{"id":"a7d4b394-5a25-486e-a96a-2af46afbdf7d","arxiv_id":"2607.05275","paper_version":1,"verdict":"CONDITIONAL","confidence":"MODERATE","novelty_score":5.0,"correctness_risk":"unknown","formal_verification":"none","parameter_count":6,"one_line_summary":"SPH simulations of AGB wind-binary interactions over 600+ orbital periods show that slow winds with massive companions produce circumbinary disks via wind Roche-lobe overflow, with disk morphology depending on eccentricity and mass ratio.","lead":"This paper uses hydrodynamics simulations to show that winds from dying stars in binary systems can form circumbinary disks, with disk shape and size depending on wind speed, companion mass, and orbital eccentricity. It matters because observed post-AGB disks are common but their formation mechanism has been unclear, and these results show a natural pathway through wind-binary interaction.","discovery_kind":"unclear","skeptic_critique":{"model":"glm-5.2","headline":"The locally isothermal EoS with H/R=0.1 is the load-bearing assumption: it maximally favors post-shock cooling and disk binding, and the binary-isothermal comparison (run 10) does not test this because it retains the same H/R and isothermal treatment.","rationale":"The reader correctly identified the thermodynamic treatment as the load-bearing concern. I agree that the locally isothermal EoS with H/R=0.1 is the weakest link, and that the binary-isothermal comparison (run 10) does not adequately address it because it retains both the isothermal assumption and the same H/R. The CONDITIONAL verdict is appropriate: the mechanism is plausible and well-demonstrated within the adopted parameter regime, the 600-orbit duration is a genuine advance, and the qualitative dependence on binary eccentricity and mass ratio is likely robust. But the quantitative claim that WRLOF 'naturally' produces circumbinary disks cannot be fully assessed without testing whether disk formation survives at the observationally motivated H/R values. The paper is honest about this limitation, which is appropriate, but the limitation is load-bearing for the central claim rather than a secondary caveat. No other concern rises to the same level of importance: the wind injection simplification is validated against Esseldeurs et al. (2023), the α_eff analysis supports (though does not rigorously prove) that spiral-driven transport dominates over artificial viscosity, and the lack of a convergence study is a secondary concern given the >10^6 particle counts. The verdict should remain CONDITIONAL with the specific recommendation that an H/R sensitivity test is the single most important additional computation.","tokens_in":15500,"tokens_out":2500,"duration_ms":61947,"concrete_test":"Rerun run 5 (or run 6) with H/R = 0.2 and H/R = 0.3, keeping all other parameters identical. If a circumbinary disk still forms via the L2 outflow mechanism with comparable surface density and radial extent, the central claim is robust. If the circumsingle disk fails to form or the circumbinary disk is substantially less dense or absent, the claim that WRLOF 'naturally' produces circumbinary disks is weakened to applying only in the low-H/R regime.","verdict_should_be":"UNCHANGED","load_bearing_attack":"The central claim — that WRLOF naturally produces circumbinary disks — depends on wind material remaining gravitationally bound after passing through shocks at the wind-disk interface and within spiral arms. The locally isothermal EoS with H/R=0.1 assumes instantaneous removal of post-shock thermal energy, which maximally favors binding. The authors themselves cite Malfait et al. (2024a) finding that no disk forms when H-I cooling is neglected, directly demonstrating that the thermodynamic treatment is decisive. Their partial mitigation (run 10, binary isothermal) does not address this concern: it changes the temperature profile shape but remains isothermal with the same H/R=0.1, so cooling is still assumed instantaneous. Observational estimates place H/R at 0.17–0.3 for post-AGB circumbinary disks (Gallardo Cava et al. 2021), and Malfait et al. (2024a) find H/R varying from 0.1 to 0.3 in simulations with realistic cooling. At H/R=0.2–0.3, the higher thermal energy content would make wind capture less efficient and could prevent circumsingle disk formation in some of the explored parameter space, which would break the L2 outflow mechanism that produces the circumbinary disk. No simulation with H/R > 0.1 is presented, so the robustness of the disk formation claim across the observationally motivated parameter range is untested.","agreement_with_reader":"agree"},"referee_report":{"model":"glm-5.2","summary":"The