{"id":"86559bd4-9abb-4de0-902b-2f8cb2c95cd7","arxiv_id":"2607.29260","paper_version":1,"verdict":"ACCEPT","confidence":"MODERATE","novelty_score":7.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":0,"one_line_summary":"Spherical Rayleigh-Taylor bubbles do not reach terminal velocity; they enter a constant-acceleration regime, and spherical geometry enhances the asymptotic acceleration relative to cylindrical geometry by a factor approaching 3.","lead":"This paper derives a mathematical model for how a Rayleigh-Taylor bubble grows on a spherical interface, showing it reaches a constant acceleration in the late nonlinear stage. The model matches numerical simulations and could improve predictions for inertial-confinement fusion and tin-droplet EUV sources.","discovery_kind":"new_application","skeptic_critique":{"model":"deepseek-v4-flash","headline":"Printed velocity potentials (6)-(7) are not harmonic: with S_L=-l (CG) and S_H=l (DG) they give r^{-l}P_l, whose Laplacian is nonzero, so equations (8)-(12) and the asymptotic law (20)-(21) are not consequences of the written ansatz.","rationale":"The paper's central claim rests on a derivation that starts from a truncated interface and a two-term potential ansatz. The critical load-bearing condition is that the ansatz actually solves Laplace's equation, because the ODEs are obtained by substituting it into (2)-(5). The printed exponents violate this condition for one fluid in each configuration: r^{-l}P_l is not harmonic for l>0. This is not a matter of disagreement with existing consensus; it is an internal inconsistency in the written derivation. A sign typo is plausible, and the DNS agreement in Figures 4-5 suggests the intended model may be correct, so the paper should not be rejected outright. However, the central asymptotic law is not actually derived from the equations as printed. The reader's weakest assumption was the θ² truncation; that is a legitimate approximation whose error is uncontrolled, whereas the harmonicity issue is an explicit mathematical error. I therefore recommend CONDITIONAL rather than ACCEPT: the paper should be accepted only after the potential exponents are corrected and the coefficients A,B,C are rechecked. The DNS comparisons provide real empirical support, but they cover only four cases (AT=0.3,0.6; l=4,8) and test the integrated model, not the printed derivation, so they cannot by themselves certify the asymptotic coefficients.","tokens_in":8622,"tokens_out":8781,"duration_ms":89565,"concrete_test":"Use a symbolic algebra system to substitute (6)-(7) with the printed exponents into ∇²φ=0 (e.g. CG, l=3). Confirm ∇²φ_L≠0. Then re-derive (8)-(9) and (11)-(12) using the harmonic exponents S_L=l for CG and S_H=-l for DG, and recompute A,B,C in (17)-(18). If the corrected ODEs and asymptotic coefficients coincide with the printed ones, the printed exponents are typographical and the central claim survives; if they differ, (20)-(21) are unsupported and the asymptotic acceleration values must be revised.","verdict_should_be":"CONDITIONAL","load_bearing_attack":"Equations (6)-(7) are introduced as 'eigenfunctions satisfying the Laplace equations (2)', but the printed exponents do not yield harmonic functions. For r^s P_l(cosθ), Laplace's equation requires s=l or s=-(l+1). Equation (7) with CG has S_L=-l, i.e. φ_L ⊃ P_l r^{-l}; its Laplacian is [l(l-1)-l(l+1)] r^{-l-2} P_l = -2l r^{-l-2}P_l ≠ 0 for l>0. Equation (6) with DG has S_H=l, i.e. φ_H ⊃ P_l r^{-l}, also non-harmonic. The added b_2 η0/r term is harmonic but angularly isotropic, so it cannot cancel this. Since (8)-(13) are derived by substituting these potentials into (2)-(5), the ODE system and hence (20)-(21) are not valid consequences of the written ansatz. The likely fix is a sign typo (S_L=l for CG, S_H=-l for DG), and the favorable DNS comparison suggests the implemented model may be the corrected one; but the paper as printed contains an unsound step in the central analytical derivation. This is more load-bearing than the θ-truncation concern: the truncation is a standard approximation, while the non-harmonic potential is an internal inconsistency.","agreement_with_reader":"partial"},"referee_report":{"model":"deepseek-v4-flash","summary":"The paper develops a potential-flow model for the nonlinear evolution of a single-mode Rayleigh-Taylor bubble in spherical geometry, covering both converging-gravity (CG) and diverging-gravity (DG) configurations. The interface near the bubble tip is represented by η(θ,t)=η0(t)+η2(t)θ², and velocity potentials are written in single-mode Legendre form with three unknown coefficients. This leads to a closed two-variable ODE system for η0 and η2: Eqs. (8)-(9) for CG and (11)-(12) for DG. An asymptotic analysis of this system gives |η̇0|/√(r0+η0) → √(2A/(B+2C)) and η̈0 → A/(B+2C), i.e., a constant-acceleration/deceleration regime rather than a planar-style terminal velocity. The model is compared with axisymmetric Euler DNS for AT=0.3,0.6 