Air makes its way from the reservoir into the radiator.

[Noci]

When a part creates flow restriction, there is higher pressure of the input water and lower pressure of the output water. The pressure drops.

This is right but the way you phrased it in your previous post I got the idea you misinterpeted.

You can see in the reviews on this site, that radiators have up to 1 PSI and CPU water blocks up to 1,5 PSI pressure drop. If you have 4 rads and two blocks in serie, you have total of 7 PSI pressure drop along the loop, if you ignore all tubing and other restrictions. That is 50 kPa difference.

I never measured those, but for the rads & blocks but these values look plausible. Only there is a wide variety of rads and blocks with different restriction levels as well as other phericals so each system has different variables, so no way of telling it applies to these numbers in this case.
Then, adding all the values as you did is not right, it doesn't work that way. Just imagine, if the pressure after leaving a rad is a little lower but them it reaches a restriction like for example the CPU waterblock it increases again.
Also there are some nuances to take in consideration, the air bubbles in the system. Liquid is not compressible, but gasses are, so the air bubbles throughout the system form (up to a certain level due to the low pressure differential over atmospheric pressure) an accumulator effect, leveling the delta-P spikes.
In my system this accumulator effect is visual on my pressure indicator as it is continuously fluctuating between 340 hPa and 361 hPa with the pump at full speed, while the flow is stable.

Overall it is way more complex than I can and want to explain, but pressure is actually not even a very important factor (just a parameter in the process) if the average flow is maintained around 100 l/h. Flow is a waaay more important factor to transport the heat away from the source (blocks) towards the heat exchanger (rads) for the efficiency.

The regular D5 or DDC pump @12V have an output pressure around the range of 300 to 500 hPa (1 hPa = 1 mBar = 0.014 psi), as we know sufficient for the regular PC water cooling system. For extreme systems with high restrictions, I asume the owner knows what he/she is doing and acts likewise.

BTW in a soda bottle you have pressure cca 250 kPa. Note the amount of gas coming out of the beverage when you pour it in a glass.

This is a completely different scenario and has not even closely an analogy with a PC closed water cooling loop or disolved gasses under atmospheric pressue.

Ok...

Maybe I was a little harsh, but considering your previous threads, posts and statements I know you're hard headed and keep on pusing misinterpretations as they seem right in your logic.

@Steevo
Regarding the evaporation through the soft tubing it is a fact, but it is really marginal even neglectible.
Especially if youre talking about molecules, just for the record one drop of water (0.05 ml) has over 1.5 sextillion (something with a shitload of zero's ) and there go 20 drops in a milliliter.

Oxygen dissolved in water, yes this is true but again not enough to create the air bubble effect in this scenario. In open water under atmospheric conditions there is @30 °C, a level of saturated DO of 9.0-7.0 mg/L.
However a closed loop is...uh closed and eventhough there is plenty of liquid movement to trigger the Oxygen absorbtion proces, no fresh air (approximately 20.9 % Oxygen in the air surrounding us) is introduced to the system.

Fun fact: in PC water cooling systems just like in central heating systems you actually want as little as possible oxygen disolved in the water as it stimulates the growth of algea, fungus and bacteria. Thats why biocides are in cooling liquids and most of the times they have Oxygen binding properties.

@ Vya Domus
Bleed the air, top-up the liquid and after a while it will slowly stabilize. If the bubble in the reservoir irritates you, the temporary second pump as suggested will speed up the proces with a flow increase. But be aware of the fact that it never will without any air bubble at all.

Now it's time for a well deserved nice cold beer, Cheers

 

[Vya Domus]

Bleed the air, top-up the liquid and after a while it will slowly stabilize. If the bubble in the reservoir irritates you, the temporary second pump as suggested will speed up the proces with a flow increase. But be aware of the fact that it never will without any air bubble at all.

It's the not air pocket in the reservoir that's the issue, it's the one in that rad that keeps coming back.

 

[thesmokingman]

This loop is an exercise in how not to build a loop.

 

[Noci]

It's the not air pocket in the reservoir that's the issue, it's the one in that rad that keeps coming back.

Believe me, it's getting smaller and smaller it just takes time.

Like mentioned before, I still notice bubbles coming into the reservoir after more than a year but it's getting less and less.

Patience is a cool virtue here

 

[Vya Domus]

This loop is an exercise in how not to build a loop.

