Question on leak testing.

Five.Stars · 21 Feb 2013 at 12:31

Hi, going to be re-doing my loop tomorrow if my tubing arrives. Obviously I plan on running the loop for 12-24 hours to see if any leaks - however, my question relates to once system is fully up and running - is it a good idea to stress the system to get fluid warmed up as a further method of leak testing give that the o-rings might expand or contract with changing temperatures?

Or os ot safe to assume that if no leaks are found on initial testing that it is unlikely to start leaking with the system up and running? I will run it woth case side off anyway for a wee while.

Also, I am not using clamps on my 1/2" barbs with 7/16" ID. It takes quite a bit of force to get tubing alllied after heating up and removal seems near impossible without slicing the tube. But wondering if there are any horror stories?

Deceptor · 21 Feb 2013 at 13:01

If you go a day without issues you'll be fine, as for temps impacting anything in the loop it won't be an issue 99% of the time. I have temp probe and really the water ranges from ambient to maybe 5-10 degrees higher (10 when my room is very toasty).

The only time something did impact my loop which was a test is when I turned off all fans (including rad) and let the system idle - heat kept gathering and the pipe went a little floppy

Although I use a smaller tubing currently.

As for barbs, not used them myself but I believe if it's trouble to get them on then it's not going to come off at all unless you try pick the tower up with the tube or something

stuie · 21 Feb 2013 at 13:02

O rings dont tend to expand and contract at the temperatures your going to experience in a cpu/gpu loop, however if you are worried about it then running just your pump for 12-24 hours will get things toasty, assuming your running a d5 they are rated to operate up to 60c. Run your leak test for half an hour with no fans and see how hot the rads get with just the pump running

Five.Stars · 21 Feb 2013 at 13:04

Thanks. And apologies for the spelling and typos in op. posting from phone while gf hogs the laptop watchig soaps! Grrrrr!

OLDPHART · 21 Feb 2013 at 13:05

Hi, if you have any leaks most of them show up as some as you boot up, l just had a slight one, but what l did for a leak test was Prime for half an hour. Then over a few days just checked before booting up after that l just check for leaks + coolent level when cleaning my dust filters once a month.

stuie · 21 Feb 2013 at 13:26

erm normal leak testing is done without the pc running, limiting damage if there is a leak

Five.Stars · 21 Feb 2013 at 14:34

stuie said:
erm normal leak testing is done without the pc running, limiting damage if there is a leak

Aye, i'm well aware of that. But what i was asking, if you had read the question, was leak testing with cold fluid is all well and good, but what the chances of the heated fluid and changes of temperature causing the o-rings to expand or contract. Thus meaning further testing with the ststem fully powered.

Pabl0 · 21 Feb 2013 at 14:46

No there's isn't enough of a water temperature difference to affect anything like rubber washers, like someone above has said we're talking 10c hotter under load compared to idle, if it was 40c difference then yes.

I test my loop all connected with the jumper socket and never had a problem, flip waiting 12 hours if it makes it to 5 minutes it's not going to leak.

OLDPHART · 21 Feb 2013 at 14:56

stuie said:
erm normal leak testing is done without the pc running, limiting damage if there is a leak

The only reason for a bad leak is that you have loose fitting, badly clamped or tightened the compression fitting enough, l'v built 5 custom loops and only had a slight weep if anything at all just put some tissue paper around the joint so if its weeping you can see it.

Just double check everything before you boot up for the first time.