What Causes Early Convective Instabilities in Supernova Explosions?

[Chronos]

Perhaps this explains the mass gap between neutron stars and stellar mass black holes.
http://arxiv.org/abs/1110.1635
Missing Black Holes Unveil The Supernova Explosion Mechanism

 

[Drakkith]

Interesting. I didn't even know there was a mass gap in neutron stars and black holes.

 

[Chronos]

Yes, and it's surprisingly large. The largest neutron stars are about 2 solar masses and the smallest black holes are about 5 solar masses - with almost nothing in between. This was first noticed around 1998, but, gathered little interest as data on neutron star and black hole remnant masses was scarce. It grew in interest over the past few years as more data became available.

 

[Drakkith]

I see. So how "accepted" or whatever is this article? Or is it too early to tell?

 

[Chronos]

I would call it a concept article with some math support. It is a back door attempt to explore core collapse supernova mechanisms from peripheral data. The lead paper is submitted for publication on ApJ - http://arxiv.org/abs/1110.1726 Compact Remnant Mass Function: Dependence on the Explosion Mechanism and Metallicity
I am a bit suspicious of their metallicity dependence assertion, but, the underlying concept looks promising. I am not aware of any observational evidence of a massive star expiring by simply morphing into a black hole, but, perhaps that is not something actively researched. It would be interesting. Gigantic stars are fairly rare in the universe, so, it would not seem hugely challenging to check to see if any have simply 'vanished'.

 

[Drakkith]

Ah ok. Thanks Chronos.

 

[twofish-quant]

I see. So how "accepted" or whatever is this article? Or is it too early to tell?

It's good work. The problem that I have with the article is that you can show by one dimensional simulations that there is just not enough convective energy to cause an "early" explosion. What you do is to take a 1-d code, move heat from the interior to the star to the shock, and even if you do it instantly, then "no-boom."

 

[twofish-quant]

Also one problem with the pre-print is that they describe the early convection as a Rayleigh-Taylor instability, when strictly speaking it isn't. An R-T instability happens when sign(dP/dr) = sign(d density/dr) which isn't the situation in the early supernova when dentropy / dr is positive.

This makes a big difference when you do stability analysis because the rate of growth for an R-T instability is the Atwood number whereas for the convectively unstable region in the supernova, you use the inverse Brunt-Vasla frequency.

On the other hand, different people will define it differently, and I remember this issue because two of the people in my dissertation committee got into a very heated argument over what R-T instability meant.

 

[phyzguy]

Also one problem with the pre-print is that they describe the early convection as a Rayleigh-Taylor instability, when strictly speaking it isn't. An R-T instability happens when sign(dP/dr) = sign(d density/dr) which isn't the situation in the early supernova when dentropy / dr is positive.

They address this specifically, and they say that there is a negative entropy gradient. Here's the quote, from page 2:
"In a collapsing star, this physical scenario is encountered just above the surface of the proto-neutron star. Neutrinos heat the turbulent atmosphere from below creating a temperature gradient, while the infalling gas from above creates a density gradient. As a result, colder and denser (low entropy) gas finds itself on top of hot and less dense (high entropy) material, and a violent displacement of layers follows."

 

[twofish-quant]

"In a collapsing star, this physical scenario is encountered just above the surface of the proto-neutron star. Neutrinos heat the turbulent atmosphere from below creating a temperature gradient, while the infalling gas from above creates a density gradient. As a result, colder and denser (low entropy) gas finds itself on top of hot and less dense (high entropy) material, and a violent displacement of layers follows."

They claim this, but I've never seen a density inversion in any simulations (including the one's I've run on my own), and I don't know of any simulations in which there is a density inversion. If they do see a density inversion in their multi-d simulations (and I haven't been keeping track so it's possible that they do), I'll retract my statement. They reference a 2007 paper by Fryer and Young, but the time scales that they are talking about aren't consistent with "density-pressure instability" to use a precise term.

Also what they say isn't consistent with the 2/3-d simulations I've seen. In all of the 2/3-d simulations that I've seen, the convection starts at the bottom of the convective region where the neutrino heating is strongest, and then moves upward. If the convection was shock driven, you'd expect to see the mixing to start near the shock and move downward.

What I've seen in simulations is that you have such a strong negative density gradient that even with neutrino heating at the base of the convective zone, you it's not enough to cause a density inversion and you get extremely strong mixing even before a density inversion has a chance to develop. You do have massive mixing, but it's thermally driven and not density-pressure driven.

Also the big problem with any convective mechanism, is that once the shock starts to revive, the shock heating is much stronger than neutrino heating which creates a stable convective zone. That limits the amount of energy that you can put into revive the shock, and that means that convection by itself is not the answer.

 

[Chronos]

Core collapse supernova do occur, that is not in dispute. The remnant mass is under scrutiny. More specifically, why the large mass gap between neutron stars and solar mass black holes? I agree the mechanisms they propose are speculative, but, not unreasonable.

 

[twofish-quant]

I agree the mechanisms they propose are speculative, but, not unreasonable.

All mechanisms for core collapse are speculative. We really don't know what causes the implosion to turn into an explosion. One thing that I find interesting about the argument is that it doesn't matter what causes the explosion. The conclusion that early convective instabilities must occur to drive the explosion is interesting, because there is this dissertation that claims that early convective instabilities *can't* drive the explosion.

http://adsabs.harvard.edu/abs/1998PhDT...8W

There is one thing that really bothers me. OK, you argue that if the explosion doesn't happen immediately, the entire star turns into a black hole. So a 35 solar mass star collapses, there is no explosion and the entire star collapses into a 5-15 solar mass black hole. Ummm... There's 20 solar mass of stuff missing here.

One problem here is that I'm not seeing how you end up with a five solar mass black hole without an explosion. If you blow off 15 solar masses during the evolution of the star, you aren't going to have silicon burning so I don't see how you end up with an iron core collapse.