Definitive Tests on the Heating and Cooling of Hub Motors

I have seen what heat pipes can do for mold cores that were running much too hot and no other way to cool them.
I think they would be very good candidates aside from cost and having to possibly re design the core spokes to accept them.
heatpipe06.gif

http://www.dme.net/catalog/DME/MUD catalog/MUD Catalog/assets/basic-html/page153.html
 
izeman said:
120C at the bearings? Did you have no temp probe installed? Windings must be cooked already at those temps.
Click to expand...
windings cooked at 120C? nahhh...

did you read the thread justin linked to at the beginning of this tread? there he posted:
justin_le said:
At this point, after 45 minutes of running a continuous 30 amps into an eZee hub, I could see that the torque was starting to plummet fast, meaning that the rare earth was demagnetizing, and the smell of "somethin' cooking" started to hang in the air.

I had to let the motor sit for about 10 minutes before I could even handle it to open up since it was too hot to touch. When I did get the side cover off, white fumes emerged for a few minutes. The windings were definitely darkened, but not burnt to a crisp. The grease in the ball bearings had fully liquefied and some of it vapourized. About 20 minutes after I stopped the test I finally got an IR thermometer on the windings and they were still nearly 150 degrees C.
file.php

So for the first 10 minutes or so, there was basically no effect that the heating had on the output torque of the magnets. But then from about 10 minutes to 40 minutes, there was a pretty steady decline, starting at 38 N-m and falling to about 34 N-m. This as the winding temperature reached over 200 Celsius. Just after this point, it seemed that the torque was about to fall quite a bit more rapidly, which coincided with things starting to burn up on the inside, so I shut off the supply not wanting to destroy the hub completely.
Click to expand...
looks to me like it reached closer to 230C!

no one recommends hub motors get hotter than 80C, but in any case, it'd be useful to know the rated and actual temp limits for various components, such as, for example:

rated/actual
bearing lube: 80C/100C?
bearing seals: 100C/120C?
halls: 150C/200C?
windings: 200C/240C?
 
GCinDC said:
izeman said:
120C at the bearings? Did you have no temp probe installed? Windings must be cooked already at those temps.
Click to expand...
windings cooked at 120C? nahhh...
Click to expand...
no. i meant: if your bearings reached 120° then the windings would have been cooked. or at least near that and be quite darkened.
my windings reach 120°C, case is 55°C and the axle is still at 30°C - i guess the bearings by then should be something between 30 and 55. so if your bearings reach 120°C - what would the windings temp be? as justin stated: 200°C+.
what is a safe temp for windings? i thought 150°C was the max for long term safety.
 
izeman said:
no. i meant: if your bearings reached 120° then the windings would have been cooked. or at least near that and be quite darkened.
my windings reach 120°C, case is 55°C and the axle is still at 30°C - i guess the bearings by then should be something between 30 and 55. so if your bearings reach 120°C - what would the windings temp be? as justin stated: 200°C+.
what is a safe temp for windings? i thought 150°C was the max for long term safety.
Click to expand...


Same here ;)
normal use: 120 Windings - 55/60 Case - Axle30/35
hill climbing: 160 Windings - 60/65 Case - Axle -/-


Motor runs fine. My Speedtest was after that torture.
(My Andersons were Melted after that last ride)
No Problem....Motor Pulls like a Bull in 28" Rim at 8Kw :p
 
GEDC0308.JPG
These are pic's of my 3525 at 80v 50amps. I wasn't on it thta long maybe at 10 min. at wot. And things got hot. The second pic shows the nylon string melted. When I open it up I found a cut on the yellow phase wire. It didn't look like it had shorted yet but i could be wrong. As it looks like every third winding looks as the string brunt there. The motor does test o.k. on the lyen tester. But it looks like the solder on the pigtail of the phase wires got over heated should I resolder the pigtail ?View attachment 2 Plus I think I will grind more room under the spot where the seal goes for more room for the phase wires as not be worn by the seal ect.
 

Attachments

  • GEDC0309.JPG
    GEDC0309.JPG
    74.9 KB · Views: 5,398
  • GEDC0313.JPG
    GEDC0313.JPG
    56.2 KB · Views: 5,398
liveforphysics said:
Cowardlyduck said:
I had the same thought about Peltiers a while back. Although my idea was to use the Regen to power them, that way it wouldn't require additional power.
http://endless-sphere.com/forums/viewtopic.php?f=30&t=44400#p646304

Still think it would be worth trying.

Cheers
Click to expand...


