Tubing sizes in relation to temps.

[Eddie3dfx]

One of the first little tidbits I read when I first started watercooling was Cathars testing on tubing and it's impact on temperature.

Let me copy it as the first part of the analysis.
What I don't understand is when another testing company did a real world test, the differences were greater than what Cathar had reported. (Read Below)
---------------------------------------------------------
6.35 (1/4") ID tubing with quick-fit fittings
8mm (5/16") ID tubing over 6mmID|8mmOD barbs
8mm (5/16") ID tubing with quick-fit fittings
9.6mm (3/8") ID tubing over 7.5mmID|3/8"OD barbs
9.6mm (3/8") ID tubing with quick-fit fittings
11.1mm (7/16") ID tubing stretched over 10.5mmID|1/2"OD barbs
12.7mm (1/2") ID tubing over 10.5mmID|1/2"OD barbs

http://img248.imageshack.us/img248/8522/catharbb6.png (http://img248.imageshack.us/my.php?image=catharbb6.png)
http://img248.imageshack.us/img248/catharbb6.png/1/w569.png (http://g.imageshack.us/img248/catharbb6.png/1/)



The intersections all are:

6.35mm quick fit = 4.45LPM flow, 0.0795 block c/w, 0.0374 rad c/w
8mm barbed = 4.75LPM, 0.0783 block c/w, 0.0373 rad c/w
8mm quick fit = 5.6LPM, 0.0770 block c/w, 0.0369 rad c/w
9.6mm barbed = 5.7LPM, 0.0768 block c/w, 0.0369 rad c/w
9.6mm quick fit = 6.2LPM, 0.0762 block c/w, 0.0367 rad c/w
11.1mm barbed = 6.3LPM, 0.0761 block c/w, 0.0367 rad c/w
12.7mm barbed = 6.35LPM, 0.0760 block c/w, 0.0366 rad c/w

Final CPU temperature is ambient (22C) + system load (114W) * radiator C/W + CPU Load (100W) * block C/W

The final CPU temperatures work out to be:

6.35mm quick fit = 34.21C
8mm barbed = 34.08C
8mm quick fit = 33.91C
9.6mm barbed = 33.89C
9.6mm quick fit = 33.80C
11.1mm barbed = 33.79C
12.7mm barbed = 33.77C

So there we have it. The differences between varying tubing sizes.

Okay, the more astute of you will point out that the block C/W is really the case-to-block C/W, and that the actual CPU-die-to-block C/W is a lot higher. Even if we triple block the C/W (which would be an absolute upper limit based upon older research), we get:

6.35mm quick fit = 50.11C
8mm barbed = 49.74C
8mm quick fit = 49.31
9.6mm barbed = 49.25C
9.6mm quick fit = 49.04C
11.1mm barbed = 49.01C
12.7mm barbed = 49.00C
-----------------------------------------------------------------------------------

Not exactly the huge difference that many people believe it to be. Though, this was based on a 100w heat load.

--------------------------------------------------------------------------------------


Here is what I don't understand. When reading an alternative review on tubing to temps (I am looking at going to 8/11 tubing) I see a very different story/analysis.

http://www.hardocp.com/images/articles/1038785961lCEaChcWhL_3_2.gif

The difference between 5/16 and 3/8 was 3 degrees Fahrenheit and at a higher clock speed, 4 degrees Fahrenheit.

That is essentially a 1.6mm difference resulting in a 3-4f difference. The numbers were also quite consistent, because 3f difference was the same even when adding additional components.

So what can I take from all of this?
Could tubing have more impact on certain pumps than others, or is the test really innacurate?
Just to be clear, Cathars tests were based on math resulting in a hypothesis, not real world data.

http://www.hardocp.com/article.html?art=MzkzLDMsLGhlbnRodXNpYXN0

[Kayin]

Interesting. I've got some 5/16 in my build using drilled compressions, I wanna see how it all works out. However, that's chipsets, and they were designed to work with less. I'm gonna pump enough through them to wash the copper out the blocks, we'll see what happens...

