PDA

View Full Version : Zones inside (normal) stars, mixing, element separation etc



Nereid
2006-Jul-13, 08:46 PM
Here on Earth, the 'light' continents 'float' and 'rise above' the 'heavier' oceanic crust, and both 'float' on the mantle.

Deep down, there is a core of iron (plus nickel, plus ...).

Liquid water is found 'on top of' the crust, and solid water on top of the liquid.

Air is found 'above' the solid water (where there is some), or liquid water (ditto), or crust.

The Earth has a radial structure that is 'sorted' by density, and also by composition (principally, different rock types).

The Sun, and 'normal' stars (i.e. not white dwarfs or neutron stars), are not solid, liquid, or gas - they are balls of plasma (OK, some of the very coolest stars have solid 'dust' in their upper atmospheres, and some may even have 'iron rain'; in the outer parts of these stars, the plasma may be so weakly ionised as to make the difference with a gas essentially irrelevant).

Do 'heavier' elements 'settle' into the cores of stars (in a manner similar to how iron 'settled' into the core of the Earth)?

Can the Helium (and other elements) produced by fusion in a star's core come up to its surface?

And how confusing is it, to an understanding of what's really going on, inside stars, to use words like 'light', 'heavier', 'float', and 'settle'?

George
2006-Jul-13, 10:01 PM
Cool questions. [no pun intended, if it is one.]

It is interesting because it comes on the heels of the thread (http://www.bautforum.com/showthread.php?t=43250) regarding the model having planets merge with their host star. The model has them sinking deep into the interior before something transpires to generate the possible flash seen in V838 Monocerotis.

I'm probably stating more than I know here, (so what's new)...

There is a very intense layer between the radiative and convection zone where reionization (or is it termed just ionization?) occurs, transitioning from plasma. This allows the gas to capture light and to become even hotter - “boiling” hot. [But not necessarily hotter due to bubbling upward action.] This may be in the region of the tachocline (http://solarphysics.livingreviews.org/Articles/lrsp-2005-1/articlesu5.html) where differential rotation begins and where the convective zone begins. The convective zone, I would guess, would allow for much mixing of all the gases. This may not be true, however, below the convective zone.


And how confusing is it, to an understanding of what's really going on, inside stars, to use words like 'light', 'heavier', 'float', and 'settle'? May I obfuscate this a little more with a colorful question (which you know) to illustrate how the simple can be unknown? ;)



[Edit is in brackets]

korjik
2006-Jul-13, 10:57 PM
The convective zone should be well mixed. Boiling is a good description.

In the radiative zone, mass separation will occur when the scale heights are different enough to allow separation. Basically, density of any species of ion should depend on Temperature and Mass. I will have to hit the books to check for sure.

Romanus
2006-Jul-14, 04:54 AM
<<Do 'heavier' elements 'settle' into the cores of stars (in a manner similar to how iron 'settled' into the core of the Earth)?>>

As already said, great question.

IIRC, elements can indeed separate by weight and opacity in "quiet" stars--that is, bright stars with non-convective envelopes. These reveal themselves by unusual metal abundances in their spectra.

Here's an example from the incomparable James Kaler's "Stars" site:

http://www.astro.uiuc.edu/~kaler/sow/alpheratz.html

Tobin Dax
2006-Jul-14, 09:38 AM
The convective zone should be well mixed. Boiling is a good description.

Not the boiling itself, since that's a completely different physical phenomenon. The convecting water within the pot, though, is a good description.

Squashed
2006-Jul-14, 12:41 PM
The convective zone should be well mixed.

I'm glad this thread was created because I have questions about star construction.

As far as the convection zone mixing the elements wouldn't it be more akin to a bowl of mixed granular sizes that is vibrated to cause the lighter, less dense, particles to float to the top?

Gravity is the main separating force and the vibrations just reduce the particle's inertial tendency to not move.

