Y TP YUKAWA INSTITUTE FOR THEORETICAL PHYSICS NPCSM 2016, - PowerPoint PPT Presentation

Neutrino-driven Mass Ejection from the Remnant of Binary Neutron Star Merger Sho Fujibayashi (Kyoto U), Yuichiro Sekiguchi (Toho U), Kenta Kiuchi (YITP), and Masaru Shibata (YITP) Y TP YUKAWA INSTITUTE FOR THEORETICAL PHYSICS NPCSM 2016, 11.9.2016, YITP, Kyoto, Japan

Evolution After Remnant of Binary NS Merger (Demorest et al. 10; Antoniadis et al. 13)

Neutrino-driven Outflow in MNS Phase ◎ MNS phase • Large neutrino luminosity from the MNS and torus (~10 53 erg/s) • The effect of neutrinos would be significant. ν ν ¯ ν ¯ Torus ν MNS

Neutrino-driven Outflow in MNS Phase ◎ Relativistic jet for short GRBs? Due to neutrino pair-annihilation heating GRB? ν ν ¯ ν ¯ Torus ν MNS

Neutrino-driven Outflow in MNS Phase ◎ Relativistic jet for short GRBs? Due to neutrino pair-annihilation heating GRB? Recent studies : Monte-Carlo method using Newtonian simulation Richers et al. 15, RHD simulations for BH-torus using equilibrium torus as initial conditions Just et al. 16 ν ν ¯ ν ¯ Torus ν MNS

Neutrino-driven Outflow in MNS Phase ◎ Relativistic jet for short GRBs? Due to neutrino pair-annihilation heating GRB? Recent studies : Monte-Carlo method using Newtonian simulation Richers et al. 15, RHD simulations for BH-torus using equilibrium torus as initial conditions Just et al. 16 Heavy Kilonova/ element Macronova? synthesis? ◎ Neutrino-driven winds Fernandez & Metzger 13, Perego et al. 14 Metzger & Fernandez 14, Just et al. 15 ν Fernandez et al. 15 ν ¯ ν ¯ Torus ν This component would contributes to heavy element nucleosynthesis and electromagnetic signals Viscosity-driven wind : Fernandez’s talk MNS

Neutrino-driven Outflow in MNS Phase ◎ Relativistic jet for short GRBs? Due to neutrino pair-annihilation heating GRB? Recent studies : Monte-Carlo method using Newtonian simulation Richers et al. 15, RHD simulations for BH-torus using equilibrium torus We simulate the MNS-torus system in fully general relativistic as initial conditions Just et al. 16 Heavy Kilonova/ manner in order to investigate the properties of neutrino-driven element Macronova? outflow from the NS − NS merger remnant. synthesis? ◎ Neutrino-driven winds ν → e − e + We consider reaction and investigate the effects. ν ¯ Fernandez & Metzger 13, Perego et al. 14 Metzger & Fernandez 14, Just et al. 15 ν Fernandez et al. 15 ν ¯ ν ¯ Torus ν This component would contributes to heavy element nucleosynthesis and electromagnetic signals Viscosity-driven wind : Fernandez’s talk MNS

Method ◎ Strategy x-y plane i) Merger of NS − NS and MNS formation by 3-D full GR simulation (Sekiguchi et al. 15) Equation of state : DD2 ( → The remnant is long-lived MNS)

Method ◎ Strategy x-y plane i) Merger of NS − NS and MNS formation by 3-D full GR simulation (Sekiguchi et al. 15) Equation of state : DD2 ( → The remnant is long-lived MNS) Average over azimuthal angles around the rotational axis after ~50 ms after the merger, when the system settles into Angle-averaged quasi-axisymmetic configuration. configuration x-z plane ii) Long-term Axisymmetric 2-D simulation using angle-averaged configuration as a initial condition MNS & innermost part of the torus have large neutrino emissivity !

Method ◎ Basic Equations - Full GR axisymmetric neutrino radiation hydrodynamics simulation • Einstein’s equation : BSSN formalism We use Cartoon method to impose axially symmetric conditions. • General relativistic radiation hydrodynamics : Leakage+ scheme incorporating Moment formalism Thorne 81 Shibata et al. 11 r α T α baryons, electrons, trapped neutrinos β = � Q (leak) β α streaming neutrinos r α T (S , ν ) β = Q (leak) β † We solve neutrino radiation transfer using Moment formalism with M1-closure. ††We do not consider the viscosity in this simulation. Thus we focus only on purely radiation-hydrodynamical effects on the system. • Lepton fraction equations

Results : Dynamics of Fluid Density color map of meridional plane = 1 c The density around the rotational axis falls rapidly. Outflow with ~0.5 c . Relativistic outflow is not seen.

Results : Dynamics of Fluid Log Pair-annihilation heating rate density [erg/s/cm 3 ] • First ~50 ms 10 ms Strong outflow due to Pair-annihilation heating • ~100 ms later 100 ms Heating rate decrease → outflow becomes weak ~ 0.2 c 300 ms Relativistic outflow is not observed in this setup

Results : Dynamics of Fluid • First ~50 ms Result w/o Pair-heating 10 ms Strong outflow due to Pair-annihilation heating • ~100 ms later 100 ms Heating rate decrease → outflow becomes weak ~ 0.2 c 300 ms ※ Effect of Pair-heating is large. − Strong outflow is not seen in the result without ν ν pair-annihilation.