I mean how so ?

 

[BoggledBeagle]

I tried to explain that you have a very long loop with 4 rads with all parts in serie that has a notable pressure drop along the way and you in the place of lower pressure with higher probabillity of bubbles coming out of the water placed a bubble catcher.

I just realised that a major factor in this "mysterious bubble collecting problem" may be fluctuations of atmospheric pressure. When the pressure drops, you may get a few bubbles come out of the water and when you collect them in one place, they may not be able to dissolve again and the air pocket slowly grows.

 

[thesmokingman]

I mean how so ?

Almost everything? You're using what a D5, with 4 rads and two blocks. Tiny ass reservoir only good for masochists. General rule is never stuff a case full of rads, voila 4 rads. Then pump is too weak to push water thru rads. Rad orientation creates air pockets cuz vertical rads have ports on bottom, making pump power absolutely crucial, which you don't have enough pump. By stuffing it full of rads it creates issues with port locations, etc etc. I could go on and on...

 

[Noci]

This loop is an exercise in how not to build a loop.

Loop order is not important to the capacity of heat dissipiation, but with what he has on cooling hardware I would have done the orientation different also.

Biggest 'no go' for me is the small reservoir as you mentioned but when it works it works. Everybody is in title to have his/her preference.

As long as the air bubbles in the rads have no negative impact on the cooling capacity I don't see any problems.

 

[Vya Domus]

I tried to explain that you have a very long loop with 4 rads with all parts in serie that has a notable pressure drop along the way and you in the place of lower pressure with higher probabillity of bubbles coming out of the water placed a bubble catcher.

The flow rate is the same at any point in the loop, the pressure will also be the same everywhere in the loop, I don't see how an air pocket would form in one rad but not the other if it was somehow related to this.

You're using what a D5, with 4 rads and two blocks.

It's actually a VPP, everything seems fine, I don't have a flow rate sensor but when I fill the loop it has absolutely no problem whatsoever pushing water through.

Tiny ass reservoir only good for masochists.

You seem oddly fixated on this. I bought his when I had a different case that couldn't fit a larger distro plate not that I would have went with a regular reservoir because they look old fashioned and ugly as hell to me. But it doesn't matter because this isn't related to my issue.

 

[BoggledBeagle]

The flow rate is the same at any point in the loop, the pressure will also be the same everywhere in the loop, I don't see how an air pocket would form in one rad but not the other if it was somehow related to this.

Water flows because of the pressure difference, maximal pressure is right at the pump outlet a drops along the way to zero in the reservoir. Water would not flow without a pressure difference.

Your loop is set up with a lot of components in one chain (in serie) - that means maximal pressure drop possible.

The propensity of degassing and bubble creation will increase towards the end of the loop and you placed one device there ideal to collect bubbles. I mentioned above that the major cause may be fluctuations of atmospheric pressure.

I do not think I will be able to think about anything else on this topic, I would personaly just get rid of the 240 rad,

 

[Vya Domus]

Water would not flow without a pressure difference.

The most pressure differential will be between the inlet and outlet but throughout the loop I don't see how any pressure value involved could cause problems on a system of this scale. What kind of pressure we'd even be talking about 2-3 PSI at most ? If that.

The propensity of degassing and bubble creation will increase towards the end of the loop

I just don't know how anyone could even test this, honestly I don't think this effect is even measurable on a tiny closed system like this. And if I removed this rad maybe the air pocket will just move to the other rad lol.

 

[Noci]

It's actually a VPP, everything seems fine, I don't have a flow rate sensor but when I fill the loop it has absolutely no problem whatsoever pushing water through.

The Alphacool APEX VPP (or VPP 755 V3) does about 340 l/h and max 425 hPa (with no restrictions), which is enough for your system, but the two vertical mounted rads will probably have the biggest air bubbles on top.
if you are able (don't know if your rads have connection points there) to connect an bleed valve on top of these two rads the problem is quickly solved. Otherwise tilting the system will help also to get rid of the trapped air.

Forget about pressure it is not important here. Degassing and bubble creation is just happening in the mind of some peeps not in your rads ,

 

[Vya Domus]

if you are able (don't know if your rads have connection points there) to connect an bleed valve on top of these two rads the problem is quickly solved

But I am 100% positive all the air is gone from all rads when I flip the case around because there is no more slushing/gurgling sound anymore. Then after a while I notice the water level goes up and sound is back, which clearly comes from that 240 rad.

