Small water cooling pump test

As some of the overclockers are looking for higher power density but are just running in to the motors thermal limitations, or if your dirt bike mates mock you ozone smell after attempting the first local big widow maker hill, or your just pumping way too many amps into the the crap motor you happen to love torturing. I thought something like this with a suitable radiator might help to prolong its pain. I did a quick test with my meanwell and icharger set to 12 volts and it pulled about .25amps or 3 watts total, it is a brushless coffee machine pump and all the parts are potted in clear it is very small, it weighs only 25 grams and is virtually silent when running.

http://www.ebay.com/itm/DC-12V-Mini-Electric-Water-Coffee-Pump-24-GPH-P-25A-/260895901599?ssPageName=ADME:L:OU:AU:3160

http://www.ebay.com.au/itm/ws/eBayISAPI.dll?ViewItem&item=120974391787&ssPageName=ADME:L:OU:AU:3160

I haven't got the answers on how to help you liquid cool your megawatt crystalyte hub or roasted rc bike you just have to use your imagination. But 1.2 liters of water a minute can remove a lot of heat!

Splinter oz put me on to this (i wonder what its for?) Thanks

Zappy

A lot of R&D for watercooling electronics has been done by the PC over clocking community. It might be worth seeing what you can cannibalise from their efforts!

That's a great little find. I would never have thought of looking for a coffee machine pump! That's what I love about this forum- the collective mind problem solving ability...

That's a cool little pump. I need to pick up a handful of those for various projects. 12V, that small, and 1.2L/min is a sweet little brushless motor pump.

"1.2 liters of water a minute can remove a lot of heat" . I came up with 84W per degree of temperature differential.

Liquid in the hubbie like the oil guys are doing is definitely the way to go for transferring the heat from the motor to the liquid. They just need to stop leaving the heat inside the motor. This little pump and a large ID bearing for room to send it out to a radiator may be just the ticket. So what if it takes more than one at only 4W and so little space.

thanks for the link. I was looking for a sub-$100 pump. Don't know if this is high enough flow rate for cooling 6-8kW continuous Cromotor through water cooling channels though..

yeah. 24 GPH does seem a bit slow. Isn't the best PC pump like 300 GPH?

See this thread for my cromotor water cooling project. It should resume within a couple months afte rI finish 2.5kWh A123 battery
http://endless-sphere.com/forums/viewtopic.php?f=2&t=43142

teaser:
[youtube]m_GofIWB1Hs[/youtube]

it looks like a small window wiper pump from the car...

it looks like a small window wiper pump from the car...

Click to expand...

Hi Labowski, its probably about half that size and brushless. I think you could use a windscreen pisser motor but i think they are pretty noisy and are not meant for continuous duty.

They just need to stop leaving the heat inside the motor. This little pump and a large ID bearing for room to send it out to a radiator may be just the ticket.

Click to expand...

Exactly John!!
When it comes to choice of to liquid or air cooling, maybe using both is the best option.
A standard magic pie can only dissipate around 300w of heat at 5km/hr all sealed up, or around 1000w at 75km/hr with lots of holes. Max 150 degree C winding temps. With around 1000w of heat at 5km/hr your up shit creek. If you look at the water flow rates of some performance motor bikes/cars etc they remove a LOT! of heat with fairly low liquid flow rates. and it can all be fairly small compact and light.
Eg. Yamaha YZ450 30kw at 20% efficiency maybe 1/2 of the 150kw worth of heat from the fuel burnt goes out the exhaust the rest is dealt with the air flow over engine casings etc. (Maybe as much as 3 times the surface area of a hub motor??) and some pretty small radiators.
There is a lot of copper surface area in the windings to give up the heat but you need a lot of air flow but it can deal with almost a kw of heat for virtually free, and maybe a small radiator could deal with the rest as we don't have an exhaust pipe to pollute out the rest of our heat.
Radiator 225g
pump 25g
water galleries and water etc 500g?
Maybe double potential continuous power?

