{"id":"b3c2cb9e-d07b-4dc0-bbd5-82a44e081d2d","arxiv_id":"2412.04419","paper_version":1,"verdict":"CONDITIONAL","confidence":"MODERATE","novelty_score":6.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":6,"one_line_summary":"A refined 3D hydrodynamics study finds that cores become surrounded by a near-hydrostatic 'contact binary' envelope that suppresses torques, yielding inspiral timescales near 1e5 to 1e6 orbits when the softening length is at most 0.1 a_b and cell size at most 6e-3 a_b.","lead":"Simulations of a binary spiraling inside a shared stellar envelope show that the inspiral slows when a dense, corotating gas structure forms around the two cores, with predicted inspiral times of 100,000 to 1,000,000 orbits. The paper also gives practical resolution and softening requirements that many current common-envelope simulations do not meet.","discovery_kind":"extension","skeptic_critique":{"model":"deepseek-v4-flash","headline":"The stall mechanism and the 1e5–1e6 P_orb timescales rest on the no-accretion assumption: without sinks, mass piles up into the quasi-hydrostatic shared envelope that suppresses torque, and Eq. (17) shows the numbers are not robust if real cores accrete.","rationale":"The paper is an internally consistent, carefully executed numerical parameter study. The softening/resolution analysis is supported by a clean symmetry argument, by fixed-orbit versus live-orbit comparisons, and by an honest exclusion of run E05.S.l.q03 via criterion (B.6). The absence of code/data artifacts and of uncertainty quantification on the convergence thresholds justifies only moderate confidence but is not the decisive scientific issue. The load-bearing assumption is the neglect of core accretion. It is load-bearing because the proposed stall mechanism is literally the growth of a mass pile-up around point masses; removing that pile-up with sinks is not a small perturbation to the torque. The authors themselves flag this in Sec. 3.2.1 and Sec. 4.3, and Eq. (17) provides the modified evolution equation, so the paper's own statements support reading the 1e5-1e6 P_orb values as conditional on zero accretion. That is exactly what the reader's CONDITIONAL verdict captures, and my stress-test does not identify a distinct, stronger objection that would move the verdict. I therefore recommend UNCHANGED. The proposed sink run would settle the matter: if tau_ab remains ~1e5-1e6 P_orb with accretion, the central claim is robust; if it drops toward ~1e3 P_orb, the abstract's quantitative claims must be restated as a no-accretion limiting case.","tokens_in":24611,"tokens_out":6734,"duration_ms":67312,"concrete_test":"Re-run the converged q=1/3 model E005.S.l.q03 at the same resolution, adding a numerical sink around each core with radius r_sink ~ 0.01 a_b (about three cells at level 8) and an accretion prescription of the Dittmann & Ryan (2021) or Dempsey et al. (2022) type, with sink timescale set to a few local dynamical times. Evolve Eq. (17) instead of Eq. (16) and record tau_ab and the symmetry metric ||nabla rho x nabla Phi||/(||nabla rho|| ||nabla Phi||) averaged over orbits 1625-2125. If tau_ab drops by more than an order of magnitude, or if the quasi-hydrostatic shared envelope fails to form, the no-accretion assumption is confirmed as load-bearing for the central claim.","verdict_should_be":"UNCHANGED","load_bearing_attack":"The central claim (Abstract, Sec. 3.2.1, Sec. 4.1) is that post-dynamical inspiral stalls because the cores become embedded in a corotating, nearly hydrostatic shared envelope whose symmetry makes the gravitational torque nearly vanish. This equilibrium is not externally imposed; it forms because mass accumulates around the point masses. Section 3.2.1 states explicitly that such mass accumulation and equilibrium is 'permitted by the absence of core mass accretion.' In real CEE the cores have finite radii and can accrete; numerical mass sinks remove the accumulated gas and break the rho = rho(Phi) symmetry that underlies Eqs. (14)-(15). If that symmetry is broken, the torque-suppression mechanism itself is not operative, not merely shifted in magnitude. Quantitatively, Eq. (16) omits mass and angular-momentum accretion, and the authors' Eq. (17) shows how accretion