How much pump is enough? How much is too much? - Page 11

bigben2k · 05-20-2007, 04:28 PM

Efficiency would indicate what electrical energy is wasted, some of which is going to be expended as heat, part of which may or may not end up in the cooling loop.

The pumping action is, IMO, the main source of pump induced heat. It can be calculated as a function of flow and pressure. In fact, you can calculate the heat induced this way, for each element within a loop, all of which will add up to the pressure differential at the pump inlet/outlet.

Maybe someone still remembers the formulae?

Roscal · 05-21-2007, 01:48 PM

Hydraulic power[W] = Flowrate[m³/s] * Pressure[mH2O] * Fluid density[kg/m³] * g[9.81 m/s²]

You know the input power (electrical), then global efficiency is the ratio between hydraulic power and input power. For instance, a Laing DDC 10 W has a peak efficiency of ~17 % where a Laing D5 is ~14 %. Quite bad, but normal for such little pumps. The bigger the pump is, the better the efficiency will be generally.

Heat induced by an element uses the same formula, but pressure become its pressure drop instead. At final : hydraulic power provided by pump = sum of all heat induced by each element due to flow (to fight against friction), all water movement is converted to heat.

bobo5195 · 05-27-2007, 11:19 AM

NB I have never actually seen a water cooling pump in the flesh. But have actually made one in my uni course (100mm diameter brass rotor with an ally body powered by a 200W motor, think the best were managing about 30% eff (real to actual hydraulic power) at 6 litres per minute and 10meters of head or something like that).

Those numbers seem reasonable. (hydraulic power (W) = delta P (pa) * Q (kg/m/s), Q being the mass flow rate).

Numbers are low due the pumps not avoiding restrictions and cavitations. Cavitation occurs due to the rotors action. At the start due to centrifugal action the water is moving slowly while at the end it is moving pretty fast. Also the flow channel is expanding. This means that you will get areas of gas build up inside the pump sucking the water backwards. Also the blade designs are often straight when you actually want to curl them so the water around. Same with the inlet where you want a nice little deathspike to smooth the flow around so the loss factor is lower.

I was thinking that I could do a nice little business in upgrading D5’s with a “tuning kit”, hehehehe

muffy · 05-28-2007, 12:47 AM

I was thinking that I could do a nice little business in upgrading D5’s with a “tuning kit”, hehehehe

yeah i have turned my D5 from a 18 watt pump to a 80watt ... one she sure kicks ass now

bobo5195 · 05-29-2007, 06:24 PM

18 to 80w might not be far of the mark actually

I was thinking heatsink / liquid cool the motor so it can be overvolted.

Add a 3D impeller, overhung volute diffuser vanes maybe even a variable geometry setup, ported output area to ½ inch tubing and watch it go. Efficiency would make it in the range of a 80 watt water cooling setup with the current tech and there would probably be less heat dump (due to pumping) and certainly less noise/ vibration to boot.

CAUTION: this post is powered by beer (budvar) to be precise hence is as inaccurate as an alcoholics ramblings

Marci · 05-30-2007, 05:01 AM

Quote:

NB I have never actually seen a water cooling pump in the flesh.

Where in the UK are you? Can send you some...

bobo5195 · 06-03-2007, 07:36 AM

Cheers it would be nice to see some of these things in the flesh. Despite my ramblings on these forums I am still yet to actually see any watercooling components with my own eyes so it would be useful to play with some stuff. An old (doesn’t have to working) waterblock would be useful, if it is not too much trouble.

I am based in London now working for a product design company, but the address is for my home in good old Wales.

We’ll have to see if I come up with a nice cunning plan for making watercooling mainstream! I think I have something devious in mind…

mwolfman · 06-29-2008, 04:01 PM

Aslong as I have a flow (sometimes it stops) I can use the Liang DCC 12 plus/ultra @8,2V= zero noise.
OK the water going IN to my CPU-cooler is very coold, around ambriant temperature (5*12cm rad).
If my CPU/GPU gets hot my NG-BIG runs the pump and the fans alot faster.

billbartuska · 07-31-2009, 11:05 AM

Quote:

Originally Posted by kostiakot

A really bad pump as a peak eff of about 11%
A normal cheaply made pump eff of 40%
a well made cheap pump with 2D rotor is 50%
i would say
regards

Sooooo......

Which pumps have a high efficiency and why is that good for WCing?

bobo5195 · 08-25-2009, 01:30 PM

Well designed centrifugal pump with curved blades. I.e what you normall have but curve the blades.

Beign well made maters more. Woudl really question 50% though, is that including drive/IC.

