{"id":"2eac3609-a7b9-4b84-a4d5-270f2b518a2a","arxiv_id":"2505.23407","paper_version":1,"verdict":"CONDITIONAL","confidence":"MODERATE","novelty_score":5.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":4,"one_line_summary":"Axisymmetric SPH-MHD simulations show accretion discs around oblique millisecond pulsars are disrupted when the inner disc radius lies well beyond the light cylinder, confirming the Ekşi and Alpar stability diagram.","lead":"This paper uses axisymmetric magnetohydrodynamic simulations to show that an accretion disc around a millisecond pulsar is destroyed when its inner edge sits far beyond the light cylinder and the pulsar's magnetic axis is tilted. The results confirm an earlier analytical stability map and suggest that transitional millisecond pulsars need sizeable magnetic obliquities to switch between disc and pulsar states.","discovery_kind":"new_application","skeptic_critique":{"model":"deepseek-v4-flash","headline":"The sign-undefined, once-flashed time-averaged Deutsch field is not a valid stand-in for the pulsar's time-dependent electromagnetic force, so the simulated 'instability' at x_d≥10 may be a transient response to an artificial static impulse rather than a persistent pulsar-disc effect.","rationale":"The reader's weakest assumption correctly identifies the time-averaged, sign-undefined Deutsch field flashed at t=0 as the foundational simplification, and the paper's own Section 2 and Section 3 acknowledge it. My stress-test sharpens two consequences of that assumption. First, the sign ambiguity is not just a detail about energy: the momentum equations depend on products like B_r B_φ, so the arbitrary sign choice changes the magnetic torque and the Lorentz force, and therefore can change whether a disc is classified as stable or unstable. The paper says the calculations 'work on an energy basis', but the MHD equations are not energy-averaged; they use the static sign-free field as an instantaneous vector field, which is not the time-average of the actual force. Second, the 'flash and deactivate' procedure means no continuous pulsar energy input is present, so the simulation captures only the transient response to a sudden turn-on. The short integration time (a few orbital periods of the outer simulated annulus) is too short to establish a long-timescale instability, especially since the far tracer groups are only a few orbits old when the 20% criterion is applied. These concerns do not invalidate the paper—they are consistent with the authors' stated scope of an 'energy approach' and 'stability right after the magnetic field is turned on'—but they do mean the headline claim about discs at x_d≥10 being 'always unstable' should be read as conditional on the static-field approximation. Since the reader already issued a CONDITIONAL verdict for compatible reasons, and since the paper explicitly identifies these as simplifications, I see no reason to change the verdict; the conditionality is appropriate.","tokens_in":23226,"tokens_out":6952,"duration_ms":74316,"concrete_test":"Re-run the x_d=25, ξ=20° case while maintaining the external field: reapply the same time-averaged field configuration every spin period (or continuously prescribe it as a boundary term) and integrate to at least 100 P_out. If the density of the 'middle' and 'far' tracer groups stops declining and the disc settles into a new quasi-equilibrium, the 'always unstable' conclusion is an artifact of the one-shot flash; if the disc continues to lose mass monotonically, the conclusion is robust. As a complementary check, compute the period-averaged Lorentz force <J×B> and torque from the full time-dependent Deutsch fields in the disc region and compare them with the static-magnitude field used in the simulations; an order-of-magnitude disagreement would demonstrate that the MHD force is misrepresented.","verdict_should_be":"UNCHANGED","load_bearing_attack":"The central stability map (Fig. 11) and the conclusion 'x_d ≥ 10, the disc is always unstable' rest on the prescription in Section 3: the external field is set to B_i(t=0)=<B_i^2>^{1/2} with 'undefined sign', flashed once, and then evolved only by advection (Eq. 19). This is not equivalent to a time-averaged pulsar magnetosphere. The MHD momentum equations (Eqs. 8–10) contain Lorentz-force terms such as B_r B_φ that depend on the relative sign of components; with sign-free magnitudes all products are positive, so the instantaneous magnetic stress is not the period-averaged stress of the Deutsch solution. The sign choice is never specified, and the torque in Eq. 10 can have either sign depending on that arbitrary choice, directly affecting angular-momentum transport and hence disc stability. In addition, because the field deactivates after t=0, the simulation measures the response to a single impulsive energy deposit, not to a pulsar that continuously radiates a wind/Poynting flux. The integration time is only a few inner orbital periods (at x_d=25, P_out≈6.7 s), far shorter than