modifies d a_b/dt. Section 4.3 concedes that sinks could increase the torque and shorten the timescale to about 1e3 P_orb. Thus the headline values of ~1e5 P_orb (q=1/3) and ~1e6 P_orb (q=1) are not a prediction for real CEE but an upper-envelope result of the no-accretion idealization. This is not internal inconsistency: the symmetry argument is valid for a barotropic hydrostatic envelope, the resolution study is careful, and the authors are transparent about the limitation. But the decisive physical question, whether the stall mechanism survives when cores accrete at non-negligible rates, is untested in the paper.","agreement_with_reader":"agree"},"referee_report":{"model":"deepseek-v4-flash","summary":"This paper presents a numerical study of the post-dynamical inspiral phase of common envelope evolution, using statically refined 3D hydrodynamic simulations that include the binary and the intraorbital region. The authors vary the softening kernel, the softening length epsilon, and the grid resolution, and find that quantities such as the inspiral timescale and volume-averaged shear rate converge only for epsilon <= 0.1 a_b and delta <= 6e-3 a_b. In all well-resolved runs, a corotating, nearly hydrostatic envelope forms around the two cores, resembling the shared envelope of a contact binary. The authors argue that this structure suppresses the gravitational torque through a symmetry argument, leading to asymptotic inspiral timescales of about 1e5 P_orb for q = 1/3 and about 1e6 P_orb for q = 1. The paper also reports that the kinetic helicity shows no segregation, making large-scale alpha-effect dynamo action unlikely, and that intermittent pressure-driven polar outflows appear even without magnetic fields.","tokens_in":25009,"tokens_out":14620,"duration_ms":146272,"significance":"If the proposed mechanism is correct, it is an important conceptual advance for common envelope evolution: it identifies the stall of the dynamical inspiral with the formation of a hydrostatic, contact-binary-like shared envelope rather than with gas corotation or reduced Bondi-Hoyle drag. The convergence thresholds for softening and intraorbital resolution are practically useful and imply that many published CEE simulations are under-resolved. Strengths include the parameter-free symmetry argument, the systematic resolution and softening survey, the detailed comparison of softening kernels in Appendix B, and a transparent discussion of the no-accretion idealization. The main quantitative conclusions, however, inherit the limitations of that idealization, and the headline timescales should be scoped accordingly.","major_comments":[{"comment":"The central stall mechanism and the headline timescales are conditional on the absence of core accretion. Section 3.2.1 states that the mass accumulation and hydrostatic equilibrium are \"permitted by the absence of core mass accretion,\" Section 3.2.3 gives Eq. (17) showing how accretion modifies da_b/dt, and Section 4.3 concedes that numerical mass sinks could increase the gravitational torque and shorten the timescale. Section 4.1 further notes that significant core accretion can shorten the timescale to about 1e3 P_orb. Since real cores have finite radii and may accrete, the quoted values of ~1e5 P_orb (q=1/3) and ~1e6 P_orb (q=1) are an upper-limit result of the non-accreting setup rather than a robust prediction for CEE. I recommend either including exploratory runs with mass sinks or explicitly reframing the abstract and conclusions to present these values as idealized non-accreting limits, with the caveat that the stall mechanism itself may be weakened if accretion disrupts the rho = rho(Phi) symmetry.","section":"Section 3.2.1, Eqs. (14)-(15)"},{"comment":"The stated antisymmetry conditions are not the symmetry of the two-core potential. Equations (14) and (15) describe central inversion about an individual core, which is not a symmetry of the binary potential for either q=1 or q=1/3 unless the companion's potential is negligible. The correct argument is reflection across the plane containing the two cores and perpendicular to the orbital plane, under which rho(x,y,z)=rho(x,-y,z) and dPhi/dphi(x,y,z)=-dPhi/dphi(x,-y,z); this is valid for any mass ratio and preserves