Don't know if this is high enough flow rate for cooling 6-8kW continuous Cromotor

Click to expand...

well our cromotors make a lot less heat than my pie at 6-8kw in. Any way hillzofvalp the pump is the easiest thing to play with you have done all the real hard work. Well done!

Zappy

Just been through all this with the greenpower racer. The next step forward in car performance will be down to using motor dissipated heat to warm up the batteries. Unfortunately on our car we have the batteries in the nose of the car & the motor at the back so we tried water as the medium to pipe the heat around.
1) what water flow is needed? Lets say we have 180W dissipation. And let's limit the coolant temperature rise to just 1 degree. Specific heat cap of water is 4200J/kgK 180/4200 is 43 grams of water per second. If we allow a more reasonable 10 degrees rise we need just 4.3grams per second; that's about 1/4litre per minute - probably a lot less than you'd think...
2) How did we get heat out of the motor into the water? Greenpower racing has a long tradition of teams failing to manage this adequately & burning out their motors.... Wrap a copper tube round the motor? no good - actual contact is only a 'point' just here & there. OK try plastic tube. This stretches & deforms to give intimate contact over quite a large area, but still no good - it's thermal resistance is way too high. I was left with the realisation that this would only work if the water were actually in contact with the motor housing. So attempt 3 was an ally water jacket clamped to the cylindrical motor with big hose clips (duct clamps) and sealed with silicone compound. Did I say sealed? water came out everywhere...... Final attempt, used a thick rubber gasket (3mm thick). The ally jacket was just a flat sheet bent round the motor and the water space was made by cutting out the middle part of the gasket. This worked perfectly - total success, the car easily won its first race by miles & we were taking batteries out 10 degrees hotter than we put them in. Happy ending - 'fraid not; it turns out "water cooling" had been banned by greenpower several years ago & we hadn't noticed... so we were disqualified
We'll try to use air to pass the heat through the car in a closed loop for next year's races, but air is useless compared to water for this purpose. So next new car will be front wheel drive.... I think we'll get some aero benefits from that too, time will tell. Whatever - I hope our experience is of interest to some

Thanks for reporting on the pump, I'd seen them on Ebay and wondered if they'd be any good.

What About good cheap radiators? I know the ones you used, keyne, but are there any others to consider? What volume radiator would I need to cool a motor that will at times produce 7kW of heat? Or say 200Wh over a time period of 30-60 minutes? Differential equations are not compatible with my frontal cortex

The radiator size almost doesn't matter, but bigger is better, so use whatever fits nicely on your bike. It doesn't matter because it'll come down to creating a good thermal connection between the windings and the coolant, rather than how much you can cool the coolant.

Don't get me wrong, obviously there are a lot of factors at work here, but if you want to give one of them a lot of attention give it to the thermal connection between winding and coolant. The others are easier to change anyway.

The radiators in the first post are only 158x120x30 and about 20 bucks... I would imagine they would do a reasonable job.

For a radiator, it may be worth experimenting to see if you can conveniently use some of the smaller pieces of tube that are part of the bicycle frame - I'm thinking rear triangle here. It should be relatively easy to plumb. The bigger tubes might become a bit too heavy full of water...

Hey Zappy,

Would you be kind enough to let me know what the ID and OD of the in and out pipes are?

Thanks,

D

I'm missing something. Why use a fluid system to move the heat from motor to cooling fins, when you could put the fins on the motor. Where they would see greater air speed than on the frame. I don't imagine the increase in moving mass would be as great as a water system.

If cooling the casing is not good enough for heat spikes, then pumping the internal fluid out to some fins seems to come next. I'm not sure about through the bearings to the frame though. Why not spin the pump and fins round within the wheel, it seems a lot less work. Certainly for proving the concept.

