{"id":"04cc2baa-c95c-4fa0-9da5-375d32bffd12","arxiv_id":"2505.06344","paper_version":3,"verdict":"CONDITIONAL","confidence":"MODERATE","novelty_score":6.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":5,"one_line_summary":"Chaotic 'diving orbits' in an axisymmetric Milky Way model can deliver stars and binaries to the central black hole faster than two-body scattering, especially for wide binaries, with ejecta preferentially launched toward the galactic poles.","lead":"This paper shows that in the flattened potential of the Milky Way's center, stars on certain low-angular-momentum orbits follow chaotic paths that can repeatedly graze the central black hole without any gravitational scattering by other stars. These 'diving orbits' may produce most tidal disruption events and hypervelocity stars, with ejecta preferentially shot out of the galactic plane.","discovery_kind":"extension","skeptic_critique":{"model":"deepseek-v4-flash","headline":"Uniform-filling test in Fig. 6 runs ~3e5 orbits, but Eq. 58's disruption timescale is ~1e2-1e3 orbits; sticky short-time occupancy could bias tD and the claimed rates.","rationale":"The reader identified uniform occupancy of the chaotic sea as the weakest assumption. I agree that this is the right area, but I sharpen it to a finite-time mixing problem: the paper's own Poisson test runs over ~3e5 orbits, whereas the rates use a per-orbit disruption probability that corresponds to mean first-passage times of ~1e2-1e3 orbits. Sticky or slowly mixing regions, which the authors themselves observe around regular islands, could make the effective short-time occupancy non-uniform and hence bias t_D, p(beta), and the derived relative rates. This is a load-bearing concern because Eq. 58 converts a phase-space volume ratio into a disruption rate, and the central claim is a quantitative dominance by orders of magnitude. I do not think the concern disproves the existence of diving orbits or the qualitative collisionless channel: the topological classification, Figures 7-8, and the long-time Poisson test are independent support for eventual ergodicity. However, the absence of a first-passage-time test, combined with no released code or data, justifies keeping the verdict at CONDITIONAL rather than moving to ACCEPT. The proposed ensemble test would settle whether the short-time occupancy assumption actually holds in the parameter region that dominates the rates.","tokens_in":43202,"tokens_out":9128,"duration_ms":104220,"concrete_test":"Integrate an ensemble of ~100 chaotic low-lz trajectories in the fiducial GC model with a point-mass (or a_BH < 1e-5 pc) MBH, for Rc = 0.5, 1, 3 pc and lz/lc = 0.001, 0.003, 0.01. For lt = 0.3, 3, 30 pc^2/Myr, record the first periapse with lp < lt, and compare the median first-passage time to t_D from Eq. 58. Also compute the empirical distribution of l_theta,p over the first 100, 300, and 1000 periapses versus the uniform prediction. If medians agree within a factor of ~2 and occupancy is uniform at 100 orbits, the concern is resolved; if first-passage times are systematically longer, the collisionless rates in Section 6.3 need downward revision.","verdict_should_be":"UNCHANGED","load_bearing_attack":"The quantitative rates rest on the assumption (Section 4.3, Eqs. 46-58) that a chaotic orbit uniformly fills its accessible (l_theta,p, theta_p) region, so the probability of a disrupting periapse per radial period is p(D) ~ lambda_t (1-|gamma_z|)/F_dive and the mean disruption time is t_D = T_c F_dive (1-|gamma_z|)/lambda_t (Eq. 58). For the systems that dominate the claimed rates (Rc ~ 0.5-3 pc; lt ~0.3-30 pc^2/Myr), lambda_t = lt/lc is small, so 1/p(D) is roughly 10^2-10^3 radial periods. The uniform-filling test in Figure 6, however, integrates a single trajectory over ~3e5 radial periods and checks aggregate Poissonian occupancy. That validates ergodicity on timescales two to three orders of magnitude longer than the mean disruption time. Sticky boundaries around regular islands are explicitly noted in Section 4.3; if a trajectory spends even a few hundred orbits in such a sticky phase before reaching low l_theta,p, the effective first-passage time can substantially exceed t_D, and the full-loss-wedge rates in Section 6.3 are overestimated. The same short-timescale occupancy controls p(beta) and the claimed polar ejection anisotropy. The paper acknowledges the assumption may be violated for realistic DFs (Section 5) but does not test the relevant short-time behaviour.","agreement_with_reader":"partial"},"referee_report":{"model":"deepseek-v4-flash","summary":"The paper studies the dynamics of low-angular-momentum orbits in an observationally benchmarked axisymmetric model of the Milky Way's inner 200 pc, focusing on chaotic 'diving orbits' that can