the intended conclusion. These equations should be corrected and the accompanying text revised, since they constitute the load-bearing derivation of the torque suppression.","section":"Section 3.2.3, Figs. 6-7, Table 1"},{"comment":"The asymptotic timescales are inferred from torque averages over the last 500 orbits, yet the implied orbital separation change over that interval is only about 0.5% for q=1/3 (tau ~ 9e4 P_orb) and about 0.03% for q=1 (tau ~ 1.5e6 P_orb). The paper does not report the time series of the torque or the variance and convergence of the 500-orbit average, so it is unclear whether these extremely small mean torques are significant above numerical noise or are influenced by rare events and transients. Please provide the torque time series and a measure of the averaging error (for example, running averages or standard deviations) to support the claimed asymptotic values.","section":"Section 3.2.3, Figs. 6-7, Table 1"}],"minor_comments":[{"comment":"The simulation labels \"E005.S.l.q033\" and \"E05.S.l.q033\" should read \"E005.S.l.q03\" for consistency with Table 1.","section":"Section 3.3 and Fig. 9"},{"comment":"The description \"r_i = sqrt(x_i^2+y_i^2+z_i^2) is the distance to the binary's center of mass\" is confusing; it should read \"r_i is the position vector of each core relative to the binary's center of mass.\"","section":"Section 2.1"},{"comment":"There is a typo \"In our in simulations\" that should be corrected to \"In our simulations.\"","section":"Section 3.3"},{"comment":"The sentence \"See Eq. (1 for an example...\" has a missing closing parenthesis after Eq. (1).","section":"Appendix A"},{"comment":"The spelling of the softening prescription is inconsistent (\"Ruffert\" and \"Ruffert\"); please standardize.","section":"Throughout"}],"recommendation":"major_revision","confidential_remarks":"The no-accretion limitation is the main reason I do not recommend acceptance in the present form; it is a real idealization that affects the central quantitative claim, although the authors are transparent about it. The symmetry-equation error in Section 3.2.1 needs correction, but the intended reflection argument is sound and easily fixed. I see no scope or novelty concerns for A&A; the paper fits the journal and will be valuable once the headline claims are scoped and the symmetry derivation is corrected."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"Short version: this deserves a real referee, and the referee will mostly be arguing with the authors' idealizations, not with the numerics. The genuinely new piece is the claim that the post-dynamical inspiral stalls because the two cores sit in a corotating, nearly hydrostatic shared envelope, and that the symmetry of that configuration kills the gravitational torque. The symmetry argument is clean and parameter-free: hydrostatic equilibrium gives rho=rho(Phi), the torque density is antisymmetric, and the integrated torque nearly vanishes. That is a real mechanism, and it is distinct from the usual \"gas corotates\" story.\n\nThe paper also does something useful on softening and resolution. The convergence study is careful: well-resolved runs with epsilon <= 0.1 ab agree, coarse runs disagree strongly, and the authors identify a necessary condition (B.6) for an asymptotic small-epsilon regime. They are appropriately skeptical of old runs that use large softening and poor intraorbital resolution. Credit where due: they walk through the different softening kernels, and their warning that many published CEE runs are effectively under-resolved looks credible.\n\nSoft spots, in order of size. The headline timescales (1e5-1e6 Porb) are outputs of a no-accretion model. The shared envelope that suppresses torque forms because mass piles up around point masses; the authors say so explicitly in Sec. 3.2.1. With mass sinks, that accumulation goes away, rho=rho(Phi) breaks, the torque returns, and the authors themselves note the timescale can drop to about 1e3 Porb. So the quantitative result is an upper envelope of the idealization, not a prediction for real CEE. That is not a hidden flaw--it is in Sec. 4.3--but it is load-bearing, so the paper's framing should keep the timescales conditional. Secondary: single idealized setup, no gas self-gravity, no MHD, no uncertainty quantification on the convergence thresholds themselves. Those are minor in context; the parameter study is internally consistent.