and l=4,8, with reported favorable agreement. The paper also claims that spherical geometry enhances bubble growth relative to cylindrical geometry with the same effective wavenumber in CG, and mitigates it in DG, with the acceleration ratio approaching 3 for large l.","tokens_in":8991,"tokens_out":15870,"duration_ms":145474,"significance":"If the derivation is correct, the result is a meaningful extension of earlier cylindrical-geometry analyses: it provides closed-form asymptotic coefficients depending only on l, AT, and g, with no fitted parameters, and it makes falsifiable predictions about the late-time scaling of the bubble velocity and acceleration. The recovery of Layzer's planar terminal-velocity limit in the large-l limit is a nice consistency check. The paper also credits and builds on prior work by Goncharov, Zhao et al., and Layzer. However, as printed, the analytical derivation contains internal errors that undermine the central claim until they are repaired: the velocity potentials in Eqs. (6)-(7) are not harmonic, and the ODEs in Eqs. (9)/(12) are dimensionally inconsistent with the coefficient definitions in Eqs. (10)/(13).","major_comments":[{"comment":"The printed potentials do not satisfy ∇²φ=0. For a mode f(r)P_l(cosθ), Laplace's equation requires f∝r^l or r^{-(l+1)}. In the CG case, φ_H with S_H=l+1 is r^{-(l+1)}P_l (harmonic), but φ_L with S_L=-l is r^{-l}P_l, whose Laplacian is -2l r^{-l-2}P_l ≠ 0 for l>0. In the DG case, S_H=l gives the non-harmonic r^{-l}P_l, while S_L=-(l+1) is harmonic. The isotropic b2 η0/r term cannot cancel this angularly dependent failure. Since Eqs. (8)-(13) are derived by substituting these potentials into (3)-(5), the ODE system and the asymptotic laws (20)-(21) are not consequences of the written ansatz. The likely correction is a sign change in the inner-region exponent (S_L=l for CG, S_H=-l for DG); the favorable DNS comparison suggests the implemented model may already use the corrected exponents, but the manuscript as printed is internally inconsistent.","section":"§II.B, Eqs. (6)-(7)"},{"comment":"There is a separate dimensional inconsistency. In Eq. (9), H1 as defined in (10a) has dimensions of length² (it contains (r0+η0)², η2², and mixed products), and [8η2-l(r0+η0)] has dimensions of length, so H1/8 [8η2-l(r0+η0)] η̈0 has L⁴/T². The final term AT g η2 has L²/T². The H2 term in (9) has L⁶/T². The same problem appears in Eq. (12) with (13). After substituting the asymptotic η2 relation (15), the terms in (9) scale as (r0+η0)³ and (r0+η0)⁴, whereas the claimed reduced equation (16) is linear in R=r0+η0 for the η̈ term and independent of R for the η̇² term. The printed equations can only be valid if factors of 1/(r0+η0)² and 1/(r0+η0)⁴ are missing from the H1 and H2 terms. This also affects the claimed recovery of the linear growth rate (1). The authors should correct the equations and re-verify the derivation.","section":"§II.B, Eqs. (9)-(13)"},{"comment":"The model rests on the near-tip truncation η=η0+η2θ² together with single-mode Legendre potentials. This is a standard Layzer-type ansatz, but the paper provides no estimate of the neglected θ⁴ terms, which could become dynamically important in the asymptotic stage. The asymptotic relation (15) and the constant-acceleration law (20)-(21) depend directly on this truncation. I recommend adding a consistency check—for example, retaining a θ⁴ coefficient and verifying that it remains small compared to η2, or comparing the predicted local interface shape with the DNS interface in the late-time regime—so that the validity of the asymptotic claim does not rest solely on the visual agreement in Figs. 4-5.","section":"§II.B, before Eqs. (8)-(12)"}],"minor_comments":[{"comment":"The quantitative support for the central claim is weak: only four parameter combinations are shown, and the agreement is assessed visually. Please report a quantitative measure of agreement (e.g., relative L2 error of η0(t) and of |η̇0|/√(r0+η0) relative to the predicted asymptotic value), and provide the grid-convergence results mentioned in the text.","section":"§III, Figs. 4-5"},{"comment":"The sign α in (20) is introduced without explanation. It would be helpful to state explicitly that α=-1 in the DG case follows from η̇0<0 for an inward-moving bubble, and that the square-root branch is chosen accordingly.","section":"§II.C, Eq. (19)-(20)"},{"comment":"The Chandrasekhar reference appears as an unnumbered entry after the conclusion; the reference list otherwise starts with Goncharov. Please format consistently.","section":"References"}],"recommendation":"major_revision","confidential_remarks":"The paper has a plausible and potentially useful model, and the DNS comparisons suggest the intended theory may be correct. However, the printed analytical derivation contains two load-bearing internal errors: the velocity potentials are not harmonic, and the ODEs are dimensionally inconsistent. These are fixable within the scope of the paper, but the authors must re-derive the equations carefully and confirm that the final asymptotic coefficients (17)-(18) correspond to the corrected system before publication."