manuscript presents SPH simulations of AGB wind-binary interactions using PHANTOM, investigating the formation of circumbinary disks in both the Bondi-Hoyle and WRLOF regimes. The key finding is that in the WRLOF regime (slow winds, massive companions), a circumsingle disk forms around the companion and transfers material through the L2 point into circumbinary orbits, producing a long-lived circumbinary disk. Simulations extend over 600 orbital periods—significantly longer than most prior work—and systematically explore the effects of binary eccentricity (e_b = 0, 0.2, 0.4) and mass ratio (q = 0.52, 1.0, 1.92). The authors find that higher eccentricities produce more extended and eccentric disks, while lower mass ratios yield lower-density disks. Angular momentum transport is shown to be dominated by spiral structures and shocks rather than the imposed artificial viscosity. The paper is clearly written and the physical picture is well-motivated.","tokens_in":16233,"tokens_out":1118,"duration_ms":68294,"significance":"The primary contribution is the long-term evolution of circumbinary disks formed via WRLOF, extending to 600 orbital periods where most prior simulations stop at ~100. The distinction between BH and WRLOF regimes via the specific energy ratio η (Eq. 2) provides a useful diagnostic. The α_eff analysis (Eq. 10) demonstrating that physical stresses dominate over imposed viscosity is a valuable check. Comparisons with Esseldeurs et al. (2023) for spiral structure and Chen et al. (2017, 2020) for disk formation provide external validation. However, the significance of the central claim is tempered by the thermodynamic treatment: the locally isothermal EoS with H/R = 0.1 is acknowledged by the authors as likely favoring disk formation, and no simulation with H/R > 0.1 is presented to test robustness across the observationally motivated range (0.17–0.3).","major_comments":[{"comment":"Section 2 and Section 4 (Discussion): The locally isothermal EoS with H/R = 0.1 is load-bearing for the central claim that WRLOF naturally produces circumbinary disks. The authors themselves cite Malfait et al. (2024a) finding that no disk forms when H-I cooling is neglected, and note that observational estimates place H/R at 0.17–0.3. The partial mitigation (run 10, binary isothermal) does not address this concern because it retains the same H/R = 0.1 and isothermal treatment—cooling is still effectively instantaneous. No simulation with H/R > 0.1 is presented. The claim in the abstract and conclusions that WRLOF 'naturally' produces circumbinary disks should be qualified to state that this result holds under the assumption of efficient cooling (small H/R), and that robustness at observationally motivated H/R values remains untested. At minimum, the authors should add a simulation at HR","section":null}],"minor_comments":[{"comment":"Table 1: The EoS column lists 'Locally Isothermal' and 'Binary Isothermal' but the distinction between these two prescriptions is only explained in Section 4. A brief footnote in the table caption would help.","section":null},{"comment":"Equation (10): The alpha_eff stress proxy is defined but the spatially averaged or rms value is not reported quantitatively in the text. The statement that typical values range between -0.3 and 0.3 is given, but a radial profile or azimuthally averaged comparison with alpha = 0.1 would make the claim more quantitative.","section":null},{"comment":"Section 2: The statement that the injection rate is 'not equivalent to the physical mass loss rate' because particles fall back is important but briefly stated. A quantitative estimate of the fraction of injected mass that is re-accreted versus escaping would clarify the effective mass-loss rate.","section":null},{"comment":"Figure 5: The y-axis label uses e_d for disk-averaged eccentricity, but the text sometimes refers to 'disk eccentricity' and sometimes to particle eccentricity e_p. Please ensure consistent terminology.","section":null},{"comment":"Section 3.1: The terminal wind velocities of ~20 and ~14 km/s are compared to Esseldeurs et al. (2023) values of 15–25 km/s. The injection velocity for runs 1–2 is 42 km/s. A brief discussion of the deceleration mechanism would help the reader assess this comparison.","section":null},{"comment":"The reference to 'S. Huang et al. 2025' in the conclusions appears to be self-referential; please ensure this is properly cited and that readers unfamiliar with that work are given sufficient context.","section":null}],"recommendation":"minor_revision","confidential_remarks":"The stress-test concern about H/R = 0.1 is valid and well-documented by the authors themselves. I judge this to be a qualification issue rather than a fatal flaw: the authors are transparent about the limitation and the mechanism (L2 outflow from circumsingle disk) is physically motivated and consistent with prior work. A minor revision requesting qualification of the 'naturally' language and ideally one additional simulation at H/R = 0.2 would substantially strengthen the paper. If the authors cannot run the additional simulation, explicit qualification of the claims is still necessary."