Unsolved Problems in modeling neutron star- neutron star and black hole- neutron star binary mergers

Transcription

1 Unsolved Problems in modeling neutron star- neutron star and black hole- neutron star binary mergers By Ma: Duez, WSU A member of the CCCW SXS collaboradon MICRA 2011

2 This Talk review of the phenomenon Problem 1: accurate inspirals + the neutron star surface Problem 2: BHNS parameter space uncharted territory Problem 3: Neutrino transport leakage and beyond Problem 4: magnedc field issues magnetospheres, MRI turbulence Throughout this talk, ea will mean et al

3 What we want to get out of simuladons GravitaDonal waves Need long inspirals, accurate bulk evoludon, accurate metric evoludon/wave extracdon Test sgrb viability Need to verify massive disk, baryon poor region Need MHD and/or neutrino radiadon for energy extracdon Ejecta effects (r- process source, a]erglow emission) Need to track/characterize ou_lows with reasonable accuracy

4 Basic stages Effects that don t seem to carry over from Newtonian sims. (for circ. orbit): unbounded ejecta core survival/muldple mass transfer events

5 Evolving the nuclear fluid ConservaDve form u = F+ S, u =, E, S, Y t x e u = F F t [ i 12 / i+ 12 / ] x+ S i+ 12 / = ( i+ 12 / L, i+ 12 / R, i+ 12 / L, i + 1/ 2 R) F F F u u Choices: FD vs FV ReconstrucDon(u R,L ): how and what Riemann Solver( ) AlternaDves: SPH, PS, DG F i- 1/2L F i- 1/2R F i- 1/2 u i u i F i+1/2l F i+1/2 F i+1/2r

6 SpEC: the Spectral Einstein Code SpEC h:p:// holes.org/spec.html Core wri:en by Kidder, Scheel, Pfeiffer Pseudospectral g μυ evoludonà extreme accuracy for smooth g μυ Generalized harmonic formalism, BH excision, comoving coords. Ma:er evolved on a separate FD/FV grid BHNS binaries: see Duez, Foucart ea (08-11) New work: Neutron star- neutron star binaries (Kaplan)

7 Problem 1: accurate inspirals Can EoS be constrained from inspiral? BHNS: only sdffest EoS for AdvLIGO (Pannarale ea 11) yes if include merger (Lackey ea, Kyutoku ea 10) NSNS: only sdff/low- M NS for early insp. (Hinderer ea 10) larger late inspiral Ddal effects (Read ea 09) Longest inspirals to date NSNS: Baion ea (10,11): 10 orbits, ½- rad phase error EOB calibradon BHNS: Muhlberger ea: 9 orbits

8 Neutron star surface errors Main source of hydro error for inspirals Drop to 1 st order; c s à 0 shocks Problem worse for realisdc NS Nasty ρ(r) profile Neutrino emission very sensidve to T Nonsmoothness in EoS tables destroys convergence of inidal data (and, hence, of evoludon) We (or nuc. phys. people) need to smooth the tables

9 Improving the NS evoludon Steepness issue SoluDon: Concentrate grid near surface Surface tracking? Moving meshes? HRSC opdons Have some effect Actual discondnuity Subcell modeling? AnD- diffusive fluxes? Ghost fluid? Other numerical techniques/grids may do be:er MulDpatch HRSC (Zink ea 08) DisconDnuous Galerkin (Radice & Rezzolla 11) see talk

10 Problem 2: BHNS parameter space Define q=m BH /M NS, s= S BH /M BH 2 Parameters: q, R NS, S BH (s & misalignment), e (clusters) q vs. s space for aligned BH, ~14km NS

11 For q=3, high s, can get very massive disks Low- q, high- s region Disk self- gravity effects Runaway instability doesn t happen (Montero ea 10) Spirals/clumping/fragmentaDon not seen Papaloizou- Pringle instability (m=1) Korobkin ea (11), Kiuchi ea (11) à disk & BH wobble à strong GW signal

12 RealisDc- q, high- s region SXS effort led by Foucart (see his poster) Studying q=5,7 Massive, extended disk for q=7 if s 0.7, C=0.14 q=7, s=0.5 q=7, s=0.7

