thesis project

[atrix007]

Hey guys,

You guys are the experts so i thought i would use some of your experience. I am a mechanical engineering student working on a thesis project that involves a water block design. I am actually planning to make the final design. I need to set criteria and from reviewing your site I saw that the maze 4 water block seems to be one of the best. I m looking to make a water block that will allow for low flow rates but will be able to extinguish a lot of wattage. Any tips you guys can give me would be greatly appreciated. I am hoping to send the final design to you guys for your expert testing.

Thanks,

Mike

[bigben2k]

Criterias:

-Should be durable/reliable and not prone to failure, either through cracks in the top, corrosion due to the use of mixed metals, etc.

-Should have a reasonable price, for its performance.

-Should either perform well with any pump, or have its own pump (if required).

-Should be able to mount it on any processor curently available, easily, and without any issues, especially due to the hoses tugging.

[atrix007]

I need a bench mark like, the best delta T achieved so far and by whom, and with what LPM. Also if someone knows the actual maximum mounting pressures recomended and maximum water block weight.

[bigben2k]

dT is a parameter that depends on the power applied. What did you have in mind?

Mounting pressures changes with the new processors that have an IHS. Specs are available at Intel and AMD's sites (but also within many threads here).

[atrix007]

yeah you are right I neglected that it will be tested on a regular AMD 3000+ 400MHz or 2600+ 333MHz I think they put out about 70W. I will have to check into that. I'm just getting into that aspect of things. Basically I want to reach high h values while the fluid is passing over the chip. The idea is to use venturi effects via the fins to increase flow rate directly over the chip. This and heat sink design will hopefully help provide more efficient cooling.

[Etacovda]

testing via c/w is a hard task, you cannot directly compare testbeds; setting up a test bed is a project in itself, and will more than likely be harder than the block itself...

This will not be an easy project - but good luck. Id suggest getting a bunch of blocks tested here and using those as a comparasion basis - ie, mcw6000, dtek whitewater, DD tdx, and making a test be to reliably get data comparable to what has been done already.

[bigben2k]

Right on Etacovda.

...but if you want to join the WBTA, I'll walk you through the whole thing. PM or email me.

BTW, I've already acquired 3 blocks for WBTA member cross-testing; they'd be at your disposal.

[HammerSandwich]

Mike, Phaestus has tested quite a few blocks that eclipse the Maze 4. Check out the interactive WB page (http://www.procooling.com/html/pro_testing.php).

[jaydee]

I need to set criteria and from reviewing your site I saw that the maze 4 water block seems to be one of the best.
Actually it is one of the worst.

[atrix007]

To tell you guys the truth an in depth analysis will probably not be possible I don't really have the time or the resources. Our university is quite poor and will not fund such an endeavor. Also this thesis has to be finished by November, and I still have to work on other classes while I'm doing it. I wanted the test to be simple, without a thermal loop. By using the on board thermal couple and one directly before the water inlet and knowing the thickness of the water block right over the diode I can calculate the heat transferred to the water. I think we will just do a comparison to the water block that I already have, just so the test procedures are the same and the discrepancies will be low.

[atrix007]

Does anyone know where I can find out what the motherboard standards are for CPU cooling ie. What the maximum dimensions, mounting pressures and weights are etc.

[Long Haired Git]

AMD specs are on AMD's site, Intel's are on Intel's.

[pHaestus]

"By using the on board thermal couple and one directly before the water inlet and knowing the thickness of the water block right over the diode I can calculate the heat transferred to the water."

Details? I would opine that you don't have enough measurements (or the correct ones) to accurately solve for W.