{"id":"47445e83-335f-44cd-8cc0-5870eb4c2c97","arxiv_id":"2502.02933","paper_version":3,"verdict":"CONDITIONAL","confidence":"MODERATE","novelty_score":6.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":4,"one_line_summary":"3D simulations predict a pre-common-envelope accretion disk around a 1.4 solar mass neutron star with mass ~5e-3 solar masses, radius ~40 solar radii, and scale height ~5 solar radii.","lead":"This paper uses 3D hydrodynamic simulations to predict what the accretion disk around a neutron star looks like just before a red supergiant companion is swallowed in a common envelope event: a disk about 40 times the Sun's radius, holding about 0.5% of the Sun's mass. The result gives common envelope modelers a concrete starting condition for accretion and jet feedback.","discovery_kind":"new_application","skeptic_critique":{"model":"deepseek-v4-flash","headline":"The factor-of-few mass uncertainty excludes the dominant thermodynamic sensitivity: the paper's own 3D γ=4/3 run changes disk mass by two orders of magnitude, so the central claim is conditional on the fixed γ=1.1, no-cooling EOS.","rationale":"The central claim is a numerical prediction with uncertainty bounds. The existing variation suite (Sections 4.1–4.4) is well designed for numerical and setup parameters, but all those runs share the same quasi-isothermal compressibility. A physical prediction can only carry the advertised factor-of-few uncertainty if the most sensitive physical parameter is varied. The paper's own γ=4/3 comparison demonstrates that disk mass changes by two orders of magnitude when the adiabatic index moves from 1.1 to 1.33—still below the monatomic value. Since real donor envelopes have effective γ set by recombination and radiative cooling, and the paper itself notes that disk formation requires γ≲1.2, the fixed γ=1.1 choice is not an innocent calibration. A 3D run at γ=1.2 and/or with a cooling function would bracket the plausible regime. If those runs confirm stability, the central claim stands; if not, the factor-of-few bound is unsupported. The 21-year duration is less concerning because the disk residence time Mdisk/Macc ≈ 1.5 yr is much shorter than the run length, so the disk is approximately in balance with the imposed mass-transfer rate; self-gravity is also negligible (Q≫1 for the quoted disk). Thus the thermodynamic sensitivity is the single most load-bearing concern, matching the reader's weakest assumption about fixed γ=1.1 and no explicit cooling.","tokens_in":24378,"tokens_out":11168,"duration_ms":109887,"concrete_test":"Run sim-0 for 21 yr with the same setup but with γ=1.2, and separately with a simple optically thin radiative cooling term added to the Euler energy equation (retaining γ=1.1), measuring Mdisk, Rdisk, and Hdisk by the same first-inflection criterion used in Figure 6. If both variants remain within a factor of ~2 of sim-0's (5.5×10^-3 M_sun, 39 R_sun), the claimed error bars survive; if either changes by an order of magnitude—as the existing γ=4/3 run does—the paper should state the disk parameters as conditional on the adopted EOS and remove the factor-of-few claim.","verdict_should_be":"UNCHANGED","load_bearing_attack":"Load-bearing concern: the claimed 'factor of a few' mass accuracy is not robust to the one physical ingredient the paper does not vary—the effective thermodynamics. The reference run uses a fixed adiabatic index γ=1.1 with no radiative cooling (Section 2.2.2). Section 4.5 shows that the same 3D setup with γ=4/3 gives Mdisk = 5.37×10^-5 M_sun and Rdisk = 15.2 R_sun, versus 5.5×10^-3 M_sun and 39 R_sun for sim-0: a factor of ~100 in mass and ~2.6 in radius, with identical injection. At γ=5/3 no disk is measurable. The robustness tests in Section 4 (injection rate, nozzle velocity, background, resolution) all stay inside the γ=1.1, no-cooling model, so they cannot bound the dominant uncertainty. The real pre-CE envelope has a temperature- and density-dependent effective gamma (partial ionization and recombination), and radiative losses