Slashdot Mirror
Planet-Gobbling Star
crymeph0 writes "BBC is carrying a story about a star that mysteriously brightened three times last year. Scientists now know why. It's been eating gas-giant planets that orbit it! I'm just glad Earth isn't a tasty gas planet, or else we'd have to start making sacrifices to Sol to play it safe." It's hard to prove things from 20,000 light years away, but this explanation is interesting.
If it's 20,000 LY away it didn't brighten 3 times last year (that we know of)... rather it brightened 3 times in one year approximately 20,000 years ago.
Read Niven's A World Out Of Time (multiple meanings in the title) for a similar idea. It's one if his first "State" books.
SPOLIERS BELOW
Basically, something else gets dropped into Jupiter. And there's some fascinating ideas on how to move a planet around.
You cannot apply a technological solution to a sociological problem. (Edwards' Law)
1) See, I asked about how much it *brightened*. Not how bright it *got*. I noted that line, too., but knowing that it is now 600,000 L_sun isn't really helpful in telling us how much brighter it is now than before. We would need to also know what its starting point was. This makes a difference: if the star brightened by 0.1%, the possible mechanisms are quite different from the star brightening by a factor of 100.
2) Um, no. When a star gets to Fe (and only very large stars do), it makes a nice little explosion adn we enrich the interstellar medium. Which is where pretty much all of the "metals" (anything heavier than helium, according to astrophysicsists) in your body, Earth, the Sun, etc. come from. So the question about metal-rich stars isn't "are they producing the metals", they would have had to leave the main sequence for one thing. The question is did the cloud that formed them have an more metals than the average, or did the metals get preferentially introduced by, say, planets smacking in to them.
No, see, as star is WAY bigger than a planet. (By definition, almost.) So a planet, particular a gas giant which is in large part hydrogen and helium (10s of percent and up, by mass) smack into the star, unless the material stays right near the surface, all of those metals will basically be so thinly spread throughout the volume of the star that you'll never see a real enrichment to within error bars.
And remember, the volume of a shell goes like the radius or the star squared, so the thickness of the shell has to be pretty thin to keep an appreciable fraction of the metals. Say we want to spread the metals out over a volume roughly equal to the volume of the original core. Uranus is mostly core, so let us use its radius as the radius of the core. (Note: much of Uranus's core is hygrogren compounds, as are all giant planet cores. This means that we're *over*estimating the volume of metals.) And lets spread it over a spherical shell on the Sun's surface.
V_Uranus = 4/3 pi r_u^3
V_shell = 4 pi r_s^2 deltaR
where r_u is Uranus's radius (2.62E9 cm) and r_s is the Sun's (6.9E10 cm), deltaR is the thickness of the spherical shell, and the Vs are volumes. Equating and cancelling, we get that deltaR = r_u^3/3 r_s^2. Plugging in numbers, that's a thickness of about 1.3E6 cm, or about 0.0018 % of the Sun's diameter. Which, when you consider that the Sun is fluid and convection does happen (although the most convective part is a bit lower down below the surface), isn't a whole lot. Confining the metals to that region would be very difficult.
This would probably be why current thinking tends more towards the "the clouds that formed star with planets were unusally rich in metals." Also, it makes sense: more metals, more stuff to actually *build* planets with.
No - you don't fuse iron. You neutron-stuff them. During the supernova, the outbound shock wave carries so much energy (and neutrinos) that neutrons are literally "shoved" into nuclei. You get ridiculous things like iron with hundreds of neutrons, which then decay down into normal elements by alpha emission and beta decay. This is r-process stellar nucleosynthesis. (There's also s-process stellar nucleosynthesis, which is also neutron stuffing, but on a much longer timescale. Essentially everything past iron is formed by r-process stellar nucleosynthesis. Check Carroll & Ostlie pp 527-528 for more info.
Stars die when they hit the iron stage because they can generate no outward pressure from fusing iron. They can't even fuse iron at all! It's actually a really complex procedure - basically, the iron starts to lose all of its electrons (from proton capture and other mechanisms) so the core rapidly loses electron degeneracy pressure, which is what was (briefly) supporting it. The inner core collapses very uniformly to a little neutron star, and the outer core decouples from the inner core, and the outer core rushes inwards at extreme velocities. The collision of the two is one of many explosions in a supernova. (Again, see Carroll & Ostlie's section on the Death of Massive Stars)
Anyway, the fuel isn't insignificant depending on what stage the star is in, and also depending on how fast the planet's orbit would decay once it's inside the photosphere. If it meets with the star's core without significantly losing mass, that could cause a VERY large brightening. Functionally it's equivalent to a nova, or the pulsing of Wolf-Rayet stars (without the mass shell shielding it).
Depends on the kind of star. The core of the star - that is, the "dense" part -
In this case, it's an F-class star. And you missed my point entirely, which was that if the star can convect the planet's hydrogen into the shell-burning zone, it can damn well convect its own hydrogren reserves down there, which are vastly in excess of what the planet could provide. So the star should never notice the miniscule addition of the planet's hydrogen.
This is all theory, of course, but unfortunately, theory doesn't quite bear out the "hydrogen compounds = gas giant planet cores".
See, that's where you're amazingly wrong. Let's review out giant planets, shall we? (If you want, I suggest you crack open Protostars and Planets IV; it's always good to actually do a bit of research.)
Jupiter May or may not have a core in the first place. If it does, it's at most around 10 Earth-masses (maybe as high as 15, but that's at the outer edge of the error bars). Mostly, it'll be hydrogen compounds with some rock and metal (real metals, not in the astrophysical sense). You need that core in standard formation models before you can accrete the hydrogen and helium gas. The metallic hydrogen is a layer right above the core, not the core itself.
Saturn Has a core, around 10-15 Earth-masses. Same as Jupiter in composition. This is easier to work out in theory because the equation of state is better understood for the interior pressures within Saturn. (Jupiter's higher pressures make things more dicey.)
Uranus and Neptune Definately have cores. Also icey with a bit of metals and rock thrown in. Again, need said core to hold on to the gas in the first place. Cores are pretty well constrained in size at around 15 Earth-masses in both planets. Given that both planets are around 18 Earth-masses in size, you bet your ass that this means that they are both mostly core. In fact, it's this that has lead some leading researchers to dub them "ice giants", in contrast to Jupiter and Saturn, the "gas giants."
I don't know where you got your "facts", but they're pretty much uniformly wrong. See Wuchterl et al. in P&P IV for more details on constraints on the present structures of these planets.