The "What's your Water-cooling System's Flow Philosophy" Poll

[Cathar]

Probably the only thing I think missing from the equations here is the different style blocks. Some are more restrictive than others and I would imagine they would affect the overall rating. My case: CPU: AquaXtreme MP-05 SP LE - high restriction impingement block, and GPU: AquaXtreme MP-1 - high restriction block also. As opposed to a Maze 4 or non-impingement style blocks. Cathar, if you are still reading this thread now and then, what's your take?

Wow! Bet you didn't expect me to come answer your question one year on!

The actual restriction of the blocks is somewhat independent. Why?

The reason is that the poll is there to determine what rest of the components are doing in terms of offering a waterblock flow. If we've got a weak pump, small tubing, restrictive barbs, and so on, then both the high-restriction and low-restriction blocks will be operating at the lower end of their flow-performance curve. Vice-versa applies. Got a strong pump, low restriction fittings and larger bore tubing, then both the water-blocks will be operating towards the upper-end of their flow-performance curve. Independent of the block, the rest of the system is designed to deliver a high level of flow to the block.

I do accept your point to a degree though. If we have a high-flow system and stick in a super-restrictive waterblock and flow gets choked from 10LPM without the block, to 1LPM, then it's no longer a high-flow system. The problem with attempting to put that as a question is that the level of restriction of most water-blocks are not published, nor even known.

I also saw that most were arguing over pumps, radiators, tubing, and so on. People in general weren't arguing so much about what blocks are used as most people already understand that it's not so much the absolute flow-rate you deliver that determines performance, but rather what you do with the flow/pressure you have, and that is a function of the waterblock's design. The poll is therefore more focused around how much flow/pressure our system is setup to deliver to some waterblock that gets introduced. We can give a low-restriction block pathetic flow, or we can really try to ram flow through a high-restriction block, and that's independent of the block itself.

It all gets a whole lot more gnarly once we start mixing high/low restriction blocks, and the impact that a high-restriction block has on overall system flow-rates and therefore the performance of a low-restriction block in the same system. I was trying to keep the poll simple enough so people could easily find common ground, rather than nit-picking and fighting with each other.

Sadly, looking at these forums 2 years on, I see that little has changed.

 

[BillParrish]

not so much the absolute flow-rate you deliver that determines performance, but rather what you do with the flow/pressure you have,

Thank you ! (already voted under another ID back in the day) but did not like the way the question was presented due the the very fact you mention above.


IMO the coolant velocity in the different parts of the loop is just as important as flow rate if not more and total system design as an intergrated matched system will have greater impact, if done well, that any particular component being swapped out "cuz its more better" .

(
Btw nice to see you still hanging around, you dont know me from Adam's cat but I have been reading your stuff for at least 10 years, lol probally 15 or more but is hazy that far back, and it never fails to impress. Best regards. )

 

[penguin]

Holy crap Cathar's back!

 

[Justin_Diduch]

Sadly, looking at these forums 2 years on, I see that little has changed.

Now where would the fun be in that?

Glad to see you posting, dont be a stranger!

 

[Top Nurse]

It all gets a whole lot more gnarly once we start mixing high/low restriction blocks, and the impact that a high-restriction block has on overall system flow-rates and therefore the performance of a low-restriction block in the same system.

So on my recent computer (the old one died when a glass of water fell in the box) I used an Aquastream clocked up to about 1048 levels. I used a Cuplex XT which is a fairly restrictive block as the first block in my loop. After that I used very unrestrictive TwinPlex (older ones and not Pro) blocks to cool the NB, GPU, and the Mosfets (did it because it looked nice and I lost the stock heatsink).

Anyway is this the best way to run the loop with the most restrictive block at the beginning or would it be better to put it at the end of the loop?

 

[madmat]

It doesn't matter where you have it in the loop, the overall flow will end up being the same. Put it where it's most convenient for your loop routing.

 

[Top Nurse]

It doesn't matter where you have it in the loop, the overall flow will end up being the same. Put it where it's most convenient for your loop routing.

But the pressure will change that the block sees, correct?

 

[madmat]

No, it's a closed loop. the pump is drawing from the other side of the blocks, it can't pull the water any faster than the blocks will allow it to flow so the flow rate will be the same no matter where the restriction is.

 

[Top Nurse]

No, it's a closed loop. the pump is drawing from the other side of the blocks, it can't pull the water any faster than the blocks will allow it to flow so the flow rate will be the same no matter where the restriction is.

Wrong terminology. What I meant to say is that the pressure changes that the block sees regardless of the flow being the same. Am I making sense here?

 

[madmat]

Wrong terminology. What I meant to say is that the pressure changes that the block sees regardless of the flow being the same. Am I making sense here?

None at all. The pressure will equalize in the loop. No matter where the restriction lies in the loop you will have negative pressure on one side and positive pressure on the other. To prove it to yourself I'd suggest getting a low pressure gauge and putting it in a T line then mounting it at various places in your loop.

 

[Cathar]

Wrong terminology. What I meant to say is that the pressure changes that the block sees regardless of the flow being the same. Am I making sense here?

Yes. It makes sense, but the answer is still that it doesn't make a difference.

The flow-rate doesn't change. The level of pressure at the entrance of some block will change, but because the flow-rate hasn't change, the level of pressure-drop across some block doesn't change.

Given a fixed flow-rate, a waterblock will "take" so much pressure away from inlet to outlet. That's the water-block's "pressure drop" for that flow-rate. The actual pressure at the inlet isn't important, so long as the pressure at the inlet is more than the pressure-drop of the block for that flow-rate.

Think of it as like adding up a series of numbers. If you jumble the numbers about it doesn't affect the final total. The running total half-way through adding up the series may be different, but the end-result is always the same, and so there's no net change.

 

[Top Nurse]

None at all. The pressure will equalize in the loop. No matter where the restriction lies in the loop you will have negative pressure on one side and positive pressure on the other. To prove it to yourself I'd suggest getting a low pressure gauge and putting it in a T line then mounting it at various places in your loop.

Funny you should mention that as I have been looking at getting such a meter just for that purpose. I'm also thinking of adding in some more temp sensors as I have 6 total and am only currently using three (water in, water out, and my HD's) so I can see how the temps change from block to block by varying the flow rates.

 

[JC634]

Funny you should mention that as I have been looking at getting such a meter just for that purpose. I'm also thinking of adding in some more temp sensors as I have 6 total and am only currently using three (water in, water out, and my HD's) so I can see how the temps change from block to block by varying the flow rates.

I have 10 temp sensors in my computer and 3 in my external water box. Temps do change due to CPU/GPU usage. Flow makes a difference, also. Since my SW auto adjusts fans based on temps, the temp differences are minimal.

However, I have found that the parameter that has the most effect on temps is ambient air temperatures. I have seen a differance of 5-6°C between hot and cool days.

Which is, of course, the limitation of any "normal" cooling: you are limited by ambient air temps. "Normal" as in not "Extreme Cooling Solutions"