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"Colleague,\n\nThis paper extends the Layzer/Goncharov potential-flow machinery to spherical RTI and makes a clean, novel claim: the bubble tip doesn't saturate to a terminal velocity; instead |η0_dot|/√(r0+η0) → const and η0_ddot → const, giving a uniform-acceleration law (20)-(21). The coefficients are derived analytically without fitting, and the DNS comparisons for AT=0.3/0.6 and l=4/8 show good trajectory agreement and clear approach to the predicted asymptotic constant. The factor-of-3 enhancement of spherical over cylindrical acceleration at equal k_eff is new and physically plausible, and the large-l limit recovers Layzer's planar tube result. Credit is due: no fitted parameters, honest DNS, parameter-free predictions.\n\nThat said, there's a load-bearing defect in the written derivation. Equations (6)-(7) are said to satisfy Laplace's equation, but they don't as printed. For CG, S_L=-l makes φ_L ∝ r^{-l}P_l, whose Laplacian is -2l r^{-l-2}P_l, not zero; for DG, S_H=l makes φ_H ∝ r^{-l}P_l, same problem. So the ODE system (8)-(13) is not a valid consequence of the stated ansatz. This is not a cosmetic typo: it's the bridge from the potentials to the equations the whole paper rests on. The likely fix is a sign error in the exponents (S_L=l for CG, S_H=-l for DG), and the DNS agreement suggests the implemented model is the corrected one. But as printed, the derivation has a hole that any referee must catch.\n\nOther soft spots are minor: only four validation cases, no error bars or quantitative agreement metric, no code/data release. The θ² truncation is standard for this model family and is tested indirectly by the DNS, so I wouldn't overweight it.\n\nBottom line: the result is probably right and worth having in the literature, but the paper needs a revised Section II with correct exponents and a re-derived ODE system before the formal claim is trustworthy. Send to peer review — a competent referee will find this immediately. I'd hold off citing until the correction appears; then I'd cite it for the asymptotic law and the geometry-dependent factor.\n\nBest.","headline":"Worth a serious look: novel spherical-geometry acceleration law with honest DNS, but the printed potential ansatz in (6)-(7) is not harmonic and the central derivation needs fixing before the result can be trusted.","tokens_in":9428,"tokens_out":9812,"would_cite":false,"duration_ms":91518,"reading_group":"maybe","serious_thinker":"no","would_accept_peer_review":true},"rs_alignment":null,"lean_confirmation":null,"pith_extraction":{"msc":[],"pacs":[],"model":"deepseek-v4-flash","headline":"This paper claims that a Rayleigh–Taylor bubble on a spherical interface does not saturate to a terminal velocity; instead, in the nonlinear regime it enters a uniform-acceleration stage governed by |η0'|/sqrt(r0+η0) → constant and η0'' → c","keywords":["Rayleigh-Taylor instability","spherical geometry","bubble growth","asymptotic acceleration","potential-flow model","converging gravity","diverging gravity","single-mode perturbation"],"falsifier":"Run a single-mode spherical RTI simulation or experiment with, say, l=10 and Atwood number 0.5 and measure the bubble-tip acceleration η0'' long after the linear stage. The paper's claim requires it to approach A/(B+2C) computed from their coefficients; observing a terminal velocity (η0'' → 0 with η0' constant) or an acceleration that drifts with radius would falsify the central claim.","tokens_in":8517,"feed_emoji":"🌀","tokens_out":5501,"duration_ms":54035,"temperature":0.7,"pith_summary":"Rayleigh–Taylor bubbles on spherical interfaces behave differently from their flat-interface counterparts: instead of settling to a constant terminal velocity, they enter a stage of uniform acceleration. The paper claims that in the nonlinear regime the bubble-tip satisfies |η0'|/sqrt(r0+η0) → constant and η0'' → constant, with the constant fixed by mode degree, Atwood number, and gravity. This matters for long-time predictions in spherical inertial-confinement-fusion configurations and other converging/diverging flows, where assuming a flat-interface terminal speed would misstate late-time bubble growth.","feed_headline":"Spherical bubbles accelerate forever, not reach terminal speed","feed_subtitle":"Late-time spherical bubbles keep accelerating at one fixed rate, beating cylindrical-growth estimates by 3×.","key_machinery":"The central object is a two-unknown local model of the bubble tip: the interface near the pole is written as η(θ,t)=η0(t)+η2(t)θ^2, and each fluid's velocity potential is represented by the single Laplace-equation eigenmode selected for the spherical mode l, with the light-fluid potential allowed to be singular at the center but used only near the tip. Imposing the kinematic