{"id":"db5487b3-26c2-4b41-b00f-8383a2656459","arxiv_id":"2602.07407","paper_version":4,"verdict":"CONDITIONAL","confidence":"MODERATE","novelty_score":6.0,"correctness_risk":"low","formal_verification":"none","parameter_count":0,"one_line_summary":"Annular equilibria of constant-vorticity Euler flows bifurcate into non-annular admissible domains at specific vorticity values, with rigidity and Neumann-stability results for three classes of free-boundary problems.","lead":"This mathematics paper proves that steady two-dimensional fluid flows with constant vorticity inside ring-shaped regions can sit in many non-circular shapes, not just the standard annulus. It also proves when only circles are possible and that circular flows survive small changes in boundary data.","discovery_kind":"new_application","skeptic_critique":{"model":"deepseek-v4-flash","headline":"Theorem 1.5's proof misidentifies the cokernel; the transversality check tests the wrong functional, so Crandall–Rabinowitz condition (H2) is not established as written.","rationale":"The Reader's weakest assumption was the Fréchet differentiability of the nonlinear operators, imported from shape calculus rather than proved. That is a reasonable concern but is standard in the shape-derivative literature and likely patchable. A more concrete, textually verifiable issue is the transversality check in the proof of Theorem 1.5: the range of the linearized operator at the bifurcation point is mischaracterized, leading to an incorrect test for Crandall–Rabinowitz condition (H2). The theorem may still be true — the correct first-component test appears numerically nonzero — but the proof as written is incomplete. Since this is a fixable gap rather than evidence of a false result, the Reader's CONDITIONAL verdict remains appropriate; however, the authors should be asked to correct the range/transversality verification. This does not undermine Theorems 1.1 or 1.3, whose scalar operators have self-adjoint structure and whose transversality checks are valid. The concrete test would settle whether the flaw is merely cosmetic or fatal to Theorem 1.5.","tokens_in":29765,"tokens_out":25263,"duration_ms":205194,"concrete_test":"Fix λ=0.5 (or keep λ symbolic) and compute M_{1,γ*_1} from (5.8). Verify that its first row is zero and its second row is (C,D) with C,D nonzero, so Range = span{(0,1)}. Then compute the vector ∂_γ M_{1,γ}|_{γ=γ*_1} · (1,1)^T and check whether its first entry is nonzero. If the first entry is nonzero, condition (H2) actually holds and the theorem survives despite the flawed orthogonality claim; if it is zero, the transversality condition fails and the bifurcation proof of Theorem 1.5 collapses.","verdict_should_be":"UNCHANGED","load_bearing_attack":"In §5.2 (proof of Theorem 1.5), the authors verify condition (H2) of Crandall–Rabinowitz by asserting that the range of L = ∂_{(η,ξ)}G(γ*_1,(0,0)) is the set of pairs (φ1,φ2)∈eY orthogonal to the kernel generator (η*,ξ*)=(α cosθ, β cosθ). From the explicit matrix M_{k,γ} in (5.8), at γ=γ*_1 one has A(k,γ*_1)=0 and B(k,γ*_1)=0 for all k, so for k=1 the linearized operator acts as M_{1,γ*_1} = [[0,0],[C,D]] with C,D nonzero. Its range on the mode-1 subspace is therefore {0}×ℝ (all second components), whose orthogonal complement is ℝ×{0}, not the orthogonal complement of span(cosθ, cosθ). Hence the range is not orthogonal to the kernel; the two subspaces are skew. The subsequent transversality computation evaluates (η*,ξ*)·[∂_γ L (η*,ξ*)] (a quadratic form mixing the first and second components), but the correct test for H2 is whether the first component of ∂_γ L (η*,ξ*) vanishes (since the cokernel is the first-component subspace). The proof as written therefore does not establish the transversality condition for Theorem 1.5. The same flaw affects the second bifurcation point γ**_1. This is a concrete gap in one of the paper's three central existence theorems, independent of the differentiability assumptions the Reader flagged.","agreement_with_reader":"partial"},"referee_report":{"model":"deepseek-v4-flash","summary":"The paper studies three steady Euler free-boundary problems with constant vorticity in two dimensions: the partially overdetermined problem (1.2), the two-phase problem (1.3), and the fully overdetermined two-free-boundary problem (1.4). For each problem the authors compute explicit radial trivial solutions, linearize via shape derivatives, identify explicit bifurcation values of the vorticity, apply the Crandall–Rabinowitz theorem to obtain nontrivial branches, and use the implicit function theorem to establish stability under small Neumann perturbations. The main results are Theorems 1.1, 1.3, and 1.5, together with Corollaries 1.2, 1.4, and 