Using Regen for cooling hub motors: Refrigerants or Peltiers

[Cowardlyduck]

Hi all,

So I had an idea to use the Regen current from our BLDC motors to cool them instead of sending the energy back to the battery. That way the motor gets nice and cool before you hit it with current again for the next take off. Could even be a better use of the Regen current than putting it back to the battery, I dunno. i.e. the efficiency gained from having a cool motor could be greater than the extra kms/miles gained from Regen to the battery?

There are 2 methods I can think of that may work, with the second being the best IMO.

1. Using a very small DC refrigeration compressor with the cooling lines routed inside the motor. This would be very complex, big/bulky and difficult to implement, so probably not the best idea, but may prove beneficial if anyone were willing to give it a go.
I can find several links to small compressors that might do the job.
http://www.danfoss.com/BusinessArea...s/DS_Direct_Current_Compressors_R134a_48V.htm

http://www.made-in-china.com/showro...r-for-Rooftop-DC-Air-Conditioning-System.html
One of the advantages/nice features if this method is the narrow part of the copper tubing (i.e. the end of the compression stage) should easily fit through the axle hole. Only the exiting tubing would still need to be of a large diameter.

2. Use Peltiers attached to the covers with small heatsinks. For those that don't know, Peltiers are a form of Thermoelectric cooling using 2 different metals. Read up on it here. http://en.wikipedia.org/wiki/Thermoelectric_effect They are often used in extreme overclocking, usually in conjunction with water cooling. They can apparently cool to 50C below ambient.
The idea is to send all the Regen current through the Peltiers.
It should be really simple to implement for someone in the know. I can only see one main difficulty. How to send the current to the Peltiers when they are mounted on the spinning side covers. Brushes maybe?
They are relatively cheap to.
http://www.frozencpu.com/products/2408/exp-01/245W_Potted_Peltier.html?tl=g30c105s187&id=JERzHGZs

2 of these secured to each side cover along with some thermal paste and short heat sinks that don't get in the way would give nearly 1000W cooling capability. The Peltiers above could only be used if the Regen voltage could be limited to ~15V. I don't know if higher voltage Peltiers exist, or if these one's can take higher voltages.

For both methods a way to switch/send the current from the motor to the cooler would need to be figured out. May require an additional line from the controller?
Would there be an easy way to send excess Regen current not being used by the cooler to the battery as well?
What are others thoughts? Anyone with expertise/experience doing similar mods willing to try either method? I would try it myself, but have limited experience/time/resources/money etc, so just offer my ideas for others to try/criticize/like etc.

Cheers

*EDIT* Actually if you ran 4 of the above Peltiers in series it could work with a 48V battery/controller.

 

[Hugues]

hi,

about peltiers,

been discussed here and there, they are not good heat sink actually. And if you make them to become good heat sink, then they become poor at producing electricity. You just can't get both. search a little around, you'll find those discussions with more details.

 

[amberwolf]

Look up Peltier in the ES search or with google site search, and you'll see some previous discussions showing their inefficiencies and why that probably wouldn't help enough to matter on a hubmotor.

You'd also need some pretty big heatsinks to help dissipate the heat they generate, along with the heat they are pulling out of the covers.

If they are on the covers, they are not going to do a lot to remove heat quickly from the motor itself. Passing the current to them would require brushes and some form of commutator, adding a fair bit of complexity to your wheel.

To make the Peltiers properly cool the motor you'd really need to run them off of more than the regen current, too, as that is simply not enough, unless perhaps you are only using them for really long downhill braking sessions, where the regen itself is what creates the heat you need to dissipate. Personally I'd rather be using that power to recharge my battery after all the power that had to be used to climb the hill or mountain in the first place.



I'm sure the refrigeration system would help, but you would need to plumb the freon(or whatever) system thru the axle and into the stator. it would be very vulnerable to damage from crashes/slides on the side of the bike, unless you built something around it to protect that. It would also be necessary to do some interesting work to make the wheel removable in the field for repairs.

You could use a liquid cooling system with a heat transfer into the

It would also be pretty power hungry. Running the little compressor in my tiny window AC unit to keep the room cool takes several KW of power in a day. I don't have a KillAWatt meter to check it's actual consumption, but it is pretty hefty. I'm not sure how much power a smaller unit intended to be powered from DC would use, but I doubt you could power it from regen--you would ahve to have a separate pack just for it, or you would have to sacrifice significant range to using it.



