant. If this is confirmed, verify the reduction of (1.5) with K≡c to the 2D Prandtl system in variables (x, y−cx, z); if the reduction holds, amend the statement to a 2D theorem with a parameter and remove the 3D novelty claim, or provide an explicit nonconstant admissible K on an unbounded domain satisfying the stated W^{m+1,∞} regularity.","verdict_should_be":"CONDITIONAL","load_bearing_attack":"The central claim depends on condition H admitting genuinely three-dimensional flows, but H is essentially empty. Since K∈W^{m+1,∞}(R²) and ∂xK+K∂yK=0, differentiate in y: with q=∂yK, q satisfies ∂xq+K∂yq+q²=0. Along characteristics, q(x)=q(x0)/(1+q(x0)(x−x0)). If q(x0)>0, q blows up at the finite value x0−1/q(x0); if q(x0)<0, it blows up at x0+1/|q(x0)|. Global boundedness on all of R² therefore forces q≡0, so K is independent of y, and the equation then gives K constant. Thus every admissible K is a constant c. In that case (1.5) becomes ∂tu+u∂xu+cu∂yu+w∂zu=∂zzu and ∂xu+c∂yu+∂zw=0, which in the variables (x, y−cx, z) is exactly the 2D Prandtl equation with the transverse variable as a parameter. Consequently Theorem 1.1 does not establish long-time well-posedness for a genuinely three-dimensional structure; it recovers the Xu–Zhang 2D result under a coordinate substitution. The local results [26,36] can tolerate nonconstant K because only local-in-x existence is needed, whereas the global R² Sobolev framework of Theorem 1.1 cannot. The title and abstract's claim of extending local 3D results to long time therefore overstates what is proved.","agreement_with_reader":"partial"},"referee_report":{"model":"deepseek-v4-flash","summary":"The paper claims long-time well-posedness for the 3D Prandtl system (1.5) under the monotonicity condition and a special structural assumption v=Ku, where K satisfies the Burgers equation ∂xK+K∂yK=0 in R² with U=1 (condition H). The proof regularizes the system, derives weighted Sobolev estimates for the vorticity, introduces linearly-good unknowns g_n, and then asserts existence, uniqueness, and stability for small perturbations of a monotone shear flow on an arbitrary time interval [0,T]. The abstract further claims a lifespan of order e^T for initial data of size e^{-T}.","tokens_in":51773,"tokens_out":5627,"duration_ms":65628,"significance":"If condition H admitted genuinely three-dimensional nonconstant functions K, the result would be a significant extension of the 2D long-time theory of Xu and Zhang to 3D. The manuscript contains substantial technical work: detailed compatibility conditions, boundary reductions in Appendix B, weighted energy estimates for the regularized vorticity, and the formal g_n transformation in Appendix C. However, the central structural assumption is much weaker than advertised: every admissible K under the stated global regularity is constant, so the system reduces by a coordinate change to the 2D Prandtl equation with the transverse variable as a parameter. The claimed new 3D content is therefore not established. In addition, the two lemmas that carry the stability and uniqueness proof are only outlined, leaving a central part of the proof unverifiable. The paper is honest about these omissions but they are load-bearing.","major_comments":[{"comment":"Condition H forces K to be constant, so the 'special structure' does not yield genuinely three-dimensional flows. Indeed, differentiating ∂xK+K∂yK=0 in y gives ∂xq+K∂yq+q²=0 for q=∂yK. Along the characteristics dx/ds=1, dy/ds=K, one has q(s)=q(0)/(1+q(0)s). Since K∈W^{m+1,∞}(R²) is globally bounded, q must vanish identically; otherwise q blows up in finite time either forward (if q(0)<0) or backward (if q(0)>0). Hence K is independent of y, and then ∂xK=0, so K(x,y)=c. Substituting K=c into (1.5) and setting X=x, Y=y−cx gives ∂tu+u∂Xu+w∂zu=∂zzu and ∂Xu+∂zw=0, which is exactly the 2D Prandtl system with Y as a parameter. Thus Theorem 1.1 is a parameter-dependent 2D result, not a long-time well-posedness result for a genuinely three-dimensional structure as claimed in the title and abstract.","section":"Section 1, condition H, and Theorem 1.1"},{"comment":"The stability and uniqueness part of Theorem 1.1 rests on Lemma 6.1 and Lemma 6.2, but neither is actually proved. Lemma 6.1 is labeled 'Outline of Proof' and ends with 'we will omit some details here', while Lemma 6.2 says its proof 'can be recovered by the standard process as what we did in Lemma 4.3'. Since these estimates control the difference of two solutions and are the only mechanism proving