bring stars or binaries arbitrarily close to the central massive black hole without two-body scattering. The authors show that such orbits are common for small conserved z-angular momentum, derive an analytic distribution p(beta) for the encounter depth assuming uniform filling of the chaotic region in periapsis phase space, compute disruption timescales t_D, and compare the resulting collisionless disruption rates with standard collisional loss-cone rates for a range of tidal angular momenta lt. They conclude that the collisionless channel can dominate the collisional one, by orders of magnitude for large lt (wide binaries, giants), and predict that disrupted material is preferentially ejected toward the galactic poles.","tokens_in":43473,"tokens_out":6691,"duration_ms":71810,"significance":"If the results hold, the paper identifies a potentially important, largely overlooked channel for tidal disruptions and hypervelocity-star production in axisymmetric galactic centers, with a distinctive polar ejection anisotropy that is testable with HVS surveys and TDE host-galaxy orientations. The analytic derivation of p(beta) ∝ beta^(-3/2) from a geometric area integral is elegant and parameter-free once uniform filling is assumed. The numerical work is substantial and reproducible in structure: 300,000-orbit integrations, Lyapunov exponent classification, Poisson-statistics checks of occupancy, and a clear comparison with Magorrian & Tremaine (1999). The sensitivity study of the NSC flattening (Section 4.5) is a useful addition. The main quantitative claims, however, rest on an ergodicity assumption whose validity on the relevant (short) timescales is not directly tested, and the headline 'orders of magnitude' statement applies to an idealized single-lt population; these points are addressed in the major comments.","major_comments":[{"comment":"The uniform-filling assumption that underpins p(beta), t_D, and all subsequent rates is validated only through aggregate occupancy over ~300,000 radial periods (Figure 6), while the mean disruption time t_D = T_c F_dive / [lambda_t (1 - |gamma_z|)] is typically 10^2-10^3 radial periods for the systems dominating the rates (Rc ~ 0.5-3 pc, lt ~ 0.3-30 pc^2/Myr). If a chaotic trajectory experiences sticky phases near regular islands before exploring the low-l_theta,p region, the first-passage time to a disrupting pericenter can be substantially longer than 1/p(D), biasing the rates in Section 6.3 upward. The paper acknowledges sticky boundaries (Section 4.3) and notes that realistic DFs may violate the assumption (Section 5), but it does not test the short-time occupancy or the distribution of waiting times between low-l_theta excursions. I ask the authors to run targeted integrations that measure the first-passage-time distribution to beta >= 1 for representative (Rc, lz, lt) values, or to provide a quantitative argument that the long-time Poisson statistics imply exponential inter-arrival times with the same mean on the t_D timescale.","section":"Section 4.3 and Eq. (58)"},{"comment":"The abstract's claim that the 'relative collisionless rate can dominate by orders of magnitude' is only demonstrated for an idealized population consisting entirely of systems with a single, large lt (e.g., lt = 30 pc^2/Myr in Figure 14). For TDE-relevant lt ~ 0.3 pc^2/Myr the enhancement is only a factor of a few, and the paper explicitly cautions (Section 6) that the single-lt calculation cannot be integrated over a realistic lt distribution without additional assumptions about the population and survivability of binaries/giants. As written, the abstract overstates the generality of the 'orders of magnitude' conclusion. I recommend the abstract and discussion be qualified to state that the large enhancement applies to systems with large tidal angular momentum (wide binaries and giants), while for main-sequence-star TDEs the collisionless channel is competitive but only moderately dominant.","section":"Abstract and Section 6.5 (Figure 14)"}],"minor_comments":[{"comment":"Equation (68) appears to be missing a division sign: the full loss-cone differential rate should scale as n(rho, gamma)/T_r(rho, gamma), not n(rho, gamma) * T_r(rho, gamma), consistent with the subsequent integration in Eq. (71).","section":"Section 6.4.1, Eq. (68)"},{"comment":"The phrase 'Most of these disruptions involve stars come from the Nuclear Stellar Cluster' contains a grammatical error; 'come' should be removed.","section":"Abstract"},{"comment":"The sentence 'Stars are are scattered ... by two-body encounters' duplicates 'are'; please correct.","section":"Section 1"},{"comment":"The text refers to 'Noether's theorum'; the correct spelling is 'theorem'.","section":"Section 2.1"},{"comment":"The sentence 'These simplifying assumption can be complicated for Hills mechanism separations' should be 'These simplifying assumptions can be complicated ...'