[Eddie3dfx]

Interesting. I've got some 5/16 in my build using drilled compressions, I wanna see how it all works out. However, that's chipsets, and they were designed to work with less. I'm gonna pump enough through them to wash the copper out the blocks, we'll see what happens...

Yeah, see what the results will be.
I think after reading those results, I'm going to skip 5/16/8mm and just stay with 7/16 tubing.
Now I'm curious what the real world difference is between 3/8 and 1/2 outside of hypothesis.

[alejo]

I'm running 11/8 MM tubing right now, on just the CPU and I get loads of about 45C during priming, but lower when I game. Idles at 30 or so. Ambient is 25.

I'm using

MCR220 (with low speed Yates)
EK Supreme
MCP350 (XSPC res top)

This is on an E8400 @ 3600 with 1.28V. It's open air at the moment. I plan on adding my 8800gtx along with an MCR120. I'll post what my temps are, but it will be a week or two.

[Eddie3dfx]

I'm running 11/8 MM tubing right now, on just the CPU and I get loads of about 45C during priming, but lower when I game. Idles at 30 or so. Ambient is 25.

I'm using

MCR220 (with low speed Yates)
EK Supreme
MCP350 (XSPC res top)

This is on an E8400 @ 3600 with 1.28V. It's open air at the moment. I plan on adding my 8800gtx along with an MCR120. I'll post what my temps are, but it will be a week or two.


Do you have 1/2" tubing and barbs to compare it to?

Any pics of your setup as well.. I am curious to how the tubing looks in a rig.

[alejo]

Yeah, I'll post a pic or two later tonight. It' just a simple loop I set up a few nights ago. I was in a hurry, so it's kinda messy. :)

I used tygon 3603 and honestly, I'm not liking it that much. Since it's only 1/16" wall, it tends to twist and fall on itself unless you keep your loop short and route the tubing carefully. I have some 5/16 id 7/16 Masterkleer I bought from McMaster that is a little stiffer than the tygon. I think it will work better.

I've never used anything bigger than 3/8 and temps were basically the same.

These are the fittings I'm using.

http://www.sharkacomputers.com/5hfcofig1.html

[Eddie3dfx]

Yeah, those were the barbs I was looking at.
Can't wait to see the pics!

[Snyxxx]

Remember that flow and pressure drop is a square function with diameter as seen by Cathar's graph. Pressure drop reduces tremendously with an increase in diameter. 5/16" ID hose has a lot more pressure drop. If you square the decimal size and compare, it will become clear.

I use 3/8" ID hose with 1/2" barbs since the ID of these barbs is about the same as the 3/8" ID hose for matched flow continuity.

Increasing hose size had diminishing returns. I think the overall conclusion was 3/8" ID hose was the optimal size when considering bend radius and such for tight turns in a case.

[RedRaider]

Nice post Eddie.:2up:

Regardless of what anyone says, 1/2" x 3/4" for me please...:D

[Chilly]

7/16 here FOR LIFE.

[Ricey]

Its necessarily subjective. You can manipulate math to suit whatever conclusions you desire.

[Xilikon]

Its necessarily subjective. You can manipulate math to suit whatever conclusions you desire.

I agree... However, the actual point of Cathar's numbers is to show that there is not much difference by picking 3/8" vs 1/2" ID and it's a matter of preferences. This is true for me because I'm a fan of 3/8" ID 5/8" OD which serve me very well in terms of routing and bending radius.

It's logical that the bigger ID, the better the temps but it's not a big difference. What should worry us the most is the pressure drop due to a smaller ID (3/8" tubing and barbs is inherently more restrictive than 1/2" as measured by Martin).