uniqueness, this is a load-bearing gap. The brief discussion after (6.4) does not show how the difficult terms, especially those involving ∂yK and the loss of tangential derivatives, are closed; a complete proof is required.","section":"Section 6, Lemmas 6.1 and 6.2"},{"comment":"The abstract states that the lifespan can be extended to any large T provided the initial perturbation has size e^{-T}, but Theorem 1.1 contains no such quantification: it only asserts the existence of some δ0 for a fixed T, with no explicit dependence of δ0 on T. The proof in Section 5 does not derive the e^{-T} relation. Moreover, the limiting passage from approximate solutions to solutions is sketched: after obtaining convergence in C^0([0,T1];C^{2,δ'}_{loc}), the manuscript asserts the convergence of w and the validity of the nonlinear system in (5.6)-(5.7) without a careful justification of the uniform estimates needed to pass to the limit in the terms w∂z(us+ũ) and ∂y(K(us+ũ)). These points need to be made precise for the existence claim to be verified.","section":"Abstract and Section 5"}],"minor_comments":[{"comment":"The statement has mismatched notation: the left side is ∥g_m(0)∥_{H^2_{\\ell'}} while the right side is C∥\\tilde w_0∥_{H^{m+2}_{k+\\ell''}}, and the proof uses ∂^n_{xy}φ_0. Please correct the norms, indices, and variables to match the definition of g_n.","section":"Lemma 4.2"},{"comment":"The displayed regularity of w, 'L∞(R_+,z); (H^m(R²)H^{-1}(R_x) ∪ H^m(R²)H^{-1}(R_y))', seems to be a typographical artifact; the conclusion should be stated cleanly as w∈L∞([0,T1];L∞(R_+;H^{m-1}(R²))).","section":"Section 5, Eq. (5.6)"},{"comment":"The hypotheses refer to 'compatibility conditions (2.3)-(2.4)', but (2.3) is the shear-flow decay bound from Proposition 2.1; the intended references are the compatibility conditions (2.4)-(2.5).","section":"Theorem 5.1"},{"comment":"There are numerous typos, including 'well-posdness' in the running header, 'initial date' instead of 'initial data', 'local-posedness', 'by virture', 'recurence', and an extra '+' in the inequality before (4.10). The paper would benefit from a careful proofreading pass.","section":"General notation and typos"},{"comment":"In the display following (4.8), the inequality contains a double plus sign before the ε terms; this should be cleaned up so that the dissipation terms are written correctly.","section":"Section 4, Lemma 4.3"}],"recommendation":"reject","confidential_remarks":"The central advertised novelty collapses because condition H admits only constant K under the global W^{m+1,∞} assumption; after the coordinate change (x,y−cx,z), the theorem is a parameter-dependent version of the 2D result [42]. The title and abstract therefore overstate the contribution. Separately, the stability proof is incomplete because Lemmas 6.1 and 6.2 are only sketched. These are load-bearing issues that cannot be fixed by local revision within the scope of the paper. I would recommend rejection, though the detailed boundary reductions in Appendix B may be useful groundwork for future work."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"Plainly: the advertised 3D result is not there. The structural assumption H, with K solving ∂xK + K∂yK = 0 and K ∈ W^{m+1,∞}(R²), forces K to be constant. Differentiate in y: q = ∂yK satisfies ∂xq + K∂yq + q² = 0, so along characteristics q(x) = q(0)/(1 + q(0)x). On all of R², boundedness forces q ≡ 0, and then K ≡ const. Thus system (1.5) is just the 2D Prandtl equation with y − cx as a parameter. The stress-test note is correct: Theorem 1.1 proves long-time well-posedness for a family of 2D problems, not for a genuinely 3D structure. The local results [26,36] can tolerate nonconstant K because they only need local-in-x existence; the global R² framework here cannot.\n\nCredit where it is due: this is a serious technical effort. The weighted-Sobolev energy method, the boundary reduction with the ϵ-corrector, and the formal transformation to g_n are carefully worked out, and the estimates plausibly close. If the authors reframed the result as long-time well-posedness for 2D Prandtl with a constant transverse drift, the core analysis would be a legitimate extension of Xu–Zhang. The citation pattern looks honest.\n\nSoft spots beyond the structural issue: Lemma 6.1 and Lemma 6.2, which underpin stability and uniqueness, are only sketched with “standard process” for a load-bearing step. The abstract promises an e^{-T} lifespan that Theorem 1.1 does not state. The limiting argument in Section 5 is compressed, and the reader's soundness score of 5 reflects that. These are real but secondary; the degeneracy of condition H is primary.