GPU sag can probably cause a leak over time, can't it?

[Ace McPlane]

I've been thinking..

GPU sag might be a point of weakness for springing leaks in the long term.

What are your thoughts on this?

 

I'm thinking I'm going to use a bracket in my next build in a year or so to prevent this in the long term.

 

What do you think?

Am I wrong about this?

If not, what are some ways other than brackets to prevent this in the long term?

If you've ever sprung a leak, was it around the GPU and was it possible that the droop effected it over time?

Just deep in thought on this and curious in a for the sake of science sort of way.

I'm sure if you plan out your loop right and do regular maintenance, this is not an issue.

[VVoltor]

Sure, it will put a little extra pressure on things, although the tubes will also help hold your card up. I don't think it's a massive concern.

 

If you don't want a bracket, just put your case on the side. Problem solved.

[00RaZoR11]

12 hours ago, stateofpsychosis said:

What do you think?

Aside from extreme cases, the PCB stays straight. Watch Jayztwocents' video on video card sagging. It will explain it all to you

 

[Mick Naughty]

Don’t see why this would be specific to the gpu. Why would it matter if it sagged? I assume you have to be referring to hardline as it would have zero effect on soft tube. Just build the loop better or support the card. All easily done. 

[Neokolzia]

I don't believe so, Hard tubes shouldn't just "slip" out from a graphics card sag if they were installed correctly,  And soft tubes wouldn't have any issues, specially if compression fittings are used, but anything shouldn't cause a leak like you described

 

Believe me that with the oring clamped down on a hard tube no amount of human force is removing it without mechanic advantage.

____________________________________

Infact I would say the opposite it true, to your question about sag, that Hard lines prevent sag as they give your system and components much needed support.

Not to mention weight is very re-distributed on a GPU block compared to a 2+ slot air cooler 

[TechyBen]

I would assume any pressure or force over time would cause damage. Rules of physix.

 

 

[edit] Though I hold the fact that "time" is relative, so it could last 100s of years before the pipe slips out.

[Neokolzia]

Just now, TechyBen said:

I would assume any pressure or force over time would cause damage. Rules of physix.

No.  I swear people on this forum.   Thats like saying a bolt on a bridge will fail.  Sure after 100-150 years.

 

Assuming the OP is talking about Hard lines there is no way it would cause a leak, there are usually double sealing Orings inside the fitting and 1 for clamping the tube down tightly inside after.  If correctly installed even at the maximum sag there is no way it would pull out far enough to break all 3 seals, or nearly 3/4 of a inch.  Not to mention you'd have to likely exert 500+ lb's of force to pull a fitting out when its clamped down.

[TechyBen]

So you are suggesting people allow structural load on hardline tubing?

[Neokolzia]

1 minute ago, TechyBen said:

So you are suggesting people allow structural load on hardline tubing?

What structural load, the 1/4 of a lb of sag force?

[TechyBen]

And we are certain there are no vibrations and the desk/PC is never knocked/wobbled?

[Neokolzia]

So your trying to tell me wobbles and vibrations will break loose hundreds upon hundreds of pounds of clamping force from the fittings.

 

My god the world we'd live in if that was the case ;p  Nuts and Bolts littering the road everywhere falling off from Cars/Planes and every which thing that vibrates and knocks and wobbles.

I mean "Rules of physix." Bro ;o

[TechyBen]

I've seen lots of cars with parts falling off, and airplanes too... They do rattle apart.

 

I don't think in any way the physical structure of the tubing, or the structural strength of the clamps will fail.

 

But I don't know who installed it. I don't know how they are using it. The installation can fail!

 

[edit] But as an example, what are the risks of cracking the threading? Or if using a gasket only, of that working it's way loose?

[Mick Naughty]

Doesn’t matter if there is force on it or not. Not gonna fail from such a stupid reason. What about the pressure from inside, guess we can just ignore that too. 