.","section":"Section 5.1.2"},{"comment":"The axis labels in Figure 2 appear garbled in the compiled PDF (e.g., '½ = log10', '² [pc2 Myr¡2]'); please verify the rendering of the LaTeX labels.","section":"Figure 2"}],"recommendation":"major_revision","confidential_remarks":"The paper is well within the scope of MNRAS and makes a solid contribution to the study of collisionless loss-wedge dynamics. The main technical concern (short-timescale ergodicity) is addressable with additional numerical experiments, and the authors' acknowledgment of the single-lt limitation is a useful starting point for the needed qualification. I would encourage the editor to seek a revision rather than rejection, as the qualitative conclusions are likely robust even if the quantitative rates shift by a factor of a few. One editorial note: the paper repeatedly refers to 'Penoyre et al. 2025, in prep' and an unpublished NSC DF by 'Vasiliev et al. in prep'; these dependencies should be clarified, for instance by making the relevant codes/models publicly available alongside the revision."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"Zeph, here is my take. The paper is a real step beyond the standard loss-cone/loss-wedge picture. The periapse-based surface of section is a good idea, and it feeds a clean analytic result: p(beta) ~ beta^{-3/2} from area-filling, steeper than the geometric beta^{-1} that orbit-averaged treatments give. Writing everything in terms of the single tidal angular momentum lt is effective, and it lets the authors cover TDEs, binaries, and giants on the same footing. The numerical campaign—hundreds of thousands of orbits, Lyapunov classification, Fdive maps—is substantial. The qualitative claim that low-lz trajectories in the S22 Milky Way model are commonly chaotic and dive to arbitrarily small pericenters is well supported, and the predicted polar ejection anisotropy is sharp and testable.\n\nThe soft spots are real but not fatal to the central idea. The quantitative rates depend on the assumption that a chaotic orbit uniformly fills its periapse phase space quickly. Figure 6 tests that with ~300,000 radial periods, while the disruption timescales that dominate the rates are ~100–1,000 periods. A sticky orbit near a regular island can take much longer to reach low l_theta,p, which would push t_D up and the collisionless rates down. The paper acknowledges stickiness and even notes that realistic DFs may not fill uniformly, but it does not measure first-passage statistics on the timescale that matters. That is the first thing I would ask for.\n\nSecond, the headline “orders of magnitude” is for a model where every system has one lt, not a population. The authors are open about this and about the survival/lifetime caveats, which keeps the absolute-rate claim honest, but it also means the central comparison is more illustrative than definitive. Third, no code or data are released. Combined with fitted DFs and a hand-tuned GMC relaxation boost, the quantitative rates are hard to audit.\n\nWho gets value: anyone working on TDE rates, hypervelocity star production, or S-star dynamics. It deserves a serious referee. I would send it out, with the request that the authors either release the integration and DF code or provide a first-passage-time test for representative sticky orbits, and then soften the “primarily collisionless” summary unless the population-level caveats are addressed.","headline":"A credible, well-scoped case that collisionless chaotic orbits dominate binary disruptions in an axisymmetric galactic center, but the headline rates rest on an ergodicity assumption tested on timescales much longer than the disruption process, and no artifacts are released.","tokens_in":44092,"tokens_out":2322,"would_cite":true,"duration_ms":25386,"reading_group":"yes","serious_thinker":"yes","would_accept_peer_review":true},"rs_alignment":null,"lean_confirmation":null,"pith_extraction":{"msc":[],"pacs":[],"model":"deepseek-v4-flash","headline":"This paper claims that chaotic 'diving orbits' in the flattened, axisymmetric potential of the Milky Way's inner 200 pc can deliver stars and binaries to the central massive black hole without any two-body scattering, and that this…","keywords":["chaotic orbits","tidal disruption events","hypervelocity stars","galactic center","axisymmetric potential","loss wedge","diving orbits","nuclear stellar cluster"],"falsifier":"Integrate a set of low-$\\ell_z$ chaotic trajectories in the paper's potential for more than $10^6$ radial periods and histogram periapse occupancy in the low-$\\ell_{\\theta,p}$ corner: if the count CDF deviates from the Poisson expectation beyond