13 Foucart ea (11) Misaligned S BH Modestly reduced- mass, Dlted disks M b /M b, s.5i0 s.5i20 s.5i40 s.5i60 s.5i80 (degrees) t=20ms t=30ms t=40ms (degrees) (t-t merger )(ms) r(km)

14 Eccentricity NSNS/BHNS collisions in clusters may contribute significantly to sgrb rate (Lee ea 10) Sims: Newtonian: Lee ea 10; GR: Stephens ea 11 Features: Large M disk even for a=0 Unbounded ejecta Core survival zoom- whirl behavior from Stephens ea 11

15 Problem 3: neutrino transport Neutrinos Cool/heat gas (à wind?) Changes Y e (à convecdon?) RadiaDon pressure annihiladon energy At MICRA 2009!! My q=3, s=0.5 BHNS merger with Shen EoS Y e =advected or assumed μ v =0 β- equilibrium

16 Problem 3: neutrino transport Neutrinos Cool/heat gas (à wind?) Changes Y e (à convecdon?) RadiaDon pressure annihiladon energy At MICRA 2011!! My q=3, s=0.5 BHNS merger with Shen EoS Y e from neutrino leakage evoludon

17 Genng our feet wet: Neutrino leakage Cooling & leptonizadon, no headng or transport GR: O Connor and O: (10), Sekiguchi (10) Both introduced at MICRA 2009 de dne In rest frame, = Q, = R dt dt T = Qu, ( ny u) = R R = fluid R local for 1 diff diff 2 n T for 1 ( T c) Must estimate = min dl e

18 Neutrinos from NSNS merger Sekiguchi ea (11) L [10 erg/s] L [10 erg/s] L [10 erg/s] AnDneutrino dominated 0 3 Detectable out to 10Mpc ( a) L e e x ( b) M ( c) H t - t merge [ms] Cooling driven HMNS collapse for low M? from Sekiguchi 11

19 Grey diffusion Farris ea (08) Next steps Truncated moment transport Shibata ea (11) Monte Carlo 6D transport Abdikamalov ea QuesDons: What do we need to avoid missing any qualitadve effect? What can we afford?

20 Problem 4: MHD issues Important progress: BHNS merger with MHD (Chawla ea 10) MagneDc winding à EM signal from HMNS (Shibata ea 11) The missing link (Rezzolla ea 11): NSNSà sgrb B amplificadon to G in axial jet, unbounded ou_low from Rezzolla ea (11)

21 Magnetospheres B 2! >> 1! force-free limit! c E + J c " B = 0 Important for Blandford- Znajek effect NSNS magnetospheres interacdon à sgrb presignal? (Hansen & LyuDkov 01) Standard MHD codes fail in this regime SoluDon: evolve force- free equadons Palenzuela al (10,11) FF code applied to binary BH, BZ effect New: coupled to MHD code inside NS

22 Trouble ahead: the MRI! MRI ~ v A! = P orb B 2" 4"# disk: P orb ~ 10 "2 s, #~10 11 g cm "3, B~10 13 G #! MRI ~ 10 4 cm HMNS: P orb ~ 10 "3 s, #~10 14 g cm "3, B~10 13 G I don t see how AMR helps. SoluDons: #! MRI ~ 10 2 cm ~ 10 "4 R! Crank up B (c.f. Duez ea 06) Local MRI subgrid model calibrated to shearing box sims Kato & Yoshhizowa (95), Ogilvie (03), Pessah ea (06) test local vs global sims: Sorathia ea (10,11)

23 Other fluid instability issues KH instability B- field amplificadon Discrepancy between SPH & Whisky simuladons Turbulence How to idendfy/characterize it? How well do we evolve it? Is numerical dissipadon/reconnecdon adequate? RadiaDon L v might approach v Eddington limit How will radiadon affect disk turbulence/inflow/jet?

24 Conclusions Lots of papers (esp. le:ers) this past year NSNS/BHNS a hot topic Challenges ahead EoS tables Hydro accuracy Cases to study Neutrino transport approximadons MulD- scale modon: instabilides and turbulence