Peltiers have no place on a hubmotor. You will not make a hubmotor cooler, but rather hotter, and a lot hotter.
Click to expand...


Peltiers, AKA Thermoelectric Modules, are excellent heat pumps. To install them on the exterior of a hub motor would pump heat from the hub to the ambient air, thus cooling the hub. The energy they consume might make them less than favourable, but they would certainly do the job. The difficult part is powering them while they are on a spinning motor.

Having said that, there is an air-bearing cooler being developed that demonstrates the ability to cool a computer processor or high-power LED through a thin layer of air, and it's quite effective. https://share.sandia.gov/news/resources/news_releases/cooler/

This might have some relevancy, were the hub motor to be designed to include a much more precise clearance between case and stator.
 
carbon.nanotube.capacitor said:
Peltiers, AKA Thermoelectric Modules, are excellent heat pumps. To install them on the exterior of a hub motor would pump heat from the hub to the ambient air, thus cooling the hub. The energy they consume might make them less than favourable, but they would certainly do the job. The difficult part is powering them while they are on a spinning motor.
Click to expand...


No, they wouldn't actually, the motor would just run hotter. Learn about them enough to do an energy transferred vs energy they add themselves and you will understand.

I will waste my time to show you.

Lets say you have a motor you're running 2500w into, and it's 80% efficient. This means to not continue to climb in temp, you require 500w of energy leaving this system (because that's the same rate you're adding thermal energy to the system).
Say you target a best optimized peltier solution for this, you're going to be looking at 8% efficiency while it's in it's best optimized temp range. To transfer that 500w of waste heat, you require a peltier array that will consume ~7,100w, and now your total thermal budget to reject has jumped from being 500w to 7,600w of heat to shed (which is virtually impossible for something the size of a bicycle wheel to do without exceeding the solder-melt joints on a peltier junction anyways).

And finally, say you can deal with this 15x heat increase and the reduction of system efficiency from ~80% to ~21%, this peak efficiency temp number actually only happens in a Peltier while it's holding a uselessly small DeltaT.

You waste money, space, efficiency, weight, and end up making your motor melt faster than ever before. Just like so many misguided folks doing peltier CPU cooling early on and actually making a CPU heater rather than anything useful to cool. This doesn't mean you can't effectively cool a CPU with a peltier, it just means the people putting say 100w peltiers on a chip with say 25w Pd were simply adding heaters, while the folks adding 300w peltiers to chips with say 15w Pd actually did manage to get a big better delta-T between core die temps and ambient than air-cooling alone could provide, albeit at a steady-power demand that is likely greater than the entire rest of the system.
 
Epic thread Justin

What I am most interested in (being the hack that I am...) is a motor with drilled side covers and a small blower mounted to the frame blowing into those holes (picturing back at an angle to work with the wind).

Any guy can get the covers off of a motor and get them drilled by hand... and any guy can get a cheap small 12V blower and hook it up to a little 4S pack.

(This next comment is sort of aimed at John in CR to deflect any debate)
I personally dont want to get into any of the debate over internal fins... as that sort of work is beyond the capabilities of myself and many DIY folks. I am simply interested in comparing the effectiveness of a certified "methods ghetto hack" that I think could have comparable cooling effects for much less work and complexity. Yes I totally get that the motor is already turning and it is a fan and all that.

I am thinking about a 20,000rpm screamer blower, over-clocked by a 4S pack, hooked up to a temperature control that just friggging BLASTS air through the holes (when needed). Noisy yes. Inefficient yes. Clunky yes. Simple yes. Easy yes.

How does it stack up? No clue... that is a question for the man with the test setup.

Easy test. :wink: :wink: :wink:

-methods



Edit: My definition of simplicity is that which is simple for me (an Electronics guy). I totally get that simplicity for a mechanically inclined fellow would be to build something mechanical. So I empathize with the idea that some people think that what is simple for me (temperature controlled fan) is actually complex.. and that internal fins are simple (which to me are complex to implement).
 
methods said:
Edit: My definition of simplicity is that which is simple for me (an Electronics guy). I totally get that simplicity for a mechanically inclined fellow would be to build something mechanical. So I empathize with the idea that some people think that what is simple for me (temperature controlled fan) is actually complex.. and that internal fins are simple (which to me are complex to implement).
Click to expand...

Too complex for me. For me, I have difficulties building fins and I have difficulties building a temp controlled fan. I would probably end up asking Justin to implement a 12V fan controller into the CAv3 Prelim 74 or something.....

