Hydraulics Question

[Player0]

I have a slight problem with the Cube. I now have four BIX radiators cooling the hot water loop. My pump sits on the bottom of the case with the resevoir, which holds maybe 20 ounces of water or so. The hot water loop holds nearly two quarts of fluid, most of which is stored in the radiators. The radiators are mounted on the top rear of the case, and all 4 sit physically higher than the resevoir.

The problem I have, is ever since adding the third and especially now with the fourth radiators, is that when the pump is shut off, water actually drains out of the radiators, and I get air bubbles inside, which I have to manually flush out every day when I turn on the system.

Im not entirely sure where the air is coming from. I beleive that all of the weight of the water in the radiators presses back down on to the pump and resevoir after the system is turned off. I suppose this is sucking air in slowly from all the fittings or other spots on the system. Its water tight, but maybe not air tight.

I have been trying to figgure out how to correct this problem. I have been thinking that if I mount a second resevoir up on the top of the cube which is above the radiators, that this would help collect any air that enters the system when it is shut off. It would also allow me to store more water in the system which should help keep water temperatures more stable. Do you guys think this idea might work? Or would putting more water in the system potentially make things worse?

[Mr. Thompson]

A reservoir at the top of the case would allow the air to seek the highest point in the system but it won’t get to the root of the problem. I know ;) you know this, but it sounds like hose clamps to me. FWIW, I have always included a small reservoir at the highest point of every watercooled rig I have built, just for convenience.

[Player0]

All the clamps are as secured as they get unfortunately. I think another problem is with the expansion of the water itself. It gets quite warm, and expands inside an already full system, I think this expands the tubes slightly, and when it cools down, the tubes contract, and this could be sucking air in somewhere. I dont think theres really a solution to the main problem.

I guess ill just try the upper resevoir. As long as the air is stored up in there and not in the radiators, I shouldn't need to reprime them everytime.

Well hopefully.

[kms]

hi P0, any chance of a quikc sketch or some piccies of your system ? is your pump in the res at the bottom ?

[Player0]

Alright, this picture should work:

http://www.liquidninjas.com/albums/p0_cube2002/hotloop_rad_connects.jpg

You can see the pump and reses on the bottom left, and the radiators on the upper right. The only difference now is that im using silicone tubing, and there are now 4 radiators on the back (two radiators sandwhiched together to make one super radiator, a set of these).

Help?

[kms]

still a little confused, is that the ehim at the very bottom of the pic in the middle ? is that 2 pumps & 2 res's i'm looking at ? stuggling to work out the plumbing any chance of a quick sketch ? p.s. did Jai email you today ?

[Player0]

The black thing with the red mark on the bottom is the eheim. Yes, there are two pumps and two resevoirs on the bottom there. The one farthest from the camera is for the hot water loop. The water flows out of the pump and in to the TEC heat exchanger (which is in the other side of the cube and you cant see it). It then flows in to this compartment, and upwards to the top radiator. It flows back down to the bottom radiator. And then from that radiator back in to the resevoir (grey and blue thing)

[Player0]

Ps...no, havent gotten her email yet.

[kms]

whats your normal bleed point ? is there a bleed on the res's themselves ? i'm thinking along the lines of air build up in the res & cumatively in the rads, let me explain, if the system is water tight its not going to be far off air tight too & at pressures we operate at my money would be on water tight = air tight (along with your thorougness i doubt its air ingress?? - but still possible, check by fairy liquiding all joints), now as i read it your water flow is down into the resevoir (i.e. opposite flow to the way the bubbles want to go naturally - up) - the same is also true for your rads, i.e. water flow is down, air wants to go up. What this may cause to occur is air build up in the res or lower rads (you must share with me here that even after a week or so of 24/7 running and even when self convinced that all air is out the system a quick shake often proves your wrong!) What i'm getting at is as your eheim appears to pull from the centre of the res there is room for air to collect above here - unnoticed. Also rads in their very design always have spots that trap air. In your system by virtue of flow direction it is possible that these air trap zones are exagerated (air bubbles have flow to compete with to get to the top) so upon power off its possible theres a surge of air back to the highest point through the rads & also the possible pocket in your res. Then again it could be a load of ####'!/ but v.possible too!? :) p.s. missing email (women grrrrr) was to say new toy finally en-route :)

[kms]

ps.

i had major air probs a few months back with an old setup and checked joints & all were ok ??? couldnt work it out as machine was on 24/7 too, then after a couple of BSOD's a while later noticed a trickle of water running down my mobo, was a pin hole in my 5x7 cpufx rad :D

[Player0]

My normal bleed point is the resevoir down at the bottom of the case. The resevoirs are sealed, but do have a bleed line. The pumps are powerful enough and do end up moving the air easily down in to the resevoir, where its replaced by water. Once all the water is out of the system that I can get, I top the resevoir off and seal the bleed hole.

If there is an air ingress some of the fittings, im not sure i'd know how to find it. Since the air is being sucked in, it wouldnt be easy for me to see any evidence of the seapage.

Even though the resevoir is on the bottom, the pump is strong enough to provide enough pressure to push all the air down in to the resevoir. Air only is a problem after i turn off the pump, which is when it starts getting in there. While the pump is running, even if I wriggle the system around, I cant get any extra bubbles to show their face.

Its possible that there are tiny air pockets here and there inside the radiators, waterblocks etc, but no where near the amount of air that ends up filling my radiators after shut off. So i definately dont think its only residual air.

With the resevoir completely full of water and the air shaken out of the system as well as I easily can, there just isnt enough air in the system to account for the problem.

Unless its not actually air im seeing, and is really some sort of vacuum....mm...nah...cant be that.

[kms]

was told by a m8 whos a hvac engineer while back (when mine was ingressing) that fairy liquid (not sure if its the same name in the states - hand dish washing detergent) is good for eliminating small joint ingress leaks in low pressure systems as it as as high film strength and if smeared around a joint will be pulled in to a point and seal. PC w/c circuits very much fall under the low pressure catagory. Worth a try though may get messy with the amount of joints you have in your system. Although a slight vacuum will be pulled on the joints when your system stands un powered it will be small, what head is it ? Almost out of ideas m8 hope you sort it, are there any components / bleeds that can act as an NRV, i.e. one way valve that under system pressure seal but pass under vaccuum ??? /scratches head

[kms]

P0 is it easy enough to add a T & valve to your highest point then assuming there is no leak you can just keep filling her up as it settles to the top each time. If its not leaking water out anywhere then it must be residual air as for it to continually pull air in but never leak water out doesnt add up? :)

[Player0]

Testing all the fittings with soap is near impossible while its all mounted in the case, i just can't see reach them all.

Im trying to avoid filling it up every day ;) Sure, water can evaporate out of the system. In fact, im not sure where the water is going, I keep adding fluid. But I dont see it leaking anywhere.

Gah...this is annoying. I could try a one-way valve so water doesnt reverse when the pump turns off. I just wonder if they will work well in this sort of system.