Results : Luminosity & Pair-annihilation heating rates 10 53 electron ◎ Luminosity anti-electron other decreases to ~10 52 erg/s in ~300ms. Luminosity [erg/s] and get quasi-stationary. 10 52 Neutrino t = 0 ms − Luminosities ( ν e ν e ν x ) 10 51 0 50 100 150 200 250 300 350 400 time [ms] Total heating rate ( � >10 10 g/cc) [erg/s] 10 51 0.1 Total Heating Rate ( ρ <10 10 g/cc) 10 50 0.01 Efficiency t = 300 ms 10 49 0.001 Efficiency 10 48 0.0001 0 50 100 150 200 250 300 350 400 Time [ms] time [ms]

Results : Luminosity & Pair-annihilation heating rates 10 53 electron ◎ Luminosity anti-electron other decreases to ~10 52 erg/s in ~300ms. Luminosity [erg/s] and get quasi-stationary. 10 52 Neutrino ◎ Total pair-annihilation heating rate − Luminosities ( ν e ν e ν x ) Z 10 51 ˙ ρ < 10 10 g / cm 3 d 3 x ˙ E pair = Q pair 0 50 100 150 200 250 300 350 400 time [ms] Total heating rate ( � >10 10 g/cc) [erg/s] is >10 50 erg/s in first 50 ms, but 10 51 0.1 Total Heating Rate decreases to ~10 49 erg/s in ~300 ms ( ρ <10 10 g/cc) 10 50 0.01 Efficiency Efficiency : 10 49 0.001 ˙ Efficiency E pair L ν , tot ~ 0.3% → 0.03 % η = 10 48 0.0001 0 50 100 150 200 250 300 350 400 Time [ms] time [ms] ˙ 2 → e ffi ciency ∝ L ν E pair ∝ L ν

Results: The Properties of the Ejecta • Unbound mass ~ 3 × 10 -4 M ☉ 3 × 10 -4 ejected mass • Kinetic energy ~ 5 × 10 48 erg M ej [M sun ] 2 × 10 -4 1 × 10 -4 Subdominant compared to w/ pair w/o pair 0 × 10 0 kinetic energy dynamical ejecta 6 × 10 48 E kin [erg] (~10 -3 M ☉ , 2 × 10 49 erg for DD2 EOS) w/ pair-heating 4 × 10 48 (Sekiguchi et al. 15) 2 × 10 48 w/o pair-heating • Average velocity 0 × 10 0 Typical velocity 0.2 ~ 0.1 − 0.2 c V ej [c] 0.1 0 ※ Effect of Pair-heating is large. 0 50 100 150 200 250 300 350 400 Time [ms] Time [ms] Without pair-heating process, we underestimate the amount and kinetic energy of the neutrino-driven outflow.

Results : Electron fraction & Entropy distribution ◎ Mass histogram of ejected material @ t=400 ms specific entropy electron fraction 400.28 ms w/o pair 10 -4 mass per bin [Msun] Histogram with pair-heating without pair-heating 10 -5 10 -6 0 0.1 0.2 0.3 0.4 0.5 10 100 1000 Y e s [k B /baryon] • Material of Y e >0.25 is mainly ejected. Typical value : ~0.4. • A small amount (~10 -6 M ☉ ) of material has very large specific entropy. ν → e − + e + Pair-annihilation process ( ) can inject energy ν + ¯ regardless of baryon density.

r-process in ν -driven outflow ◎ Estimate following Hoffman et al. 97. (assuming τ exp ~ 50 ms) 275.02 ms Mass distribution in Entropy – Electron fraction plane 0.5 -5 A>130 A>130 Mass per bin [Msun] 0.4 A>200 A>200 0.3 Y e nuclei are produced -6 0.2 via the r-process 0.1 0 -7 10 100 1000 s [k B /baryon] In the most of the neutrino-driven outflow, heavy nuclei of A>130 are hardly produced via r-process. Detailed nucleosynthesis study → Next work

Pair-annihilation Heating by Ray-tracing Method Current treatment of neutrino transfer: Moment formalism with M1-closure relation (Shibata et al. 11) This method cannot treat the crossing of two beams. Pair-annihilation heating rate should be compared to more Ab initio calculation. Calculate the pair-annihilation heating rate by ray-tracing method using snapshots of the simulation. (Ruffert et al. 97) ν = 1 C 1 + C 2 � 0 Z Z ν ) 2 d Ω 0 I ¯ ν [ h ✏ i ν + h ✏ i ¯ ν ](1 � cos ✓ ν ¯ Q ν ¯ d Ω I ν 4 c ( m e c 2 ) 2 3 d 3 x 0 d Ω I ν = Q e ff Log Total neutrino emissivity [erg/s/cm3] ν π | x − x 0 | 2 *We ignore general relativistic effects.

Pair-annihilation Heating Rate by Ray-tracing method Pair-annihilation heating rate along z-axis using snapshot at t = 100ms t = 100.000 ms 33 Q pair (M1) log heating/cooling rate [erg/s/cc] 32 (ray-tracing) total cooling 31 Ray-tracing 30 29 28 Moment formalism 27 26 Cooling 25 24 10 15 20 25 30 35 40 z [km] Heating rate estimated with (simple) ray-tracing method ~10 times larger that that with moment formalism. Heating rate would be underestimated.