the day/yeartimescales over which tMSP state transitions occur, so the observed 20% density declines in the tracer groups could be transient oscillations following the initial flash rather than evidence of persistent instability. Finally, the agreement with Ekşi & Alpar (2005) is not an independent check because both approaches use the same time-averaged Deutsch energy density as their starting point.","agreement_with_reader":"agree"},"referee_report":{"model":"deepseek-v4-flash","summary":"This paper presents axisymmetric SPMHD simulations of a thin accretion disc around a millisecond pulsar, using the time-averaged RMS components of the Deutsch vacuum solution as the external magnetic field. The disc inner radius is varied from x_d = R_in/R_LC = 0.5 to 25 and the magnetic inclination from ξ = 0° to 30°. Stability is assessed by the evolution of tracer particle groups, with a 20% relative density decline criterion. The main result is that discs with x_d ≳ 10 and ξ > 0° are severely altered and classified as unstable, while discs with x_d ≲ 1 remain stable; this is compared with the analytical stability line of Ekşi & Alpar (2005). The authors further argue that tMSP state transitions can be triggered by inner-radius fluctuations by a factor of 3–4 for ξ ≳ 20°, and discuss implications for LMXBs and supernova fallback discs.","tokens_in":23570,"tokens_out":7654,"duration_ms":72737,"significance":"If the stability map is correct, the paper provides a first simulation-based test of the Ekşi–Alpar energy-density criterion and offers a plausible, falsifiable mechanism for tMSP state transitions through inner-radius variability. The paper includes genuine strengths: the numerical code is public; resolution convergence (Sec. 5.1.2) and resistivity sensitivity (Sec. 5.1.1) are explicitly tested; and the central claim is crisply stated. The main limitation is that the external magnetic field is represented by unsigned RMS components flashed once at t=0, so the simulations test the energy-model assumptions rather than the full time-dependent pulsar magnetosphere; the agreement with Ekşi & Alpar is therefore a consistency check of that shared energy model, not an independent validation.","major_comments":[{"comment":"The initial condition B_i(t=0)=⟨B_i^2⟩^{1/2} with 'undefined sign' is not equivalent to a period-averaged Deutsch magnetic field for the MHD equations being solved. The Lorentz-force terms and the stress tensor in Eqs. (8)–(12) involve products such as B_r B_φ, whose sign depends on the relative signs of the components; the period average of such a product is not the product of the RMS values. With all components assigned positive signs by construction, the azimuthal torque in Eq. (10) acquires a convention-dependent sign, which directly affects angular-momentum transport and hence the stability classification. The manuscript must specify a sign convention for each component and show that the Fig. 11 stability map is invariant under admissible sign choices, or justify why an energy-only treatment is consistent with solving vector MHD equations.","section":"§3, Eqs. (8)–(10), (19)"},{"comment":"The statement that 'the averaged pulsar radiation flashes the disc at t=0 and subsequently deactivates' means that the external Deutsch field is not maintained; after the initial flash, only the gas-advected field evolves. A real pulsar continuously supplies a rotating electromagnetic field and a Poynting/wind flux, so the simulated response is to an impulse rather than to persistent irradiation. The integration times are only a few inner orbital periods (P_out ≈ 6.7 s for x_d=25 in Table 2), far shorter than the day-to-year timescales of tMSP transitions, and the density traces in Fig. 7 oscillate. Therefore the §7 statement that 'at sufficiently large values of the inner radius, x_d ≥ 10, the disc is always unstable' needs additional support that the observed 20% density declines are secular rather than transient post-flash oscillations; a continuous-field run or an analytic persistence argument would address this.","section":"§3, Eq. (19); §5.1; §7"},{"comment":"The simulation grid has x_d = 0.5, 1, 2, 6, 25, but no run at x_d = 10 or at any radius between 6 and 25. The quantitative claim in §7 that the unstable region begins at 'x_d ≳ 10' is an extrapolation from a single far-out run at x_d=25; the location of the boundary is therefore not determined by the simulations. Adding at least one run at x_d ≈ 10, and preferably at x_d ≈ 8–15, is needed to support the claimed location of the stability boundary in Fig. 11.","section":"§6, Fig. 11; Table 2"},{"comment":"The paper itself acknowledges in §3 that the reduction to axial symmetry is 'an ad hoc artificial constraint' and that hydromagnetic instabilities are better represented in 3D. Because the central claim is a global stability statement for discs ('the disc is always unstable' for x_d ≥ 10), the axisymmetric setup excludes non-axisymmetric modes that could either destabilize the