I can drill holes. I just need some recommendation what kind of holes and where to put them. I can also point a fan somewhere. Whether that's an internal hubfan or an external one, I need guidance as well.
 
Peltiers are awful heatpumps in efficiency terms - listen to LFP. And on a scale too small for the heat we work with.

I do wonder if aluminium flywire sheets aren't of some use externally, or internally to aid heat transfer to a fluid such as ATF or peanut oil (high burn temp). They are a cheap form of heatsink that is easily cut with scissors and moulded to shape, unlike the costly machining of traditional aluminium heatsinks. Those solar popcan heaters are tending to be made of mesh these days for the same reasons.
http://www.builditsolar.com/Experimental/DeepMeshCol/120116Test.htm

My AUD$0.02.
 
hjns said:
... and I have difficulties building a temp controlled fan....I can drill holes. ... I can also point a fan somewhere....
Click to expand...

Not to get too far off topic - but I should not have even mentioned the temperature control issue. All a guy needs to do is find a 12V blower and hook it up to a 4S Hard pack. No temp control needed. A typical fan will draw maybe an amp - so a 4S hard pack should last 4 hours+ (long enough to outlast any main pack under heavy load). The higher 14.8V of a 4S pack gets a lot more umpf out of a 12V blower.... and you can just let it run for the whole ride.

Not knowing the outcome of the test that I hope Justin will do.... here is what my blind (never tired it) advice would be

* big holes, lots of them
* big fan, blowing lots of air
* Lots of zip ties
* wear a full face helmet
* Buy an IR temp gauge (to measure temp through the holes) or instrument a CA-V3 with 10K NTC.

That is the methods recipe for success :mrgreen:

-methods
 
Hi Methods, something like thishttp://www.bunnings.com.au/products...ryobi-18v-one-unit-only-cbl1802_P3380380.aspx is maybe what you are envisaging. Toolman2 uses one of these to get rid of a couple of KW of waste heat from his large RC out runners. It is very noisy, and would probably wear a bit thin after awhile. These blower type fanshttp://www.jaycar.com.au/productResults.asp?MID=1&SSUBID=722&SUBCATID=981&keyform=CAT2#1 are a lot quieter and you can just series wire 4 or so of these together to run off the mother ships pack (I over volt my fans a bit and they work even better) . They would flow over 100cfm(as much as a vacuum cleaner) and are quite light. They are also available at different voltages from other suppliers. You could mount a stationary disk inside the side plate so air pressure does not back flow out your side ports that are open or make a plenum to cover all the side ports(getting less easy now)
I did put up some instructions for making some vanes at the bottom of herehttp://endless-sphere.com/forums/viewtopic.php?f=30&t=48518. Maybe some one can contribute a design for the popular nine continent/ cro motor covers? They are not that hard to make, these were a simple design so you can just dock them off with a drop saw or just use tin snips. And are only ever meant to be a reasonable improvement at commuter type speeds.
If we seek better power density we need to get a lot more active cooling. How much power can your car motor constantly produce if you cut the radiator hose?
Maybe 1/20th, maybe not even enough to turn its massive frictional losses over without seizing?
Is duct taping /zip tying 1 KG of shrieking blower improving power density? Or would we better spending that weight on extra copper? air gap diameter? reduction gearboxes? water cooling? or All of the above?
Yeah, thought so.

Zappy
 
Many of us seem to be forgetting the kiss principle.

Having done all sorts or tests on trying to cool my motors, I reckon that air cooling your hub is a great idea and Justin has done a good job at trying to quantify which holes work and which don't.

Oil cooling is messy and a pita. Just about the only solution for a geared hub like my little mac. Keep it simple and just do a oil splash system. Very effective but no matter how well you try, still suffers from little leaks.

Bottom line- fans, peltiers (one of the strangest suggestions I've seen and no they wouldn't work), people running alongside with a leaf blower pointed at your hub etc are just making things unnecessarily complicated.

If you air cool your 9C and it still gets hot, then buy a bigger motor. Cro or scooter hub. If they get hot, air cool them. If they get hot still, then maybe it's time to go to an electric motorbike with proper brakes, suspension etc and a sorted powertrain.
 
Spicerack said:
Many of us seem to be forgetting the kiss principle.

Having done all sorts or tests on trying to cool my motors, I reckon that air cooling your hub is a great idea and Justin has done a good job at trying to quantify which holes work and which don't.

Oil cooling is messy and a pita. Just about the only solution for a geared hub like my little mac. Keep it simple and just do a oil splash system. Very effective but no matter how well you try, still suffers from little leaks.

