3/8" tubing streached on 1/2" id fittings

Let me start by saying that this thread isn't to start a debate about big tubing vs. small tubing. There's enough of that over at proforums right now. I don't want an argument but rather would like some insight by people who have tried this before and have definate results. Please, if you don't have physical experience with this then please don't input opinion.

After examining the different fittings involved with watercooling, I noticed that the internal diameter of most of them was 3/8". I'm sure almost everyone has noticed this. This size means that only so much liquid can travel through the fitting at a given time - less than a fitting with 1/2" id. Also, I think it can be assumed that the amount of flow in a system is limited to the flow through the point of most resistance, which could be the small id fittings.

The obvious choice would be to use bigger fittings, right? Wrong. Some fittings can't be replaced as the tap'd holes can't be enlargened at all. This is evident when you look at cpu, northbridge, and gpu block's fittings. Other fittings can be made bigger: tee if used, pump inlet and output.

Now, if I were to streach 3/8" id tubing over the 3/8" id fitting, wouldn't I in theory get the same amount of flow? I am aware of a few people on this forum that have actually done this and my question is this: Is there any performance hit when using 3/8"id tubing streached over 1/2"id (3/8" interlan fitting diameter) fittings?

I think the overall volume of water would be better, but the biggest issue is at the inlets on the cpu, gpu or northbridge, it would still be 1/2, and 3/4 of water flow might cause a bottle neck at the inlets, but I know I have 1/2 ID stretched over 3/4 on my pump and HC.

well, dont forget that 3/8 ID tubing will also somewhat limit the total amount of flow you may have.
AFAIK a 3/8 ID tubing over 1/2 OD barbed system will not have the same flow as a 1/2 ID tubing over 1/2 OD barbed system. Just as a 1/2 ID tubing over 3/4 OD barbed system will have different flow characteristics as an all 1/2 system.
also...

The obvious choice would be to use bigger fittings, right? Wrong. Some fittings can't be replaced as the tap'd holes can't be enlargened at all

Click to expand...

using bigger fittings where possible would yield more flow.

SenC.

Well with 3/8" ID tubing over 1/2" OD fittings, you would have a true 3/8" system. 1/2" ID tubing over 1/2" OD fittings does not yeild a true 1/2" system (in my book) as you have found already that 1/2" OD fittings are roughly 3/8" ID. With that said, you will lose a bit, flow wise with the 3/8" ID tubing over the 1/2" tubing. As to what that breaks down to in terms of temp. gain, well that is dependant on many variables but, I would imagine it would be on the order of less than 1C or so.

Bongo has 5/8ths fittings on everything in his setup, and uses 1/2id tygon so hes got a true 1/2 setup

My "mad scientist" experiments have led me to firmly believe that increasing diameter of tubing or fittings will help performance; how much depends on far more variables than I really comprehend or can test.

But from the silly things I have tried, which included running a block with 1/8" ID barbs with 1/8", 1/4", 3/8" and 1/2" ID tubing...

each increase in tubing size gave me a slight increase in performance even though the block remained 1/8" and pump and rad were not changed (1/4" pump, 3/8" rad).

The difference between 1/8" and 1/2" was clearly noticeable and a pretty decent improvement.

My barely educated theory for this is that the reduction of resistance in the larger tube is responsible.

Note that I mention this because my experimentation with fitting sizes pretty much follows the same pattern, with one major exception:
water blocks with a central inlet may perform BETTER with a smaller barb size- this creates a "jet" similar to the nozzle used on many of the newer blocks.

Other than that, bigger barbs generally provide better flow and improved performance, although, as stated by others, just how much difference it makes depends on many things, and may not even be noticeable in a system.

(Sorry I have no facts and figures to quote but those experiments were done quite a while back and I have LONG since lost the data.)