1.6.","tokens_in":30106,"tokens_out":18267,"duration_ms":164656,"significance":"If correct, the paper gives the first systematic local bifurcation construction of nontrivial compactly supported constant-vorticity Euler flows with closed streamlines, with explicit spectral computations and checkable nondegeneracy conditions. The shape-derivative approach is a genuine alternative to the Hanzawa-transformation methods used in earlier work. The main weakness is that the proof of Theorem 1.5, one of the three central existence results, contains a concrete error in the verification of the Crandall–Rabinowitz transversality condition.","major_comments":[{"comment":"The range of L=∂_{(η,ξ)}G(γ~,(0,0)) is misidentified. At γ~=γ1* of (5.11), M_{1,γ1*}=[[0,0],[C,D]] with C,D≠0, so on the k=1 mode the range is {0}×R and its annihilator is the first-component subspace, not span{(η*,ξ*)}. Hence the statement that the range consists of all pairs orthogonal to (η*,ξ*) is false, and the subsequent quadratic-form computation tests the wrong functional. H2 requires ∂_γ L(η*,ξ*)∉R(L); for γ1* this is equivalent to the first component being nonzero, which is true (the kernel condition forces α1=(λ²−1)β1, so α1+β1=λ²β1≠0), but this is not what is shown. The same error occurs at γ1**. Theorem 1.5 is therefore not proved as written, although the gap appears repairable by a direct annihilator computation.","section":"§5.2 (Eq. (5.8), proof of Theorem 1.5)"},{"comment":"The bifurcation and stability arguments require the solution map η↦ψ_η (or (η,ξ)↦ψ_{η,ξ}) to be at least C² into the appropriate Hölder spaces. This is asserted by citing [35, Thm 5.3.2] and Schauder theory. For problem (1.3) the map is for a transmission problem with piecewise-constant coefficients across the fixed interface ∂B_λ; the cited reference is formulated for smooth single-phase shape variations, and the transmission case needs a separate justification (e.g. flattening plus the implicit function theorem). Since all three existence theorems and the stability corollaries depend on this differentiability, a precise lemma or an exact reference covering the transmission case should be supplied.","section":"§3.2, §4.1, §5.1"}],"minor_comments":[{"comment":"The proof of Corollary 1.4 is omitted with only a reference to Corollary 1.2. If the argument is truly parallel, please state the relevant linearized operator and the nondegeneracy condition explicitly, or mark the corollary as a remark rather than a numbered result.","section":"§4 (after Theorem 1.3)"},{"comment":"The remark states that the bifurcation value γ~ in Theorem 1.5 lies in (−∞,−4) for all λ∈(0,1). This is false for γ1**=4/(2λ²lnλ+λ²−1): for λ=1/2 this value is ≈−3.65. The interval in Theorem 1.5 already gives the correct range; please correct the remark.","section":"Remark 1.3(i)"},{"comment":"The vector multiplying M_{k,γ} is written with β_k sin(kθ), but the Fourier expansion in (2.13) uses cos(kθ). Please correct the typo.","section":"§5.1, display after Eq. (5.8)"},{"comment":"The assertion that f_de(0.2483,2)≈0 and hence f_de(λ,k)<0 for all λ∈(0,0.2483) and k≥2 is supported only by a numerical check, not by a proof. Please either prove the required monotonicity in λ or state this as a numerical observation rather than a rigorous claim in the remark.","section":"§3.2, Remark 1.1(i)"}],"recommendation":"major_revision","confidential_remarks":"The central issue is localized to §5.2, where the cokernel of the linearized operator is misidentified. The authors should be asked to rewrite the range/transversality verification for Theorem 1.5 and to add a precise statement covering the differentiability of the two-phase solution map. If these repairs are made, the paper would likely be suitable for publication."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"Colleague,\n\nMy take: this is a serious piece of local bifurcation theory for closed-streamline Euler flows with constant vorticity. The genuinely new part is treating vorticity as the bifurcation parameter—previous work used the inner radius and Hanzawa transformations—and doing it systematically for three free-boundary problems: partially overdetermined, two-phase, and fully overdetermined. The explicit spectral computations are heavy but are laid out in enough detail to be checked, and the main existence, rigidity, and stability claims for Theorems 1.1 and 1.3 look sound. That is real content, not filler.\n\nThe soft spots are real but mostly minor. Corollary 1.4 is asserted without proof; the simplicity of the higher-mode bifurcation values in Remark 1.3(i) is only numerical; and the surjectivity step in Corollary 1.6 is sketched rather than shown. The Fréchet differentiability of the solution map is imported from [35] and elliptic regularity rather than proved; for the two-phase transmission problem that is a non-negotiable ingredient, though a standard one.