are absent; the paper itself concedes that more reasonable cooling physics would need to be included. Consequently the central numbers—especially the factor-of-a-few mass estimate—are conditional on near-isothermal thermodynamics, and the stated 15% radius accuracy has no support outside that regime. The Shakura-Sunyaev comparison in Section 3.1 is circular (α=0.1 and Σ=Mdisk/Rdisk^2 are assumed, not derived) but is not the primary issue.","agreement_with_reader":"agree"},"referee_report":{"model":"deepseek-v4-flash","summary":"The paper presents 3D hydrodynamic simulations of Roche-lobe overflow mass transfer from a 7 M_sun red supergiant onto a 1.4 M_sun neutron star, targeting the phase immediately before common-envelope (CE) entry. A 1D MESA simulation supplies a 30,000-year mass-transfer history, while the 3D MEZCAL simulations follow only the last 21 years around the accretor, using a fixed quasi-isothermal equation of state with gamma=1.1 and no explicit cooling. The central claims are that the pre-CE disk has mass ~5e-3 M_sun, radius ~40 R_sun, scale height ~5 R_sun, and accretion rate ~4e-3 M_sun/yr, and that these values are robust to a factor of a few in mass and 15% in radius. The paper includes resolution, injection-rate, nozzle-velocity, background, and adiabatic-index sensitivity tests, as well as a comparison of the measured accretion rate and temperature profile with Shakura-Sunyaev theory.","tokens_in":24712,"tokens_out":3532,"duration_ms":35597,"significance":"If the central numbers hold, this would be a valuable quantitative constraint on accretion feedback in the poorly understood pre-CE phase, with direct consequences for studies of jet feedback and CE outcomes. The paper's strengths are its explicit stress testing: the resolution study (Section 4.4) shows convergence with only 12% mass variation across levels, and the parameter scans in Sections 4.1-4.3 cover several numerical and physical choices. The authors also openly state many limitations, including the absence of cooling and the short simulation window. However, the stated uncertainty budget does not include the dominant thermodynamic sensitivity, and the Shakura-Sunyaev comparison is circular; these issues bound the significance of the claimed numbers unless the claims are narrowed or the physics is extended.","major_comments":[{"comment":"The claimed robustness expressed in the abstract and Section 5 (\"reasonable within a factor of a few for the mass and 15% for the radius\") is not supported for the dominant physical uncertainty. The paper's own 3D run with gamma=4/3 gives Mdisk=5.37e-5 M_sun and Rdisk=15.2 R_sun, versus Mdisk=5.5e-3 M_sun and Rdisk=39 R_sun for sim-0 -- a factor of ~100 in mass and ~2.6 in radius -- and the gamma=5/3 run yields no measurable disk. All variations in Sections 4.1-4.4 keep gamma=1.1 and no explicit cooling, so they cannot bound the sensitivity to the effective thermodynamics. The paper should either add simulations that vary the equation of state/cooling in a physically motivated range, or explicitly restate the central claim as conditional on a near-isothermal, no-cooling model, removing the factor-of-few and 15% statements as global robustness claims.","section":"Section 4.5 and Figure 19"},{"comment":"The agreement between the simulated accretion rate and the Shakura-Sunyaev prediction is not an independent validation: Eq. (11) uses the simulation's own Mdisk, cs, and H, assumes Sigma = Mdisk/Rdisk^2, and adopts alpha=0.1 arbitrarily. The predicted rates are therefore partly determined by construction, making the comparison circular. The text should either derive alpha from the measured turbulent stress or shock dissipation in the simulation, or clearly label this as a consistency check rather than a prediction.","section":"Section 3.1, Eq. (11)"},{"comment":"The extrapolation that the 21-year disk mass is \"likely only a factor of a few smaller than if the entire phase were modelled\" is not directly supported. The cumulative mass of sim-0 is still increasing