It would seem to be the same within a star because convective currents both rise and fall and so it would seem that the heavier ions would rise less and fall more which would eventually cause a migration of the heavier ions towards the center of gravity.

Tobin Dax
2006-Jul-14, 06:48 PM
I'm glad this thread was created because I have questions about star construction.

As far as the convection zone mixing the elements wouldn't it be more akin to a bowl of mixed granular sizes that is vibrated to cause the lighter, less dense, particles to float to the top?

Gravity is the main separating force and the vibrations just reduce the particle's inertial tendency to not move.

It would seem to be the same within a star because convective currents both rise and fall and so it would seem that the heavier ions would rise less and fall more which would eventually cause a migration of the heavier ions towards the center of gravity.

The cells are convective: the rise and fall are due to the temperature of the cell vs. its surroundings. The composition of the cell itself isn't going to matter as much as temperature since the whole cell is composed of a mix of different ions. They also follow currents, not vibrations. The convection cells move a large distance within the sun before cooling or dissipating.

Squashed
2006-Jul-14, 08:18 PM
... The convection cells ...

Cells - can you describe these more completely or detailed (I'm not an expert on solar anatomy).


They also follow currents, not vibrations.

I just used the vibratory bowl as a metaphor for mass sorting by gravity ... and so the metaphor within the metaphor is vibration=convection currents.

94z07
2006-Jul-14, 08:30 PM
When I was in grade school I got to see Carl Sagan present slides showing the developement of the Voyager craft and their missions. Afterwards there was a Q and A session. (Only the much older kids got to ask questions and it was clear they had been picked in advance and they used note cards.) I will never forget Sagan saying that, excepting H and He, the elements we find in our solar system were created in the hearts of dying stars. He then speculated that when stars capture debris that the stars fision the material into H and He once more. I've never seen the bit about fision anywhere else. Anyone know if that was an idea who's time has passed?

Tobin Dax
2006-Jul-15, 01:45 AM
Cells - can you describe these more completely or detailed (I'm not an expert on solar anatomy).
Blobs of plasma. A convection cell is a (relatively) small volume in the sun where the temperature is hotter and the density less than the surrounding region (or cooler & denser instead). Because they're less dense, they rise (or fall if more dense).


I just used the vibratory bowl as a metaphor for mass sorting by gravity ... and so the metaphor within the metaphor is vibration=convection currents.
I understand your metaphor, but it's not right. It's more like tossing a salad. (Vibrations are too small - the scale over which convection happens is much bigger.) If the salad was already well-mixed to begin with, tossing it isn't going to cause it to differentiate. The same is true with convection.

Ken G
2006-Jul-16, 12:42 PM
Indeed, the salad mixing analogy is apt because even if the salad starts out differentiated, the act of tossing it will cause it to mix. So it is with convection. Also note that the nice description linked to by Romanus points out that in some stars that are not convective in their outer layers, you can get something more along the lines of the "vibration" effect, but it's not just gravity that causes the sorting, it's a combination of gravity pulling some particles downward more strongly than others, while radiative forces push other particles more strongly upward. Collisions between the particles prevent complete separation, but at low enough density and a strong enough radiation field, you can still get substantial separation. The way to get low density and strong radiation is to have a large, massive, and very luminous star (Apheratz is a B subgiant).

Another point to keep in mind in the vibrational sifting analogy is that the mass of an individual particle never directly affects its motion due to gravity-- all that directly matters is its velocity. The vibration gives the particles their speeds, but the accompanying collisions cause the higher mass particles to give up some of their speed to the lower mass particles (that's where the mass comes in, in determining the speeds after a collision), and that's what allows them to fall preferentially into gravity's clutches. So collisions play a role in the differentiation, but note that too many collisions can prevent the differentiation too-- by making it take too long to occur so that you never actually have time to see it. In most stars, the densities are so high that the collisions are so frequent there just isn't time to get the differentiation. (How long it would need to take isn't immediately clear to me.)

