Millisecond Magnetars as GRB Central Engines Todd Thompson The Ohio State University B. Metzger, N. Bucciantini, E. Quataert, P. Chang, J. Arons son et al. 2004 er et al. 2007, 2008a,b antini et al. 2006, 2007, 2008, 2009 1992 son 1994 er et al. 2000
37
Embed
Millisecond Magnetars as GRB Central Engines Todd Thompson The Ohio State University B. Metzger, N. Bucciantini, E. Quataert, P. Chang, J. Arons Thompson.
This document is posted to help you gain knowledge. Please leave a comment to let me know what you think about it! Share it to your friends and learn new things together.
Transcript
Millisecond Magnetars as GRB Central Engines
Todd ThompsonThe Ohio State University
B. Metzger, N. Bucciantini, E. Quataert, P. Chang, J. Arons
Thompson et al. 2004Metzger et al. 2007, 2008a,bBucciantini et al. 2006, 2007, 2008, 2009
Usov 1992Thompson 1994Wheeler et al. 2000
“A nagging question in all these models is what produces the ‘observed’ ultra-relativistic flow? How are 10-5 Msun of baryons accelerated to an ultra-relativistic velocity with ~ 100 or larger? Why is the baryonic load so low? Why isn’t it lower? There is no simple model for that. An ingenious theoretical idea is clearly needed here.”
Piran’s questions circa 1999.
The Basic Character of GRBs
• What is required?– Lorentz Factor: 100 < < 1000 (if internal shocks)
Piran:“Why is the baryonic load so low? Why isn’t it lower? There is no simple model for that.”
&E = &Mc2
The Answer The characteristic mass-loss rate in the wind
driven from the surface of a newly-born neutron star is (Duncan et al 1986; Qian & Woosley 1996)
for a timescale of 10 -100 seconds. If a millisecond magnetar, the wind has kinetic luminosity of 1050 ergs/s, with ~ 102 - 103, (or magnetization of > 102 - 103).
&M ~ 3×10−7 Lν, 51
5 /2 Me s−1
The Progenitor
Instability& Collapse
Collapse
Bounce, shock formation
Prompt Shock Stalls (~ ms)R ~ 100 km
Shock Revival & Explosion
~ 0.5 - 1 s
KH cooling epoch, wind.
Bounce
CoolingEpoch
€
e− p →ν en
€
e− p → ν en
€
tKH ~ 10−100s
€
Eν ~ 1053.5ergs
300 km
Burrows et al. (1995)
Collapse Bounce Stall Explosion Wind Cooling
QuickTime™ and aYUV420 codec decompressor
are needed to see this picture.
Wind Emergence: Common to All SN Explosions
Supernova Shockwave
Wind
Required physics:
νen↔ e−p νep↔ e+n+ NSgravitational potential.
Scheck et al. (2006)
&M ~ 3×10−7 Lν , 51
5 /2 Me s−1
Qian & Woosley (1996)
What about rotation and magnetic fields?
Espanek, 1999
Initially, not magnetically-dominated, only ~10s later.
Magnetar-strengthfields dominate dynamics in the cooling/wind epoch!
Rotation?
C. Thompson (1994); T. Thompson (2003), T. Thompson et al. (2004)
What we’ve done• We solve the free 1-D equatorial non-relativistic neutrino-heated
magneto-centrifugal wind problem for a ms-proto-magnetar.
– Derive wind structure (T, , V, X ), the spindown rate, the mass loss rate -- given a monopole field geometry.
(Thompson et al. 2001, 2004; Metzger et al. 2007, 2008a,b)
• Solve the free 1-D and 2-D axisymmetric relativistic MHD wind problem in the regimes < 1 (non-relativistic) and > 1 (relativistic, magnetically dominated).
– Assess 1-D assumptions!
– Calculate spindown explicitly when >> 1, but with significant mass-loss (spindown is not “force-free,” not “vacuum” dipole, is not infinity, not Usov 1992!).
– Calculate interaction of wind with the overlying progenitor.
(Bucciantini et al. 2006, 2007, 2008, 2009)
Evolutionary Calculations (Metzger et al. 2007)
€
1015G€
1016G
€
˙ E
€
1015G
€
1016G
=B2
4πρc2
RL
→ γ ∞
Evolutionary Calculations (Metzger et al. 2007)
€
1015G€
1016G
€
˙ E
€
1015G
€
1016G
=B2
4πρc2
RL
→ γ ∞
Conclusion:Millisecond proto-magnetars drive winds with
Exactly in the range needed for GRBs.
What about asymmetry?
,σ > 100 and &E ~ 1050 ergs s-1
on 10 -100s timescales.
Jet Production
• Wind hits spherical SN shockwave in ~ few seconds.
• Creates reverse shock, contact discontinuity.
