Dismiss Notice
Join Physics Forums Today!
The friendliest, high quality science and math community on the planet! Everyone who loves science is here!

I Core collapse of a supernova: the "void" left by the collapsed core?

  1. May 31, 2018 #1
    When core collapse occurs, a ~10000km diameter core of the star collapses into neutron star or a black hole. Let's look at NS case here: the resulting neutron star is on the order of 20 km across.

    And this happens in a few seconds.

    The upper layers of the star cannot immediately fall onto the NS in just a second. Effectively, the collapsed core must be leaving an Earth-sized "void" where it used to be.

    It won't be really empty: there won't be a sharply defined spherical surface dividing the core and the rest. The picture should be something like this, if we freeze "core just collapsed" star in time: the surface of the star and many hundreds of thousands of kilometers of upper layers didn't feel a thing yet, but somewhere about 5000-10000km distance from the center the density suddenly starts to decrease, and the gas is no longer stationary - it falls inward. And then there is ~20km diameter ultra-dense, ultra hot ball of neutrons in the center.

    This should result in some interesting dynamics wrt release waves on the inner surface of the "void" and radiative pressure on its walls. As gas from the surface expands inward, it exerts an outward pressure on the upper layers.

    And newly formed neutron star has a tremendous luminosity, on the order of 10^19 solar. Radiation pressure alone is not negligible from it, but just like in a thermonuclear bomb, even larger effect should be that this light from NS (gamma rays really) should heat up the walls of the "void", causing gas from the walls to expand inward faster, which increases outward pressure.

    Has this been modeled in the supernova simulations? (I suppose it was - I can't be the first one to think about it...)
     
  2. jcsd
  3. May 31, 2018 #2

    phyzguy

    User Avatar
    Science Advisor

    Of course it has been modeled. There is a whole industry of people working on these simulations, with probably thousands of papers. When the free falling outer layers hit the dense neutron star core (called a "proto-neutron star" in the literature), they bounce and drive a shock wave back out into the star outer layers. Historically the (mostly 1D) simulations show that this shock wave stalls due to the rush of infalling outer material and the star fails to explode. Despite more than 20 years of work, I think it is still not completely clear what drives the explosion. The pressure of the huge flux of outgoing neutrinos is part of the answer, and doing the simulations in 3D rather than 1D or 2D is more likely to result in an explosion because apparently the explosion is not even close to being spherically symmetric. If you Google
    "core collapse supernova simulations" you will find more stuff than you can read.
     
  4. May 31, 2018 #3
    Bouncing off a neutron star means rising from a surface with about 10 billion G surface gravity. No wonder it's not that efficient.

    It looks to me that a (another?) shock may be generated: a shock may be created by the photons impinging on the wall of the "void" left by just-collapsed core. This "wall" is not significantly moving yet, and it is dense (compared to the "void"), thus it will absorb EM radiation from NS, heat up and try to expand. It can easily expand inwards (because the "void" is below it). This will create a reaction force, like in rocket engine, pushing even higher layers outward. Newly born neutron star has a very high EM luminosity, right?
     
  5. May 31, 2018 #4

    phyzguy

    User Avatar
    Science Advisor

  6. May 31, 2018 #5
    Thanks! Excellent. Exactly what I looked for!

    According to the article, the idea they pursue is that the shock propagates in the material which falls on the proto-neutron star. The difficulty is that it does not work with simple calculations, and scientists work on making the calculations which account for more factors, use smaller grids, do not simplify to 1D or 2D, etc.

    I have a slightly different idea. What if _that_ shock does not indeed "succeed" to separate and propagate outward? What if that material successfully accretes on the PNS, without blowing up? (And let's suppose the star is not massive enough for black hole, we do end with a NS).

    The infalling material does not fall uniformly - the material in "outer core" (see article) falls quickly since it's close to the center, material farther away from center falls progressively slower, up to star surface which does not fall at all. That's what forms what I termed "the void" in place of former core, a low-density region ~10000 km is diameter between NS and the rest of the star matter.

    My theory is what if the explosion results from NS pumping this region full of photon gas?
     
  7. Jun 1, 2018 #6

    Ken G

    User Avatar
    Gold Member

    Photons are already in the model, so what you're saying should be in the models too. So it's importance will have been considered. To be a new effect not already included, you would need to include some new physical effect that the simulations either leave out, or are incapable of resolving accurately.
     