Bottom line- fans, peltiers (one of the strangest suggestions I've seen and no they wouldn't work), people running alongside with a leaf blower pointed at your hub etc are just making things unnecessarily complicated.

If you air cool your 9C and it still gets hot, then buy a bigger motor. Cro or scooter hub. If they get hot, air cool them. If they get hot still, then maybe it's time to go to an electric motorbike with proper brakes, suspension etc and a sorted powertrain.
Click to expand...

im not sure we just "forgot" to keep it simple,
we have tried motors up to 15kg that completely dominate the feel of the bike and still only produce about half of the torque we need, whilst overheating cos it turned around half of the batteries energy into heat, -not sure which direction we should head there really, bigger motor agian? :roll:
total kiss returns you around 200w of continuous output power per kg of hub motor, just like in 1880 when engines had no active cooling you could nearly get a couple of horsepower from a 200kg engine.
i agree, Volkswagen went completely mad introducing ridiculously complex forced fan cooling in the beetle in about 1950, then it just got silly with liquid cooling and actively computer controlled electric fans, that all drivers have to put up with today.. :lol:
with an effective cooling system you can get motors up to 5000w continuous output per kg, so maby have another quick look at what power and weight you want on the the rear axle of your bike?

i consider my slackarsed $59 leafblower cooling setup to be crude 1950's style technology, but even the 100w it uses (and only when at full tilt) to be quite ok, cos when you pin it to say 5000w in to a hubbie off the mark you get over 4000w of ir loss, now with cooler copper you will save over 1000w :!: -good plan methods.

so to me keeping it simple is completely optional for everyone, and yep these tests are right on the mark to then very simply and easily improve your average hub motor out for a huge number of people.
 
liveforphysics said:
Yet, on the other end of the spectrum, even the x5304 I used on an early gen of deathbike raced at the Tuscon deathrace wasn't nearly big enough to handle the 400amp 120v setup I was hitting it with. It just melted down in 3 laps in a smoking mess of disappointment, yet in Arlo's BMX, the exact same type of motor delivered a thrilling overly-wheelie-prone experience that was like the ultimate cross-town short wheel base maneuverable hard accelerating vehicle to rip through traffic with an it had no cooling issues at all in this usage model, so being larger would just be more weight for nothing but a tiny bit more range perhaps from efficiency improvements, but overall not worth the extra 5lbs to carry.
Click to expand...
You forgot I was able to push it like that all the time because I have 5 little fans blowing strait on the windings inside at 2x their rated voltage! The fans were tested with and with out one night for a blast strait up a steep hill for a ride that was about 2 min long and they made a huge difference. Its in the Cooling X5 thread some where....
 
I just want to reiterate that this thread's conclusions are important to me. The most frustrating thing with my ebike rides are the hills in my home city. Too much hill, too much distance, inevitably results in a shutdown of my motor - a Heinzmann 500 watt geared, which has a thermistor, back to the controller. So, I'm stuck walking my bike, the very thing I'm trying to avoid. I'm looking for that summary post and conclusion of what's the best way to cool the motor, taking into account the KISS principal, effective heat dissipation, costs, etc. Its really quite beyond me, so I reaaalllllyyyy appreciate the thought & effort being put in here.
 
toolman2 said:
im not sure we just "forgot" to keep it simple,
we have tried motors up to 15kg that completely dominate the feel of the bike and still only produce about half of the torque we need, whilst overheating cos it turned around half of the batteries energy into heat, -not sure which direction we should head there really, bigger motor agian? :roll:
total kiss returns you around 200w of continuous output power per kg of hub motor, just like in 1880 when engines had no active cooling you could nearly get a couple of horsepower from a 200kg engine.
i agree, Volkswagen went completely mad introducing ridiculously complex forced fan cooling in the beetle in about 1950, then it just got silly with liquid cooling and actively computer controlled electric fans, that all drivers have to put up with today.. :lol:
with an effective cooling system you can get motors up to 5000w continuous output per kg, so maby have another quick look at what power and weight you want on the the rear axle of your bike?

i consider my slackarsed $59 leafblower cooling setup to be crude 1950's style technology, but even the 100w it uses (and only when at full tilt) to be quite ok, cos when you pin it to say 5000w in to a hubbie off the mark you get over 4000w of ir loss, now with cooler copper you will save over 1000w :!: -good plan methods.

so to me keeping it simple is completely optional for everyone, and yep these tests are right on the mark to then very simply and easily improve your average hub motor out for a huge number of people.
Click to expand...

