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

Watercooling flow rate question...

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


Jan 10, 2002
Dublin, Ireland
I am going to put a Waterblock on my cpu aswell as my NorthBridge. The CPU one is 1/2" and the NorthBridge one is 1/4".

If I use a Y splitter to split the flow to go through the two waterblocks, i.e. a parallel flow, would 66% of the flow go throught the CPU one, and 33% through the other one?

I know in theory, assuming they have the same resistance to flow, that is how it would be, but are there any other factors I should know of before I go parallel which would screw this up???

I'd be using the Eheim 1250 pump, and a Heatercore as a rad, with no res. in the system.
In theory, that's NOT how it would flow. If you do the calculations on head loss, you'll see that the ratio of wetted surface to flow volume is what is going to effect the head loss the most. With a smaller diameter hose, there is a higher ratio of surface to flow volume, thus there is a higher percentage of the flow subjected to a higher friction force along the wall.

Add in the differences between the total length of the flow path, different fittings and their associated head losses, and different losses through different blocks, and it's not as simple to predict as figure ratios of cross sectional areas.

BTW, you're biggest mistake is that you are trying to relate diameter to diameter, not area to area. A 1/4" diameter tube will have an area 4 times smaller than a 1/2" diameter tube, not a 1:2 ratio as you suggest.
Listen to Aesik. You do have a great pump so you will get
enough flow thru the northbridge cooler.

You could build a distribution manifold to distribute and
adjust the flows to what you want.

Far easier: Get a 1/2 inch northbridge cooler. :) Then run
them in series, CPU first.

How would you go about constructing a suitable manifold?

I've tried to find good commercial water manifolds, but I can't find any to suit my pump (3/4" ID).

Any suggestions?

with either system in your sig, aircooling would do the trick on the nb, but that doesnt sound like its what you want to do, soooo... you can basically disregard calculations, as aesik explained, its just too complicated to be worthwhile.

as you can also tell from what he said, there will be a MUCH greater amount of restriction in the nb block so you are making a good choice imo on going with the nb in parrallel. The amount of flow through the nb is not very important, it will be sufficient. The large majority of the flow will be going through the waterblock, which is good, because it would be more useful there anyways. So to directly answer your question, your setup will be fine like that.

Before i got it all put together though i would leave the northbridge cooling (or lack there of, whatever the case may be) the way it is temporarily and get the system running just with the cpu wb, heatercore, and pump and run toast (or your preferred program) to see what your load temps are. You can then go ahead and install the nb wb and include that in the loop to see what the load temps are then with it included. you may find that there is no noticeable difference in cpu temp, or you may be surprised. If there is a noticeable difference, it would likely be of more benefit to you to go with air cooling on the nb. But I would feel better knowing one way or the other if it were me. :) This kind of test would be far more useful than the calculations you were first considering. good luck. :)