\n\nThe stress-test note finds a genuine flaw in the proof of Theorem 1.5, and I think it is right about that. The paper claims the range of the linearized operator is the set of pairs orthogonal to the kernel generator. At γ*_1 the mode-1 matrix is [[0,0],[C,D]], so the range is {0}×R; its orthogonal complement is the first-component subspace, not the orthogonal complement of the kernel. The quadratic-form computation that follows tests the wrong functional and does not establish H2 by itself.\n\nBut the note overreaches if it implies transversality fails. The fix is short: at γ*_1 the kernel vector is (cosθ, cosθ), and the first component of the mixed derivative ∂_γ∂_{(η,ξ)}G applied to it is 2a cosθ with a = (2λ² lnλ + 1 − λ²)/(2 lnλ(λ²−1)) ≠ 0. Since the range is {0}×R, that vector is not in the range, so H2 actually holds. The same repair should work at γ*_**.\n\nBottom line: Theorems 1.1 and 1.3 are in good shape; Theorem 1.5 is probably right but needs a corrected H2 argument, and Corollary 1.4 needs a proof or an explicit deferral. This deserves a serious referee and a revision, not a desk reject. I would cite it for the partially overdetermined result and would bring it to a reading group on overdetermined elliptic problems or free-boundary Euler flows.","headline":"Solid bifurcation paper with genuinely new results; Theorem 1.5 has a real proof gap in the H2 verification, but the gap is repairable and the result looks true.","tokens_in":30663,"tokens_out":11143,"would_cite":true,"duration_ms":90765,"reading_group":"yes","serious_thinker":"yes","would_accept_peer_review":true},"rs_alignment":null,"lean_confirmation":null,"pith_extraction":{"msc":["35B32","35N05","51M10"],"pacs":[],"model":"deepseek-v4-flash","headline":"For steady constant-vorticity flows in compact domains, non-annular vortex shapes exist, yet the circle remains the only rigid shape.","keywords":["steady Euler flows","constant vorticity","free boundary","annular domain","bifurcation","shape derivative","overdetermined elliptic problem","stability"],"falsifier":"Fix λ=1/2, compute γ*_1 from the explicit formula, and solve the free-boundary problem numerically with boundary perturbation η=s cosθ for γ near γ*_1. If no non-annular solution branch appears at that vorticity, or if the leading angular mode is not cosθ, then Theorem 1.1's bifurcation claim fails. Alternatively, check the transversality condition ∂_{γη}G(γ*_1,0)[1,cosθ]≠0 numerically; if it vanishes, the one-dimensional kernel is degenerate and the described curve does not exist.","tokens_in":29599,"feed_emoji":"🌀","tokens_out":7130,"duration_ms":70946,"temperature":0.7,"pith_summary":"This paper studies steady, two-dimensional, incompressible Euler flows with constant vorticity whose fluid region is compact and has one or two free boundaries. It proves that for each of three overdetermined free-boundary problems there are special vorticity values at which non-circular, non-annular solution domains branch off from the standard circular annulus, so the circle is not the only shape a constant-vorticity vortex can occupy. It also proves rigidity: under natural positivity or curvature conditions the only admissible domains are circular, and the bifurcating nontrivial solutions must be sign-changing. Finally, it shows via the implicit function theorem that the circular solutions respond smoothly and locally uniquely to small perturbations of the Neumann data, with explicit linear response coefficients. If correct, closed-streamline configurations with constant vorticity are simultaneously flexible and rigid in a controlled way, opening a path beyond the laminar strip-like water-wave setting.","feed_headline":"Non-annular steady flows exist for constant vorticity","feed_subtitle":"Three free-boundary Euler problems admit non-circular vortex cores, with rigidity and stability near the circles.","key_machinery":"The central mechanism is the shape derivative of the stream function with respect to normal boundary perturbations η (and ξ for the inner boundary). Restricted to the trivial annulus, the linearized shape-derivative operator is diagonalized by the Fourier modes cos(kθ); for each mode the spectral condition becomes an explicit scalar dispersion relation (σ_k(γ)=0 for the first problem, μ_k(γ_2)=0 for the two-phase problem, det M_{k,γ}=0 for the fully overdetermined problem). The zeros of these functions give the bifurcation vorticities, and a standard local bifurcation theorem for one-dimensional kernels, with transversality checked explicitly, produces the nontrivial branches. The same inver","core_discovery":"For any inner radius λ∈(0,1), the Bernoulli