Can to "MUCH" flow hurt thermal performance?

[newls1]

Does anyone know first hand if having to much water flow will hurt performance? I guess thinking about it logically in my brain, having to fast of flow will not let water in rad long enough to dissipate heat effectively?. I was wanting my GPU loop to match my CPU loop, and that loop uses a dual D5 EK pump housing, and I have a spare dual D5 EK pump housing sitting here and was thinking about adding that to my GPU only loop so both loops will be identical. I have 3 D5's laying around along with that Dual D5 housing so I figured why not! But i certainly dont want to hurt performance. If anyone knows this answer please make me smarter and let me know!
Thanks

[Avacado]

5 minutes ago, newls1 said:

having to fast of flow will not let water in rad long enough to dissipate heat effectively?. 

That's not how it works. If you "Slow" your loop down to allow for "more time in the radiator" to cool it, then you are conversely allowing that same coolant in your CPU/GPU block to absorb heat as well. The answer is no. There comes a point where having higher flow may not help reduce temperatures anymore, but the notion that high flow inhibits cooling is ridiculous.  D5's do dump heat into a loop because they rely on the coolant to lubricate and cool the motor, but this is negligible with a high flow and rad fin density. Not having enough radiator surface area certainly hurts your cooling more than high flow ever would. 

[Sir Beregond]

Exactly - unless you are really lacking in radiators, the heat from a D5 shouldn't be anything to worry about really. I am not saying you should necessarily set to 100%, in fact I don't recommend that unless you need to for flow reasons. Basically there definitely is a point of diminished returns on where more flow will give you more cooling performance, but less flow will always hurt your cooling. In your thinking that less flow gives more time in the radiator for dissipation, you have to think the exact opposite is happening on any of the blocks on heat generating parts which will just more easily saturate the loop with heat. You have to remember that overall a loop equalizes and the mechanism by which heat is removed from the blocks and dissipated by the rads is via the constant flow through both continually facilitating heat transfer and dissipation respectively. Slow that down and now you are hurting that mechanism.

[newls1]

Thanks for the replies guys, much appreciated.  Not sure why you guess thought I was concerned about the heat dump of a D5, but i could care less about that and my post says nothing about that.... I have PLENTY of surface area to compensate the 20 or so watts of heat a D5 adds.  Just was wondering if high flow would ever hurt performance, but im guessing not.  Thanks y'all. Ill go ahead and add a Dual D5 pump setup in the gpu loop for shits and giggles 

[Sir Beregond]

30 minutes ago, newls1 said:

Thanks for the replies guys, much appreciated.  Not sure why you guess thought I was concerned about the heat dump of a D5, but i could care less about that and my post says nothing about that.... I have PLENTY of surface area to compensate the 20 or so watts of heat a D5 adds.  Just was wondering if high flow would ever hurt performance, but im guessing not.  Thanks y'all. Ill go ahead and add a Dual D5 pump setup in the gpu loop for shits and giggles 

Oh just saw Avocado cover that, so added my $0.02 on D5. 

 

Anyway, yeah high flow will not hurt your performance. Have fun.  

[UltraMega]

It would be physically impossible unless higher flow lead to some kind of bubble issue, but assuming there is no air in the tube, no because physics. 

[Avacado]

42 minutes ago, newls1 said:

Thanks for the replies guys, much appreciated.  Not sure why you guess thought I was concerned about the heat dump of a D5, but i could care less about that and my post says nothing about that.... I have PLENTY of surface area to compensate the 20 or so watts of heat a D5 adds.  Just was wondering if high flow would ever hurt performance, but im guessing not.  Thanks y'all. Ill go ahead and add a Dual D5 pump setup in the gpu loop for shits and giggles 

Whenever flow and heat are used in a question, NOT talking about the fact that increased pump speed causes increased heat dispersion in the loop would be an oversight. No matter how negligible the excess heat would be. Thats why we talked about it. 

[UltraMega]

Just for the sake of being thorough and not having anything else to do at the moment... 

 

If the water flows quickly enough, even without any air in the tube, at a certain point there can be enough pressure the cause the molecules to separate into gases. That could be what happens in the waterblock mentioned on the OCN thread you made for the same topic, but it's also highly possible that is just where tiny bits of air get settled. I think it would be very hard to create a scenario where that would bring down the overall temps even if it does allow for a hotspot to exist. 

 

But assuming there is not enough pressure to cause that anywhere in your system, it would be impossible for the temps to be higher due to more flow unless the ambient temp was higher than the water temp in the system. 

