Ferris wheel effect, and pressure head

so i've had a few discussion with people on this site about pump head pressure in a closed loop something (something we all use)

i just couldn't understand/believe that the pressure going up one side of the loop cancelled out the other and this meant you could actually make really high build heights in relation to the pump. the only way i'm going to be convinced of this was to build it.

items:
1x D5 pump + pump top
2x flow indicators
8x barb fittings
1x 1/4 valve
2x hose joining barb fittings
1x battery charger (for powering pump)
10m clear vinyl tubing


the D5 pump:
taken from http://www.acousticpc.com/laing_d5_water_pump.html
pump specs as follows

Nominal voltage
12 V DC

Operating voltage range 6 to 14 VDC
Nominal power (@ 12 V)
24 W

Nominal current (@ 12 V) 2 amps
Motor type Brushless, microprocessor controlled
Maximum head
10 ft (3.1 m)

Maximum discharge
~ 317 GPH (1200 LPH)

Connection size
½" barbs

Maximum pressure 50 PSI (3.5 BAR)
Temperature range 32 °F to 140°F (0 °C to 60 °C)
Electrical connector
Molex 4 pin

so the part high lighted in bold basically says if you put a res on one side and a hose on the other the pump will push water straight up for 3.1m (this figure is 3.9m on the ek website)




The build

so i went about setting up a loop out off my balcony.



loop its self was simple, i installed the flow meters circled in red, i did it slightly offset so i knew i'd be able to see one through the glass at the top.
at the top i setup a Tee piece with a 1/4" valve and a few fittings and a piece of pipe to act as a bit of a res as i tried to bleed the loop. this was a bit bit tricky but i found you fill the pipe and squeeze/pulse the tubes to get the air out as it basically developed an air lock.




For power i used my super cheap auto battery charger as it fitted the requirements for power, 12v @ 1.6a




so the height i started off at was close enough to 4m from ground to top of the loop. This is basically the height a D5 should rated to run at, which still seems really quite high for such a small pump.




connected up the power and:


i fiddled with the pump speed going from 1 - 5 and it made no difference, other then slowing down the flow as indicated by the wheels spinning slower. so no matter the speed the pump was still able to move water around 4m



so i added some more tubing taking the total height to 4.8m (appologies for blurry photo) this is getting close to 1m past what the pump should be rated for in static head pressure. the ruler was used to spread the hoses apart as they where kinking while handing there



turned power on and:



so at 4.8m a d5 has no trouble moving water around. i would love to have gone even taller but it would involve getting on the roof and buying more tubing.


so whats happening

The Science

The term pump head, has nothing to do with height. It is a term that is used that is the pumps ability to overcome the friction that is created when fluid flows through a pipe. Once a system is filled with fluid – the pump merely has to overcome this friction. An easy way to think about it is that as water goes up one side of the loop, it also comes down the other side of the loop. The two cancel each other out.

A good illustration is like a Ferris wheel. One car going up balances the other car coming down. The cars are in balance and the motor only has to overcome the friction in the bearings and we are off and spinning. The pump in a closed loop system in simple terms is like the motor on a Ferris wheel. All the pump has to do is overcome the friction that is created when the water rubs against the pipes – and that friction has nothing to do with height. In fact we could take a ten story building, size a pump for it, knock the building over on its side and we would still need the same exact pump to move the water around.


In a closed loop system, you only have to consider the dynamic pressure head only. i.e The pressure that the pump has to develop to circulate the fluid overcoming the friction loss and fitting pressure loss (aka items in the loop)


so where does the 3.1m head pressure come into it

what the manufacturer is saying, depending on the height of your res, the pump will move water 3.1m above that.... so if your res is located 5m above your pump, as the water level will level out equally over the pipe work, the pump can move water 3.1m above the point for a total of 8.1m (3.1m + 5m)

in a closed loop that all changes because there is no difference in res height because its a properly bled system so as mentioned you only need to over come the losses caused by the pipe work and fittings. (which is really quite small) this is a reason why water cooling computer components are more concerned about flow and pressure rather then pump head height




The conclusion

well I've definitely convinced my self that there is nothing to worry about with how tall you want to make a build and that the ferris wheel effect works.

any thoughts?

i know i'm just proving a concept that alot of people would have already known to be true, but at least a thread like this may convince the "nay sayers" like my self

 

Yep, science works. Now, the weather, how we doing Ollie?

 

Conclusion.... You need a taller house...

 

Not quite, pump head refers to the pumps ability to push water up in an open loop system.

What does this mean? It means that if you did the same experiment again, this time with a bucket on the ground next to the pump full of water to draw from, you will find the maximum height you can push water up the hose (open at the top to spill out freely) will be your head pressure height (at a given flow rate).

Otherwise, very cool experiment, this should clear a few things up for some noobies here

 

The science is there is always static head and dynamic head when designing a pumping system. In this instance you negated static head due to bled loop.

In a real pumping system also need to consider suction head. Try and suck over 9.81m and the water "boils."

I have stood in a plumbing supply store and argued with the guy behind the counter on how to calculate head losses in a system. In the end I demanded he give me what I asked for..... it worked perfectly.

 

if the experiment was changed to a bucket reservoir at the bottom and a open tube am i right in thinking the tube would fill to the rated head height?

if the tube outlet was at half the rated head height would the pump only give half the rated flow?

 

Yeah you are right a pump rated for 5.1m can only deal with 5.1m of head.

The flow rate will be dependent on the design of the impeller (or multiples of in some pumps) and resulting curve. Some pumps run well until about 90% of their max head and then it just becomes about wasting power. Others offer a more linear response but typically in higher pressure applications.

I haven't seen a curve for the shitty little pumps we run in computers so am not sure.

 

Just found this. Well done for testing it and proving many of us right

 

a cheaper version of this experiment uses a garden hose and filling it with water. then stand with the hose ends up on the balcony and pour water in one end of the hose (with a funnel attached) and watch the water 'magically' flow out the other side. no pump involved ! just the weight of the water in the funnel and it pushs the water the whole distance.

crappy stick man diagram

Code:

\/
 |    |
 |    |
 |    |  
 |    |
 |____|

 

or you can open up dwarf fortress, dig a nice big fort then put in a section like this

Code:

##########R##
########   ##
<   #### ####
### #### ####
### #### ####
###      ####
#############

# = wall
R = river
< = rest of fort

water will flood your entire fort.

 

Yeah but wouldn't have helped prove just how much head a D5 pump can push around, helps to stop some of the newer guys to water cooling wondering if a D5 pump is enough for their needs

 

Or that stupid swimming pool loop question that went for far too long. Or any other time that people have ended up with their knickers in a knot about this.