at t=21 yr (Figure 6, bottom-right panel), and the sim-mdot-2 and sim-mdot-3 runs start at different times and run for different durations without reaching a secular plateau. The paper should quantify the disk growth rate at the end of the simulation and either show evidence of approach to saturation or reframe the CE-entry mass as an upper limit.","section":"Section 2.2.2, Table 5, Figure 6"}],"minor_comments":[{"comment":"There is a duplicated word in \"We then use use the 3D hydrodynamic numerical code\"; please correct.","section":"Section 2.2"},{"comment":"The caption states \"Tbg = 105 cm s–1\" for what should be a temperature; the units should be K, and the same typo appears in the legend context.","section":"Figure 15 caption"},{"comment":"The Shakura & Sunyaev (1973) reference is listed twice with identical bibliographic data; please consolidate.","section":"References"},{"comment":"The phrase \"only slightly higher\" for gamma=4/3 compared to gamma=1.1 is imprecise; a change from 1.1 to 1.333 is substantial in compressibility, so the sentence should be rephrased.","section":"Section 4.5"},{"comment":"The text uses alpha both for the Shakura-Sunyaev viscosity parameter and for the temperature power-law exponent; this is confusing and should be disambiguated with different symbols.","section":"Section 3.1, Figure 9"},{"comment":"The conservation test modifies the boundary conditions and removes the inner inflow boundary, so it does not verify conservation in the production setup; the discussion should acknowledge that this is a simplified check.","section":"Appendix 1.1"}],"recommendation":"major_revision","confidential_remarks":"The manuscript is within scope for a stellar-astrophysics journal and the simulations are useful. The main editorial question is whether the authors are willing to narrow their uncertainty claims or invest in additional thermodynamic sensitivity runs. The paper already includes many honest caveats, so a revision that repositions the claims as conditional on the chosen equation of state could be viable; the circularity of the Shakura-Sunyaev check is secondary but should still be corrected."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"The short version: this paper gives the first concrete quantitative prediction of a pre-CE accretion disk (Mdisk ~ 5e-3 M_sun, Rdisk ~ 40 R_sun, H ~ 5 R_sun) for a 7 M_sun red supergiant plus 1.4 M_sun neutron star, with the 3D run fed by a MESA mass-transfer history. That fills a real gap: common-envelope and jet-feedback simulations rarely have pre-CE disk initial conditions to work from. The numerical work is solid: convergence tests give 12% mass change between 3 and 5 refinement levels, and the injection-rate, nozzle-velocity, and background variations keep the disk radius within ~15% and the mass within about a factor of two. The authors are also candid about what is simplified.\n\nThe load-bearing soft spot is thermodynamics. Every robustness test stays within the fixed gamma=1.1, no-cooling model. The one physical parameter they do vary is the adiabatic index: in 3D with gamma=4/3, the same injection produces Mdisk=5.4e-5 M_sun and Rdisk=15 R_sun, two orders of magnitude less mass; with gamma=5/3 there is no measurable disk. The real pre-CE envelope has an effective gamma set by partial ionization and radiative losses, which the simulations do not include. The paper itself concedes that proper cooling physics is needed. So the 'factor of a few' mass estimate is really conditional on near-isothermal thermodynamics, and the stated 15% radius accuracy does not transfer to other gamma. That caveat needs to be front and center, not in the conclusions.\n\nThe Shakura-Sunyaev consistency check in Section 3.1 is circular (alpha=0.1 and Sigma=M_disk/R_disk^2 are assumed), but it is not what the central claim rests on. Minor quibbles: no code or data shipped, and the 21-year window is short, though the sim-mdot-3 run suggests a longer evolution would stay within a factor of a few.