\n\nBottom line: I would not cite this as a 3D result, and the title/abstract overstate. But the technical machinery deserves a referee's time, and the flaw is subtle enough that a serious editor should not desk-reject. Send it to review with a note to check the Burgers degeneracy. The authors either need to find a nonconstant admissible class of K (say, local in x, or with controlled growth but not bounded on all of R²) or honestly revise the framing to the parameter-dependent 2D theorem.","headline":"The special structure v=Ku degenerates: Burgers equation forces K constant, so the theorem is a 2D result with a parameter, not a genuine 3D extension.","tokens_in":52189,"tokens_out":3707,"would_cite":false,"duration_ms":39638,"reading_group":"maybe","serious_thinker":"yes","would_accept_peer_review":true},"rs_alignment":null,"lean_confirmation":null,"pith_extraction":{"msc":["35M13","35Q35","76D10","76D03","76N20"],"pacs":[],"model":"deepseek-v4-flash","headline":"The 3D Prandtl boundary layer equations are solvable on any time interval [0,T] under a monotonicity condition plus a special secondary-flow structure, provided the initial perturbation around a shear flow is exponentially small.","keywords":["3D Prandtl equations","long-time well-posedness","energy method","special structure","monotonic shear flow","weighted Sobolev space","parabolic regularization","lifespan"],"falsifier":"Take monotone initial data that are small around the same shear profile but violate the structural assumption, so that $\\partial_z(v/u)$ is not identically zero; the linearized system for $(\\partial_z u, \\partial_z v)$ then contains the terms $(\\partial_z v)\\partial_y u - (\\partial_z u)\\partial_y v$ that the paper's assumption removes. If an energy calculation or numerical experiment shows such data also stay smooth for arbitrarily long times, the structural assumption is unnecessary and the theorem is not sharp; if exponentially growing modes appear instead — as the 3D ill-posedness analysis this paper cites indicates — then the assumption is genuinely load-bearing. A second, softer check is whether the lifespan bound $T \\sim \\ln(1/\\delta)$ holds quantitatively by comparing maximal existence times for data of size $\\delta$ and $2\\delta$.","tokens_in":51215,"feed_emoji":"🌊","tokens_out":14354,"duration_ms":118446,"temperature":0.7,"pith_summary":"This paper claims that the three-dimensional Prandtl boundary layer equations — the viscous-flow equations describing the thin layer of fluid near a solid wall — admit unique smooth solutions on any prescribed time interval $[0,T]$, not merely on a short local existence time. The hypotheses are the standard monotonicity condition $\\partial_z u > 0$ together with a special structural assumption: the second tangential velocity is a fixed multiple of the first, $v = K u$ (equivalently $\\partial_z(v/u) \\equiv 0$), where the multiplier $K(x,y)$ solves the two-dimensional Burgers equation $\\partial_x K + K\\partial_y K = 0$. Under these assumptions the full 3D system collapses to a single scalar equation, and an energy method in polynomially weighted Sobolev spaces closes; the price is that the initial perturbation around a monotone shear flow must be exponentially small, of size $e^{-T}$. A sympathetic reader would care because in general 3D the Prandtl system is ill-posed in Sobolev spaces even under monotonicity, so this is the first construction showing that the long-time behaviour known in 2D survives in three dimensions within a natural solution class.","feed_headline":"3D Prandtl layer solved for arbitrarily long times","feed_subtitle":"New energy estimates take the 2D long-time theory into 3D: any horizon T is reachable if data shrink like e^(−T).","key_machinery":"The central mechanism is the linearly-good unknown $g_n = \\big(\\partial^n_{xy}\\tilde{u}/(u^s_z + \\tilde{u}_z)\\big)_z$, the vertical derivative of the ratio of a tangential derivative of the perturbation to the total shear rate; this is the 3D analogue of the quantity that cancels the loss of one tangential derivative in the 2D theory. Around it, the paper sets up a parabolic regularization (adding $\\epsilon(\\partial_x^2 + \\partial_y^2)$ and a corrector), derives the vorticity equation (3.8) for $\\varphi = \\partial_z \\tilde{u}$, and then performs the formal transformation (4.1) so that the highest-order