George
2006-Jul-16, 07:15 PM
Leaving the convective zone for the core, wouldn't things be a lot different in there? IIRC, the greatest activity is slightly outside the very center. Perhaps, photon pressure would push larger atoms toward the very center. [Just guessing, however.]

Ken G
2006-Jul-16, 10:02 PM
AFAIK, the effects of gravity "filtration" are only relevant near the surface-- the density is just too high everywhere else and the gas is all tightly coupled together by collisions. More massive stars are not convective near their surface, they are low density, and they are more luminous, all of which tends to assist the chemical peculiarities found near the surface of Alpheratz. Perhaps you are thinking of another type of differentiation that can occur in stellar interiors, which has to do with the nuclear burning history. In nonconvecting layers, the abundances of various nuclei depends on the temperature history of that layer, and what has been able to fuse there. But that has nothing directly to do with the mass of the nuclei, it depends on the fusion temperatures.

ngc3314
2006-Jul-17, 01:24 AM
When I was in grade school I got to see Carl Sagan present slides showing the developement of the Voyager craft and their missions. Afterwards there was a Q and A session. (Only the much older kids got to ask questions and it was clear they had been picked in advance and they used note cards.) I will never forget Sagan saying that, excepting H and He, the elements we find in our solar system were created in the hearts of dying stars. He then speculated that when stars capture debris that the stars fision the material into H and He once more. I've never seen the bit about fision anywhere else. Anyone know if that was an idea who's time has passed?

It's roughly as hard to destroy heavy nuclei as to create them (until you get to the really heavy ones that take care of that themselves). A few light elements (deuterium and lithium come to mind) can be fused at very low temperatures and thus destroyed, and in interstellar space some cosmic "ray" particles have enough energy to break these light and fragile nuclei apart (spallation). At the other end of the environmental scale, in the hot cores of very massive stars (possibly only the first-generation, initially nearly metal-free Pop III stars) one process which can bring on core collapse is the sort of reverse nucleosynthesis called photodisintegration. This happens when the temperature is so high that many of the ambient photons are gamma rays of high enough energy to splinter heavy nuclei apart, undoing all the work the stellar core has spent its whole lifetime doing. This saps the radiation field's energy very quickly and leads to core collapse. As far as I know, this remains theoretical on a stellar scale; we have yet to see any Pop III stars, much less watch them blow up and work out which of several competing processes did the deed.

Ken G
2006-Jul-17, 02:17 AM
Someone on here, might have been trinitree, said that the endothermic process that leads to core collapse is generally the induced fusion of very heavy nuclei, possibly coupled with direct neutrino escape, and ngc3314 has pointed out that in pop III stars it might be the induced fission of ligher nuclei. Of course there are also good old radiative diffusion and neutrino escape, so which process of removing core heat is thought to dominate the initiation of card-carrying, pop I, O star core collapse?

Tim Thompson
2006-Jul-17, 03:19 AM
The core collapse in a supernova explosion is a result of the photodissociation of 57Fe nuclei by the photon bath in a stellar core where the temperature is roughly 3x109 Kelvins (give or take a few). You can find it in any text book on stellar physics that includes supernovae, or try Explosion mechanism, neutrino burst and gravitational wave in core-collapse supernovae (http://adsabs.harvard.edu/cgi-bin/nph-bib_query?bibcode=2006RPPh...69..971K&amp;db_key=PHY&amp;d ata_type=HTML&amp;format=&amp;high=4366fa465129717), Kotake, Sato & Takahashi, Reports on Progress in Physics 69(4): 971-1143, April 2006.