  8. Jun 1, 2018 #7

    phyzguy

    User Avatar
    Science Advisor

    As Ken G said, the effects you are talking about are already in these simulations. You can't do these things qualitatively, you have to model what is happening as exactly as you can using detailed computer simulations. If you want to understand more about it, you would have to dig into the details of these simulations. Many of the codes are open source and publicly available.
     
  9. Jun 1, 2018 #8
    They may well be leaving it out. They are concentrating on modeling the inner region, basically NS and its neighborhood only, maximum to ~1000km out from center. (Not surprising, of course, since making grid larger costs a lot in compute resources.)
     
  10. Jun 1, 2018 #9

    phyzguy

    User Avatar
    Science Advisor

    I strongly doubt they are leaving out the effects you mentioned, but you would have to dig into the simulations to be sure.
     
  11. Jun 2, 2018 #10
    I think nikkkom may be on to something here. All simulations concentrate on the inner region surrounding the neutron star in an attempt to find a way to revive the stalled shock. Neutrino transport and magnetohydrodynamic effects are mentioned repeatedly yet no mention is made of the energy released in gamma rays from the formation of the neutron star. In any case, the shock, even if revived, cannot bridge the vacuum of the gap that nikkkom points out is left by the formation of the neutron star. It could very well be that the supernova explosion occurs after the formation of the neutron star rather than being caused by it!
     
  12. Jun 2, 2018 #11

    phyzguy

    User Avatar
    Science Advisor

    Like I said, I strongly doubt that the people who are doing these simulations are leaving out these effects. Also, there is not a "vacuum" left behind when the neutron star collapses. The very high pressure gas behind will follow the surface of the collapsing neutron star. There may briefly be a lower pressure region, but it won't last long. But feel free to set up your own simulation. If you and nikkkom are right, there is a major publication to be had!
     
  13. Jun 3, 2018 #12
    You are right. It's far from being a vacuum. I'd wager it will not even fall as low as ~Earth's air density. However, there definitely will be a lower-density region:

    The collapse starts when the very center of the star reaches conditions when "neutronisation" (p+e -> n+ve) causes loss of electron degeneracy pressure. So, the very center (a small region, maybe 50 km across) starts shrinking (due to pressure of overlaying matter, not due to its own gravity, which is relatively negligible).

    Layers above the center start "falling" onto the shrinking central region - which _inevitably_ means that the density above these layers decreases.

    Tempting, but I'm not planning a sudden career change from Linux hacker to physicist... I love my current job/hobby...
     
  14. Jun 3, 2018 #13

    Ken G

    User Avatar
    Gold Member

    There are a few claims here I don't understand. First of all, radiation is generally included in simulations, so I don't understand why you would say the gamma rays are not included. Instead, I think it is found they are not important. Also, I don't see why you would say the shock needs to bridge any gaps-- the point of the shock is to turn downward falling gas into upward moving gas, so by the time the shock would reach very low density gas it would have already done its job, no need to "bridge" anything. Finally, the energetics required to blast the material out into space at high speed can only be provided by the gravitational energy of forming the neutron star, so I cannot see why it could be claimed that the formation of the neutron star does not "cause" the supernova.
     
  15. Jun 3, 2018 #14
    I'm sure gamma rays are included... as long as they are in the simulated volume in the center of the star. What simulation assumes to happen when they escape from this volume?

    We know the rest of the star is transparent to neutrinos which are leaving the simulated volume (thus simulation can just ignore them), but it's definitely not true for gamma gays.

    Not necessarily. Shocks cause material to accelerate in the direction of the shock, yes, but here material is infalling. If shock is not strong enough, it may fail to reverse the material's velocity as it passes through it. The material may still fall down, just slower.
     
  16. Jun 3, 2018 #15
    We may be talking about the same thing here. Correct me if I'm wrong, but when you talk about the shock wave, do you mean a shock wave formed by the material falling onto the neutron star? All numerical simulations of supernova explosions are concerned with shockwav e es moving outward from within the forming neutron star as the means for the explosion. Early simulations had found that the "accretion shock" of infalling material from the envelope would stall above the surface of the neutron star and would not move outward in radius to provide an explosion, rather piling up on the surface of the neutron star and eventually forming a black hole. So the gap I was referring to is the void left by the contraction of the core into a neutron star. The overlying layers of the envelope do not infall immediately even though they are no longer supported against gravity, they are supported by radiation pressure from infrared radiation released by the gravitational binding energy of the neutron star, similarly to the layers pushed out in the formation of a red giant star. A shockwave needs a medium to propagate in, without it the shockwave produced by the formation of the neutron star finds the void produced by the formation of the neutron star insurmountable. Now while the energetics of the neutron star's gravitational binding energy is comparable to the energy of the supernova explosion, they aren't necessarily connected. For instance, in the aforementioned simulations, nuclear reactions in the infalling material, incredible as it may seem, were not included. It could be that there were computing constraints, but I can only surmise.
     
