{"id":"ec4a6ad6-b5f9-4abf-8642-9ac3e05f445a","arxiv_id":"2511.03171","paper_version":4,"verdict":"CONDITIONAL","confidence":"MODERATE","novelty_score":8.0,"correctness_risk":"low","formal_verification":"none","parameter_count":1,"one_line_summary":"For axisymmetric Euler flows without swirl with anti-parallel, one-signed vorticity, the radial moment satisfies P(t)[log t]^{5/2}/t^{3/2} -> infinity and the vorticity maximum reaches a fixed fraction of t on (1-eta) of every large dyadic time interval.","lead":"The paper proves that in axisymmetric swirling fluids, two colliding ring vortices can push the strongest vorticity outward so efficiently that, on most long time intervals, the vorticity maximum grows at least linearly in time. It also proves the first radial-moment growth bound faster than linear, advancing a 2008 conjecture by Childress.","discovery_kind":"new_method","skeptic_critique":{"model":"deepseek-v4-flash","headline":"No significant objection identified: the central two-moment argument and its dyadic-to-full-time conversion hold up; only minor display-level typos were found.","rationale":"The reader's weakest assumption correctly targets Proposition 3.3: the coercivity of the second mixed moment is the technical linchpin that converts the L1-in-time dissipation D into the radius-dependent energy localization (3.28), which in turn drives every subsequent transfer estimate. I checked the algebra behind (3.14)–(3.21): with the stated g(ϱ)=ϱ/arsinhϱ, the matrix H has positive diagonal, detH≥w for λ0 sufficiently large, and the asymptotic of Θ is bounded (the displayed O(1/ℓ) term in (3.18) is harmless and the true expansion approaches 1 from below). Lemma 3.5's spherical-harmonic computation is correct: the n=0 mode vanishes by the flux identity and each n≥1 mode decays at least as C^{-3}, so the C^{-3/2} bound follows. The near–far decomposition in Lemma 3.7 correctly handles the logarithmic kernel via Lemma 2.1, whose p-dependence was independently verified from the tail estimate. The dyadic increment argument for (1.18) is compressed but sound, since every t can be written as 2T with T=t/2 and P(t) dominates the increment over [t/2,t]. The only real defect is the sign error in the logarithmic exponent in (3.72); this affects an intermediate display rather than the central claims, since the corrected exponent still gives (1.28). The dependence on [25, Thm 1.1] is limited to the p<2 patch bounds and does not enter Theorem 1.1 or the primary Theorem 1.4 statements. No circularity was found; the paper explicitly and honestly records in Remark 3.8 the limitation that κ cannot be sent to infinity with T without extra decay information. The reader's CONDITIONAL verdict remains appropriate because of the (3.72) typo and the terse full-time conversion; no adjustment is needed.","tokens_in":66,"tokens_out":50845,"duration_ms":870887,"concrete_test":"Independently recompute the determinant computation in Proposition 3.3: symbolically or numerically evaluate Θ(ϱ)=d_g^2/(4w)-g'-g/ϱ on a dense grid ϱ∈[10^{-6},10^6] and verify it is bounded above by 1, so that λ0=3 yields detH≥w. If any point had Θ>1, the coercivity of the second mixed moment would fail and the energy localization argument would collapse; a pass confirms the central mechanism.","verdict_should_be":"UNCHANGED","load_bearing_attack":"No significant objection identified. Independent re-derivation of the key steps — the multiplier identity (3.1), the dissipation coercivity in Proposition 3.3 (including the determinant computation detH ≥ w under λ0 ≥ 3), the exterior harmonic estimate Lemma 3.5 with C^{-3/2} decay, and the Hölder bookkeeping in Lemma 3.7 — does not uncover a flaw that threatens Theorem 1.1 or the linear-growth conclusion of Theorem 1.4. The full-time limit (1.18) follows from the dyadic increment bound because for arbitrary t one may set T=t/2 and use P(t) ≥ P(t)-P(t/2); the proof is terse but valid. The only genuine mathematical typo is in (3.72), where the logarithmic exponent should be 1/(2p)-2 rather than 2-1/(2p); the corrected bound still yields (1.28) by choosing κ large. The asymptotic expansion around (3.18) is also slightly imprecise, but the boundedness of Θ — the actual load-bearing point of Proposition 3.3 — is unaffected. The external upper bound [25, Thm 1.1] is used only for the secondary p<2 patch family (1.27a).","agreement_with_reader":"agree"},"referee_report":{"model":"deepseek-v4-flash","summary":"The paper studies long-time vortex stretching for three-dimensional axisymmetric Euler flows without swirl in the anti-parallel class. It introduces two monotone mixed moments: one with weight Φ(r,z)=∫_0^z √(r²+s²)ds, yielding a uniform bound on ∫ rz[-ω], and one with the logarithmically weighted multiplier Ψ=λ₀Φ+zϱ/arsinhϱ, whose