sensei ! how to choose good waterblock ya? is there anyone here ever done custom made waterblock? i'm sorry if the topic is already open. i'm doing my final year project on waterblock alone. so i'm looking oso for other current trend of waterblock for evaluation. this thesis gimme headache already.

suggesstions and information deeply appreciate. thank you!

Check out waterblock reviews, that would be a good place to start.

Btw, what kind of course are you doing? Thats one cool final year project..

ask Metalzone, he's the guru for watercooling

hi there, hope this helps

http://www.xtremesystems.org/forums/showthread.php?t=67211

i did this design for my fren:



could run at 4Ghz...before could do 3.6Ghz but at temps of 70C. Now temps is idling at 44C anf 50C on load.

manufacturing course. this project only consider machining the waterblock. i'm just about to start searching for waterblock materials. nanti, i hv to do the rig myself, which i'm not so sure how to build one

thanks a lot. informative to me!

walla~ so u did it on what ey? can u gimme the size of the block u design there?

i'll prolly ask more in the future. anyway, thanks in advance

This post has been edited by bubblewrap: Jul 13 2007, 09:19 AM

material is pure copper.....2" in diameter ( enough to cover the surface area on all Proc ). water channels are made using a bench drill and dremel tool kit to smoothen the drilled edges. the mounting braket is made of acrylic.

ooh is it? u use manual? cool. i wonder if can i do the same design. is it already copyright?
thanks for the useful info. i might consider one for my block analysis then!
the block is still in use now? how you assemble the mounting? you use screws only?

anyone, wants to share any custom blocks here?

is it ok if i want to consider using brazing or welding to assemble the blocks? coz i'm considering using whole copper as the waterblock

nope...no copyrights, u can use my design . I built this block specially for me fren since he complain PC hot and yet doesn't wanna buy a WC setup. I posted some pics here ( # post 172) :

http://forum.lowyat.net/topic/41021/+160

I use acrylic as mounting bracket and screws the secure it. My whole block is copper...yup, brazed the whole thing up. But make sure you have good brazing skills.

oh, so this was where Allngap told u to ask me... lol. i replied ur pm's adi.

anyway, to seal the waterblock, since u have access to CNC machines, i suggest you just mill the O-ring grooves and put a rubber O-ring there to seal it (plus screws outside it to hold them together of course). then you no need to braze and can use other materials for the top, as well as being able to open up the waterblock whenever you wish. you don't actually need to use all copper for the whole block as you probably know how difficult it is to machine copper. the only important part that should be copper is the base plate, in which all your whatever cooling design goes into. the top cover/plate or any additional mid plates don't contribute to the cooling performance so you can use other softer and easier to machine materials like acetal (or delrin acetal).

if i make a waterblock size 2" x 2" (like extremeocer dsgn), how much power of pump shoud i u use then? is it the smaller the inlet/outlet tube, the lower power pump i should use? i'm afraid i might leak the waterblock soon enough if the block cannot sustain the pressure inside.

another question, do these people ever done analysis on their waterblock? so far haven't seen one through journals only testing base on temp decreament

This post has been edited by bubblewrap: Jul 18 2007, 04:17 PM

Actually the smaller the inlet/outlet, the more powerful a pump you will need. Small openings constrict water flow, reducing cooling performance.

Analysis on waterblocks? What do you mean? Your question is a little too general..

I suppose what you mean is the pressure drop (or flow restriction) of the waterblock. The physical size of the waterblock doesn't tell you anything about flow restriction (aside from the size of the barbs). It is the design of the internals. Check the pm I previously replied you. I explained the design differences between a high restriction(relying on flow turbulence) and low restriction(relying on surface area). That is what dictates the flow restriction of the waterblock.

The higher the flow restriction of the waterblock, the more pumping power is needed for it to perform efficiently.

The waterblock won't leak if it is built properly. I've used pumps up to 6 metre head pressure and I haven't had any waterblock leak (btw, these are o-ring sealed ones, and with acetal tops).
In fact, the funny part is the joint between the tubing and the barbs leaked first, while i was testing the Storm waterblock with that 6m Hmax pump. Had to clamp kao kao to stop the leaking.

What kind of waterblock analysis are you talking about? Performance/flow/pressure drop tests? Or design analysis?

heat transfer and finite element analysis. therefore, need to study the pressure drop inside (based on pump that i will use), heat transfer efficiency, how the water flow and such. not sure if different channel design will give different? considering using microchannel gonna use CFX and ProMechanica. if possible, gonna try for ABAQUS. aiyoh, need to study more. my fundamentals not good enough

nway, thanks mr MetalZone and lohwenli. i'll work on the design. i might show some of them.

ah cool... Computational Fluid Dynamics. the way the engineers have been doing it. yes of course different designs will give you different results. hey you're gonna be more pro and than any of us soon enough already.

Lee garbutt has done some of the best water cooling tests: http://www.systemcooling.com/review_index.html#liquidcooling
Might be a good read for you.
he really put a lot of effort and money into his testing and test bench. the VWR thermistor thermometers he uses costs RM1500 each already.
you might want to start off with the storm waterblock to help you understand a high pressure drop pure-impingement(relies on flow turbulence) waterblock's flow vs thermal resistance characteristics: http://www.systemcooling.com/swiftech_storm-01.html

---

Added on July 19, 2007, 11:35 amHere's another from procooling, testing the Swiftech MCW6002-A:
http://www.procooling.com/index.php?func=a...s&type=2&cat=18
here you can see how the more restrictive 3/8" barbed inlet on the MCW6000-A actually performed better than the 1/2" barbed inlet of the MCW6002-A. This is a very simple example of how accelerated flow (impingement) can bring improvements to cooling efficiency. The D-TEK FuZion's performance could in fact be further improved with the addition of an an nozzle at the inlet (another example of impingement).
but to take advantage of a high pressure drop waterblock utilising impingement, you need a more powerful pump (something like at least 3m head is good enough).
Haha, yea, okay... i'm a sucker for impingement waterblocks.

---

Added on July 19, 2007, 4:03 pmbro... oddly enough... i didn't get your pm... something seems to be wrong. coz someone else also pm'ed me and i didn't receive it.

This post has been edited by MetalZone: Jul 19 2007, 04:03 PM

oo okay! will look into it. so many infos. i want to cry
if i don't want to consider jet impingement method, relying on surface area is enough then?

haven't got my own design to analyze

here`s how i DIY it , just some infos and idea.

http://forum.lowyat.net/topic/474837

hope this will help

Hmm..same project as mine?

Writen a thesis on this.. just shoot me a Q if u need help

This post has been edited by aleck: Jul 24 2007, 12:28 AM

^ this is the real WC guru here...

oh really? yeay! what is ur title then? hmm. i think i heard ur name from my senior. i wonder is it alex? so u grad from UM? if its true, tell me ur thesis title so i can find the book in the library

right now, i'm discussing with my supervisor on the material and cutting tool. i need to determine the cutting tool diameter so i can design channel according to tool. gonna use MasterCAM to create the tool path for CNC machine.

so u use G-Code?