 

[damric]

From looking at the loop, yes the small reservoir is the problem.

Pumps need net positive suction head to operate properly. For a D5 that means a pretty large reservoir. The lack of adequate NPSH is causing the pump to cavitate when speed ramps up, stirring up whatever air is in the reservoir and drawing it into the loop to be trapped in your radiator.

So yes the loop could have been engineered a bit better. In the meantime you can probably just run the pump at a slower steady rate which doesn't cause cavitation. The solution to run the pump at full speed is a large reservoir.

I can make a video with my D5 at full speed and various levels of coolant in the reservoir to show you exactly what cavitation looks like and at which level the cavitation stops. Please do eventually fix the issue as cavitation will overheat your pump and thus shorten its life.

 

[Noci]

But I am 100% positive all the air is gone from all rads when I flip the case around because there is no more slushing/gurgling sound anymore. Then after a while I notice the water level goes up and sound is back, which clearly comes from that 240 rad.

As I mentioned before, I suspect it's micro bubbles recirculating. Just keep on bleeding air and topping up the liquid in time it will stabilize. To actually hear slushing/gurgling in a rad there must be quite a air bubble inside, when it gets less audible doesn't mean it's free of air.

Patience my friend .

 

[Outback Bronze]

This air could be coming from your blocks.

Many moons ago I was getting air travelling all around the loop, didn't stop. Pulled everything apart, put it back together, same thing. Even pulled the GPU blocks apart (HD6950's) and still nothing.

Did this many times and still couldn't fix the issue. Then said fuk it, im going to pull the CPU block apart too.

This was the problem.

Junk has started accumulating into the fins of the CPU block and must have been creating air bubbles.

Once I cleaned all the junk from the block fins (there was a fair bit) It solved all my problems. No more air bubbles going through the loop.

The bubbles could be so minutiae that you would not even see them but eventually will create a sizable air bubble within your rads.

Just my experience from water cooling and getting unknow bubbles when it's impossible coming from the res.

Good luck

 

[Vya Domus]

Junk has started accumulating into the fins of the CPU block and must have been creating air bubbles.

Once I cleaned all the junk from the block fins (there was a fair bit) It solved all my problems. No more air bubbles going through the loop.

I actually have cleaned both CPU and GPU blocks when I built the loop in the configuration that it is now, so a couple of months ago.

Pumps need net positive suction head to operate properly. For a D5 that means a pretty large reservoir. The lack of adequate NPSH is causing the pump to cavitate when speed ramps up, stirring up whatever air is in the reservoir and drawing it into the loop to be trapped in your radiator.

I know what you are talking about, there is cavitation if I run the thing at full speed but I never do anyway, I run it 50% max and there is no visible cavitation.

 

[Noci]

From looking at the loop, yes the small reservoir is the problem.

Pumps need net positive suction head to operate properly. For a D5 that means a pretty large reservoir. The lack of adequate NPSH is causing the pump to cavitate when speed ramps up, stirring up whatever air is in the reservoir and drawing it into the loop to be trapped in your radiator.

So yes the loop could have been engineered a bit better. In the meantime you can probably just run the pump at a slower steady rate which doesn't cause cavitation. The solution to run the pump at full speed is a large reservoir.

I can make a video with my D5 at full speed and various levels of coolant in the reservoir to show you exactly what cavitation looks like and at which level the cavitation stops. Please do eventually fix the issue as cavitation will overheat your pump and thus shorten its life.

You're right about the fact that most centrifugal pumps need a positive suction head. In this case I believe it is on the edge due to the small reservoir.
But cavitation however is caused by vacuum bubbles that implode, I have the same with my pump if I run over 4200 rpm, despite a large reservoir. When the vacuum bubbles implode they do not create air bubbles as far as I know, but because of the split second vacuum the possibility exsists that some liquid evaporates due to the low vapour pressure. That would also create micro bubbles in the system.

This is something I have to look up in the technical literature as in industrial pump systems I never encountered this problem, just damaged impellors due to cavitation. I was also wondering about the orientation of the pump, like in this situation it is mounted vertically instead of the ideal position; horizontally. But then again, according to Alphacool these pumps are designed to also be used like this due to their ceramic bearing design. But I always say: "gravity don't discriminates"

As just mentioned by Outback Bronze algea, fungus and bacteria also excrete gasses as they digest/grow. Not sure in which quantities but could have some impact as it makes perfect sence. But since Vya domus said he cleaned everything a couple of months ago it seems less viable,

Man this thread is fun from a scientific/technological point of view, never too old to learn.