\n\nWho this is for: anyone modeling CE in-spiral with accretion or jet feedback, and anyone needing concrete pre-CE disk parameters. It deserves a serious referee. I would send it to review, with the expectation that the authors either add a 3D gamma=1.2 run plus some cooling treatment, or substantially soften the factor-of-a-few claim to a conditional one.","headline":"Solid 3D pre-CE disk simulation, but the factor-of-few mass claim is hostage to the fixed gamma=1.1 no-cooling thermodynamics; deserves review but needs that caveat front and center.","tokens_in":25292,"tokens_out":3999,"would_cite":false,"duration_ms":37354,"reading_group":"maybe","serious_thinker":"yes","would_accept_peer_review":true},"rs_alignment":null,"lean_confirmation":null,"pith_extraction":{"msc":[],"pacs":[],"model":"deepseek-v4-flash","headline":"A 3D simulation predicts that just before a red supergiant and neutron star enter a common envelope, the neutron star hosts an accretion disk of about 0.005 solar masses, about 40 solar radii across, accreting at about 0.004 solar masses…","keywords":["accretion disks","common envelope","Roche lobe overflow","neutron star","hydrodynamics","numerical simulations","red supergiant","mass transfer"],"falsifier":"Run the same binary through a 3D simulation with explicit radiative cooling and a realistic depth-dependent adiabatic index out to the onset of the common envelope: if the disk mass falls outside roughly $1.5\\times10^{-3}$ to $1.5\\times10^{-2}\\,M_\\odot$ or the radius outside $34$ to $46\\,R_\\odot$, the paper's stated uncertainties are wrong. Alternatively, identify an observed X-ray binary with a supergiant donor overfilling its Roche lobe and measure its accretion rate from X-ray luminosity; a rate more than an order of magnitude away from $4\\times10^{-3}\\,M_\\odot\\,\\mathrm{yr}^{-1}$ would challenge the prediction.","tokens_in":24202,"feed_emoji":"🪐","tokens_out":9484,"duration_ms":77030,"temperature":0.7,"pith_summary":"This paper asks what the accretion disk around a neutron star looks like in the decades immediately before its binary companion's envelope swallows it. The authors simulate the Roche-lobe overflow from a 7-solar-mass red supergiant onto a 1.4-solar-mass neutron star in 3D, using a long 1D binary evolution calculation to set the mass-transfer rate. They predict that just before the common-envelope phase the disk holds roughly $5\\times10^{-3}\\,M_\\odot$, extends to about $40\\,R_\\odot$, has a scale height near $5\\,R_\\odot$, and accretes onto the neutron star at about $4\\times10^{-3}\\,M_\\odot\\,\\mathrm{yr}^{-1}$. These numbers matter because this disk may power jets that help decide whether the common envelope is ejected or the neutron star merges with the giant's core.","feed_headline":"A neutron star's pre-engulfment disk would span 40 solar radii","feed_subtitle":"3D simulations put its mass near 0.005 solar masses and its accretion rate near 0.004 solar masses per year.","key_machinery":"The machinery is a compact-object-centred, co-rotating 3D hydrodynamic simulation with an L1 mass-injection nozzle: gas enters through a small elliptical nozzle at a rate interpolated from a 1D stellar evolution calculation covering 30,000 years, while a low-density background surrounds a spherical inner boundary at $1.3\\,R_\\odot$. The equation of state is quasi-isothermal with $\\gamma=1.1$, chosen because earlier work found that RLOF and CE disks form only for $\\gamma\\lesssim1.2$; no self-gravity, magnetic fields, or explicit cooling are included. Disk radius is defined operationally as the first inflection point of the cumulative enclosed-mass profile, vertical density profiles define the scale height, and a specific-energy versus specific-angular-momentum diagram is used to show that the dense disk gas is bound and moving on low-eccentricity orbits.","core_discovery":"The central claim is that a pre-common-envelope disk around a neutron star fed through the inner Lagrange point should have $M_{\\rm disk}\\sim 5\\times10^{-3}\\,M_\\odot$, $R_{\\rm disk}\\sim 