[alejo]

Here are a couple of pics of the simple cpu only 11/8 mm loop. Sorry for th e cruddy pics...iphone and bad lighting. One additional thing, which program has more reliable temps? Realtemp is reporting 5 degrees less than Everest on both cores. So at load, Everest was reporting 44*C while Realtemp was reporting 39. Once again, this is an E8400 @3600 and 1.28V.

I think eventually I'm going full Aquacomputer 8MM OD, 6MM ID to see how that works out. I asked Gary to order me some tubing and Plug N Cool fittings. I will also try 7/16 ID, so I can try stuff for MYSELF and not take anyone else's word for it.

http://www-personal.umich.edu/%7Ealexo/images/photo1.jpg

http://www-personal.umich.edu/%7Ealexo/images/photo2.jpg

[Eddie3dfx]

Man that looks really nice.
Looks like it would super easy to route.

download speedfan.

[Eddie3dfx]

Remember that flow and pressure drop is a square function with diameter as seen by Cathar's graph. Pressure drop reduces tremendously with an increase in diameter. 5/16" ID hose has a lot more pressure drop. If you square the decimal size and compare, it will become clear.

I use 3/8" ID hose with 1/2" barbs since the ID of these barbs is about the same as the 3/8" ID hose for matched flow continuity.

Increasing hose size had diminishing returns. I think the overall conclusion was 3/8" ID hose was the optimal size when considering bend radius and such for tight turns in a case.

Cathars shows that the difference between 5/16 and 3/8 is .19c.
Very minute..

While the real world data tests show it's a difference of 2.22c

Thats a pretty big difference.
I'm searching the web for more tests, but can't find any.

[Kayin]

Could the difference come down to greater wattage of heat source?

[Snyxxx]

Cathars shows that the difference between 5/16 and 3/8 is .19c.
Very minute..



Yes, but look at Cathar's flow vs. pressure drop for the 5/16 and 3/8". Big difference. Though the temp difference may be nil, the squared relationship with diameter is still there.


RRR,

I use to be a 1/2" ID 3/4" OD fan, but saw no real difference in temps and the routing looked uglier in my loop with those "fat" pipes. I only use 3/8" ID 5/8" OD now and would discourage 1/16" wall thickness tubing for everyone.

[RedRaider]

1/2" x 3/4" is MAN size tubing...

[Xilikon]

1/2" x 3/4" is MAN size tubing...

Only if you compensate for a lack of manhood :2pts:

[Septim]

lolz, don't worry about RRR, he has his big Case to fill up...

[HESmelaugh]

You have to keep in mind that the length of tubing in your loop also makes a big difference for tubing sizes. In a loop with lots of tubing, the difference in flow between smaller and larger diameters can be huge but if the loop is tight and uses only little tubing, the differences can become minimal and have almost no impact on temperatures.

I've tested both and saw an increase in flow of 50% in the large loop but an increase of only 16% in the smaller loop. The step up in tubing sizes was the same in both cases.

[Septim]

Shane maybe you'd care to join any of our distributed computing teams?

[Xilikon]

You have to keep in mind that the length of tubing in your loop also makes a big difference for tubing sizes. In a loop with lots of tubing, the difference in flow between smaller and larger diameters can be huge but if the loop is tight and uses only little tubing, the differences can become minimal and have almost no impact on temperatures.

I've tested both and saw an increase in flow of 50% in the large loop but an increase of only 16% in the smaller loop. The step up in tubing sizes was the same in both cases.

That's correct, tubing length has a impact as well. That's why plumbers try to have a bigger diameter for very long runs so it doesn't kill pressure. My HTPC case only have about 3 feets of tubing for the whole loop so it's very efficient.

Another factor to consider as well is pumping force : If you own a Iwaki, it's a better idea to use 1/2" ID tubing due to sheer pressure/flow of those Iwakis. A smaller ID could cause cavitation and pump damage and that's why people try to have the EK reservoir modded so the intake is 5/8" ID direct to the pump to give as much water as possible.

[zlojack]

I'm using 3/8"-1/2" in my compact little build.

We'll see how it goes.