 

[Steevo]

My central heating system has a device has a float to leak off gas; is there anything simillar for these small systems?

Maybe a T with a transparent tube sticking up to catch gas might be a nice addition.



Heat some water and one can see the dissolved oxygen being released long before any boiling... which leads to the idea of boiling the coolant before use...

My closed loop had bubbles for years, I would add a few Oz of water a time, the fill port was the highest part, I tried looping through a reservoir for days to get the air out.

My geothermal ground loop was a closed loop, but heating and cooling and the plastic itself confirmed by multiple HVAC Geothermal techs confirmed that it absorbed oxygen and transpired water caused the pressure (a whopping 5-15PSI on over a mile of loop) to change and the eventuality was adding a rechargeable pressure vessel. First a hose fitting to use house pressure to charge, then an air reservoir that needs charged once every few months with a few pumps of air. And once every few years 64Oz of water to keep the fluid level up. It has NO leaks.

 

[BoggledBeagle]

This air could be coming from your blocks. Junk has started accumulating into the fins of the CPU block and must have been creating air bubbles. Once I cleaned all the junk from the block fins (there was a fair bit) It solved all my problems.

The junk created big water flow restriction and big pressure drop which knocked the air from the water.

 

[Noci]

The junk created big water flow restriction and big pressure drop which knocked the air from the water.

The light is on, but there is nobody home .

 

[BoggledBeagle]

The light is on, but there is nobody home .

You developed a habit of attacting my person...

 

[Noci]

You developed a habit of attacting my person...

Dear BoggledBeagle, I've abolutetly nothing against you as a person.
I actually admire your curiosity for the unknow/dillema's and the engergy that you put in to solving problems or even improving products.

But if someone is posting ideas as being the truth by bending the laws of physics or applying the wrong ones and thus posting misinformation, it triggers something in me to react. A lot of times I can resist that urge, but you seem to have the talent to hit my sensitive spot . Especially if things are blown out of proportions, pretending significant impact while in reallity it is just a fraction of big complexity of interacting mechanisms with low to neglectible impact on the whole. So yeah, if BS is posted I do tend to develop a minor form of OCD which urges me to counter react .

You actually have the abillity to more than once pin point the right physical mechanisms, but for whatever reason misinterpeting them totally wrong by assumptions instead understanding them. As far as I've red your posts, it's 99% PC watercooling related and that happens to be a hobby of me. This as a spin off from working for decades with small up to very large industrial pump systems in a wide variety of settings.
So not to brag, but I do know a little bit about transporting liquids by means of pumps and the use of heat exchangers. And yes, I do make mistakes too and don't have all the wisdom so don't take everything I say for granted, stay assertive.

But this, for God sake, is a tech site were peeps come for advise/understanding/solving the PC problem they are dealing with and misinformation will only increase confusion.
Imagine some young guy/girl with his/her first PC which he/she bought after struggling to get the funds together comes here and follows up some wacky advise resulting in hardware damage of which he/she is not able to afford a replacement...
Would you want to be the one that gave the wrong advise?

So I hope you see that it is not you as a person. Get your facts right and you've got my full support.

 

[Keullo-e]

Had to fight against bubbles as well as I did some rearrangements on my loop. Still not perfect but I'm pretty sure I got most of the air out at least.

Would be interesting though to know how much temps would improve if I'd get all the air out of the loop.

 

[BoggledBeagle]

Dear BoggledBeagle, I've abolutetly nothing against you as a person.
...
But if someone is posting ideas as being the truth by bending the laws of physics or applying the wrong ones and thus posting misinformation, it triggers something in me to react. ... So yeah, if BS is posted I do tend to develop a minor form of OCD which urges me to counter react.
...
So I hope you see that it is not you as a person. Get your facts right and you've got my full support.

YOU DO ATTACK MY PERSON and there is no point in explaining your motivation for doing so.

If you think that a clogged CPU block is not a major water flow restriction, that it does not cause pressure drop and that the air dissolved in water after a pressure drop does not come out of water in bubbles, you do not understand basic laws of physics.