shot noise for $\\ell_z/\\ell_c \\lesssim 0.1$, the uniform-filling assumption fails and the derived rates and $\\beta$ distribution change. Observationally, a kinematically unbiased sample of hypervelocity stars with three-dimensional velocities showing no excess toward the galactic poles would contradict the predicted injection anisotropy.","tokens_in":42936,"feed_emoji":"🕳️","tokens_out":7341,"duration_ms":70716,"temperature":0.7,"pith_summary":"This paper claims that the flattened, axisymmetric potential of the Milky Way's inner 200 pc has its own, collisionless route for feeding stars and binaries to the central black hole. Orbits with near-zero vertical angular momentum can be chaotic, and chaotic 'diving' orbits randomly sample periapse angular momentum, so a system eventually passes inside the tidal radius without needing two-body scatterings. For a fixed tidal angular momentum, the paper derives a probability distribution for the 'diving factor' $\\beta=(\\ell_t/\\ell_p)^2$ that scales as $\\beta^{-3/2}$, and folds it into a model of the Galactic Center to compute disruption rates. The central quantitative claim is that this collisionless channel can dominate the collisional two-body scattering rate: by orders of magnitude for binaries and giants, and by a factor of a few for main-sequence tidal disruption events. If true, most Milky Way disruptions would arrive nearly perpendicular to the disk and eject debris or hypervelocity stars preferentially toward the galactic poles.","feed_headline":"Chaotic diving orbits may dominate Milky Way star disruptions","feed_subtitle":"Flattened center lets orbits randomly plunge to the black hole, out-producing scattering rates by orders of magnitude.","key_machinery":"The central object is the 'diving orbit': a chaotic trajectory in an axisymmetric potential whose polar angular momentum $\\ell_\\theta$ varies under the non-spherical torque while the vertical component $\\ell_z$ is conserved, allowing periapse to wander down to the minimum permitted by $\\ell_z$. The argument is carried by the periapse surface of section: periapsides of a chaotic orbit uniformly fill the accessible $(\\theta_p,\\ell_{\\theta,p})$ region, so probabilities of disruption reduce to area integrals. That yields the key analytic identity $p(\\beta)\\propto\\beta^{-3/2}$ for the diving factor $\\beta=(\\ell_t/\\ell_p)^2$, steeper than the geometric $\\beta^{-1}$ scaling, and the 'loss wedge' description of the vulnerable region $|\\ell_z|<\\ell_t$.","core_discovery":"On the paper's own terms, the discovery is that in an observationally benchmarked axisymmetric model of the Milky Way center, the loss wedge $|\\ell_z|<\\ell_t$ is fed primarily by chaotic orbits rather than by relaxation. Any trajectory that reaches the minimum pericenter permitted by its conserved $\\ell_z$ is chaotic; regular orbits cannot dive. Because chaotic trajectories fill their accessible periapse phase space roughly uniformly, the probability per periapse passage of disruption is $p(D)=\\lambda_t(1-|\\gamma_z|)/F_{\\rm dive}$, and the disruption time is $t_D\\approx T_c F_{\\rm dive}(1-|\\gamma_z|)/\\lambda_t$, producing encounter properties with $p(\\beta)\\propto\\beta^{-3/2}$, a steep preference for shallow, grazing, near-parabolic encounters that are strongly misaligned with the disk. Integrating over the Galactic Center distribution function, the collisionless relative rate $\\bar\\Gamma_d$ exceeds the collisional $\\bar\\Gamma_s$ by orders of magnitude for large tidal angular momentum (binaries, giants) and by a factor of a few for small $\\ell_t$ (solar-type TDEs).","pith_inferences":["Beyond the paper: the steep $\\beta^{-3/2}$ distribution implies partial disruptions are far more common than full ones; if repeated weak encounters accumulate, they may measurably alter the spin, mass, and orbital energy of surviving stars before the final disruption, an effect the authors flag but do not model.","Beyond the paper: the same diving mechanism can supply extreme-mass-ratio inspirals to the central black hole without two-body relaxation; the authors mention EMRIs as a motivation but do not compute their rates.","Beyond the paper: if the assumed uniform phase-space filling is replaced by a sticky chaotic distribution, the relative rate ordering may shift; a direct test is to compare the observed sky distribution of hypervelocity stars with the predicted polar excess.","Beyond the paper: triaxiality, which the authors note may be present in the inner galaxy, would likely enlarge the chaotic sea and strengthen the collisionless channel further; quantifying this for observed bar and bulge shapes would be a natural next step."],"forward_implications":["For a Milky Way composed entirely of solar-type stars, the model gives a