• Residual Bphi squeezes (pinch force) flow.
• Net polar acceleration.
• Analogous to models of pulsar wind nebulae by Li & Begelman ‘92; Chevalier & Luo ‘93. See also Konigl & Granot ‘02; Komissarov & Lyubarsky 03,04; Uzdensky & MacFadyen ‘07; Nagataki ‘09).
Jet Production (Bucciantini et al. 07, 08, 09)
Wind shocks on the outward-going SN shockwave at ~104 km, toroidal B exerts a pinch force that drives jet. Models are Poynting
flux-dominated at breakout from progenitor.
jet 20
wind ~ 20
jet
supernova
Jet Production (Bucciantini et al. 09)
• Magnetic dissipation in outflow is uncertain: consider both high & low magnetization limits.
– Low- : wind(~RL) ~ jet(~ Rstar)
– High- : If (~RL) ~ 60, then jet(~ Rstar) ~ 10 -15 (but, accelerating), magnetic field ~5 times equipartition.
– Jet production robust in either case.
• All of the spindown energy goes to the jet, virtually none goes to the spherical “supernova.”
• For weaker winds, the reverse shock propagates back into the outflow, coupling the wind & bubble dynamics: not a “free” wind (see also Uzdensky & MacFadyen; Komissarov & Barkov).
• Worries: 3D stability, entrainment, etc. (e.g., McKinney & Blandford)
Nucleosynthesis (Bucciantini et al. 09)
• For most cases, not hot enough for significant excess 56Ni production. However, exact prediction is sensitive to timing of wind-bubble interaction.
• However, we expect 10-2
Msun of Ne & O & C with v ~ 0.1 - 0.2c along the poles.
• The power of the wind (B, P) is directly connected with abundance of nucleosynthetic products, their velocity distribution.
Compare w/ Nagataki et al; Maeda et al; Mazzali et al
• For magnetar B & ms P, everything works for GRBs. How?– Weak physics sets baryon loading of the wind.– B & P set Lorentz factor, magnetization, & kinetic luminosity.– Neutrino diffusion time sets characteristic timescale.– Interaction of wind with progenitor leads to collimation.
• Robust to most uncertainties (e.g, magnetic dissipation)• Nucleosynthesis connected to timing, wind power, etc.
– Association with supernovae is free. – Non-Association with supernovae? Maybe AIC of WD.
• This model predicts well-defined (testable) correlations between these quantities in an average sense. They are not independent.
• The model appeals to the idea that neutron stars are probably born with a range of B & P, implying a connection with normal supernovae (e.g., Soderberg). Diversity: B(P).
• If you believe in the “Collapsar” mechanism, you (should) believe the ms-magnetar mechanism: angular momentum requirements.
Summary
• Wind & Spindown:
– Couple physics more realistically to the supernova explosion, feedback effects on wind and magnetar torque in weak-wind limit
– Does time dependence of outflow at magnetar surface imprint itself on the outflow dynamics of the jet at the progenitor surface?
• Jets: Stability in 3D? Entrainment?
• Lightcurves:
– Use Lorentz factor, magnetization, opening angle, kinetic luminosity to predict prompt lightcurves (model-dependent). Try to make contact with observations. See Metzger et al. 2008: SGRBEEs!)
• Late-time activity: power-source for plateau? Flares? (Dall’Osso et al. today; talk by Margutti)
• LIMITATIONS: Energetics (e.g., Cenko et al. today): Maximum energy ~ few x 1052 ergs (e.g., 090926A)
The Present & Future
The observation of a long-duration GRB does not imply black hole birth
(necessarily).
EnergyLorentz FactorTimescale
Collimation
Lightcurves?
• Picture: WD/WD merger or AIC of a WD to a rapidly rotating magnetar.
• SGRB: ~ 0.1Msun disk accretes on ~ 0.1-1s. Polar GRB as in NS-NS merger calculations. (e.g., Dessart et al. 07)
• EE: Remaining magnetar spins down, drives wind.
A Model for Short-Duration GRBs with “Extended Emission”
Gehrels et al. 2006Gal-Yam et al. 2006
GRB 060614
Proto-magnetar winds can account for EE:s
Short GRBs with Extended Emission
Metzger et al. (2008a)
• Characteristics of the model:– As time increases, degree of relativity increases.– As time increases, energy loss rate decreases.– A magnetar is left behind. Late-time activity? Flares?– Maximum total GRB energy: ~ 5 1052 ergs. Gravity waves.
(e.g., Cenko et al. 2009)
• Some questions:– Magnetic dissipation? – What is the nature of SN-less long-duration GRBs? AIC?– What is the nature of the short-duration GRBs with extended
emission? AIC?– Can we predict prompt lightcurves? (model-dependence)– With energy & Lorentz factor & jet opening angle, can we calculate