stable cases or saturate the unstable ones. The scope of the conclusion should be restricted to axisymmetric perturbations unless a 3D test (or a linear stability argument for the relevant modes) is provided.","section":"§3, axisymmetry assumption; §7"},{"comment":"The classification into stable and unstable in §6 uses a 20% relative density decline in the 'middle' or 'far' tracer groups, with no significance level or error bar. The B=0 control checks in §4 already show density fluctuations at the ~10% level, so the threshold is only a factor of two above the numerical noise floor, and the time series in Fig. 7 show oscillatory rather than monotonic behavior. The robustness of the Fig. 11 classification to the threshold value and to the choice of tracer regions should be quantified.","section":"§6, stability criterion; §4"}],"minor_comments":[{"comment":"In Eq. (19), the integrand is written as r_{ij} B_i dt, but Eq. (15) has dB_i/dt = Σ_j r_{ij} B_j; the index in the integrand should be B_j.","section":"§3, Eq. (19)"},{"comment":"There is a stray word 'radius' immediately after the paragraph ending '...not a plausible mechanism to drive the tMSP state transitions.' It appears to be a leftover fragment and should be removed.","section":"§6.1"},{"comment":"Papitto & de Martino (2022a) and (2022b) are listed as two separate references with identical titles and identical article pages; if they are distinct chapters or versions they should be distinguished, otherwise the duplicate should be merged.","section":"References"},{"comment":"The caption of Fig. 8 states that resistivity values are shown 'from bottom to top' but the main text lists them as (0.2, 1, 5); please make the ordering of panels consistent with the caption.","section":"Fig. 8"},{"comment":"The axes of Fig. 11 are not labeled; the text refers to the ξ–x_d diagram but the figure should show the parameters and units on both axes.","section":"Fig. 11"},{"comment":"The data availability statement says the code is available at 'Axis-SPHYNX download i'; the hyperlink appears to be missing or malformed.","section":"Data availability"}],"recommendation":"major_revision","confidential_remarks":"The comparison with Ekşi & Alpar (2005) is presented as an agreement, but because the same time-averaged Deutsch energy density is used as input, this is a consistency check rather than an independent validation. The sign ambiguity and the impulse-like treatment of the external field are the main technical concerns; I would ask the authors to address those and to add a run near x_d ≈ 10 before reconsidering the paper."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"First, the useful part: this is the first SPMHD simulation suite aimed at the tMSP disc-magnetosphere interaction in the regime R_in > R_LC, and it pushes the inner radius out to x_d = 25, farther than previous numerical studies. The disc models use realistic Shakura-Sunyaev structures, and the resolution and resistivity checks are sensible. The qualitative pattern is clear: oblique rotators with the inner disc far outside the light cylinder disrupt the innermost resolved disc within a few orbital periods, while discs at or inside R_LC remain largely intact. That is a genuine step beyond the analytical estimates.\n\nThe soft spots are where the paper makes its strongest claims. The magnetic field is the time-averaged Deutsch magnitude with an undefined sign, flashed once at t=0 and then only advected. That is not the same as a continuously radiating pulsar field, and because the Lorentz terms in Eq. 10 contain products like B_r B_phi, the arbitrary sign matters for angular-momentum transport even though the visible effect is mostly ohmic heating. The run times are a few inner orbital periods, so the simulation measures the response to an impulsive energy deposit, not a persistent pulsar wind. I would not call this fatal, but it means the stability classification is tied to this averaged-field prescription.\n\nSecond, the conclusion that x_d >= 10 is 'always unstable' overshoots the grid: the runs are at x_d = 0.5, 1, 2, 6, 25, with no x_d = 10 to pin the boundary. The 20% density-decline threshold is a defensible but ad hoc diagnostic, and the axisymmetry, acknowledged as an ad hoc constraint, prevents non-axisymmetric instabilities from developing. The agreement with Ekşi & Alpar is a useful consistency check but not independent, since both analyses start from the same time-averaged Deutsch energy density. The authors are honest about several limitations, yet the abstract and conclusions state the x_d >= 10 line without those caveats.\n\nThe observational discussion is the most valuable part: if the obliquity is large, R_in fluctuations by a factor of 3-4 can cross the stability boundary, which fits the stable L_X of tMSPs in the disc state. That is a concrete, testable suggestion. Reproducibility is mediocre: code is public, data are available on request, but there is no commit hash or parameter files.