Not point going 3D with water pumps as not benefit

05-20-2007, 04:28 PM	#251
bigben2k Responsible for 2% of all the posts here. Join Date: May 2002 Location: Texas, U.S.A. Posts: 8,302	Re: How much pump is enough? How much is too much? Efficiency would indicate what electrical energy is wasted, some of which is going to be expended as heat, part of which may or may not end up in the cooling loop. The pumping action is, IMO, the main source of pump induced heat. It can be calculated as a function of flow and pressure. In fact, you can calculate the heat induced this way, for each element within a loop, all of which will add up to the pressure differential at the pump inlet/outlet. Maybe someone still remembers the formulae? __________________ WBTA (Water Block Testing Alliance) Nordic Hardware WC 101

05-21-2007, 01:48 PM	#252
Roscal Cooling Savant Join Date: Apr 2002 Location: North of France Posts: 198	Re: How much pump is enough? How much is too much? Hydraulic power[W] = Flowrate[m³/s] * Pressure[mH2O] * Fluid density[kg/m³] * g[9.81 m/s²] You know the input power (electrical), then global efficiency is the ratio between hydraulic power and input power. For instance, a Laing DDC 10 W has a peak efficiency of ~17 % where a Laing D5 is ~14 %. Quite bad, but normal for such little pumps. The bigger the pump is, the better the efficiency will be generally. Heat induced by an element uses the same formula, but pressure become its pressure drop instead. At final : hydraulic power provided by pump = sum of all heat induced by each element due to flow (to fight against friction), all water movement is converted to heat. Last edited by Roscal; 05-21-2007 at 01:54 PM.

05-27-2007, 11:19 AM	#253
bobo5195 Cooling Savant Join Date: Aug 2005 Location: uk Posts: 400	Re: How much pump is enough? How much is too much? NB I have never actually seen a water cooling pump in the flesh. But have actually made one in my uni course (100mm diameter brass rotor with an ally body powered by a 200W motor, think the best were managing about 30% eff (real to actual hydraulic power) at 6 litres per minute and 10meters of head or something like that). Those numbers seem reasonable. (hydraulic power (W) = delta P (pa) * Q (kg/m/s), Q being the mass flow rate). Numbers are low due the pumps not avoiding restrictions and cavitations. Cavitation occurs due to the rotors action. At the start due to centrifugal action the water is moving slowly while at the end it is moving pretty fast. Also the flow channel is expanding. This means that you will get areas of gas build up inside the pump sucking the water backwards. Also the blade designs are often straight when you actually want to curl them so the water around. Same with the inlet where you want a nice little deathspike to smooth the flow around so the loss factor is lower. I was thinking that I could do a nice little business in upgrading D5’s with a “tuning kit”, hehehehe

05-28-2007, 12:47 AM	#254
muffy Cooling Neophyte Join Date: Jun 2003 Location: Auckland, New Zealand Posts: 40	Re: How much pump is enough? How much is too much? I was thinking that I could do a nice little business in upgrading D5’s with a “tuning kit”, hehehehe yeah i have turned my D5 from a 18 watt pump to a 80watt ... one she sure kicks ass now

05-29-2007, 06:24 PM	#255
bobo5195 Cooling Savant Join Date: Aug 2005 Location: uk Posts: 400	Re: How much pump is enough? How much is too much? 18 to 80w might not be far of the mark actually I was thinking heatsink / liquid cool the motor so it can be overvolted. Add a 3D impeller, overhung volute diffuser vanes maybe even a variable geometry setup, ported output area to ½ inch tubing and watch it go. Efficiency would make it in the range of a 80 watt water cooling setup with the current tech and there would probably be less heat dump (due to pumping) and certainly less noise/ vibration to boot. CAUTION: this post is powered by beer (budvar) to be precise hence is as inaccurate as an alcoholics ramblings

06-03-2007, 07:36 AM	#257
bobo5195 Cooling Savant Join Date: Aug 2005 Location: uk Posts: 400	Re: How much pump is enough? How much is too much? Cheers it would be nice to see some of these things in the flesh. Despite my ramblings on these forums I am still yet to actually see any watercooling components with my own eyes so it would be useful to play with some stuff. An old (doesn’t have to working) waterblock would be useful, if it is not too much trouble. I am based in London now working for a product design company, but the address is for my home in good old Wales. We’ll have to see if I come up with a nice cunning plan for making watercooling mainstream! I think I have something devious in mind…

06-29-2008, 04:01 PM	#258
mwolfman Cooling Savant Join Date: Jun 2004 Location: Sweden Posts: 152	Re: How much pump is enough? How much is too much? Aslong as I have a flow (sometimes it stops) I can use the Liang DCC 12 plus/ultra @8,2V= zero noise. OK the water going IN to my CPU-cooler is very coold, around ambriant temperature (512cm rad). If my CPU/GPU gets hot my NG-BIG runs the pump and the fans alot faster. __________________ Very old specs below, but I keep it for sentimental values :-/ Gigabyte GA-7N400 Pro2, Barton 2500@3200* -11201MHz, X800 XL , 1 GB Cosair 3200 (6-2-2-2) @400MHz in dual channel TurboBlock Rev. 1.0~0,15452756±0,004125251°C/W , HL120.3* with 3 Papst 4412@7volt , Hydor L30II

08-25-2009, 01:30 PM	#260
bobo5195 Cooling Savant Join Date: Aug 2005 Location: uk Posts: 400	Re: How much pump is enough? How much is too much? Well designed centrifugal pump with curved blades. I.e what you normall have but curve the blades. Beign well made maters more. Woudl really question 50% though, is that including drive/IC. Not point going 3D with water pumps as not benefit

Currently Active Users Viewing This Thread: 5 (0 members and 5 guests)