The convective layers of a star are well mixed for the same reason that the convective layer of Earth's atmosphere (http://en.wikipedia.org/wiki/Earth's_atmosphere) (the troposphere (http://en.wikipedia.org/wiki/Troposphere)) is also well mixed. There is strong convective (http://ww2010.atmos.uiuc.edu/(Gh)/guides/mtr/cld/dvlp/cnvct.rxml) & advective (http://ww2010.atmos.uiuc.edu/(Gh)/guides/mtr/af/adv/adv.rxml) transport on virtually all spatial scales, and in all directions, which prevents heavier elements from settling. Transport in the radiative region of a stellar interior is dominated by diffusion, and I think one should expect some amount of settling under such circumstances. But I don't know off hand what the diffusion time scale is for nuclear particles in a stellar interior.

Ken G
2006-Jul-17, 03:27 AM
Thanks for the info. Heck we could look up any of this stuff, but it's more fun to talk about it. So what does the 57Fe photodissociate into, and where do the neutrinos come in? Does the energy get eaten primarily by the daughter nuclei in the endothermic induced fission, or does it escape via neutrinos? Sounds like photon diffusion is irrelevant, naturally since that takes very long.
Transport in the radiative region of a stellar interior is dominated by diffusion, and I think one should expect some amount of settling under such circumstances. But I don't know off hand what the diffusion time scale is for nuclear particles in a stellar interior.
That is indeed the issue. I think it must be long, even on stellar evolutionary timescales, because otherwise the core of even solarlike stars would be iron!

Tim Thompson
2006-Jul-17, 04:16 AM
There are two simultaneous processes.

1) Photodissociation of 56Fe (I had 57, which is wrong) into alpha particles (also known as 4He nuclei) sucks up ~124 Mev of energy, removing the thermal support for the core.


gamma + 56Fe -> 13alphas + 4neutrons - 124.4 Mev

2) Electron capture by 56Fe removes mechanical support by eating up electrons, which otherwise provide support against collapse by virtue of the Pauli Exclusion Principle. The neutrinos (nue) come from this.


56Fe + e- -> 56Mn + nue

The energy in the photodissociation gets eaten up by the process of tearing apart the 56Fe nuclei. The collapsing core goes so fast it goes too far & rebounds, expanding rapidly outward on the rebound when the infalling matter hits it. Boom. The outward propagating shock loses energy by dissociating the nuclear material falling through it. Old supernova models stopped here, because the shock always ran out of energy due to this dissociating activity, stalled & fell back. No Boom. But the material is so dense that it is opaque to neutrinos. So newer models include energy deposited by coherent scattering of neutrinos, and are also fully 3D, which was not the case before because of limitations on available computing power. Boom. Supernovae models now explode. The actual core collapse event takes ~0.1 second.

Ken G
2006-Jul-17, 09:14 AM
Ah, that's quite interesting. Thanks. For those keeping score at home, Tim's two processes describe how gravity circumvents its two main adversaries, gas pressure (which had been preventing core collapse in evolved low-density high-mass cores) and electron degeneracy pressure (which prevents core collapse in high-density low-mass cores). The last remaining adversary is neutron degeneracy pressure, which in some cases does come into play but too late to prevent the supernova (that's when you get a neutron star. If even that fails to halt the collapse, you get a black hole.) It all comes from the attributes of iron, as the temperature keeps rising due to the fact that iron itself cannot be fused to release new energy to prevent the core contraction.

However, I do still have a question. There must be about 26 electrons per iron nucleus even in a pure-iron core, by charge neutrality. But process (2) above can only swallow one electron per iron nucleus, so this must only be kind of a detail about at what point the collapse can occur. Isn't it more fundamentally important that if you get a large enough mass in the iron core, electron degeneracy pressure cannot save you even if all the electrons are still there?

korjik
2006-Jul-18, 04:23 PM
I would imagine that as the electrons are captured, the rate of capture should go up very quickly.

Ken G
2006-Jul-18, 08:47 PM
But the capture process given still only captures one electron per iron nucleus, which is only 1/26 of the total.

Tim Thompson
2006-Jul-19, 04:59 AM
But the capture process given still only captures one electron per iron nucleus, which is only 1/26 of the total.
That's how you get neutron stars. One way or another, those electrons get jammed onto the protons in all those alpha particles, and make neutrons.

snarkophilus
2006-Jul-19, 05:55 AM
The actual core collapse event takes ~0.1 second.