\n\nI would send this to a referee, asking for a run at x_d=10, a longer evolution with a continuous external-field update, and a more cautious statement of the boundary. The paper deserves referee time; it should not appear with the current abstract.","headline":"First SPH-MHD survey of outer pulsar-disc interaction with a plausible qualitative stability map, but the x_d>=10 boundary is overreached and the once-flashed averaged-field setup limits the inference.","tokens_in":24144,"tokens_out":5333,"would_cite":true,"duration_ms":51675,"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":"This paper claims that a millisecond pulsar's tilted magnetic field destroys an accretion disc whose inner radius lies far beyond the light cylinder, unless the spin and magnetic axes are exactly aligned, and that modest inner-radius…","keywords":["pulsar-disc interaction","transitional millisecond pulsars","accretion disc stability","magnetohydrodynamic simulations","smoothed particle hydrodynamics","light cylinder","magnetic obliquity","Deutsch solution"],"falsifier":"Measure the magnetic obliquity of a transitional millisecond pulsar such as PSR J1023+0038 from its pulse profile while it is in the disc state; if the angle comes out below about 10 degrees and the source still undergoes disc-to-pulsar transitions, the claim that sizeable obliquity is required for transitions would be contradicted. Alternatively, a 3D MHD run with the full time-dependent rotating field acting continuously could show whether discs at $x_{\\rm d} \\geq 10$ with $\\xi > 0$ survive, which would overturn the energy-averaged stability map.","tokens_in":22979,"feed_emoji":"🌀","tokens_out":6585,"duration_ms":63212,"temperature":0.7,"pith_summary":"This paper uses axisymmetric magnetohydrodynamic simulations to ask when an accretion disc around a millisecond pulsar survives the pulsar's magnetic field. Modelling the field with the time-averaged vacuum Deutsch solution, the authors find that a disc whose inner edge sits well beyond the light cylinder ($R_{\\rm in} \\gtrsim 10\\,R_{\\rm LC}$) is always heavily disrupted unless the magnetic and spin axes are exactly aligned. At smaller radii the disc is stable up to a magnetic inclination that depends on distance, and the simulation-based stability map matches the analytical boundary of earlier energy arguments. The authors conclude that transitional millisecond pulsars can switch between disc and pulsar states through modest changes of the inner disc radius, provided their magnetic obliquity is at least about 20 degrees.","feed_headline":"Tilted pulsar fields shred discs beyond the light cylinder","feed_subtitle":"MHD stability map shows how transitional millisecond pulsars flip between disc and pulsar states.","key_machinery":"The central object is the time-averaged magnetic field of the Deutsch solution, the analytical vacuum field of an obliquely rotating, perfectly conducting sphere: each component is replaced by its period-averaged square root, $\\langle B_i^2\\rangle^{1/2}$, which removes the rapid pulsar rotation while retaining the spatial variation of the field's strength. This averaged field is flashed onto a relaxed thin-disc model built from the Shakura–Sunyaev solution and evolved with an axisymmetric smoothed-particle MHD code in which ohmic dissipation converts magnetic energy into heat, driving ablation of the disc's surface. The outcome is mapped in the $(\\xi, x_{\\rm d})$ plane and compared with the analytical stability line obtained by equating the electromagnetic energy density (which changes slope from $r^{-6}$ to $r^{-2}$ across the light cylinder) with the energy density of the orbiting gas.","core_discovery":"The paper claims that the fate of an accretion disc around a millisecond pulsar is set, in an energy sense, by two parameters: where the disc's inner edge sits relative to the light-cylinder radius, $x_{\\rm d} = R_{\\rm in}/R_{\\rm LC}$, and the angle $\\xi$ between the magnetic and spin axes. When $x_{\\rm d} \\gtrsim 10$, the time-averaged electromagnetic field of the pulsar heats and evaporates the inner disc within a couple of orbital periods for every non-zero $\\xi$, while the exactly aligned case leaves the disc essentially intact. For smaller $x_{\\rm d}$ the disc grows more robust, with stability requiring $\\xi$ below roughly 10–20 degrees depending on radius; at $x_{\\rm d} \\approx 0.5$, even $\\xi = 30^\\circ$ leaves the disc unaltered. The simulation outcomes, classified by a density-decay criterion, sit on the same side of the analytical stability boundary of Ekşi & Alpar (2005) as the theory predicts.","pith_inferences":["If the same stability criterion holds for transient supernova fallback discs, discs around young neutron stars with circularization radii above $x_{\\rm d} \\approx 10$ should be short-lived unless the newborn pulsar's obliquity is tiny, which would select for aligned rotators in the fallback-disc population.","The axisymmetry assumption excludes non-axisymmetric modes such as the