It sounds like the leaf blower is a great solution to cooling. I prize the silence of electrics too much to go that route. That's not my real issue though. I have a problem with any rig that pushes a motor anywhere close to max power. Whether you're able to get the heat out or not is irrelevant, because that means pushing a motor into saturation, since absolute max power occurs at 50% efficiency. You start pushing your motor to such low efficiency is just wasteful and makes you carry extra battery weight for the same range.

That's the reason for going for a bigger motor. I fully appreciate that you don't want big weight in wheel for your king of riding, so simply put it in a better place.

John
 
methods said:
Leaf blower :pancake: ?

I like this idea... especially the dirty talk about a shroud being all close up on the inlet holes.

-methods
Click to expand...

The simple solution for that is to use a scooter motor with a drum brake. Then you've got a stationary cover about 5" in diameter and relatively well sealed to the side cover with is perfect for pumping in your big air flow. That's the easiest. Then there's extra sized bearings like Zappy has done, or an extra large hollow axle like Farfle has done. Both are great, though they do require precision tools.

John
 
John in CR said:
toolman2 said:
im not sure we just "forgot" to keep it simple,
we have tried motors up to 15kg that completely dominate the feel of the bike and still only produce about half of the torque we need, whilst overheating cos it turned around half of the batteries energy into heat, -not sure which direction we should head there really, bigger motor agian? :roll:
total kiss returns you around 200w of continuous output power per kg of hub motor, just like in 1880 when engines had no active cooling you could nearly get a couple of horsepower from a 200kg engine.
i agree, Volkswagen went completely mad introducing ridiculously complex forced fan cooling in the beetle in about 1950, then it just got silly with liquid cooling and actively computer controlled electric fans, that all drivers have to put up with today.. :lol:
with an effective cooling system you can get motors up to 5000w continuous output per kg, so maby have another quick look at what power and weight you want on the the rear axle of your bike?

i consider my slackarsed $59 leafblower cooling setup to be crude 1950's style technology, but even the 100w it uses (and only when at full tilt) to be quite ok, cos when you pin it to say 5000w in to a hubbie off the mark you get over 4000w of ir loss, now with cooler copper you will save over 1000w :!: -good plan methods.

so to me keeping it simple is completely optional for everyone, and yep these tests are right on the mark to then very simply and easily improve your average hub motor out for a huge number of people.
Click to expand...

It sounds like the leaf blower is a great solution to cooling. I prize the silence of electrics too much to go that route. That's not my real issue though. I have a problem with any rig that pushes a motor anywhere close to max power. Whether you're able to get the heat out or not is irrelevant, because that means pushing a motor into saturation, since absolute max power occurs at 50% efficiency. You start pushing your motor to such low efficiency is just wasteful and makes you carry extra battery weight for the same range.

That's the reason for going for a bigger motor. I fully appreciate that you don't want big weight in wheel for your king of riding, so simply put it in a better place.

John
Click to expand...

in this case john, you could not be more mistaken, the bike uses less power to trot alongside any of our others at any speed so far, and peak efficiency is yet to arrive until 2 or 3 times this power level, with no sign of saturation and allowing prolly double your battery capacity for the same total system weight.
the 50% at max power thing is only when a: you have no current limiting (and we always do) AND b: you have also loaded the motor to 50% of its free speed, have another read of chapter 3 in electric motors and drives. :wink:

but i take your point about the noise etc (1950's teck compared to water cooling) i just got the 20v laptop supply to drive the leaf blower from the main pack and the blower only needs to be cranked up an noisy when over 7kw output, and things are already out of hand, and yep in general i agree 99% of folks would be happier with a nicely vented hub, using the great info here. so if we want to cover more of the maxed out forced air stuff im happy to, but maby elsewhere to keep this thread as intended?
 
You must log in or register to reply here.

Similar threads

vladrov
Commuter E-Bike: Grin All-Axle Hub Motor with Detachable VESC & 27s3p P45B Battery Pack in One Box
Replies
19
Views
2,821
tigcross
tigcross
Captain Qwark
    • Doin' Science
    • Like
  • Question
Simple 500w 24v build not working
Replies
17
Views
1,895
MxusMick
M
L
Joe's E-bike build
2 3
Replies
60
Views
11,825
lightbulb010
L
T
Multistage axial flux motor design and prototyping
2
Replies
25
Views
6,488
larsb
L
A
Non-silicon and/or soft-switched controller topologies?
2 3
Replies
63
Views
7,473
ARod1993
A
Back
Top