You have two hollow metal spheres, one slightly smaller. They both open into hemispheres. Magnets are arrayed along the inside of the larger, and on the outside of the smaller. When the smaller sphere is placed inside the larger, the magnets repel each other and the smaller sphere floats without contact. If ice is placed in the smaller sphere and all air is (somehow) evacuated from between them, will the ice remain frozen indefinitely? Would heat from the outside affect the smaller sphere with no air to conduct it?
Unless I’m misunderstanding your question or physics, I would assume the ice will melt because air is not necessary for heat transference. After all, there’s no air in space but we still get thermal radiation from the sun. I wouldn’t be surprised if it took longer though.
No, because heat can be transferred by radiation.
Aren’t you essentially describing a “perfect” Thermos flask? As others have said, it would keep the ice cold for some time, but radiated heat would still melt it eventually.
Exactly.
Heat transfers by radiation and by conduction.
Radiation is a lot less of a heat loss, but it is real - glass thermos bottles for example have a mirror coating to help reflect back some of the radiation. (And reflect back outside radiation trying to get in, depending on whether the thermos is holding a hot or cold load.)
So your example is probably the most efficient thermos bottle, but nothing’s 100% insulated. the ice will melt eventually. Heat radiation from the outside shell will hit the inside shell.
I suppose you have to allow for radioactive background, cosmic rays, etc. Off hand I have no estimate how much of a difference that makes, other than “not a lot”, since you can keep liquid He close to absolute zero in the real world with the right insulation.
This is basically how a Thermos bottle works. There are two glass tubes coated with silver, and fused at the ends in a near-vacuum*. This creates a situation just like what you describe, using glass instead of metal. It will keep the ice frozen for a long time but not forever.
My father used to work for Thermos. I got to go on a plant tour to watch how they do this.
There would also still be some kind of mechanical coupling between the outer and inner layers - even if magnetic - so vibrations to the outer shell could transfer to the inner one and some of that kinetic energy would end up being converted to heat through friction etc.
Right. The usual triumvirate for heat exchange is “conduction, convection, and radiation”. Convection is simply currents caused by heat differences, so it’s really not a transfer between two separate things, it’s mixing of heat differential in a liquid or gas. md2000 isn’t wrong; I just thought I’d complete the list from a more general context.
That’s worth considering. I would guess that to be vanishingly small factor, since the thermal motion of one set of magnets would be very non-coherent, and all that would be imparted to the outer set would be changes in averages in regions. That is, as one dipole moves out, another nearby is likely to be moving in, and they’d nearly cancel each other out. Regardless, there would still be some small component that gets transferred. My guess is that radiation would be far more significant.
Incidentally, the magnet arrangement as described won’t work. There’s no arrangement of permanent magnets and/or static electric charges which will stably hold any object in place.
What actually would happen if you put a sphere covered in magnets with north poles facing outwards inside a larger one covered in magnets with north poles facing inwards?
Not to mention, you’d need pretty beefy magnets to levitate a metal sphere pulled down by gravity.
Well I’ll be . . . I did not know that Thermos was a brand name. Thank you!
One more to add to the list (scotch tape, kleenex, popsicle, . . . )
I went to school with a Thermos.
Title edited to better indicate subject.
Colibri
General Questions Moderator
If you think about the classic lines of force diagram of an ordinary bar magnet - the field lines come out of one pole, then curve back around (in all dimensions) and sweep back in at the opposite pole.
Those lines still want to (have to) do that regardless what you do with the magnet - so if you cluster the magnets seamlessly together into a sphere with all like poles facing inwards, the magnetic fields will interfere with each other - I think to the effect that some magnets will be weakened while they remain in that configuration.
Incidentally, when you pour liquid nitrogen into a thermos-style Dewar flask, you get a series of events. First there’s intense bubbling as the cold LN hits the room-temperature interior of the flask and cools it down, but then things seem to equilibrate as the “hard boiling” finishes up and the interior cools down. But after a minute or two, there’s suddenly another period of apparently unprovoked boiling again.
What happens is that there’s a thin layer of vapor at first between the inner wall of the thermos and the liquid nitrogen, even while it seems quiescent. The inner glass wall has been partially cooled down, but it’s still a lot hotter than the liquid nitrogen (although very cold by human standards). Eventually radiation across the thermos walls cools the inner wall enough so the vapor barrier breaks down, and then suddenly the liquid nitrogen is in direct contact with the inner wall, causing that second stage of boiling.
I used to work with liquid nitrogen-cooled systems a lot in grad school, and was fascinated and perplexed by this secondary boiling, until I stumbled across an article in American Journal of Physics that broke down the heat transfer history of LN in a Dewar flask into five distinct stages.
A curious fact which I learned recently:
Thermal conductivity in a gas is nearly constant down to around 1/1000 of an atmosphere. You would think that evacuating 99.9% of the air from a vessel would give you a very strong vacuum, but in fact it’s quite mild by scientific standards. Only once you get to around 1/1000000 atm does the conductivity drop to a small fraction of the atmospheric level, so if you’re building a Thermos-like vessel you need to get to that range.
PS: I don’t suppose there are any serious vacuum experts hanging around? I have some questions myself…