Flow Rate and Temperature

[Carlof2XOD]

Does anyone know of any benchmarks or real numbers showing flow rate in a water cooling system compared with temperature?
I found a great thread (Flow Rate Impact in Water Cooling) but they didn't show any real numbers.

I recently put my first WC system together and I want to better understand how I can manipulate & control my temperatures.

I used the MCP655 w/ Speed Controller with the hope that I could control the noise and temperatures, just as I can with my speed controller on my fans.
pump specs:
Maximum head - 10 ft (3.1 m)
Maximum discharge - ~ 317 GPH (1200 LPH) == maximum setting

They have a nice chart on that page that shows flow rates at the different settings.
The flow rate range is from ~350 LPH to 1200 LPH for the lowest and highest setting.


I would guess that going from the lowest setting to the highest setting would have a fairly significant impact on cooling performance, but I do not see any change when I actually tested this theory.

Here is my testing method: (quick and dirty)
1) Pump is at lowest flow rate (~350 LPH)
2) Run prime95 to increase cpu usage to 100%. this should give me the highest possible temps in the system.
Wait 5 minutes for system to reach highest temp.
3) Wait another 5 minutes to see stable temps at lowest flow rate. Record avg temp (26.5 C)
4) Change pump to highest speed setting (1200 LPH) wait 5 minutes for temperatures to stabilize. Record avg temp (26.5 C)

Here you can see this happen on speedfan:


Am I not waiting long enough for the temperatures to change? I imagine the water cooling system may need a little longer to get full equilibrium with the heat exchange between the cpu and rad, but shouldn't I see at least some kind of change between the lowest and highest flow rate??

I will be repeating this test overnight for the next 2 nights. The first night I will log the temperatures all night at the lowest flow, and the second night with the highest flow. The other motherboard temperature sensors will give a good idea of the ambient temperature as a reference point between the nights. I hypothesize that the results will be the same, but how can it be explained??

[alacheesu]

Some people have tested temp vs flow rate for a few blocks. Both swiftech and thermochill have data for rad performance vs flow rate. In general, increasing flow rate much over 1-1.5gpm seems to have very limited effect for most components. If you're concerned, the best you can do is test a few different settings for yourself and see what you get. You shouldn't be surprised if you don't see much difference, or none at all.

[sirheck]

Have you looked around here?
http://www.skinneelabs.com/MartinsLiquidLab/

[Carlof2XOD]

I hadn't seen that site or known that flow rate had such a small effect on temps. Good to know!

I guess it really doesn't matter if I'm on the high or low setting because either way it's over 1.6 GPM. Now I can run totally silent with a piece of mind!

360 liters per hour = 1.58 gallons per minute
1200 liters per hour = 5.28 gallons per minute

Martins Liquid Lab had a great chart demonstrating what you said:
http://www.skinneelabs.com/MartinsLi...Estimator.html

Thanks!

[MpG]

I hadn't seen that site or known that flow rate had such a small effect on temps. Good to know!

I guess it really doesn't matter if I'm on the high or low setting because either way it's over 1.6 GPM. Now I can run totally silent with a piece of mind!

360 liters per hour = 1.58 gallons per minute
1200 liters per hour = 5.28 gallons per minute

Martins Liquid Lab had a great chart demonstrating what you said:
http://www.skinneelabs.com/MartinsLi...Estimator.html

Thanks!

Be aware that those flowrates you're quoted are maximum - you'll never actually see them in a watercooling system, because the blocks/components create restriction. Use that flowrate estimator to give you a rough idea of where your flow would be, and see where you stand. Ultimately, you should just try your system on different settings and see what you get. It's quite possible that the highest setting doesn't give any advantage over the lower settings, but it will vary from system to system. For testing, I'd give the system at least 15 minutes to warm up, just to make sure that the liquid reaches equilibrium.

[AndreaBZ]

http://www.xtremesystems.org/forums/...d.php?t=206542

[HESmelaugh]

And another one. (link, that is)
You'll find the flowrates to temperatures graph towards the bottom of the second post.