\n\nAll in all: the paper is honest, the mechanism is physically interesting, and the resolution findings will be useful to anyone running or interpreting CEE simulations. The citation pattern is fine; this builds directly on their own earlier models and on circumbinary-disk resolution work, and the new elements are identifiable.\n\nRecommendation: send it to peer review. The referee should push on how much of the stall mechanism survives accretion, but that is a revision-level issue, not a desk-reject-level one.","headline":"Careful, honest numerical study whose resolution criteria are solid and whose stall mechanism is real but conditional on no core accretion; worth a serious referee.","tokens_in":25516,"tokens_out":2727,"would_cite":true,"duration_ms":75958,"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 the post-dynamical inspiral of a common-envelope binary stalls because the two cores become wrapped in a corotating, nearly hydrostatic shared envelope whose symmetry cancels the gravitational torque, leaving a slow…","keywords":["common envelope evolution","post-dynamical inspiral","gravitational torque","gravitational softening","hydrostatic equilibrium","contact binary","kinetic helicity","polar outflows"],"falsifier":"Run the same binary-plus-envelope setup with numerical mass sinks that let the cores accrete: if the quasi-hydrostatic shared envelope fails to form, or the gravitational torque remains large and the inspiral timescale drops toward $\\sim10^3$ orbital periods, then the no-accretion assumption is what produces the reported stall. Alternatively, a well-resolved simulation satisfying $\\epsilon\\le0.1\\,a_\\mathrm{b}$ and $\\delta\\le6\\times10^{-3}\\,a_\\mathrm{b}$ that does not develop the contact-binary-like structure would falsify the mechanism.","tokens_in":24338,"feed_emoji":"⭐","tokens_out":11894,"duration_ms":105102,"temperature":0.7,"pith_summary":"Post-dynamical common envelope inspiral is usually studied with simulations that stop early because of limited resolution and artificial softening of the gravitational potential around the two cores. This paper builds three-dimensional hydrodynamical models of that late phase with statically refined meshes and several softening prescriptions, and asks what actually slows the inspiral down. The answer is a shared, nearly hydrostatic envelope: within a few tens of orbits, the cores are wrapped in a corotating gas structure like a contact binary, and its symmetry suppresses the gravitational torque that would otherwise continue shrinking the orbit. For well-resolved runs, the inspiral timescale converges to about $10^5$ orbital periods for mass ratio $q=1/3$ and about $10^6$ orbital periods for $q=1$. Convergence requires a softening length at most $0.1$ orbital separations and a mesh spacing at most $6\\times10^{-3}$ separations, which suggests that many published simulations are effectively under-resolved.","feed_headline":"Inspiral stalls when cores share a corotating gas envelope","feed_subtitle":"Well-resolved simulations give 100,000–1,000,000 orbit inspiral times; coarser runs miss the torque-suppressing cocoon.","key_machinery":"The load-bearing object is the corotating quasi-hydrostatic shared envelope that forms around the two cores, analogous to the common envelope of contact binaries. It works through a symmetry argument: hydrostatic equilibrium implies $\\rho=\\rho(\\Phi)$, and the torque density $s\\rho\\,\\partial\\Phi/\\partial\\varphi$ is antisymmetric about the plane perpendicular to the orbital plane that contains the line joining the cores, so the net gravitational torque from the symmetric structure nearly vanishes. The quantitative criteria that let this structure survive in a simulation are the softening length $\\epsilon\\le0.1\\,a_\\mathrm{b}$ and the mesh spacing $\\delta\\le6\\times10^{-3}\\,a_\\mathrm{b}$, because a larger softening radius weakens the hydrostatic support of the gas and a coarser grid fails to resolve the pressure gradients inside the softened regions. Among the tested prescriptions, the spline-softened