free-boundary problem admits a C^1 curve of solutions (γ(s),η(s)) branching from the standard annulus at the explicit vorticity value γ*_1 = 4/(λ^2−2λ^2 ln λ−1), with leading boundary perturbation η(s)=s α1 cosθ+o(s); hence non-annular admissible domains exist. The same structure appears for the two-phase problem at γ_2*=γ_1 and for the fully overdetermined problem at two explicit values γ*_1 and γ**_1. The bifurcating solutions are necessarily sign-changing under the rigidity theorems that force positivity to yield radial symmetry. Moreover, the trivial annular flows are locally unique and stable under small Neumann perturbations, with explicit le","pith_inferences":["The explicit dispersion relations can be read as ready-made predictions: pick a λ, compute γ*_1, and run a numerical continuation from the annulus to look for a cosθ-symmetric branch; a clean detection at that value would confirm the mechanism, while its absence would point to the imported regularity assumption failing.","Because the bifurcating solutions are sign-changing, the corresponding physical flows likely contain an interior stagnation curve (where ψ changes sign) separating counter-rotating regions—a topological feature that could be probed in experimentally realizable vortex-core flows.","The stability amplitudes 1/(2√Q_γ σ_k) imply a resonance-like amplification as γ approaches a bifurcation value; near higher-mode bifurcation points the response to forcing at that mode should blow up, so the stable regime is precisely the complement of the bifurcation set—a duality the paper does not spell out.","For the fully overdetermined problem, the matrix M_{k,γ} resembles objects that appear in shape-optimization resonance problems; studying det M_{k,γ} as a function of γ for all k could classify all possible bifurcation vorticities, analogous to a spectral trace formula."],"forward_implications":["For every λ∈(0,1), the partially overdetermined problem has non-annular admissible domains for γ near γ*_1<−4; for λ small, higher-mode bifurcations give additional branches with positive vorticity.","Rigidity: any positive solution of the first problem with γ>0 is radially symmetric on an annulus, and the two-phase problem's concentric-disk configuration is the only one when the curvature-type condition ∂_ννψ_2=m holds; hence the new branches are sign-changing.","Stability: for γ avoiding the finitely many resonant values, every sufficiently small Neumann perturbation ρ has a unique nearby admissible domain, with explicit leading-order amplitude formulas such as η(ρ)≈Σ τ_k cos(kθ)/(2√Q_γ σ_k).","The choice of vorticity as the bifurcation parameter yields branches for every λ∈(0,1), and the method is claimed to extend to higher dimensions and to affine vorticity functions γ(ψ)=βψ."],"fun_headline_variants":["Constant vorticity flows break annular symmetry","Non-circular vortex cores emerge at critical vorticity","Bifurcation yields non-annular Euler flows","Annular flows stable, non-circular branches exist"],"cache_read_input_tokens":2304,"weakest_assumption_plain":"The proof leans on the assumption that the map from a boundary perturbation to the normal derivative of the resulting stream function is twice continuously differentiable (C^2) near the trivial annulus; this regularity is imported from external shape-derivative theory and, in the two-phase problem, must survive across an interface where the vorticity jumps.","fun_headline_variants_meta":{"raw":{"variants":["Constant vorticity flows break annular symmetry","Non-circular vortex cores emerge at critical vorticity","Bifurcation yields non-annular Euler flows","Annular flows stable, non-circular branches exist"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.000501,"raw_usage":{"total_tokens":2269,"prompt_tokens":712,"completion_tokens":1557,"prompt_tokens_details":{"cached_tokens":256},"prompt_cache_hit_tokens":256,"prompt_cache_miss_tokens":456,"completion_tokens_details":{"reasoning_tokens":1508}},"tokens_in":456,"tokens_out":1557,"duration_ms":10718,"temperature":1.0,"reasoning_tokens":1508,"cache_read_input_tokens":256,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-03T03:37:44.489162+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"Fix λ=1/2, compute γ*_1 from the explicit formula, and solve the free-boundary problem numerically with boundary perturbation η=s cosθ for γ near γ*_1. If no non-annular solution branch appears at that vorticity, or if the leading angular mode is not cosθ, then Theorem 1.1's bifurcation claim fails. Alternatively, check the transversality condition ∂_{γη}G(γ*_1,0)[1,cosθ]≠0 numerically; if it vanishes, the one-dimensional kernel is degenerate and the described curve does not exist.","supporting_citations":[],"review_version":2}