},"author_rebuttal":{"model":"glm-5.2","summary":"We thank the referee for a careful and constructive report. The referee's central concern regarding the thermodynamic treatment and its potential impact on disk formation is well-taken. We address it below.","responses":[{"response":"We agree with the referee that the thermodynamic treatment is the most important caveat in our work, and we appreciate the referee pushing us on this point. We will make the following changes in the revised manuscript.","revision_made":"yes","referee_comment":"The locally isothermal EoS with H/R = 0.1 is load-bearing for the central claim that WRLOF naturally produces circumbinary disks. The authors cite Malfait et al. (2024a) finding no disk when H-I cooling is neglected, and note observational estimates place H/R at 0.17–0.3. The partial mitigation (run 10, binary isothermal) does not address this because it retains H/R = 0.1. No simulation with H/R > 0.1 is presented. The claim that WRLOF 'naturally' produces circumbinary disks should be qualified, and at minimum a simulation at H/R > 0.1 should be added."},{"response":"We agree. The word 'naturally' in the abstract and conclusions overstates what our simulations demonstrate, given that H/R = 0.1 assumes efficient cooling and sits at the lower end of the observationally inferred range. We will revise the abstract to read that wind-binary interactions 'can generate' rather than 'can naturally generate' diverse circumbinary disk morphologies, and we will add an explicit caveat in both the abstract and the conclusions stating that our results assume efficient cooling (H/R = 0.1) and that robustness at the observationally motivated range H/R ~ 0.17–0.3 has not yet been tested. The Discussion (Section 4) already acknowledges this limitation; we will strengthen the language there as well to make clear that H/R = 0.1 likely favors disk formation.","revision_made":"yes","referee_comment":"Qualify the claim in the abstract and conclusions that WRLOF 'naturally' produces circumbinary disks to state that this result holds under the assumption of efficient cooling (small H/R), and that robustness at observationally motivated H/R values remains untested."},{"response":"We agree that this is the most direct way to address the concern. We will add a new simulation with H/R = 0.2 (within the observationally inferred range of 0.17–0.3, citing Gallardo Cava et al. 2021 and Malfait et al. 2024a), using the same binary parameters as run 5 (e_b = 0, q = 1.92, eta = 0.781, Gaussian injection). Given computational constraints, we will run this simulation for at least 200 orbital periods, which is sufficient to determine whether a circumsingle disk forms around the companion and whether material is transferred through L2 into circumbinary orbits — the key qualitative question. We will present the results in a new figure and discuss them in the context of the referee's concern. If disk formation is suppressed or significantly weakened at H/R = 0.2, we will state this explicitly and further temper our conclusions. If a disk still forms, we will note that the result is robust to at least this value of H/R, while acknowledging that the full parameter space remains to be explored. We note that a full 600-orbit simulation at higher H/R would be computationally expensive and is beyond what we can complete within the revision timeframe, but the shorter run will directly address whether the qualitative disk-formation mechanism survives at a more realistic disk thickness.","revision_made":"yes","referee_comment":"At minimum, add a simulation at H/R > 0.1 to test robustness."