dissipation controls both spherical velocity components with scale-dependent weights. These tools are combined with positive-kernel energy estimates, an exterior harmonic-regularity estimate, and a dyadic-in-time argument. The main results are: P(t) grows at least like t^{3/2}/(log t)^{5/2}; the support radius R_ω(t) is at least linear on a (1-η) fraction of every large dyadic interval; for unit-strength relative-vorticity patches the vorticity maximum, equal to the outer radius, has the same linear growth on a large proportion of dyadic intervals and a full-time t^{3/4}/(log t)^{5/4} lower bound; and L^p vorticity norms grow for all 1≤p≤∞. The proof is explicit and does not use the previously essential vertical moment Z.","tokens_in":35662,"tokens_out":25329,"duration_ms":202724,"significance":"If the advertised results stand, this is a substantial advance on the lower-bound side of Childress's t^{4/3} conjecture. In particular, it gives the first radial-moment lower bound with polynomial exponent greater than one, and the first linear-scale radial support growth for this class. The method is new: the two mixed moments convert conservation of kinetic energy into radial escape, and the exterior harmonic estimate transfers exterior velocity energy to exterior vorticity. The constants are universal or data-dependent, and the logarithmic exponents are forced by the estimates rather than fitted. The external dependencies are clearly identified, and the key algebraic and analytic steps are checkable. The main limitation is secondary: the p<2 patch family relies on the external upper bound of Egamberganov–Yao, but the core linear-growth theorem does not depend on it.","major_comments":[],"minor_comments":[{"comment":"The displayed logarithmic exponent in (3.72) appears to be a typo. The preceding line correctly factors the lower bound as (R_T/2)^{1-s_T/p}(log T)^{-s_T/p}, which gives (log T)^{1/(2p)-2}, not (log T)^{2-1/(2p)}. The subsequent proof of (1.28) uses the corrected exponent, so the error is local and does not affect the theorem.","section":"Eq. (3.72)"},{"comment":"The justification of the bound (r̄r)^{(1-2a_T)/2}(z ̄z)^{a_T} ≤ R^{1-4a_T}[(rz)(r̄ ̄z)]^{a_T} is abbreviated. The reader must divide by (z ̄z)^{a_T} and then use (r̄r)^{1/2-2a_T} ≤ R^{1-4a_T}, which follows from r ̄r ≥ R² and 1/2-2a_T<0. Adding this line would improve readability.","section":"§3.4, around (3.59)"},{"comment":"The asymptotic expansion of Θ(ϱ) at infinity is presented very tersely. Since boundedness of Θ is the load-bearing point for the choice of λ₀, a short derivation of Θ = 1 - 1/ℓ + O(ℓ^{-2}) from the displayed formulas for w, d_g, and g' would be helpful.","section":"Proposition 3.3 / (3.18)"},{"comment":"In the estimate (3.40), the factor ∫ rη_H is obtained after absorbing a factor m₀ into the constant. This is correct, but the absorption is not stated. A brief remark would prevent confusion.","section":"§3.3, Step 3 of Proposition 3.6"}],"recommendation":"minor_revision","confidential_remarks":"The paper is technically strong and the main claims are supported. The only dependency that could shift in principle is the external upper bound [25, Thm 1.1] used for the secondary p<2 patch family; the central linear-growth theorem does not rely on it. I recommend minor revision to correct the display typo in (3.72) and improve the cited exposition points."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"The short version: this paper delivers what it promises, and the proof holds up under scrutiny. It proves P(t) ≥ t^{3/2}/log^{5/2} for the radial moment and, for patches, that the vorticity maximum (which equals the outer radius) exceeds cη t on a (1−η) fraction of every large dyadic interval. Both are firsts: prior pointwise lower bounds for the vorticity maximum sat at t^{1/2} or t^{3/4}, and the radial-moment exponent one was the record. The linear-on-large-sets conclusion is new even for general data, not just patches.\n\nThe new machinery is a pair of mixed radial–axial moments. The first, built on Φ, gives the crucial rz-weighted control that replaces the old vertical-moment input. The second, with the logarithmically weighted multiplier, produces a dissipation that dominates both spherical components of the velocity with scale-dependent weights. The determinant computation behind the coercivity is the heart of the argument, and it checks out — I re-derived the key identities, including the harmonic exterior estimate and the Hölder bookkeeping in Lemma 3.7. No circularity: the constants are universal or data-dependent, and the logarithmic exponents are forced by the estimates, not tuned.