  17. Jun 3, 2018 #16
    Just formed neutron star's emissions peak not in the IR, rather somewhere at about 10 MeV, which are gamma rays.

    Nuclear reactions are releasing (or in some cases, absorbing) on the order of 1%-0.1% of rest mass-energy of their ingredients, whereas released gravitational binding energy during NS formation is ~15% of its rest mass-energy.
     
  18. Jun 3, 2018 #17

    stefan r

    User Avatar
    Gold Member

    The statement is ambiguous. Clearly a lot of mass falls down. The overall average is a drop.

    This video uses the example of a stacked basketball, bouncy ball, and golf ball. The stack's mass is lower after the bounce. The golf ball, however, can be seen up at 8.5 meters.
     
  19. Jun 4, 2018 #18

    Ken G

    User Avatar
    Gold Member

    Yes, but that shock would have nothing to do with any gap left behind by the infalling gas. The trick is to get the infalling gas to turn into outflying gas, which is what they need the shock to do. What happens to the gas above the infalling gas is of little energetic significance, the energy scale is set by the neutron star gravity and is all happening very deep down where the density is very high.
    They have to be connected, they are the only energies at that scale in the entire problem.
    The nuclear physics in this process is neutronization and photodisintegration of the nuclei, which is a sink for energy, not a source, and is included in the simulations. You can think of it as basically undoing all the fusion the star has undergone up to that point. The energy for that undoing, as well as the energy from the explosion, all comes from the gravitational energy of creating the neutron star. It's the only energy of significance.
     
  20. Jun 4, 2018 #19

    Ken G

    User Avatar
    Gold Member

    The action of importance is happening where the energy scale is important, which is way down deep where the neutron star forms and the density is high. What is happening much farther out in the gravitational potential is a paltry energy in comparison, and is of no particular significance to getting an explosion.
    Yes, but all of that is happening to the infalling gas, not to any void left behind by the infalling gas.
     
  21. Jun 4, 2018 #20
    Yes.

    No. If formation of NS fails to create an outgoing shock wave and movement of matter from its surface, it's not the end of the story. Now we need to examine what happens to the rest of the star. Maybe there is _another_ process which will make the rest of the star explode?

    Let's look at "paltry energy".

    Luminosity. The newborn NS temperature is upwards from 100 billion Kelvins. This means that surface luminosity, assuming blackbody radiation, is (1e11/6000)^4 times higher than Sun's surface. The surface of NS is smaller though. NS radius is ~10km, Sun is ~700k km, ratio of surface area is (700000/10)^2. Thus, NS luminosity is (1e11/6000)^4 / (700000/10)^2 = 1.57e19 solar, if NS is at 100 billion K. If it's at the upper end of the range, 1000 billion K, then its a ten thousand times more. That's quite large amounts of "paltry" energy.

    Radiation of these intensities drive radiation-dominated shock waves ("Marshak waves") in matter by itself - it does not need any "rebound shock wave" to do so. Essentially, it's a photon shock wave, where matter is merely "along for the ride". It will propagate away from NS (much) faster than local speed of sound through a low-density region, until it impacts on a not-yet disturbed, dense layers of the star (walls of the "void").

    (On a much smaller energy scale, this is used in thermonuclear weapons to drive implosion of the secondary. The X-rays from the primary, of paltry 100 million K, impinge on a cylinder of uranium, evaporate its surface, creating ablation-driven shock, about 50-75 times stronger than radiation pressure alone. The achievable shock velocities are on the order of 500 km/s.)

    Marshak wave velocity is approximately proportional to T^2. To get about 50000 km/s wave velocity, the temperature of the wave hitting the walls of the "void" needs to be at least ~1 billion K. Looks very plausible to me.

    See
    http://nuclearweaponarchive.org/Nwfaq/Nfaq3.html (particularly section 3.5.6 Thermal Waves with Hydrodynamic Flow)
    http://nuclearweaponarchive.org/Nwfaq/Nfaq4-4.html
     
    Last edited: Jun 4, 2018
Share this great discussion with others via Reddit, Google+, Twitter, or Facebook

Have something to add?
Draft saved Draft deleted