and dynamic boundary conditions yields a closed pair of ordinary differential equations for η0 and η2. Their large-time integration produces the asymptotic balance (r0+η0)A = (r0+η0)η0''B + (η0')^2 C, from which the constant-acceleration result follows. All of the geometry and density-ratio dependence is carried in the t","core_discovery":"Spherical geometry introduces an expanding (converging-gravity) or shrinking (diverging-gravity) intrinsic length scale, so the natural asymptotic quantity is the ratio of bubble-tip velocity to the square root of the instantaneous radius, |η0'|/sqrt(r0+η0), and this ratio—not the velocity itself—approaches a constant. Equivalently, the tip acceleration tends to a positive constant A/(B+2C), meaning outward bubbles keep accelerating and inward bubbles keep decelerating without saturation. The coefficients A, B, C are explicit algebraic functions of the mode number l, Atwood number, and gravity for the two configurations. The same local-expansion model reproduces exponential linear growth and","pith_inferences":["If the 3:1 sphere-to-cylinder asymptotic ratio persists at higher modes, late-time bubble growth in implosion geometries could be significantly faster than cylindrical prototypes suggest; extending the DNS to l=12–16 and larger Atwood numbers would test that directly.","The model omits surface tension and viscosity; these may cut off the acceleration when the bubble tip approaches the center in the diverging-gravity case, since the singular light-fluid potential is only a local approximation. A regular inner-region model could reveal where the acceleration law breaks.","A practical consequence left implicit: for double-cone ignition targets, estimates of bubble-spike impact timing near the axis should be revised relative to planar estimates, because the bubble keeps accelerating instead of coasting.","The constant-acceleration law suggests a simple late-time similarity: bubble-tip displacement grows like t^2 with a coefficient set by A/(B+2C); this could be used as a reduced-order input for mix models."],"forward_implications":["Long-time bubble-radius or mixing-width estimates for spherical RTI should use uniform acceleration, not a terminal velocity; flat-interface models would understate late-time growth in converging gravity and overstate it in diverging gravity.","For the same effective wavenumber, the sphere's asymptotic acceleration is larger than the cylinder's, and the ratio approaches 3 for high modes independently of Atwood number, so spherical convergence is a stronger late-time driver than cylindrical convergence.","The model unifies linear and nonlinear stages, recovering the exponential growth of spherical RTI in the linear phase and the new acceleration law after several e-folding times.","In the high-mode limit the accelerating solution reduces to the classical constant bubble-velocity law for a tube, so the new result is consistent with the flat/tube picture at large l."],"fun_headline_variants":["Spherical RTI bubbles never stop accelerating","Bubble growth in spheres: constant acceleration, no terminal velocity","Spherical bubbles keep accelerating at fixed rate","In spherical RTI, bubbles accelerate without limit","Spherical geometry makes RTI bubbles accelerate forever"],"cache_read_input_tokens":2304,"weakest_assumption_plain":"The load-bearing premise is that the bubble-tip interface remains adequately described by the truncated local form η0(t)+η2(t)θ^2 with single-mode Legendre potentials; if higher harmonics become dynamically important in the asymptotic stage, the closed two-equation model and the constant-acceleration conclusion would not hold.","fun_headline_variants_meta":{"raw":{"variants":["Spherical RTI bubbles never stop accelerating","Bubble growth in spheres: constant acceleration, no terminal velocity","Spherical bubbles keep accelerating at fixed rate","In spherical RTI, bubbles accelerate without limit","Spherical geometry makes RTI bubbles accelerate forever"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.000152,"raw_usage":{"total_tokens":989,"prompt_tokens":639,"completion_tokens":350,"prompt_tokens_details":{"cached_tokens":256},"prompt_cache_hit_tokens":256,"prompt_cache_miss_tokens":383,"completion_tokens_details":{"reasoning_tokens":292}},"tokens_in":383,"tokens_out":350,"duration_ms":3752,"temperature":1.0,"reasoning_tokens":292,"cache_read_input_tokens":256,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-03T10:32:37.138037+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"Run a single-mode spherical RTI simulation or experiment with, say, l=10 and Atwood number 0.5 and measure the bubble-tip acceleration η0'' long after the linear stage. The paper's claim requires it to approach A/(B+2C) computed from their coefficients; observing a terminal velocity (η0'' → 0 with η0' constant) or an acceleration that drifts with radius would falsify the central claim.","supporting_citations":[],"review_version":1}