40\\,R_\\odot$, and $H\\sim 5\\,R_\\odot$ at the moment the in-spiral begins, with an inner-boundary accretion rate of a few $\\times10^{-3}\\,M_\\odot\\,\\mathrm{yr}^{-1}$. The disk rotates at nearly Keplerian speed near the inner boundary and at 80 to 90 percent of Keplerian at its outer edge, where pressure support matters. Its mid-plane temperature falls as $T\\propto r^{-1.1}$, steeper than the $T\\propto r^{-3/4}$ law of the standard Shakura-Sunyaev thin-disk model, and the authors attribute the difference to their quasi-isothermal equation of state without explicit cooling. Based on simulations that vary the mass-injection rate, the nozzle velocity, the background density and temperature, and the mesh resolution, they argue the disk mass is reliable to within a factor of a few and the radius to within about 15 percent.","pith_inferences":["If the disk is destroyed upon entering the common envelope, its measured mass and angular momentum set the starting budget for any disk that reforms around the in-spiralling neutron star, so these numbers anchor CE-feedback models even if the disk itself does not survive.","The simulations omit magnetic fields, and the paper notes that Gauss-level fields could be amplified by the magneto-rotational instability within a few orbits; if that happens, magnetic stress may raise the accretion rate above the measured value.","A comparison run at $\\gamma=4/3$ already produces a five-times-smaller and hundred-times-less-massive structure, so if the real envelope contains zones of reduced compressibility, disk formation in this pre-CE phase could be marginal rather than assured.","Observing a candidate pre-CE system in unstable mass transfer and comparing its inferred accretion rate and disk signatures with these predictions would test whether the simulated 21-year window is representative of the full 30,000-year RLOF phase."],"forward_implications":["The neutron star enters the common envelope carrying a disk of about $5\\times10^{-3}\\,M_\\odot$, giving the in-spiral an immediate reservoir of mass and angular momentum that can feed jets or outflows.","Because the inner accretion rate is roughly a hundred times smaller than the L1 injection rate, most of the transferred mass is stored in the disk or leaves the domain rather than being accreted promptly.","A jet launched from the inner boundary of this disk could carry a mechanical luminosity near $3\\times10^3\\,L_\\odot$, and a jet launched close to the neutron star surface could reach about $10^8\\,L_\\odot$, enough to affect envelope unbinding.","The disk's steeper-than-standard temperature profile implies that radiative cooling must be included before the Shakura-Sunyaev $T\\propto r^{-3/4}$ prescription is used for pre-CE disk temperatures.","The convergence of disk mass, radius, scale height, and accretion rate across refinement levels indicates the predicted properties are numerical, not resolution artifacts."],"supporting_citations":[{"why":"Establishes the 3D RLOF disk-formation setup and the gamma <= 1.2 disk-formation threshold used to choose the equation of state.","marker":"Makita et al. (2000)"},{"why":"Supplies the optically-thin L1 nozzle-area prescription whose scale-height dimensions set the mass-injection region.","marker":"Jackson et al. (2017)"},{"why":"Gives the Bernoulli-based L1 stream velocity used for the subsonic nozzle injection speed.","marker":"Lubow & Shu (1975)"},{"why":"Provides the Roche-lobe radius formula used by the 1D evolution model to set the onset of mass transfer.","marker":"Eggleton (1983)"},{"why":"Provides the alpha-disk accretion-rate formula and the T proportional to r^-3/4 profile that the simulation's temperature is compared against.","marker":"Shakura & Sunyaev (1973)"},{"why":"Argues disk formation inside a CE depends on the thermal properties of the envelope, supporting the gamma ~ 1.1 choice.","marker":"MacLeod & Ramirez-Ruiz (2015)"},{"why":"Shows the pre-CE and CE disk phase is transitory and connects the adiabatic index to disk