 

If something fails, more common than not, it’s human error. Same with bridges or airplanes seeing in how that’s being brought up for some reason. 

[TechyBen]

I totally agree. Most of my stuff lasts forever. Only breaks if I drop it 2 stories. But other people... I don't know how they manage for everything they touch to spontaneously explode! 

[Ace McPlane]

On 4/21/2018 at 9:10 AM, Mick Naughty said:

Doesn’t matter if there is force on it or not. Not gonna fail from such a stupid reason. What about the pressure from inside, guess we can just ignore that too. 

 

If something fails, more common than not, it’s human error. Same with bridges or airplanes seeing in how that’s being brought up for some reason. 

I am talking about human error.

I'm wondering if this could happen in a poorly planned out loop or just due to bad luck as well.

I'm not asking if it's common, likely, or whether or not it can't happen to someone who does a perfect job.

I'm just curious as to whether or not this is a possible point of failure at all

 

[Mick Naughty]

45 minutes ago, stateofpsychosis said:

I am talking about human error.

I'm wondering if this could happen in a poorly planned out loop or just due to bad luck as well.

I'm not asking if it's common, likely, or whether or not it can't happen to someone who does a perfect job.

I'm just curious as to whether or not this is a possible point of failure at all

 

Any thing is possible with human error. Can’t assume anything with an actual build with a picture. 

[Ace McPlane]

7 minutes ago, Mick Naughty said:

Any thing is possible with human error. Can’t assume anything with an actual build with a picture. 

Yea, just another thing to consider right?

When you like troubleshooting and helping people, it's good to ask questions like this so you have ideas for when you're helping people later on.

If this has caused a leak for anyone, I'm hoping they comment.

That's the data I really want.

[Mick Naughty]

Just now, stateofpsychosis said:

Yea, just another thing to consider right?

When you like troubleshooting and helping people, it's good to ask questions like this so you have ideas for when you're helping people later on.

If this has caused a leak for anyone, I'm hoping they comment.

That's the data I really want.

Again, so little detail. Gpu sag with soft tubing would have zero effect what so ever.  Hard tube would be a install issue, as the cards barely sag when a water block is installed. So if it has happened, I doubt it could be narrowed down to sag being the culprit. 

[Ace McPlane]

5 minutes ago, Mick Naughty said:

Again, so little detail. Gpu sag with soft tubing would have zero effect what so ever.  Hard tube would be a install issue, as the cards barely sag when a water block is installed. So if it has happened, I doubt it could be narrowed down to sag being the culprit. 

Of course I'm talking about rigid tubing not soft tubing.

It's probably unlikely either way and yes it's probably hard to determine if it was the cause, but it can't hurt to ask to see what other people's experiences were like with their leaks right?

[Mick Naughty]

Just seems specific for gpu sag, my card can’t sag as they use a bridge and have blocks on them. Doubt many have experiences from gpu sag leaks. 

[Ace McPlane]

19 minutes ago, Mick Naughty said:

Just seems specific for gpu sag, my card can’t sag as they use a bridge and have blocks on them. Doubt many have experiences from gpu sag leaks. 

Yea, I like to leave no stone unturned though.

I wouldn't have discovered that monitoring GPU power usage causes compositor frame drops in VR if I wasn't so obsessive with chasing after every possibility. 

It has it's benefits.

I'll wait and see if anyone else comments with their experiences  

[For Science!]

10 minutes ago, stateofpsychosis said:

 

I'll wait and see if anyone else comments with their experiences  

The thing is, GPU sag isn't something that worsens over time (since the card doesn't gain in weight over time...). And so a loop build with a saggy card should have the sag accounted for.

 

Also I actually haven't really seen any saggy waterblocked cards, perhaps since the weight is more equally distributed.

[derr12]

i had no issues with any leaks due to gpu sag over the year my loop has been in service. the hardtube actually buttresses the card a bit (from the bottom) and reduces sag a tad.