And when you consider the scale on which this is happening... those particles are moving fast!

Which brings up a question... how fast? How much does the core collapse? Do relativistic effects come into play at this point? I suppose they must, to some extent, just because of the high gravity. But are they important because of the speeds of the particles involved?

George
2006-Jul-19, 12:20 PM
That's how you get neutron stars. One way or another, those electrons get jammed onto the protons in all those alpha particles, and make neutrons.
Is it an on-going, reversible process though? How would you get such enormous mag. fields from a neutron star otherwise?

Tim Thompson
2006-Jul-19, 02:45 PM
The collapsing core is no bigger than Earth, and at that size the collapsing material will only move about 1/2 the speed of light. That's not fast enough for any serious relativistic effects, like time dilation & etc., but you do have to include everything in any reliable physical model of the collapse.

The neutron star magnetic field comes primarily from the parent start. The magnetic field is frozen into the collapsing core, so magnetic flux conservation concentrates the field into a far smaller volume, making the surface magnetic field that much stronger. There are other effects going on as well, that I am less familiar with, such as amplification of a seed field by differential rotation, but I don't know how important they are.

Ken G
2006-Jul-19, 04:02 PM
That's how you get neutron stars. One way or another, those electrons get jammed onto the protons in all those alpha particles, and make neutrons.
Indeed, but the issue behind my question is one of cause and effect. It is not necessary to swallow up the electrons in order to suppress their pressure so as to obtain the high densities, it may proceed the other way: high densities caused by gravity defeating electron degeneracy pressure causes the electrons to get eaten up. My understanding is that it tends to be the latter not the former. If so, then if the electrons did not get swallowed into neutrons, the collapse would always skip the neutron star and go right to a black hole. But maybe the idea is that when the core has just under the mass it would need to collapse, a few electrons get eaten by the process you describe and suddenly the core has enough mass to collapse without needing to actually add any more. So that is the trigger, rather than the cause, if you will, of the collapse. I can buy that.

George
2006-Jul-19, 05:13 PM
The neutron star magnetic field comes primarily from the parent start. The magnetic field is frozen into the collapsing core, so magnetic flux conservation concentrates the field into a far smaller volume, making the surface magnetic field that much stronger. There are other effects going on as well, that I am less familiar with, such as amplification of a seed field by differential rotation, but I don't know how important they are.
It's too far over me for now; I still want active charges producing a field. It would make a great separate thread, though; we may be getting a bit off topic, perhaps.

George
2006-Jul-22, 10:40 PM
From Fundamentals of Solar Astronomy (Bhatnager and Livingston) it is stated that flash spectrums show the metallic lines are confined to the lower region of the chromosphere. Hydrogen, ionized calcium, and magnesium are found in the higher regions.

[Edit: read the next two to eschew obfuscation.]

Ken G
2006-Jul-22, 11:34 PM
One must be careful, though, to distinguish temperature and ionization effects from separation effects. It sounds to me like they are saying that the last three are only ionized or excited sufficiently to give emission lines in the upper chromosphere, whereas the metallic lines are coming from the denser regions in the lower chromosphere. I doubt we are looking at mass separation here.

George
2006-Jul-23, 03:13 AM
It sounds to me like they are saying that the last three are only ionized or excited sufficiently to give emission lines in the upper chromosphere, whereas the metallic lines are coming from the denser regions in the lower chromosphere. I doubt we are looking at mass separation here.
Yep. Reading further in the book, it confirms your thoughts.

"Now we know that the distribution of emission lines with height is not a true indicator of the actual stratification of the elements in the chromosphere, but depends on the following two condistions: 1) Intrisic strength of the transition responsible for the emission of the line, and 2) Condition for excitation and ionization of atoms, which in turn depends on the temperature and the gas and electron pressures."