Making a 12V pump

[nicozeg]

This is the first stage of an experiment I'm making: Build a 12V pump with rpm monitoring.

The volunteer for this was a 60mm fan with some broken blades. After tearing it apart I submerged the electric parts in waterproof epoxy; then lathed it to fit inside the ring magnet.

[nicozeg]

I leaved a wax drop over the wire leads to keep the area free from epoxy. After cleaning the wax I soldered back the wires and sealed the area with silicone.

[nicozeg]

After taking out all the fan blades the bare hub rotates at about 10.000 rpm.

Here's at 9.700 rpm inside a glass of water.

[nicozeg]

I'm leaving it some days working UW for testing. I want to know if the ball bearings can survive this environment. A sleve bearing one should be better for this, but high performance fans come with ball ones

For the next stage of this I'm going to need some help: The impeller design.

There's too much variables to just guess, and I want to find an efficient design to get the highest water pressure possible from this high rpm, low torque motor. Anyone with experience? or maybe a good link?

[hara]

After taking out all the fan blades the bare hub rotates at about 10.000 rpm.

You mean 10 or 10000 ?

[theetruscan]

He means 10,000 (. and , are interchangeable as long as you are consistent throughought).

[mrmoustache]

where in mexico are 10,000 and 10.000 the same number

[dr.chaos]

You should have a look to this ZZZ number: http://zzz.com.ru/119.html

BTW read the comments too, since there are nice infos about Tesla Pumps (I'm currently studying a way to build one using 8/10 mm copper discs)

[nicozeg]

Originally posted by hara
You mean 10 or 10000 ?

In spanish we use ", " to separate decimals. Did'nt know the rest of the world was using the wrong expression.

You should have a look to this ZZZ number: http://zzz.com.ru/119.html

BTW read the comments too, since there are nice infos about Tesla Pumps (I'm currently studying a way to build one using 8/10 mm copper discs)

Nice link, very similar to what i'm making

[Puzzdre]

!!!

Even if this assembly is maybe too weak for the pump, just make the flow indicator and use it to shutdown the rig if there is no flow...

Bravo!!!

[bigben2k]

I was thinking this week...

I work for an oil drilling company, and how they get the power down to the drilling tool: they pump a fluid, and through a special lobe configuration, of a "rotor" and "stator", the drill bit is allowed to turn.

Think of it as a drill bit (which typically has two "lobes" or grooves). The "stator" has one more groove or lobe than the rotor, and the pressure differential forces the rotor to turn.

Of course this company used much larger tools!

I've often wondered why they don't use electrical power, but then I found out that the hole depth can reach several miles.

I'd be curious to see if this could be applied in a waterblock... for, I don't know, say, an active turbulator?


As for your impeller, you'd need to know what kind of force this motor is capable of. Your best bet at this point, would be to try a few basic designs:
#1: the straight paddlewheel
#2: the curved paddlewheel

You'll have to try them at various sizes.

Off-hand, because this motor has just enough power to move a fair quantity of air, I would expect it to be able to move water at the same mass rate.

Now how much denser is water, compared to air again?

[nicozeg]

Even if this assembly is maybe too weak for the pump, just make the flow indicator and use it to shutdown the rig if there is no flow...

Yes, one of my goals is to add protection against pump failure. But this type of monitoring can't be done in passive mode, it has to be the pump, not just a sensor.

The fan was rated at .36 amp at 12v (4.3w) Is on the weak side for a pump, but it can be a lot more efficient than common ac pumps as it rotates always in the same direction. If this one is successful, I'm going to make a second one with the most powerful fan with sleeve bearings I can find.

As for your impeller, you'd need to know what kind of force this motor is capable of. Your best bet at this point, would be to try a few basic designs:
#1: the straight paddlewheel
#2: the curved paddlewheel

You'll have to try them at various sizes.

That 4.3 w should say something, but I don't know the tipical efficiency of this type of motor.

#1 is discarded. Is only good for bidirectional rotation, but not very efficient.

#2 is what I want, but look at the variables:

-Impeler diameter
-Number of blades
-height " "
-Lenght " "
-Angle " "

Suppose I choose 3 values for each one. That makes 3^5 options; 243 impellers to make!

I prefer to make some calculations first and build only one, but don't know how.

[bigben2k]

Maybe you should look at a typical impeller of a similarly powered pump: that should give you a general guideline.

[Volenti]

Check this out, it's a review of that cpufx crap water cooler, but it shows the detail of the 12v pump design in detail (also based around a 12v fan motor)

may be able to point you in the right direction.

I would personally, based on my own experience with modding pumps, use a flat disk between 1.5-2'' in diameter with small vanes on it, on the order of 1/16'' high, I'll draw a diagram later.

(basically similar to how the impellers on the high head iwaki's are set up)

[MadDogMe]

Do the vanes cross over in the middle or are they around the edge only like a centrifugal fan?, I'd guess they have a hollow space in the middle...