tangential-derivative estimates on $\\varphi$ become weighted $L^2$ estimates on the $g_n$'s that are independent of $\\epsilon$. Three further ingredients carry the argument: the structural assumption $v = Ku$ with $\\partial_x K + K\\partial_y K = 0$, which collapses the 3D system to the scalar equation (1.5); the shear-flow profile $u^s(t,z)$ whose monotone gradient decays polynomially ($\\partial_z u^s \\sim \\langle z\\rangle^{-k}$), supplying the coercivity of the weighted norms; and a reconstruction argument that computes the high-order boundary values of the approximate solutions, which are needed because integration by parts in $z$ produces boundary terms that the equations alone do not determine.","core_discovery":"On the paper's own terms, the central claim is Theorem 1.1: for the reduced system (1.5), obtained from the 3D Prandtl equations by imposing the structure $v = Ku$, with the outer flow normalized to $U = 1$ and $K$ a time-independent solution of the Burgers equation, any initial datum $u_0 = u_0^s + \\tilde{u}_0$ that is a sufficiently small perturbation of a monotone shear profile satisfying the decay conditions (1.6) and the compatibility conditions admits a unique solution $(u,w)$ on $[0,T]$ with $u - u^s \\in L^\\infty([0,T]; H^m_{k+\\nu-\\delta'})$ and $w \\in L^\\infty([0,T]; H^\\infty(\\mathbb{R}_+; H^{m-1}(\\mathbb{R}^2)))$, and the solution is stable with respect to the initial data. The quantitative heart is the lifespan statement: any large $T$ can be reached provided the initial perturbation is smaller than $e^{-T}$, the exact 3D analogue of the 2D long-time theorem this paper extends. The proof works because the structural assumption removes the secondary-flow terms that make the general 3D problem ill-posed in Sobolev spaces even under monotonicity, leaving a single scalar equation whose remaining derivative loss is controlled through a new linearly-good unknown and a boundary-data reconstruction that supplies the missing high-order boundary conditions.","pith_inferences":["The structural assumption $\\partial_z(v/u) \\equiv 0$ means the ratio of the two tangential velocities is a function of $(t,x,y)$ alone, so the solution class is essentially a family of tilted two-dimensional boundary layers; a natural relaxation the paper does not treat is $v = K(t,x,y)u$ with slowly time-varying $K$, where the proof's use of $\\partial_t K = 0$ and the Burgers equation would need ","The lifespan law $T \\sim \\ln(1/\\delta)$ is stated through the smallness condition of size $e^{-T}$, but the paper does not ask whether it is optimal; a numerical study in the 2D model, where the method originated, could test whether larger or more structured data buy a longer lifespan.","Because the outer flow is normalized to $U = 1$, the theorem avoids the pressure term entirely; extending the argument to non-uniform outer flows ($\\partial_x P \\neq 0$) would be the direct test of how much of the structure is truly needed."],"forward_implications":["Any prescribed horizon $T$ becomes reachable: taking the initial perturbation smaller than $e^{-T}$ yields a solution on all of $[0,T]$, so the local existence theorem of the predecessors is truly extended in time rather than merely re-proved.","Stability with continuity in the data follows: two initial data differing by a small amount produce solutions that stay close in $H^{m-3}_{k+\\nu-\\delta'}$, and uniqueness of the solution for each datum is the case of identical data.","The 3D result now stands on the same Sobolev footing as the 2D long-time theory, with the same polynomial weight structure and the same loss $\\delta'$ in the decay exponent, so the two-dimensional barrier that previously stopped at local time is removed in the special-structure class.","The normal velocity inherits full horizontal regularity ($w \\in L^\\infty([0,T]; H^\\infty(\\mathbb{R}_+; H^{m-1}(\\mathbb{R}^2)))$), so the reconstructed $w$ is not merely a byproduct but part of the well-posedness statement."],"supporting_citations":[{"why":"Supplies the 2D long-time well-posedness method — shear-flow profile estimates, the formal transformation, and the polynomial-weight energy scheme — that the paper adapts to three dimensions.","marker":"[42]"},{"why":"Introduces the special structure $v=Ku$ for the 3D Prandtl equations and proves local well-posedness; the present theorem extends that local result to arbitrarily long times.","marker":"[26]"},{"why":"The other local well-posedness result for 3D Prandtl with the same structure, whose lifespan