Nuclear Explosions on Rapidly Rotating Neutron Stars

1
Hydrodynamics of type-I bursts
Spitkovsky, L., Ushomirsky 02 L., unpublished
Strohmayer et al 98 Muno et al 03
Hydrodynamic puzzle 1 short rise
times Hydrodynamic puzzle 2 coherencelarge
drifts
2
Heyl 04 Bildsten Piro 05, 06 Strohmayer Lee
05 Lee 05 Cooper Narayan 06,07
Popular paradigm

• Spreading hot spot during the rise
• (b) Giant m1 waves (buoyant Rossby
• waves) during the tail

Pros
1. Coherence!
2. Fractional amplitude increases with energy
3
Cons
Waves go too fast!!
0.1g
1/2
V(1/3) (g H) 12 Hz
Asymptotic frequency is not spin by at least a
few Hertz. Aql-X1 is a strong counter-argument!
Have to cut-off the drift
5 Hz
Proposal Rossby waves
Crustal interface waves
Piro Bildsten 05
1 kg
This does not work coupling is too small.
Crustal-interface wave
Rossby wave
Berkhout L., 08
0.1g
0.1g
4
Spitkovsky, L., Ushomirsky 02 L., unpublished
Different waves?
Cumming 05, 08
Observational issue QPOs seem connected in the
rise tail for a significant fraction of
bursts where both are detected. But rise QPOs are
(presumably)? not waves.
5
Burst hydrodynamics
Spitkovsky, L., Ushomirsky 02 L. unpublished

• Kelvin-Helmholtz stable!!
• Baroclinic unstable but weak.
• Heat conduction a la Niemeier,
• but across a huge interface!

30m
hot
cold
3m
3 km
Rossby radius
Notable features 1. Horizontal velocity is
sonic, 10000 km/sec
2. Initial vortex twists 60 times
6
Shallow-water simulations
Spitkovsky, Ushomirsky (unpublished)?
Jupiter red spot? Not seen here.
7
Idea try magnetic fields
Initially vertical, 100 MG
AMXP QPOs, Chakrabarty et al 03
Initial vortex does 50 revolutions!
Alfven
couplingnuclear burning time,
so B marginally
dynamically important
deflagration.

Magnetic rope is
formed and dragged
with the backward flow.
8
Idea coherent structure--magnetic fields
Initially vertical, 100 MG
Lovelace, Kulkarni, Romanova 07,
based on Cumming
Bildsten 00
It is the rope which is dragged!-YL.
Pictures from Lovelace et al
Pictures from Lovelace et al 07
9
Initially horizontal field normal non-pulsing
LMXB
10
Assume 10 G this is not buoyant yet (Cumming,
Zweibel, Bildsten 02)?
11
Twist initial vortex 50 times! Final field 510
G 1. The field floats and forms a neutron-star
spot 2. The spot is dragged a la Strohmayer et
al, Cumming Bildsten 3. The spot is not
anchored to the bottom, and drifts fast
Possible positive features 1. produce thermal
bump and reconnection 2. Rise and tail
connected
10
Conclusions

• 1. Need hydrodynamics to understand deflagration,
  burst
• rise. This is close at hand.
• 2. Need MHD simulations to understand QPOs.
• This is a great challenge.