potential gives the most accurate gradient outside the softening sphere.","core_discovery":"The central discovery is that the dynamical inspiral of a binary inside a common envelope does not end because the gas corotates with the orbit, but because the two cores become embedded in a corotating, nearly hydrostatic structure resembling the shared envelope of a contact binary. In hydrostatic equilibrium the density is a function of the potential alone, $\\rho=\\rho(\\Phi)$, while the gravitational torque density is antisymmetric with respect to the plane that contains both cores and is perpendicular to the orbital plane, so the product integrates to almost zero net torque. With that torque suppressed, the orbit contracts on a secular timescale of about $10^5$ orbital periods for $q=1/3$ and about $10^6$ orbital periods for $q=1$, rather than continuing the rapid dynamical plunge. The same simulations show that this quasi-hydrostatic state is only maintained when the gravitational softening length satisfies $\\epsilon\\le0.1\\,a_\\mathrm{b}$ and the intraorbital grid spacing satisfies $\\delta\\le6\\times10^{-3}\\,a_\\mathrm{b}$, and that softer potentials and coarser grids create artificial asymmetries that mimic a stronger torque. The paper also argues that kinetic helicity is not segregated by hemisphere, so large-scale magnetic fields are unlikely to grow through the usual $\\alpha$-effect, yet pressure-driven polar outflows appear even without magnetic fields.","pith_inferences":["If real cores accrete gas at non-negligible rates, the quasi-hydrostatic shared envelope would be drained and the torque-suppressing symmetry broken; the paper's own discussion suggests the inspiral timescale could then shorten toward roughly $10^3$ orbital periods. This is an editorial extrapolation from the no-accretion setup, not a result the simulations demonstrate.","The contact-binary analogy points to effects the paper does not model: the shared envelope could redistribute heat between the cores, and shear between the corotating envelope and non-synchronously spinning cores could drive enhanced magnetic activity, akin to the elevated activity observed in contact binaries.","A decisive numerical test would be to repeat the setup with mass sinks or subgrid accretion on the cores; if the hydrostatic envelope fails to form and the inspiral timescale drops, the no-accretion assumption is the load-bearing cause of the $10^5$–$10^6$ orbital-period numbers.","Because the paper's model grid intentionally covers only moderate mass ratios and circular orbits, the $q$ dependence and the residual envelope geometry could differ for extreme mass ratios or eccentric orbits, a regime the simulations do not address and which would need separate runs."],"forward_implications":["Post-dynamical inspiral lasts on the order of $10^5$ to $10^6$ orbital periods, so post-common-envelope binaries emerge through a long quasistationary contraction rather than a fast continuation of the plunge.","Softening radii larger than $0.1\\,a_\\mathrm{b}$ or grid spacings larger than $6\\times10^{-3}\\,a_\\mathrm{b}$ change more than the numerical accuracy: they generate artificial torques and qualitatively different envelope ejection, so published simulations in that regime need re-evaluation.","The transition from dynamical to post-dynamical inspiral is tied to the formation of the torque-suppressing hydrostatic envelope, giving a concrete physical criterion for where one phase ends and the other begins.","A large-scale magnetic dynamo through the $\\alpha$-effect is unlikely in this phase, while centrifugally collimated, pressure-driven polar outflows can appear without any magnetic field.","Final orbital separation increases with mass ratio: $q=1$ binaries halt further out and shrink about ten times more slowly than $q=1/3$ binaries of the same total mass."],"supporting_citations":[{"why":"Supply the idealized post-dynamical inspiral setup that this paper extends by including the intraorbital region and the cores themselves.","marker":"Gagnier & Pejcha (2023, 2024)"},{"why":"Provides the contact-binary shared-envelope description used as the analogy for the corotating hydrostatic structure.","marker":"Lucy (1968)"},{"why":"Shows that