Plastics in the cold

Stewie007 · 10-26-2009, 07:10 PM

I'm still not able to build because of lack of finances, however I'm always busy designing.

I've decided that not only am I going to chill the cpu, but 3 GPUs too. I've decided that with the size and complexity of the system, it will be more efficient to take as much load as possible. I believe that 500W at -10 to -15F is pretty doable.

So the question is, I don't see any GPU blocks that are made out of much other than POM or Acetal. What kind of experience do you guys have with this stuff? Keep in mind that not only do I need some decent chill resistance, my pumps are putting out a lot of head pressure. 20GPM through your typical VGA block might decrease tolerance of the block.

I suppose I could have a machine shop cut a copper piece for me that is just like the plastic piece....

TheGreenFoX · 10-27-2009, 12:32 PM

Bump... I have the same question... Im doing a dry ice cooled setup so would like to know if anybody know how the plastic part will handle the cold

Plan.B · 10-27-2009, 01:39 PM

At this coolant temp a delrin block (apogee gt) developed numerous hairline cracks around the barbs which resulted in coolant seepage after about 2-3 weeks of almost 24/7.

TheGreenFoX · 10-27-2009, 02:13 PM

I guess acetal is a no go... What coolant did you use?

Stewie007, FYI: I think ill go for this GPU cooler: http://www.ybris-cooling.it/product_...2d630837babb7d

Plan.B · 10-27-2009, 03:48 PM

I used a mix of denatured alcohol and distilled water. Been looking at gpu blocks also, I'm starting to think that a machine shop is the way to go

Utnorris · 10-27-2009, 08:45 PM

You will also need to braze the block together instead of using a o-ring to seal it if you are going into the negative range. Standard o-rings will not handle the cold very well.

flak-spammer · 10-27-2009, 11:13 PM

Quote:

Originally Posted by Plan.B

I used a mix of denatured alcohol and distilled water. Been looking at gpu blocks also, I'm starting to think that a machine shop is the way to go

Hmm I thought that was a known issue with alcohol and acetal. Pretty sure it's pretty well documented that alcohol and plastic don't seem to mix well. Should be other threads around here mentioning that (unless I'm just dreaming).

Edit: Okay nevermind I was dreaming, I'm thinking of Acrylic can't say why you'd have problems with acetal then.

ChielScape · 10-28-2009, 06:37 AM

alcohol and acrylic is a no-go, acetal is weak against cold. so you cant really use either.
you'll want to look at full metal blocks. or use antifreeze isntead of alcohol and use acrylic if you have to.

TheGreenFoX · 10-28-2009, 06:58 AM

Okay...

Last question

What pump to use at those temperatures?

Stewie007 · 10-28-2009, 02:20 PM

Brazing is not a problem for me

. I do it for Carrier corporation every day. Whatever I use I'm going to be brazing.

The problem with the machine shop option is that it will be expensive for multiple cards.

TheGreenFoX · 10-28-2009, 02:38 PM

Did you consider the block that i linked to?
If not please tell why

Stewie007 · 10-29-2009, 03:50 PM

If they make them for the 5000 HD series then I may indeed.

TheGreenFoX · 10-30-2009, 02:06 AM

It matches the 58xx series mounting holes AFAIK.

I had coolerkit.dk (danish reseller of EK bloks) doing some research for me, regarding the aceal in EKs products:
http://solutions-in-plastics.info/nl...ACETAL%20C.pdf

So -50 is the lowest temperature for the acetal in EK blocks according to that datasheet, but i might check it with Eddy before buying the blocks i need

Stewie007 · 10-30-2009, 06:56 AM

Unfortunately, though, all of the data sheets I have read state that the lower limits are short term. I believe even as much is so for Delrin. So I think I might not be using the plastics. Seems too risky.