Thermal properties of Argon

[Azriel]

I was considdering using Argon gas as the atmosphere in my sealed box prodject. It's used as a shielding gas for welding. I'm just not sure if it will transfer heat as well as the Nitrogen/Oxygen hybrid we call air. This is to combat condensation, btw.

[Azriel]

Or straight Nitrogen, I can get either very easily.

[surlyjoe]

I would imagine argon or nitrogen would work well, anything to displace the air with the moisture, if its truly sealed, even dried air would work, as long as it was free of moisture. I was thinking about doing this as kind of a "CPU dome" with a tupperware bowl and the gas over a pelted CPU w/watercooling. it sounds like a great idea!

[Azriel]

yup, truely sealed..

No drives are going in the main box. All external SCSI. Lots of gromets.

[donny_paycheck]

That sounds like a good idea. At a place I used to work, we used helium for a shielding gas for CO2 laser cutters and welders, but it did the same thing for the same reasons - the gas being inert and all.

As for thermal conductivity I imagine it would be the same as air or very similar. I think just about any gas would. I think the way to increase thermal conductivity would be to increase the pressure of the gas so the atmosphere inside your case would be "thicker". Do you know what I mean?

My first post here! w00t!

[Azriel]

I don't think that would work, as gas has to expand to absorb heat. (How freon works) It has to do with the distance between molecules. A very slight negative pressure may help, though. I could do it myself, but I am considdering having a "friend" of mine weld up the skeleton. He has owed me $75 for almost 6 months for upgrading his 486/25sx to a 233mmx, 4meg permidia, SB 512, 64meg, 2gig, 32x, 56k, 10/100. He got one hell of a deal, if you do the math.

(That's where processor #2 ended up SurleyJoe. Thanks again.)

while I'm thinking of it, if anyone wants to play with a Tillamook 266 Pentium (socket 7) PM me. It's free. You'd have to do a voltage mod for the I/O voltage. This processor was only intended for industrial SBC's, and laptops. Intel claims it does not need a heatsink. Definite OC candidate for anyone who can figure out how to lower the I/O voltage on the board. I got it to ID on an FIC PA-2013 as a 233, but it never made it past the post screen.

[surlyjoe]

[i]
(That's where processor #2 ended up SurleyJoe. Thanks again.)

[/B]

HAHAHA sweet!!

[Kaitain]

I don't think that would work, as gas has to expand to absorb heat. (How freon works)

Freon works by a combination of the following:
*liquid to gas phase change. Energy is thus absorbed at constant temperature (this is the latent heat of vapourisation).
* Joule Thompson effect. When a gas is passed through a point of throttling (eg an orifice plate), at the point of throttling some of its internal energy as heat is converted to kinetic energy (as turbulence). Accompanying this is a temperature drop. The gas as the same energy content before and after the orifice, but a lower temperature.

The most significant part of the cooling is the liquid/gas transition. A gas doesn't have to expand to absorb heat. If it's expansion is prevented, it merely becomes a hot gas at higher pressure.

A very slight negative pressure may help

You'd want a slightly positive pressure inside the case to keep your gas pure - even microscopic fractures in the metalwork will allow gas to pass.

I'm just not sure if it will transfer heat as well as the Nitrogen/Oxygen hybrid we call air.

Thermal conductivities of selected gases:
Argon: 0.0177 W/(mK)
Carbon Dioxide: 0.0166 W/(mK)
Nitrogen: 0.0260 W/(mK)
standard air: 0.0263 W/(mK)

all these are accurate at 300K, however there's huge variation with temperature.

Dry air is about the best gas you can use - seal up your case and put some dehydrated silica gel in there. Perfect

Hope that's of use

[Random Nonsense]

azreil, what are you gonna be using that will cause condensation probs? TEC's + water? or vaporphase?

[Azriel]

Peltier air conditioning.

[Azriel]

Kaitain...

Your ability to nitpick a simple answer into an encyclopedia volume never ceases to amaze me. Thank you, however, for the portion of your post that answered my initial question.

[surlyjoe]

Originally posted by Azriel
Kaitain...

Your ability to nitpick a simple answer into an encyclopedia volume never ceases to amaze me. Thank you, however, for the portion of your post that answered my initial question.

\


thats what makes kaitain ..kaitain!!! :-D

[Random Nonsense]

if you are doing it how i think, all the incoming air goes through a chiller? then the air will be dehumidified by this alone, me and spode were discussing it the other day, if you incorporate a drain into the little air chiller, then any condensation will drip out. remember the computers components will be at a higher temp than the now chilled air in the case... so no probs there

[Kaitain]

...except that the air coming off the chiller will be "saturated" (100% humidity at that temperature). If the system runs for a long time, then non heat-generating components (the frame, bare bits of mobo etc) will be cooled down to air-temperature and may still allow condensation to form, which may still run on to water-sensitive components.

Tho' the risks are reduced.

To use the chilled-air method of dehumidifying reliably, you need to sub-cool the air below the temperature you intend to use, then reheat it to a few degrees above that temperature, so that the air is less than saturated.