I seriously doubt this will have the power/torque needed nico. The one way action with an offset outlet does'nt make that much difference. You'd need a 20watt+ fan to start with I recon :shrug: ...

But your epoxy work is the Bogs Dollox! ...

[Volenti]

Originally posted by MadDogMe

I seriously doubt this will have the power/torque needed nico. The one way action with an offset outlet does'nt make that much difference. You'd need a 20watt+ fan to start with I recon :shrug: ...

But your epoxy work is the Bogs Dollox! ...

yea we have to keep the low torque issue in mind when designing the inpeller, we need to take advantage of the strength of the motor, that is it's speed.

infact it's fast enough to use a multi disc impeller, no blades, only stacked plates with a small gap between them (with a common central hole) though only 2 discs in this case because of the low torque.

[nicozeg]

Check this out, it's a review of that cpufx crap water cooler, but it shows the detail of the 12v pump design in detail (also based around a 12v fan motor)

Very detailed review; but unfortunately we all know that "optimal" is a word the designers of that crap don't know at all.

GRRRRR, Looking at that makes me anger, how people with all that manufacturing resources can waste them in such way!

Do the vanes cross over in the middle or are they around the edge only like a centrifugal fan?, I'd guess they have a hollow space in the middle...

Did'nt understand what are you saying

I seriously doubt this will have the power/torque needed nico. The one way action with an offset outlet does'nt make that much difference. You'd need a 20watt+ fan to start with I recon

This is only a test, I'm not trying to build a high performance systen arround this. But I hope this pump can perform similar to an eheim 1046.

But your epoxy work is the Bogs Dollox!

Thanks!

My interest is to get the highest head possible from this pump, that makes me think in iwaki pumps. They come in two types, high flow or high head; I guess the difference is only in the impeller.

Volenti, have you looked at your iwaki's guts? maybe is time to open it for some cleaning

[bikr]

nic , i'm kind of confused. do you have more pictures of possibly the assembly going into the ring. Wouldn't the silicone not allowe it to spin freely and be a problem.. Obviousley not from your picture of the whirpool haha but let me know if you have more shots of it and where i can find them.. thanks --Josh

[bikr]

nevermind i think i'm getting it now .. you used a hard seal epoxy and then used the lathe. I just had to stare at it for a bit , lathed it to fit inside the magnet i gotcha.. still want more pictures though --Josh

[MadDogMe]

yea we have to keep the low torque issue in mind when designing the inpeller, we need to take advantage of the strength of the motor, that is it's speed.

But will it reach those speeds without the torque?, once the waters 'load' is added?...

[Volenti]

Originally posted by MadDogMe
But will it reach those speeds without the torque?, once the waters 'load' is added?...

yea for sure when it's loaded up the speed will drop, but if it's doing 9700rpm with the bare hub in the water then if we aim for around 4000-5000rpm loaded up then we should be able to pump a reasonable amount of water.

here is a concept drawing of what I think will work;



blue is water passege, green is impeller , red is the 12v motor, the impeller is made up of 2 discs of 2'' in diameter, and about 1/12'' thick, with a similar gap between them, the disc closest to the intake has a hole in the middle, make this around the same dia as the intake, that disk can be attached to the main disk by a couple of pins or similar.

edit, forgot to add, this design is based on the principle that a centrifical pump works by spinning the water in the impeller chamber, which then forces it's way out of the tangental exit of the pump, which creates a vacume that sucks more water into the intake.

edit again, the impeller assembly of the high head iwaki pumps is very similar in principle to this, it's where I got the idea from.

[nicozeg]

That's called a tesla pump, was menctioned previously in this post, but did'nt have any reference to it's flow/head performance. I'm gona try that, seems easy to make.

[Althornin]

Originally posted by nicozeg
That's called a tesla pump, was menctioned previously in this post, but did'nt have any reference to it's flow/head performance. I'm gona try that, seems easy to make.

It supposedly will also move air, so you can test it out of water.
I dont know if the gap between the plates must be adjusted for air performance or not, nor what the optimal value for the gap is for water pumping.

This book looks very interesting:
Tests with various disk spacing

This site might also be usefull:

Tesla Turbine Club

[nicozeg]

Originally posted by Althornin

This book looks very interesting:
Tests with various disk spacing

This site might also be usefull:

Tesla Turbine Club

Good links, I didnt know too much about tesla, but seems to have several "fan clubs" arround. A simple search gives tons of info. I even found links about making pumps with cdroms!

[MadDogMe]

Any idea how they 'stack up' [doh!] compared to a simular wattage powered 'finned' centrifugal pump?. Is there an advantage to Tesla pumps?(can you stack as many 'discs' as you like?), for some reason I'm guessing that 'finned' impellors will give more torque/head/pressure than a tesla, but a tesla might run faster with a 'light load' scenario. What do you think?, or KNOW!? ...