magneto-rotational instability; including them could shift the quantitative boundary, but the qualitative conclusion that obliquity dramatically lowers the threshold for disc destruction remains a direct consequence of the energy balance.","The factor-3-4 prediction for $R_{\\rm in}$ fluctuations in high-obliquity tMSPs is directly testable: archival X-ray/optical monitoring of the double-peaked emission lines in PSR J1023+0038 could be searched for systematic changes in line separation that track the inner disc edge during the months before a state transition."],"forward_implications":["Discs truncated near the corotation radius ($x_{\\rm d} \\approx 0.5$) remain stable up to the largest obliquity tested, $\\xi = 30^\\circ$, so the persistent X-ray (accretor) state of a millisecond pulsar is confined to roughly $0.5\\,R_{\\rm LC} \\lesssim R_{\\rm in} \\lesssim R_{\\rm LC}$.","At $x_{\\rm d} \\gtrsim 10$ the disc is always unstable for any non-zero obliquity, so a slightly oblique pulsar whose inner disc is pushed well beyond the light cylinder will quickly evaporate or eject its innermost region.","A factor of 3–4 inward-to-outward change in $R_{\\rm in}$ (e.g. from $x_{\\rm d} \\approx 2$ to $x_{\\rm d} \\approx 6$) is enough to move a disc with $\\xi \\gtrsim 20^\\circ$ from the stable to the unstable region, enabling a disc-to-pulsar state transition.","Because the spin–magnetic angle cannot change on transition timescales, the authors attribute tMSP state transitions to fluctuations of the inner radius and infer that tMSPs must have sizeable obliquities, $\\xi \\gtrsim 20^\\circ$.","Low-mass X-ray binaries that show strong X-ray variability in quiescence are inferred to have nearly aligned axes, $\\xi \\lesssim 10^\\circ$, since their discs remain stable over wide changes of $R_{\\rm in}$."],"supporting_citations":[{"why":"Supplies the analytical electromagnetic energy density and the stability boundary in the ($\\xi$, $x_{\\rm d}$) plane against which the simulations are compared.","marker":"Ekşi & Alpar (2005)"},{"why":"Provides the vacuum oblique-rotator solution whose time-averaged field components drive the simulated disc heating.","marker":"Deutsch (1955)"},{"why":"Gives the full expressions of the Deutsch fields used to compute the numerical time averages at arbitrary polar angle.","marker":"Michel & Li (1999)"},{"why":"Provides the thin-disc analytical model from which the initial disc density and temperature profiles are constructed.","marker":"Frank et al. (2002)"},{"why":"The axisymmetric SPMHD code (Axis-SPHYNX) used to run the pulsar-disc interaction simulations.","marker":"García-Senz et al. (2023)"}],"fun_headline_variants":["Tilted pulsars disrupt discs beyond the light cylinder","Disc stability hinges on pulsar tilt and inner edge","Aligned spin keeps disc, tilt triggers evaporation","MHD energy map predicts disc fate in millisecond pulsars"],"cache_read_input_tokens":3200,"weakest_assumption_plain":"The simulations replace the real, time-dependent rotating pulsar field with a sign-undefined time-averaged field that illuminates the disc once at $t=0$ and then decouples, and they impose axial symmetry, so a continuously rotating three-dimensional field could in principle lead to a different stability classification.","fun_headline_variants_meta":{"raw":{"variants":["Tilted pulsars disrupt discs beyond the light cylinder","Disc stability hinges on pulsar tilt and inner edge","Aligned spin keeps disc, tilt triggers evaporation","MHD energy map predicts disc fate in millisecond pulsars"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.000452,"raw_usage":{"total_tokens":2279,"prompt_tokens":953,"completion_tokens":1326,"prompt_tokens_details":{"cached_tokens":384},"prompt_cache_hit_tokens":384,"prompt_cache_miss_tokens":569,"completion_tokens_details":{"reasoning_tokens":1261}},"tokens_in":569,"tokens_out":1326,"duration_ms":12699,"temperature":1.0,"reasoning_tokens":1261,"cache_read_input_tokens":384,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-07T12:46:21.309184+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"Measure the magnetic obliquity of a transitional millisecond pulsar such as PSR J1023+0038 from its pulse profile while it is in the disc state; if the angle comes out below about 10 degrees and the source still undergoes disc-to-pulsar transitions, the claim that sizeable obliquity is required for transitions would be contradicted. Alternatively, a 3D MHD run with the full time-dependent rotating field acting continuously could show whether discs at $x_{\\rm d} \\geq 10$ with $\\xi > 0$ survive, which would overturn the energy-averaged stability map.","supporting_citations":[{"cited_title":"C., Li H., 1999, @doi [ ] 10.1016/S0370-1573(99)00002-2 , https://ui.adsabs.harvard.edu/abs/1999PhR...318..227M 318, 227","cited_arxiv_id":null,"evidence_quote":"Gives the full expressions of the Deutsch fields used to compute the numerical time averages at arbitrary polar angle."}],"review_version":1}