}],"tokens_in":15115,"tokens_out":1278,"duration_ms":43788,"standing_objections":[]},"desk_editor":{"model":"glm-5.2","letter":"The main thing to know: this paper runs SPH simulations of AGB wind-binary interactions for 600+ orbital periods, far longer than prior work (Chen et al., Lu et al., Scherbak et al. all ran <100 orbits). That long-term evolution is the real contribution. The parameter scan over eccentricity and mass ratio is useful and the disk morphologies look physically reasonable — eccentric binaries produce more extended, eccentric circumbinary disks; lower mass companions produce lower-density disks. The distinction between BH and WRLOF regimes via the specific energy ratio η is clean and well-motivated. The α_eff analysis showing spiral-driven stresses dominate over imposed viscosity is a nice check that the results aren't artifacts of the artificial viscosity prescription. The spiral structure in the BH regime matches Esseldeurs et al. (2023) benchmarks reasonably well, including terminal wind velocities. Credit is due for the systematic approach and honest discussion of limitations in Section 4 — the authors flag their own soft spots, which is good practice but doesn't substitute for fixing them. The soft spot is real and the stress-test note lands it correctly: the locally isothermal EoS with H/R = 0.1 maximally favors post-shock cooling and disk binding. The authors acknowledge this and cite Malfait et al. (2024a) finding no disk forms without H-I cooling. Their mitigation — run 10 with a binary isothermal EoS — does not address the concern. It changes the temperature profile shape but keeps the same H/R = 0.1 and isothermal treatment, so cooling is still instantaneous. Observational estimates place H/R at 0.17–0.3 for post-AGB circumbinary disks. No simulation with H/R > 0.1 is presented. This means the central claim — that WRLOF naturally produces circumbinary disks — is demonstrated only at the most favorable point in thermodynamic parameter space. Whether disks form at H/R = 0.2 or 0.3 is untested and could go either way. This is a minor-to-moderate concern: the mechanism is physically plausible and consistent with prior work, but the quantitative disk properties and the robustness of disk formation across the observationally motivated range are not established. A resolution convergence study is also absent. This paper is for researchers working on post-AGB binary evolution and circumbinary disk formation. It provides a useful framework and long-term benchmarks even if the thermodynamic simplification limits direct comparison to observed systems. It deserves a serious referee. The referee should push hard for at least one run at H/R = 0.2 and a resolution convergence test before acceptance.","headline":"Long-term (600+ orbit) SPH simulations of circumbinary disk formation through WRLOF — a genuine advance in timescale, but the thermodynamic treatment is load-bearing and under-tested.","tokens_in":16367,"tokens_out":644,"would_cite":false,"duration_ms":49301,"reading_group":"maybe","serious_thinker":"yes","would_accept_peer_review":true},"rs_alignment":null,"lean_confirmation":null,"pith_extraction":{"msc":[],"pacs":[],"model":"glm-5.2","headline":"Slow winds and massive companions forge circumbinary disks","keywords":["circumbinary disk","AGB star wind","wind Roche-lobe overflow","binary mass transfer","post-AGB stars","smoothed-particle hydrodynamics","angular momentum transport"],"falsifier":"If simulations with self-consistent radiative transfer and realistic cooling (H/R ≈ 0.2–0.3) fail to produce circumbinary disks in the WRLOF regime for the same binary parameters, the formation pathway demonstrated here would be an artifact of the simplified thermodynamics.","tokens_in":15503,"feed_emoji":"🌀","tokens_out":1182,"duration_ms":22599,"temperature":0.7,"pith_summary":"This paper uses smoothed-particle hydrodynamics simulations of AGB star binaries, run for over 600 orbital periods, to show that circumbinary disks form naturally when a slow stellar wind from an aging giant star interacts with a sufficiently massive companion. In this regime, called wind Roche-lobe overflow (WRLOF), the companion captures wind material into a small disk around itself, and that disk transfers angular momentum to the outflow, letting material escape through the outer Lagrange point (L2) and accumulate into a larger circumbinary disk. The authors demonstrate that the resulting disk morphology — its radial extent, eccentricity, and density — depends sensitively on the binary's eccentricity and mass ratio: higher eccentricity produces more extended and more eccentric disks, while lower-mass companions yield thinner, more slowly growing disks. They also find that angular momentum transport within the circumbinary disk is driven primarily by spiral arms and shocks from the binary-wind interaction, rather than by the imposed numerical viscosity.","feed_headline":"Slow AGB winds build circumbinary disks through companion capture","feed_subtitle":"Simulations over 600 orbits show that wind Roche-lobe