That is the smallest diaphragm pump I have ever seen. I might get one to. As an ornament

Hey friendly1uk,

friendly1uk said:

I'm missing something. Why use a fluid system to move the heat from motor to cooling fins, when you could put the fins on the motor. Where they would see greater air speed than on the frame. I don't imagine the increase in moving mass would be as great as a water system.

Click to expand...

The thermal coupling from the stator to the case of a hubbie is very poor. If you are running water around the stator and have done it properly i.e. have low thermal resistance between your copper windings and the water, you will be able to remove significantly more heat.

friendly1uk said:

If cooling the casing is not good enough for heat spikes, then pumping the internal fluid out to some fins seems to come next. I'm not sure about through the bearings to the frame though. Why not spin the pump and fins round within the wheel, it seems a lot less work. Certainly for proving the concept.

Click to expand...

For my application it is not about the heat spikes, more about total heat removal. That is increasing the constant power level of my motor. I am running a RC motor and not a hubbie. Although the principles are no different.

All a bit off topic but I hope this helps :|

D

Would you be kind enough to let me know what the ID and OD of the in and out pipes are?

Click to expand...

The dimensions are on the link. Inlet 9mm , outlet 7mm
Zappy

Thanks Zappy,

I did check the link but didn't scroll down enough

Thanks, appreciate it.

Regards,

Danny

I am building a bike and using 2 x 120mm radiator from pc. Cant remember how much I have paid but it wasnt that much. I will cool my 12 fet controller, just to see how effective it is.


I am using pc pump but this little pump shure beats it by a mile:
[youtube]OhtzsJYEHNw[/youtube]

And controller:

Wow, beautifully executed agniusm.

Thanks for sharing.

Good luck with it!

$15 on this link

http://www.ebay.com/itm/Mini-Water-...all_Kitchen_Appliances_US&hash=item3a603d23d9

D

Rad!

Are you logging temperature? What's the suggested upper limit for the mosfets temperature wise?

I'm logging temperature with a 10k ohm thermistor with a ring terminal housing. This then is mounted to the screw holding the mosfet to the heatsink. Just curious what others are doing....

I just purchased a MXUS hub motor through the group buy and am planning on incorporating water cooling. I just ordered the replacement axle and cooling line setup from Linukas. His setup looked fairly simple so I thought this would be a good starting point. I am thinking about using the thermistor in the motor to control the cooling, by cycling the pump on and off, but not sure if it would be worth the hassle or not. It could save a little power per ride, but maybe not enough to make it worth spending the time to make the control circuit. A simple on off switch may be easier.

Not sure if cycling pump on and off or just running it all the time is better. Running it all the time would make the motor run more efficiently which may offset any gains by turning off the pump.

Just talking out load here. Tell me to shut up if its not relevant.

Ed

Danny Mayes said:

Hey friendly1uk,

friendly1uk said:

I'm missing something. Why use a fluid system to move the heat from motor to cooling fins, when you could put the fins on the motor. Where they would see greater air speed than on the frame. I don't imagine the increase in moving mass would be as great as a water system.

Click to expand...

The thermal coupling from the stator to the case of a hubbie is very poor. If you are running water around the stator and have done it properly i.e. have low thermal resistance between your copper windings and the water, you will be able to remove significantly more heat.

friendly1uk said:

If cooling the casing is not good enough for heat spikes, then pumping the internal fluid out to some fins seems to come next. I'm not sure about through the bearings to the frame though. Why not spin the pump and fins round within the wheel, it seems a lot less work. Certainly for proving the concept.

Click to expand...

For my application it is not about the heat spikes, more about total heat removal. That is increasing the constant power level of my motor. I am running a RC motor and not a hubbie. Although the principles are no different.

All a bit off topic but I hope this helps :|

D

Click to expand...

Is your motor not spinning then?
If it were, I would oil fill to move the heat from the stator to the case, then fin the case. I see no reason to add a pump and pipework to move the fins to somewhere with half the air speed. The extra fluid volume involved with doing so would be good for transients, but no better for long term power increases.

I think I meant to say this before, but it sure didn't sound like it.