Water block in reverse

[dith]

Hello hello,

I'm trying to design a water block which will be sandwiched between the cold sides of two peltiers (Think Swiftech MCW-CHILL). I've alread dealt with all of the heat distribution issues using pelts entails, now I'm trying to design the actual block to be sandwiched between them. I was thinking of using a crosshatched-channel design similar to the Swiftech's, but for what I'm doing I really don't care about flow rate (I need 0.3 GPM at the very most). I just need to chill a tiny bit of water moving relative slowly as quickly as I possibly can. Specifically, if I can drop ambient water coming in by 10 degrees C as it comes out, I'll be a very happy man.

So far, what I've thought of doing is:
-Make the water block bigger than the peltier surface to maximize the time the water is in contact with cold metal and maximize the internal surface area of the water block
-Install swirl inserts in the feed pipe to maximize turbulent flow
-Mill the block itself down so that it is as thin as possible (shortening the time it takes for the entire block to cool and minimizing energy wasted cooling copper)

I'm at a bit of a loss as to how to go about this, as what I want to do seems like the exact opposite of what most people are going for in their water block designs.

Can anyone help?

[ricecrispi]

-Make the water block bigger than the peltier surface to maximize the time the water is in contact with cold metal and maximize the internal surface area of the water block
-Install swirl inserts in the feed pipe to maximize turbulent flow
-Mill the block itself down so that it is as thin as possible (shortening the time it takes for the entire block to cool and minimizing energy wasted cooling copper)
===================
If you want to have a thin base plate with a surface area that extends pass the pelt it wont work. Basically you don't allow enough time for the coldside of pelt to spread cooling throughout the baseplate.

A very detail explanation why.
http://forums.procooling.com/vbb/sho...73&postcount=4

http://www.overclockers.com/articles305/
If the liquid is also cooled, then the water block's temp is ..whatever.. and a thinner 1/8" spacer is perfectly acceptable.

WB pattern that will also perform well.
http://forums.procooling.com/vbb/att...3&d=1093346558

[dith]

I should have remembered that heat spreads in a conical fashion. Right now I'm thinking I'll get a 3/8ths inch block, mill out 2/8ths of it into a crosshatched pattern like the one you linked to, and then cap it off with a 1/8th" thick bar so there's 1/8th" of copper in contact with each side of the peltier sandwich.

I'm wondering if the peltier sandwich is the best idea. Perhaps placing my two peltiers side-by-side would work better? That way, I could manufacture one side of the waterblock out of lexan, and the waterblock would be twice as long, doubling the time that the water's in contact with cold copper. Then again, I would have to insulate the side of the waterblock that isn't in contact with the peltiers, which would introduce inefficiency. Perhaps the peltier sandwich with an elongated waterblock really is the best way to go (does anyone have any idea what the optimal length of such a setup would be?)

Thanks for your input ricecrispi!


(sorry to anyone reading this for editing this multiple times, but it seemed like a better idea than cluttering up the thread)