Flow rate questions for watercooling.

[Tim]

I'm wondering what flow rates are considered "ideal" if there is such a thing, and how what I have now compares to this number I should be looking for. Using my logic (which doesn't always serve me right), If the flow rate is too slow, it isn't as effecient in whisking the heat away from the copper blocks... then again if it's too high, it's not staying in the radiator long enough for the heat to be whisked away from the water. Correct me please if this logic isn't correct.

Now, moving along, I've input the following into the excel worksheet that's up @ martinsliquidlab.com.
The sheet assumes pure water (which is what I'm using), and 1/2" barbs for all curves (which I'm also using)

EK supreme waterblock
EK NB-S Max waterblock
Swiftech MCR-320 Radiator
Laing DDC-3.2 w/ XSPC Reservoir top
(Guesstimate here, I didn't measure!) 5' of 7/16ID tubing.
1 DD Derlin 90 elbow
1 DD Derlin TEE.

It comes up with an estimate based on my parts of 1.33GPM
If I take out the DD Derlin 90 elbow, it estimates 1.34GPM
If I take out the DD TEE on top of that it estimates 1.35GPM
If I replace the Swiftech MCR-320 rad with a thermochill pa120.3, it esimates 1.36GPM, and without the DD fittings, 1.38GPM

Lets just assume here for arguments sake that these estimates were the real deal as far as these flow rate numbers go. How close am I too a supposed 'ideal' flow rate (like I said in the beginning, if there is such a thing). Also, of the changes I listed, are their real benefits in making any of those changes seeing as the estimated flow rate only slightly increases, or is that slight increase more important that I'm realizing.

Lastly, on an off topic aside... where in the hell is derwinartos at? I haven't seen him in a bit.

 

[Billb2]

Originally posted by: theplaidfad
Using my logic (which doesn't always serve me right), If the flow rate is too slow, it isn't as effecient in whisking the heat away from the copper blocks... then again if it's too high, it's not staying in the radiator long enough for the heat to be whisked away from the water. Correct me please if this logic isn't correct.

Your logic is wrong........

No matter what the flow rate is the water spends the same amount of time in the blocks and rad. Twice the flow rate and it just comes back twice as often. Half the flow rate and it takes twice as long to get back.

The importance of flow rate is to cause turbulence at the boundary layer between the water and the copper (in the block and the rad tubes).

Hold your hand up about 4 inches from your face and blow on it....cooler, huh? And that's blowing 90* air!

That's why blocks have evolved into pin configurations and impingement configurations...to increase the flow right at the boundary layer.

So, yes, increased flow rate will cause additional cooling since it just makes the blocks pins or impingement work even better. But there are diminishing returns. You really have to look at the CW data for your block(s) and rad to get the numbers to do the calculations on just how much additional cooling you can get from how much additional flow. The data is all available on Swiftech's site.

Martin has done a fantastic job that eliminates all the work we used to have to do to calculate flow rates from the data. He's put it all together for us.

 

[aigomorla]

lol...

Your getting 2 concepts confused.

in a closed loop system, the time spent in the radiator is important, however the number of passes is also important.

On a CPU block, heat can only be released into the water at a set amount. Meaning water can only pick up X amount of heat at the source each pass.

Now flow rate determines how many times the water will circulate in your system. This means how many passes per min on each block.

If you hold X constant, then you can see having more passes yields greater efficiency. However, b4 you can get full efficiency, you'll lose it on the block end.

The block can only dish out so much heat that having too much flow wont help. However if you have a very efficient block, you'll see that having more flow helps a lot.

If your playing with flow, you'll be short handed really quickly. I ran into roadblocks extremely fast. This is why im shifting to accelerators and turbulance. You guys know my favorate variable in h2o cooling.

HEAD PRESSURE HEAD PRESSURE HEAD PRESSURE!!!


A simple anology. The water loop is a school bus. Kids are heat, and school is the radiator. Having more kids at school will not help anything. How fast you can move the kids from home to school is what you need to look at, and vice versa.

 

[Tim]

Thank you, that's what I needed to hear.

 

[aigomorla]

http://www.xtremesystems.org/f...29&highlight=innovotek

This is more of a flame thread :\

However linus and others tell it very nicely. Click on the sli forum link also, Grim went down there and did a nice confusing mathmatical explaination of it.

And yeah you'll notice me at the end of that post on sli forums as well.