the paper pushes to any large $T$ under exponentially small data.","marker":"[36]"},{"why":"Establishes that general 3D Prandtl equations are ill-posed in Sobolev spaces even under monotonicity, the fact that makes the structural assumption necessary for the energy method.","marker":"[25]"},{"why":"Supplies the cancellation mechanism in the convection terms under monotonicity that the weighted energy estimates rely on at the top derivative levels.","marker":"[29]"},{"why":"Foundational local well-posedness in weighted Sobolev spaces whose weighted-space setup and direct-estimate approach the proof inherits.","marker":"[1]"}],"fun_headline_variants":["3D Prandtl long-time stability achieved via special structure","Arbitrarily long horizons for monotone 3D Prandtl solutions","3D Prandtl lifespan blown up to any T with e^-T data","Special 3D Prandtl flows: well-posed for any time horizon"],"cache_read_input_tokens":3200,"weakest_assumption_plain":"The proof rests on the structural assumption that the second tangential velocity is a fixed multiple of the first ($v = Ku$, with $K$ time-independent and solving the Burgers equation, and outer flow $U = 1$), which reduces the 3D system to a single scalar equation; nothing in the paper justifies this structure physically, and without it the general 3D problem is known to be ill-posed in Sobolev spaces even under monotonicity.","fun_headline_variants_meta":{"raw":{"variants":["3D Prandtl long-time stability achieved via special structure","Arbitrarily long horizons for monotone 3D Prandtl solutions","3D Prandtl lifespan blown up to any T with e^-T data","Special 3D Prandtl flows: well-posed for any time horizon"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.000233,"raw_usage":{"total_tokens":1562,"prompt_tokens":1084,"completion_tokens":478,"prompt_tokens_details":{"cached_tokens":384},"prompt_cache_hit_tokens":384,"prompt_cache_miss_tokens":700,"completion_tokens_details":{"reasoning_tokens":395}},"tokens_in":700,"tokens_out":478,"duration_ms":5043,"temperature":1.0,"reasoning_tokens":395,"cache_read_input_tokens":384,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-12T20:00:49.838608+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"Take monotone initial data that are small around the same shear profile but violate the structural assumption, so that $\\partial_z(v/u)$ is not identically zero; the linearized system for $(\\partial_z u, \\partial_z v)$ then contains the terms $(\\partial_z v)\\partial_y u - (\\partial_z u)\\partial_y v$ that the paper's assumption removes. If an energy calculation or numerical experiment shows such data also stay smooth for arbitrarily long times, the structural assumption is unnecessary and the theorem is not sharp; if exponentially growing modes appear instead — as the 3D ill-posedness analysis this paper cites indicates — then the assumption is genuinely load-bearing. A second, softer check is whether the lifespan bound $T \\sim \\ln(1/\\delta)$ holds quantitatively by comparing maximal existence times for data of size $\\delta$ and $2\\delta$.","supporting_citations":[{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"Supplies the 2D long-time well-posedness method — shear-flow profile estimates, the formal transformation, and the polynomial-weight energy scheme — that the paper adapts to three dimensions."},{"cited_title":"Liu, Y.G","cited_arxiv_id":null,"evidence_quote":"Introduces the special structure $v=Ku$ for the 3D Prandtl equations and proves local well-posedness; the present theorem extends that local result to arbitrarily long times."},{"cited_title":"Qin, X.Q","cited_arxiv_id":null,"evidence_quote":"The other local well-posedness result for 3D Prandtl with the same structure, whose lifespan the paper pushes to any large $T$ under exponentially small data."},{"cited_title":"Liu, Y.G","cited_arxiv_id":null,"evidence_quote":"Establishes that general 3D Prandtl equations are ill-posed in Sobolev spaces even under monotonicity, the fact that makes the structural assumption necessary for the energy method."},{"cited_title":"Masmoudi, T.K","cited_arxiv_id":null,"evidence_quote":"Supplies the cancellation mechanism in the convection terms under monotonicity that the weighted energy estimates rely on at the top derivative levels."},{"cited_title":"Alexandre, Y","cited_arxiv_id":null,"evidence_quote":"Foundational local well-posedness in weighted Sobolev spaces whose weighted-space setup and direct-estimate approach the proof inherits."}],"review_version":1}