It would be much better simply to use a more efficient motor system for the terrain/path taht you need to ride the bike on, so it doesn't generate as much heat. Then use the regen to recharge the pack.

 

[Cowardlyduck]

Ah well...was worth a try.
Didn't expect this would have been covered before...shoulda searched, sorry.

So sticking to the refrigeration method...
What about directly attaching the piston for compression to the motor and using a clutch? You could also just use air, not refrigerant, to simplify things.
Then rig up a switch, linked to the brake to drop the clutch and engage the pump, compressing air into a small external chamber when needing to slow down. Then when you take off again, all that air could be dumped/pushed through the motor. Not only will the moving air take away heat, but it's expansion will make it nice and chilly.

Haven't searched, but I doubt that's been thought of before. What do others think?

Cheers

 

[amberwolf]

Haven't seen that one, but it sure sounds like a lot of work for a little bit of cooling.

 

[John in CR]

The beauty of electrics is that they don't need all those complications. A well executed passive ventilation approach works fine Most of those with heat problems on hubmotors simply don't have enough motor for what they're trying to do. Typically the wheel size puts the gearing too high, and guys exacerbate the issue by running too much current which pushes the motor well into saturation resulting in higher peak power but killing efficiency so heat generation skyrockets.

If your riding includes lots of very low speed accelerations, whether it's picking your way through difficult trails off road or even stuck in stop-n-go traffic, you may need to add some low power fans or blowers inside the motor to both increase flow through the motor as well as create more cooling turbulence inside the motor. I've had great success with passive ventilation, and Zappy and Arlo1 have had great success including fans in their strategy. All 3 of us run extreme power with no heat related failures.

The latest fad is simply adding oil inside a sealed motor. It has some issues, but helps those running modest power levels have a cooler motor. Without some kind of pump and exterior radiator, it can't enable a hubbie to go to extreme power, because all the heat still has to be transferred to the environment from the surface of the motor shell. Adding alot of surface area to the AL covers would increase that limit, but no one has done that yet.

John

 

[Sancho's Horse]

I am working on an oil cooled hub with a passive radiation system which dramatically increases surface area. However, my hubby is a Chinese scooter hubby literally on the slow boat from China. Slow boat is only valuable for volume shipping, but I anticipate volume for other non-bike related business, and working through the kinks in the chain is more valuable to me than ES chest thumping.

However, on topic, I am some day going to use the heat to run a small Sterling engine, but that is mostly for demonstration purposes, and an excuse to buy a lathe. I really do think there is a good market for a micro generator, and suddenly a little propane turns your hybrid into a long distance, smaller battery, serious utility vehicle.

 

[bionicdan]

When I looked at the price of those small peltier plates it said if there is enough difference between ambient and the item itself it can PRODUCE 5v 1amp output. That would be cool if you could utilise heat into added range somehow.

I believe f1 and car makers have toyed with the concept due to car producing about 90% heat over motion.

One thing I think I read that would stop you wanting to use them under regen only is that they reverse direction when powered off. So when you turn the little portable car fridge off it draws heat right inside the fridge as fast as it cooled it. I hadnt observed this with my peltier fridge but sure it was in the manual warnings.

I really might do an electronics course so I actually get this stuff properly lol

 

[amberwolf]

Well, at least with teh little fridges, the big problem is that there's no "one way valve" for the heat, and there's this giant heatsink on there that can soak up heat from ambient air, then pass it back into the much colder interior fairly easily. If you used a liquid-coolant pump that shut off when the Peltier was also off, and the heatsink is not otherwise connected tot he interior, this heat transfer would happen much more slowly, as it must first warm up the liquid thru the whole tube to get from heatsink to interior; the longer this is the longer it will take, assuming the tubing is itself insulated and not heat-conducting metal.

It also generates a voltage across the Peltier while there is a difference in temperature, but unless the circuit controlling it is poorly designed this shouldn't cause it to help the heat to move back inside (and maybe not even then; I'm not sure).


With a hubmotor that is already much hotter than ambient air, I don't think you'd get any noticeable heat transfer back into the motor from any heatsink it was attached to, even if that setup was directly bonded to the motor rather than first passing thru a liquid heat pipe. It should in fact continue to pass heat out of the motor, even with the Peltier shut off, simply becuse of the temperature differential.