Pump GPM issue

[TheRedViper]

So I have two Aquacomputer PWM D5s, one in each loop (one loop for CPU, one for GPU). Within Aquaero program called Aquasuite, I can see the pumps RPM and apply a %preset to them from within a data range (min rpm to max rpm transfered in a %). Right now, the pumps are going full speed which is about 4800rpm and it is quite loud. According to these pumps specs, the pump delivers at max 1500lh of flow which is about 6.6 gpm. I know from looking at my pump speed without pwm feedback that theyre both going at 4800rpm which I say is the max speed. 

 

Now knowing that, from looking at some reviews, it seems GPM over 3 within a 12mm OD loop has negligible gains in coolant temperature drop (3-5%). That being sad, correct me if im wrong, but now that I know the ideal GPM which would allow me to reduce unnecessary noise from the pumps, I can take that data and apply it to pumps formulas:

 

The displacement of a pump = 231 x GPM divided by RPM = in3 per rev

 

We can now isolate displacement by taking the GPM at max speed (6.6 gpm) and the max RPM (4800): 231 x 6.6 gpm divided by 4800 rpm = 0.317625 in3 per rev

 

The flow in GPM of a pump = RPM x displacement divided by 231

 

If we take a GPM of 3 ------ 3 = RPM x 0.317625 in3 per rev divided by 231 ------- RPM = 2181 RPM

 

So I would need to drop my preset value until I hit 2181 RPM to have my pumps provide 3GPM instead right?

 

[W-L]

6 minutes ago, TheRedViper said:

-SNIP-

Flow rates are not a linear calculation it depends on the restriction, pump tops, and speed/power the pump is rated at the easiest solution would be to just have a inline flow meter to determine the exact amount but here is a rough idea of what you should expect from the D5. 

https://martinsliquidlab.wordpress.com/2011/04/03/koolance-pmp-450-d5-vario-pump/

[TheRedViper]

Not sure how to use these to calculate an approximative 3gpm?

[W-L]

5 minutes ago, TheRedViper said:

Not sure how to use these to calculate an approximative 3gpm?

It's complicated since you need more information to get a detailed calculation with pressure, and velocity of the fluid at a given point to get the flow rate for the loop you have designed as it's not solely based on the pump but more so what the system the fluid is going through. 

 

That's why I suggested a flow meter as that just gives you a direct answer. 

[TheRedViper]

13 minutes ago, W-L said:

It's complicated since you need more information to get a detailed calculation with pressure, and velocity of the fluid at a given point to get the flow rate for the loop you have designed as it's not solely based on the pump but more so what the system the fluid is going through. 

 

That's why I suggested a flow meter as that just gives you a direct answer. 

I see, i guess il just drop it by a 1000 rpm and call it a day.

[W-L]

11 minutes ago, TheRedViper said:

I see, i guess il just drop it by a 1000 rpm and call it a day.

Flow rates usually doesn't cause a huge difference in performance I would run a simple test at full speed and incrementally step it down to see if you get any noticeable change in temps or where you feel it is comfortable in terms of noise to get a good balance. 

[For Science!]

20 minutes ago, TheRedViper said:

I see, i guess il just drop it by a 1000 rpm and call it a day.

I have quite a restrictive loop (7+ 90 degree fittings + 360 mm HWlabs GTS + 240 mm HWlabs GTS + 2x gpu blocks + 1 x cpu block), and I need to run my single D5 pump at ~90% PWM to achieve 1 gpm. This was established using a koolance flow meter, so should be fairly accurate.

 

I would say don't worry about it so much if you are fine with your temps and acoustics. Personally I found the D5 quieter at higher rpm than lower rpm due to the frequency of the noise.

[TheRedViper]

11 hours ago, W-L said:

Flow rates usually doesn't cause a huge difference in performance I would run a simple test at full speed and incrementally step it down to see if you get any noticeable change in temps or where you feel it is comfortable in terms of noise to get a good balance. 

 

10 hours ago, For Science! said:

I have quite a restrictive loop (7+ 90 degree fittings + 360 mm HWlabs GTS + 240 mm HWlabs GTS + 2x gpu blocks + 1 x cpu block), and I need to run my single D5 pump at ~90% PWM to achieve 1 gpm. This was established using a koolance flow meter, so should be fairly accurate.

 

I would say don't worry about it so much if you are fine with your temps and acoustics. Personally I found the D5 quieter at higher rpm than lower rpm due to the frequency of the noise.

My cpu loop has 1x ek 360 xe = delta 1 psi at 2gpm, 1x ek 240 xe = delta 0.87 psi at 2 gp and 1x cuplex kryos next = delta 3.5 psi at 2gpm. I also have about 8 fittings.

 

Pump has 3.7m head = 12.14 feet

Pump max flow = 1500 lh = 6.6 gpm