 

Maybe in a lab it would be possible to create a situation where more water flow did lead to more heat because water does heat up from friction, but the amount at which it heats up from friction alone is so small that it's totally overpowered by other cooling effects. I suppose if you had water that was the same temp as the ambient conditions and you started to add friction to the water, you could heat it up to some measurable degree. 

[Bones]

Actually it's possible to reduce cooling efficiency but that's in extreme cases only.

Normally the rad can remove heat quickly enough it's not a problem to even consider.

It's the same effect if for example you remove the center from a thermostat in a car, although water is flowing all the time it flows so fast the radiator cannot remove all the heat it needs to, leading to a slow but sure heat buildup in the cooling system. I've ran into that before so I know it's true but an auto is different than a PC, you're not dealing with the same amount of heat energy with a PC even though temps can be about the same when in operation at times.

Thing is it always takes time to remove enough heat energy so you don't have heat buildup in the coolant - That being key to the problem and the sticking point of it if a problem exists.

In PC cooling systems it's possible but extremely unlikely.
If it ever happens it can be offset by installing a small restriction in the tubing somewhere to slow it down if you must. Be it an actual restriction in it somewhere, maybe a smaller piece of tubing/hose somewhere in the loop, whatever works to slow the flow is the way to go with it if such occurs.
If your pump speed is variable you can always turn it down a little too.

Also know larger tubing can be used too throughout the system, that tends to soak up some of the flow volume because it increases the volume of coolant flow and creates a slowing effect to the coolant overall, based on pump speed being the same.

So... Take the pump and try it.
I believe it will be fine but if not, a larger/extra rad inline for extra cooling area or stronger fans to the rad may offset it.

Edited June 9, 2022 by Bones

[UltraMega]

9 minutes ago, Bones said:

It's the same effect if for example you remove the center from a thermostat in a car, although water is flowing all the time it flows so fast the radiator cannot remove all the heat it needs to, leading to a slow but sure heat buildup in the cooling system. I've ran into that before so I know it's true but an auto is different than a PC, you're not dealing with the same amount of heat energy with a PC even though temps can be about the same when in operation at times.

I don't doubt that there are situations where more water flow can cause problems for a car's cooling system, but if that is the case the cause would not be the water flow itself from a physics perspective. It would have to be something else like air in the system to begin with which is a lot more common for cars compared to computers. 

 

I don't think it is possible in a PC cooling system. In theory it could be but you would need a much more powerful pump and the system would have to be able to tolerate high pressure. Basically, the thing would break long before it got hotter. 

 

Asking if more water flow can cause higher temps is different than asking if there is a flaw in the way water flows through a waterblock. 

Edited June 9, 2022 by UltraMega

[Bones]

9 minutes ago, UltraMega said:

I don't doubt that there are situations where more water flow can cause problems for a car's cooling system, but if that is the case the cause would not be the water flow itself from a physics perspective. It would have to be something else like air in the system to begin with which is a lot more common for cars compared to computers. 

Nope - Excessive waterflow can cause heat buildup in it's cooling system and yes, as said I've ran into that before and it took installing a good thermostat to restore efficiency.

I've also seen some guys install a simple washer with a given sized hole (Has to be of a certain size) as a quick fix too until they could get a proper thermostat - The washer thing is also a drag racer's trick to get more heat in the engine to help combustion without it getting too hot before the run is done.
As long as they don't mess around much getting a run in it works but know it's a low-tech trick, normally a "Hotter" thermostat is used.

I worked on autos and heavy equipment (Diesels) for a living in the field out of a service van (Not in a shop) so I've seen it before happening for more reasons than you'd believe possible yet it did.

As for PC's as I said above, I don't think they'll have any issues trying the pump they have.
 

Edited June 9, 2022 by Bones

[UltraMega]

8 minutes ago, Bones said:

Nope - Excessive waterflow can cause heat buildup in it's cooling system and yes, as said I've ran into that before and it took installing a good thermostat to restore efficiency.

I've also seen some guys install a simple washer with a given sized hole (Has to be of a certain size) as a quick fix too until they could get a proper thermostat - The washer thing is also a drag racer's trick to get more heat in the engine to help combustion without it getting too hot before the run is done.
As long as they don't mess around much getting a run in it works but know it's a low-tech trick, normally a "Hotter" thermostat is used.

I worked on autos and heavy equipment (Diesels) for a living in the field out of a service van (Not in a shop) so I've seen it before happening for more reasons than you'd believe possible yet it did.

As for PC's as I said above, I don't think they'll have any issues trying the pump they have.
 

Like I said, I don't doubt it can happen in a car but the root cause wouldn't be the flow itself. It would be some other flaw in the system that causes uneven flow. Also, a cars cooling is designed for high pressure which makes it a totally different ball game. For a non-pressurized system like a PC cooler it's just not possible.