• Welcome to Overclockers Forums! Join us to reply in threads, receive reduced ads, and to customize your site experience!

objective water block testing project

Overclockers is supported by our readers. When you click a link to make a purchase, we may earn a commission. Learn More.

athlonnerd

Member
Joined
Apr 14, 2002
Location
missouri city, tx
I’ve seen a lot of debate about what the best style water block is on these and other forums, yet none of it can be proven, mainly because the results are based on differing control variables. I think I want to start a project where I make models of most all water block designs, and compare the results.

Many have been working on computer modeling, but personally I think it’s high time that someone conduct some actual tests, with undisputable results.

What I was thinking:

Use aluminum; mainly because its easy to machine, it’s cheap, and is a good conductor, although not many would think of making their actual blocks of aluminum, it will be fine for comparison.

The only part I will change in the tests is the water block, that way, the results will be comparable.

I think I’m going to make a die simulator, this should not be too difficult, I was thinking of 100W, over a 1cm^2 area, and a built in thermometer.

I’m going to test a wide variety of water blocks, although, I’m not going to be able to make all, due to machining capability. I don’t think I will test commercial blocks, unless I make duplicates out of aluminum.

I don’t think ill use thermal interface material, because it introduces way too much variance with each new application. Ill run the tests at school, it has a much less changing room temperature.

I expect and am hoping for a lot of criticism from y’all; I’m not entirely sure about every aspect. Although, please don’t post your opinion of, no this will never work, unless you are able to give researched reason as to why not.

Has anyone tried doing this before?

My aim screen name is athlonnerd if anyone wants to discuss this.
 
So, ill start a list of designs to test:

Channeled (many variations there of)
Cross drilled (many variations there of)
Capillary style (couple of variations there of)
Spiral (if I can get a rotary milling table)
Hoot’s the wave (if I can get a rotary milling table)

Please give suggestions of other design, this may be a never ending project as new designs are developed.

I'm planning on getting an X-Y milling table, and have end mills, and a hefty drill press at school

I thing I will have a basic base thickness which all designs will incorporate.
 
when you setup the test bench, make sure the following variables are IDENTICAL:

pump, heatercore, and especially tubing. make sure the fitting is the same on each block, and make sure the lenght of tubing is the same. also, where possible, keep everthing on a level surface and keep the tubes runnign in as similar directions as possible. might want to invest in a pair of high quality threaded quick disconnects for the blocks, so you dont wear off the tubing and have to cut it down constantly...
 
yes, i realize that nothing must change, i think ill even use the same water, good point wiht the quick disconnects.

i may use a controlled automatic water chiller, so theat the water does not change temp at all.
 
athlonnerd said:
I thing I will have a basic base thickness which all designs will incorporate.

Well this is the wrong thing to do. Base-plate thickness is critical to the performance of most any design. Compare for example the Swiftech MCW-462U, which uses a 9mm thick base-plate (from memory), whereas my design uses a base-plate thickness less than 1mm 'cos that's the best thickness to use with my design.

The base-plate thickness cannot be kept the same for all the different designs.
 
what Cathar said is what I was thinking. along with flow rate.

the two biggest things is base plate thickness and flow rate - some blocks work optimally with low/high flow, and as well some blocks work well with thick/thin bases.

I'm not sure how you would test all these differences, as it would be very time consuming.....
 
ok, heres what i will do, make every block wiht like a 1/2" base thickness, and test it, then shave off 25 mills, test it again, shave 25 mills off again, etc.

that way, well find the optomum thickness for each desings, adn we can compare the optimal thickneses.
 
If you want to perform a test like this, see if some people are willing to send you there block (unmodified of course), and then test those. I wouldn't try to mill them out of aluminum. There's some famous designs like the Spir@l, TC-4, The Swifty's, etc.
 
i think the point of this test is to find the best general desdings first, not nesesarily products, although, that could be a later addition
 
athlonnerd said:
i think the point of this test is to find the best general desdings first, not nesesarily products, although, that could be a later addition

Ohh, so you wanna test things like a spiral design, zig zag etc? If so, then I guess alu would work - to test the success of each design, you'd want the same base height, wall sizes, etc. on the block I would think... :)
 
heres what i hear about swifties, with the big open area, use a thicker base? i dont get that, im thinking about picking one up somewere and milling it out so it has a thin base...
 
maskedgeek said:
heres what i hear about swifties, with the big open area, use a thicker base? i dont get that, im thinking about picking one up somewere and milling it out so it has a thin base...

well, these are the kind of disputes we can settle by conducting these tests.

so i guess the list is now:

Channeled (many variations there of)
Cross drilled (many variations there of)
Capillary style (couple of variations there of)
Spiral (if I can get a rotary milling table)
Hoot’s the wave (if I can get a rotary milling table)
swiftec style (many variations there of)
 
Back