overflow naturally forms disks whose shape tracks binary eccentricity and mass ratio.","key_machinery":"The specific energy ratio η = e_max / (e_max − e_min) governs whether wind material remains bound to the binary. When 0 ≤ η ≤ 1, a fraction of the wind is captured, enabling the WRLOF → L2 outflow → circumbinary disk formation chain.","core_discovery":"The central mechanism is a two-step angular-momentum transfer chain: the AGB wind fills the companion's Roche lobe, forms a circumsingle disk, and that disk feeds material through the L2 point into circumbinary orbits. Whether this chain activates depends on a dimensionless energy ratio η that compares the specific energy of injected wind particles to the binary's gravitational potential. When η < 1 (slow winds, massive companions), enough material remains gravitationally bound for the WRLOF pathway to operate and build a circumbinary disk. When η > 1 (fast winds or low-mass companions), the wind stays unbound and only produces a spiral density pattern. Over 600 orbits, the circumbinary disk","pith_inferences":["If the disk aspect ratio H/R is as large as observations suggest (0.17–0.3), the WRLOF disk-formation pathway may operate only in a narrower region of parameter space than these simulations imply, potentially requiring even slower winds or more massive companions.","The sensitivity of disk properties to binary eccentricity suggests that observed circumbinary disk eccentricities could be used to infer or constrain the orbital eccentricity of the underlying binary, even when direct orbital parameters are difficult to measure.","Since the companion supplies all angular momentum to the disk, systems with very low mass ratios may produce disks too tenuous to be detectable, potentially explaining why some post-AGB binaries lack observed circumbinary disks."],"forward_implications":["The diversity of observed circumbinary disk morphologies around post-AGB stars can be explained by varying binary eccentricity and mass ratio, without invoking additional formation channels like common-envelope ejection.","Extended low-density disks around post-AGB systems like AC Her may form directly from low-density material without passing through a compact high-density phase, consistent with the low mass-ratio simulations.","Angular momentum transport in these disks is dominated by binary-induced spiral shocks rather than turbulent viscosity, which affects how one models disk evolution and accretion in post-AGB systems.","The polar-aligned circumbinary disk observed in AC Her cannot be explained by this co-planar wind-binary interaction mechanism, indicating an additional physical process is needed for polar disk alignment."],"fun_headline_variants":["How binary stars build circumbinary disks from AGB winds","Circumbinary disk formation hinges on wind speed and companion mass","Slow winds and massive companions trigger circumbinary disk growth","AGB wind capture by companion feeds circumbinary disk over 600 orbits","Energy ratio decides whether AGB winds form disks or spirals"],"cache_read_input_tokens":0,"weakest_assumption_plain":"The simulations adopt a locally isothermal equation of state with a disk aspect ratio of H/R = 0.1, which assumes efficient cooling and relatively cool gas. Observations and other simulations suggest the real value could be 0.17–0.3, and higher thermal energy would make wind capture less efficient and disk formation harder. The authors acknowledge this likely biases their results in favor of disk formation.","fun_headline_variants_meta":{"raw":{"variants":["How binary stars build circumbinary disks from AGB winds","Circumbinary disk formation hinges on wind speed and companion mass","Slow winds and massive companions trigger circumbinary disk growth","AGB wind capture by companion feeds circumbinary disk over 600 orbits","Energy ratio decides whether AGB winds form disks or spirals"]},"model":"glm-5.2","effort":"high","cost_usd":0.0,"raw_usage":{"total_tokens":695,"prompt_tokens":623,"completion_tokens":72,"prompt_tokens_details":null},"tokens_in":623,"tokens_out":72,"duration_ms":9766,"temperature":1.0,"reasoning_tokens":null,"cache_read_input_tokens":0,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-07-07T20:08:48.021203+00:00","model_set":{"reader":"glm-5.2"},"falsifier":"If simulations with self-consistent radiative transfer and realistic cooling (H/R ≈ 0.2–0.3) fail to produce circumbinary disks in the WRLOF regime for the same binary parameters, the formation pathway demonstrated here would be an artifact of the simplified thermodynamics.","supporting_citations":[],"review_version":1}