Ovenizing vs. compensation

as

s so

rem

even

t for

transfer is remarkably fast. The vapour pressure of the vapour in the cham ber is set by the coolest point, and the working fluid condenses there and wicks away to the hotter points.

, heat transfer through that would probably be the limiting factor, but the vapour is is usually condensing into a woven copper cloth wicking layer.

um line when I encouraged a vacuum distillation process with a hot air gun.

the vapour flow a lot. At Iasys we used to test our heat pipes for rapid h eat transfer at close to room temperature to make sure that the manufacture r had got all the air out before putting in the water - initially we were s ending back quite a few, but they got better.

eat differentials close to room temperature, but that was only true of badl y made heat pipes.

rocess

Yeah the partial pressures don't interact, but the atoms do! You can get a lot more scattering, which reduces the mean free path of the atom/molecule.

George H.

e as

io,

t.

is so

eorem

n even

ast for

at transfer is remarkably fast. The vapour pressure of the vapour in the ch amber is set by the coolest point, and the working fluid condenses there an d wicks away to the hotter points.

nt, heat transfer through that would probably be the limiting factor, but t he vapour is is usually condensing into a woven copper cloth wicking layer.

cuum line when I encouraged a vacuum distillation process with a hot air gu n.

ow the vapour flow a lot. At Iasys we used to test our heat pipes for rapid heat transfer at close to room temperature to make sure that the manufactu rer had got all the air out before putting in the water - initially we were sending back quite a few, but they got better.

heat differentials close to room temperature, but that was only true of ba dly made heat pipes.

process

Correct, but probably not the right way to think about it.

With just the condensable vapour in the heat pipe, the gas pressure is pret ty much the vapour pressure at the cold end. The vapour boiling off a the h ot end will produce a small local over-pressure, just enough to drive the g as flow to the cold end.

You can work out gas-flow rate per watt, and estimate the pressure drop alo ng the heat pipe requires to drive that gas flow, but it isn't much.

With a non-condensable contaminant gas you've got to have gas circulation t o get the non-condensable gas away from the cold end after it's dumped the condensable vapour coming from the hot end - messier and much more pressure drop to sustain the same flow rate of condensable vapour.

What is interesting is that water, even though it is much less dense than the metals, still has a higher volumetric heat capacity, nearly twice that of aluminum and 20% more than even iron even though iron is 8 times as dense. Water is pretty amazing stuff.

Heat pipes were notoriously unreliable. Maybe they are better nowadays.

Out of two laptops, I've had to replace a heatsink assembly once. It's still going strong. Perhaps that one will die in a few years, but by then I'll buy a whole new laptop.

Doesn't seem like the ticket for hi-rel stuff, but they do work nicely while they do.

Tim

That's also why the thermal conductivity of gas is nearly constant over a wide pressure range; and why a thermocouple gage works fine for vacuum, not so well for atmospheric pressure. It has to be a pretty hard vacuum, before the lack of heat-carrying molecules dominates the modulation of the mean free path (product of molecule count and path length roughly determines the heat-transfer rate).

The 'inert' gas doesn't have heat picked up when it was evaporated, in general, so you want the evaporated stuff to actually travel from hot to cold side (dissimilar masses, elastic collisions, means the heat doesn't get quickly shared in two-particle collisions).

It's a matter of price. In the PC world, they couldn't spend enough to achieve reliability.

Stuff made for space and military uses lasts forever. For instance, see Thermacore:

Joe Gwinn