hydrostatic equilibrium near a binary suppresses the gravitational torque, supporting the stall mechanism.","marker":"Kim (2010)"},{"why":"Defines the spline softening whose gradient becomes exactly Newtonian outside the softening scale and which the paper adopts for most runs.","marker":"Hernquist & Katz (1989)"},{"why":"Defines an alternative softened potential whose gradient inside the softening sphere is much shallower and which the paper finds less suitable.","marker":"Ruffert (1993)"},{"why":"Establishes that torques are sensitive to the shape of the softened potential, motivating the systematic softening comparison.","marker":"Dong et al. (2011)"},{"why":"Represents the earlier corotation-based explanation for the end of the dynamical inspiral that the paper's hydrostatic-envelope mechanism replaces.","marker":"Reichardt et al. (2019)"},{"why":"Supplies the hydrodynamics code used to run the simulations.","marker":"Stone et al. (2020)"}],"fun_headline_variants":["Inspiral stalls when cores get a corotating gas cocoon","Contact-binary-like envelope ends dynamic plunge","Coarse grids fake rapid inspiral; resolution matters","No dynamo, but polar outflows survive in common envelope"],"cache_read_input_tokens":3200,"weakest_assumption_plain":"The load-bearing premise is that the two cores do not accrete any gas, so mass can pile up around them and build the nearly hydrostatic shared envelope that cancels the torque; if real cores swallow gas at a non-negligible rate, the stall mechanism and the quoted $10^5$ to $10^6$ orbital-period timescales would not hold.","fun_headline_variants_meta":{"raw":{"variants":["Inspiral stalls when cores get a corotating gas cocoon","Contact-binary-like envelope ends dynamic plunge","Coarse grids fake rapid inspiral; resolution matters","No dynamo, but polar outflows survive in common envelope"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.000951,"raw_usage":{"total_tokens":4168,"prompt_tokens":1169,"completion_tokens":2999,"prompt_tokens_details":{"cached_tokens":384},"prompt_cache_hit_tokens":384,"prompt_cache_miss_tokens":785,"completion_tokens_details":{"reasoning_tokens":2933}},"tokens_in":785,"tokens_out":2999,"duration_ms":23152,"temperature":1.0,"reasoning_tokens":2933,"cache_read_input_tokens":384,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-11T21:23:55.559551+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"Run the same binary-plus-envelope setup with numerical mass sinks that let the cores accrete: if the quasi-hydrostatic shared envelope fails to form, or the gravitational torque remains large and the inspiral timescale drops toward $\\sim10^3$ orbital periods, then the no-accretion assumption is what produces the reported stall. Alternatively, a well-resolved simulation satisfying $\\epsilon\\le0.1\\,a_\\mathrm{b}$ and $\\delta\\le6\\times10^{-3}\\,a_\\mathrm{b}$ that does not develop the contact-binary-like structure would falsify the mechanism.","supporting_citations":[{"cited_title":"& Pejcha , O","cited_arxiv_id":null,"evidence_quote":"Supply the idealized post-dynamical inspiral setup that this paper extends by including the intraorbital region and the cores themselves."},{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"Provides the contact-binary shared-envelope description used as the analogy for the corotating hydrostatic structure."},{"cited_title":"2010, , 725, 1069","cited_arxiv_id":null,"evidence_quote":"Shows that hydrostatic equilibrium near a binary suppresses the gravitational torque, supporting the stall mechanism."},{"cited_title":"& Katz , N","cited_arxiv_id":null,"evidence_quote":"Defines the spline softening whose gradient becomes exactly Newtonian outside the softening scale and which the paper adopts for most runs."},{"cited_title":"1993, , 280, 141","cited_arxiv_id":null,"evidence_quote":"Defines an alternative softened potential whose gradient inside the softening sphere is much shallower and which the paper finds less suitable."},{"cited_title":"A., De Marco , O., Iaconi , R., Tout , C","cited_arxiv_id":null,"evidence_quote":"Represents the earlier corotation-based explanation for the end of the dynamical inspiral that the paper's hydrostatic-envelope mechanism replaces."}],"review_version":1}