\n\nSoft spots are minor. The display (3.72) has a wrong logarithmic exponent: it should be 1/(2p)−2, not 2−1/(2p). The proof text just above it makes the intended bound clear, and choosing κ large still gives the stated density-one conclusion (1.28), so this is a typo, not a flaw. The conversion from dyadic increments to the full-time limit in (1.18) is terse — the reader is left to notice that P(t) ≥ P(t) − P(t/2) with T = t/2 — but it is valid. For the patch family 1 ≤ p < 2, the paper leans on the external upper bound [25, Thm 1.1]; that dependency is clearly stated and only enters a secondary estimate. The slight imprecision in the asymptotic expansion around (3.18) does not affect the boundedness of Θ, which is the load-bearing point.\n\nThe paper is also honest about its limits: Remark 3.8 explicitly notes the logarithmic factors are not asserted sharp, and the Childress t^{4/3} conjecture remains open. No code or data, but none is needed for this type of result.\n\nThis deserves a serious referee. I would send it out, expect minor revisions around the typo and some expanded exposition in the dyadic-to-full-time step, and be surprised if the main theorems change. For researchers working on vortex stretching, this is worth reading, and I will cite it.","headline":"A genuine advance in lower bounds for anti-parallel axisymmetric Euler: first super-linear radial-moment growth and linear-on-large-sets vorticity maximum, with an argument that appears correct despite a display typo.","tokens_in":36226,"tokens_out":2176,"would_cite":true,"duration_ms":21295,"reading_group":"yes","serious_thinker":"yes","would_accept_peer_review":true},"rs_alignment":null,"lean_confirmation":null,"pith_extraction":{"msc":["35Q31","76B47"],"pacs":[],"model":"deepseek-v4-flash","headline":"This paper proves that the maximum vorticity in a head-on collision of two coaxial vortex rings grows at least linearly on almost every large dyadic interval, with the radial moment at t^{3/2} up to logarithms.","keywords":["axisymmetric Euler equations","vortex stretching","anti-parallel vorticity","vortex-ring collision","radial-moment growth","mixed moments","vorticity maximum","t^{4/3} conjecture"],"falsifier":"Run a high-resolution numerical simulation of the inviscid head-on collision of two equal, opposite axisymmetric vortex rings with the paper's sign data, and measure the Lebesgue time measure in each dyadic window [T,2T] during which the maximum vorticity exceeds cT for a fixed small c. The theorem predicts this fraction is at least 1-eta for every eta if the window is large (T>=T_eta); a persistent deficit, say a fraction below 1/2 at all large T, would falsify the linear-growth claim.","tokens_in":35243,"feed_emoji":"🌀","tokens_out":8742,"duration_ms":77515,"temperature":0.7,"pith_summary":"The paper studies inviscid, swirl-free axisymmetric flows of an ideal fluid — the geometry of two coaxial vortex rings moving toward each other. It proves that the radial moment of the vorticity, a measure of how far vorticity has been carried outward, grows at least like t^{3/2} (up to logarithms), and that the maximum vorticity itself reaches the linear scale t on an arbitrarily large fraction of every sufficiently large dyadic time interval. For initial data that are vortex patches, this vorticity maximum is exactly the outer radius of the patch. If correct, this is the first lower bound on the radial moment with polynomial exponent greater than one, and it supplies the strongest known evidence that vortex stretching in this class proceeds without saturation. The mechanism is a pair of monotone 'mixed' moments that convert conservation of kinetic energy into quantitative radial escape and axial compression.","feed_headline":"Vorticity maximum grows at least linearly in colliding vortex rings","feed_subtitle":"Best-known growth rates: radial moment at t^{3/2} (up to logs); vorticity max above c t on nearly every large window.","key_machinery":"The argument hinges on two new monotone moments of the vorticity distribution in the upper half-plane. The first, with weight Phi(r,z)=integral_0^z sqrt(r^2+s^2) ds, is non-increasing and gives sup_t integral rz(-omega) < infinity, which forces escaping vorticity to lie near the symmetry plane. The second, with weight Psi = lambda_0 Phi + z * rho/arsinh(rho), is also monotone and its dissipation D(t) controls both independent meridional velocity components with specific radius-dependent weights; the universal constants in this control follow from a determinant positivity condition (lambda_0 >= 2 + sup Theta(rho)) and a logarithmic-kernel estimate. The energy localization then proceeds throug","core_discovery":"Within the class of compactly supported, odd-in-z, sign-definite axisymmetric swirl-free Euler data, the paper proves the pointwise-in-time lower bound P(t)[log(2+t)]^{5/2}/(1+t)^{3/2} -> infinity for the radial moment P(t), and consequently ||Omega(t)||_{L^inf}[log(2+t)]^{5/4}/(1+t)^{3/4} -> infinity. More sharply, for every 0<eta<1 there are c_eta>0 and T_eta>1 such that for all T>=T_eta the set of times t in [T,2T] with the vorticity maximum at least c_eta t has measure at least (1-eta)T; for vortex patches the vorticity maximum equals the support radius. The proof also gives a quantitative energetic picture: on a density-one set of times, a vorticity truncation carrying arbitrarily close","pith_inferences":["The t^{4/3} conjecture remains unproven, but this linear-on-most-times result narrows the range of possible scalings: the true exponent for the vorticity maximum, if it exists, lies between 1 and 4/3. A natural next step would be to construct initial data that push the linear support estimate toward t^{4/3} by concentrating energy at a slowly growing radius.","The mixed-moment coercivity is general enough that the same two-weight mechanism may apply to other symmetric fluid models (e.g., axisymmetric MHD or two-and-a-half-dimensional flows) where radial transport plays the role of stretching; the exterior harmonic projection lemma is symmetry-agnostic.","The logarithmic factors are tied to the L^1-integrability of the dissipation D; if the exceptional set could be made to decay exponentially instead of merely o(T), the t^{3/2} moment bound might improve to t^{3/2} without logs, and one might test numerically whether the true radial moment in head-on collisions has clean t^{3/2} scaling.","A numerical check of the near-far decomposition would be valuable: for a simulated head-on collision, track the time spent with ||Omega||_inf >= c t in dyadic windows; the theorem predicts this fraction approaches 1-eta for any prescribed eta, which is a sharp, testable signature of the mechanism."],"forward_implications":["The radial moment P(t) grows faster than t/log t, the previous best, and (up to logarithms) at the rate t^{3/2}; the vorticity maximum grows at least at the t^{3/4} rate in the full-time pointwise sense.","On a proportion 1-eta of every large dyadic interval the vorticity maximum is at least c_eta t; for vortex patches this is a statement about the outer radius of the transported patch.","Every vorticity L^p norm, uniformly in 1<=p<=infinity, grows: inf_p ||Omega(t)||_p [log(2+t)]^{25/12}/(1+t)^{1/4} -> infinity, upgrading the previous t^{1/4} liminf phenomenon to a pointwise-in-time statement up to logarithms.","The proof yields a quantitative Eulerian form of the collision picture: the part of the vorticity generating almost all the kinetic energy is simultaneously at large radius and in a shrinking neighborhood of the collision plane, on a density-one set of times.","No patch assumption or regularity of the patch boundary is needed; the same estimates hold for all compactly supported initial data in the class."],"fun_headline_variants":["Vorticity max grows linearly for colliding vortex rings","Linear vorticity growth proven for axisymmetric Euler flows","Vortex rings: vorticity max hits linear scale almost always","Radial moment grows faster than t: new t^{3/2} bound"],"cache_read_input_tokens":2304,"weakest_assumption_plain":"For the proof to work, one fixed constant lambda_0 must make a certain 2x2 dissipation matrix positive definite at every radius, so that the dissipation integral is finite in time; if the worst-case ratio ever escapes control, the energy-localization step fails.","fun_headline_variants_meta":{"raw":{"variants":["Vorticity max grows linearly for colliding vortex rings","Linear vorticity growth proven for axisymmetric Euler flows","Vortex rings: vorticity max hits linear scale almost always","Radial moment grows faster than t: new t^{3/2} bound"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.000387,"raw_usage":{"total_tokens":2034,"prompt_tokens":1052,"completion_tokens":982,"prompt_tokens_details":{"cached_tokens":256},"prompt_cache_hit_tokens":256,"prompt_cache_miss_tokens":796,"completion_tokens_details":{"reasoning_tokens":910}},"tokens_in":796,"tokens_out":982,"duration_ms":8944,"temperature":1.0,"reasoning_tokens":910,"cache_read_input_tokens":256,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-04T06:52:27.339627+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"Run a high-resolution numerical simulation of the inviscid head-on collision of two equal, opposite axisymmetric vortex rings with the paper's sign data, and measure the Lebesgue time measure in each dyadic window [T,2T] during which the maximum vorticity exceeds cT for a fixed small c. The theorem predicts this fraction is at least 1-eta for every eta if the window is large (T>=T_eta); a persistent deficit, say a fraction below 1/2 at all large T, would falsify the linear-growth claim.","supporting_citations":[],"review_version":2}