formation.","marker":"Murguia-Berthier et al. (2017)"},{"why":"Identifies the magneto-rotational instability as a possible viscosity source that could amplify Gauss-level fields.","marker":"Balbus & Hawley (1991)"},{"why":"Provides the magnetically regulated accretion-rate estimate used to argue the required inner-rim field is about 28,000 G.","marker":"Wardle (2007)"}],"fun_headline_variants":["Pre-common-envelope disk around neutron star: 40 solar radii across","Neutron star's pre-CE disk holds 0.005 solar masses of gas","3D simulations size up accretion disk before common envelope","Disk feeding neutron star before engulfment: 0.005 Msun, 40 Rsun","Birth of a pre-CE disk: mass and radius from 3D hydrodynamics"],"cache_read_input_tokens":3200,"weakest_assumption_plain":"The load-bearing premise is that a 21-year, quasi-isothermal ($\\gamma=1.1$) 3D run without explicit cooling or self-gravity, driven by a 1D model of the donor's mass loss, captures the disk's state at the moment the common envelope begins; if the real envelope cools or compresses differently, or the donor responds to mass loss differently, the disk's mass and radius could move outside the stated factor of a few.","fun_headline_variants_meta":{"raw":{"variants":["Pre-common-envelope disk around neutron star: 40 solar radii across","Neutron star's pre-CE disk holds 0.005 solar masses of gas","3D simulations size up accretion disk before common envelope","Disk feeding neutron star before engulfment: 0.005 Msun, 40 Rsun","Birth of a pre-CE disk: mass and radius from 3D hydrodynamics"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.000956,"raw_usage":{"total_tokens":4194,"prompt_tokens":1183,"completion_tokens":3011,"prompt_tokens_details":{"cached_tokens":384},"prompt_cache_hit_tokens":384,"prompt_cache_miss_tokens":799,"completion_tokens_details":{"reasoning_tokens":2906}},"tokens_in":799,"tokens_out":3011,"duration_ms":19090,"temperature":1.0,"reasoning_tokens":2906,"cache_read_input_tokens":384,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-09T10:31:50.917991+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"Run the same binary through a 3D simulation with explicit radiative cooling and a realistic depth-dependent adiabatic index out to the onset of the common envelope: if the disk mass falls outside roughly $1.5\\times10^{-3}$ to $1.5\\times10^{-2}\\,M_\\odot$ or the radius outside $34$ to $46\\,R_\\odot$, the paper's stated uncertainties are wrong. Alternatively, identify an observed X-ray binary with a supergiant donor overfilling its Roche lobe and measure its accretion rate from X-ray luminosity; a rate more than an order of magnitude away from $4\\times10^{-3}\\,M_\\odot\\,\\mathrm{yr}^{-1}$ would challenge the prediction.","supporting_citations":[{"cited_title":"2000, Monthly Notices of the Royal Astronomical Society, 316, 906","cited_arxiv_id":null,"evidence_quote":"Establishes the 3D RLOF disk-formation setup and the gamma <= 1.2 disk-formation threshold used to choose the equation of state."},{"cited_title":"2017, The Astrophysical Journal, 835, 145","cited_arxiv_id":null,"evidence_quote":"Supplies the optically-thin L1 nozzle-area prescription whose scale-height dimensions set the mass-injection region."},{"cited_title":"H., & Shu, F","cited_arxiv_id":null,"evidence_quote":"Gives the Bernoulli-based L1 stream velocity used for the subsonic nozzle injection speed."},{"cited_title":"2015, The Astrophysical Journal, 803, 41","cited_arxiv_id":null,"evidence_quote":"Argues disk formation inside a CE depends on the thermal properties of the envelope, supporting the gamma ~ 1.1 choice."},{"cited_title":"2017, The Astrophysical Journal, 845, 173","cited_arxiv_id":null,"evidence_quote":"Shows the pre-CE and CE disk phase is transitory and connects the adiabatic index to disk formation."},{"cited_title":"2007, Ap&SS, 311, 35","cited_arxiv_id":null,"evidence_quote":"Provides the magnetically regulated accretion-rate estimate used to argue the required inner-rim field is about 28,000 G."}],"review_version":1}