total TDE rate of a few $\\times 10^{-5}\\,\\mathrm{yr}^{-1}$, with the collisionless channel responsible for the majority.","Binary and giant-star disruptions (the Hills mechanism and S-star formation) may be boosted by one to three orders of magnitude relative to scattering estimates, so observed hypervelocity star and S-star populations may be substantially fed by chaotic diving.","Encounter geometry makes ejecta anisotropic: hypervelocity stars and tidal debris should preferentially emerge along the galactic poles rather than isotropically.","The process is generic: any galaxy with a flattened nuclear cluster and a central massive black hole should have a comparable or stronger collisionless disruption channel, since chaotic diving appears even for nearly spherical nuclear cluster shapes."],"supporting_citations":[{"why":"Sets up the standard loss-cone theory for collisional star disruptions that this paper compares against.","marker":"Frank & Rees 1976"},{"why":"Formulates two-body relaxation filling of the loss cone, part of the collisional rate baseline.","marker":"Lightman & Shapiro 1977"},{"why":"Provides the empty loss-cone Fokker-Planck solution used for the collisional rate in the empty regime.","marker":"Cohn & Kulsrud 1978"},{"why":"Introduces the loss wedge and the orbit-averaged map for collisionless disruption in axisymmetric nuclei; this is the paper's direct baseline and method comparison.","marker":"Magorrian & Tremaine 1999"},{"why":"Supplies the fiducial self-consistent potential and distribution function of the Galactic Center used for all orbit integrations and rates.","marker":"Sormani et al. 2022"},{"why":"Previous collisionless loss-wedge calculation, including relaxation-time scalings and regular-orbit precession, which the paper extends to chaotic diving orbits and larger tidal radii.","marker":"Vasiliev & Merritt 2013"},{"why":"Establishes Hills-mechanism binary separation outcomes and the caveat that moderate $\\beta$ can still disrupt, informing the disruption criterion.","marker":"Sari et al. 2010"},{"why":"Argues massive perturbers such as giant molecular clouds dominate collisional scattering at large radii; used for the massive-perturber test of the rate comparison.","marker":"Perets et al. 2007"},{"why":"Cited as the source of the caveat that a galactic-center distribution function might violate the uniform-occupancy assumption on which the analytic probabilities rest.","marker":"Thomas et al. 2004"},{"why":"Provides the relaxation-time formulas and the loss-cone rate formalism used in the collisional comparison.","marker":"Merritt 2013a"}],"fun_headline_variants":["Chaotic orbits drive star disruptions in Milky Way center","Chaotic plunges dominate star disruption in Galactic center","Chaotic orbits outpace scattering for star deaths in Milky Way","Chaotic paths boost star disruption rates in Galactic center","Milky Way's chaotic stars dive into black hole disruptions"],"cache_read_input_tokens":3200,"weakest_assumption_plain":"The load-bearing premise is that each chaotic diving orbit fills its allowed periapse region uniformly, so the probability of a deeply disruptive passage is simply the area of the tidal region divided by the area of the chaotic sea.","fun_headline_variants_meta":{"raw":{"variants":["Chaotic orbits drive star disruptions in Milky Way center","Chaotic plunges dominate star disruption in Galactic center","Chaotic orbits outpace scattering for star deaths in Milky Way","Chaotic paths boost star disruption rates in Galactic center","Milky Way's chaotic stars dive into black hole disruptions"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.000771,"raw_usage":{"total_tokens":3448,"prompt_tokens":1014,"completion_tokens":2434,"prompt_tokens_details":{"cached_tokens":384},"prompt_cache_hit_tokens":384,"prompt_cache_miss_tokens":630,"completion_tokens_details":{"reasoning_tokens":2355}},"tokens_in":630,"tokens_out":2434,"duration_ms":16569,"temperature":1.0,"reasoning_tokens":2355,"cache_read_input_tokens":384,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-15T22:44:55.923572+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"Integrate a set of low-$\\ell_z$ chaotic trajectories in the paper's potential for more than $10^6$ radial periods and histogram periapse occupancy in the low-$\\ell_{\\theta,p}$ corner: if the count CDF deviates from the Poisson expectation beyond shot noise for $\\ell_z/\\ell_c \\lesssim 0.1$, the uniform-filling assumption fails and the derived rates and $\\beta$ distribution change. Observationally, a kinematically unbiased sample of hypervelocity stars with three-dimensional velocities showing no excess toward the galactic poles would contradict the predicted injection anisotropy.","supporting_citations":[],"review_version":1}