Definitive Tests on the Heating and Cooling of Hub Motors

[parabellum]

Is such shock therapy even good, spraying cold water on hot winding? Can it be used on daily basis without damaging the motor?

 

[cal3thousand]

Is such shock therapy even good, spraying cold water on hot winding? Can it be used on daily basis without damaging the motor?

I don't plan on spraying a 'hot' motor myself. It would be a spray when I get home and then running it until dry. Only on the occasion that I feel dirt ingress occurred.

Is that a bad idea?

 

[gogo]

Is such shock therapy even good, spraying cold water on hot winding? Can it be used on daily basis without damaging the motor?

I don't plan on spraying a 'hot' motor myself. It would be a spray when I get home and then running it until dry. Only on the occasion that I feel dirt ingress occurred.

Is that a bad idea?

Dunno, but compressed air might be better?

 

[Punx0r]

Sand + water = grinding paste?

 

[spinningmagnets]

The stator laminations must be steel (of course), but the core of the Crown motor and the Cromotor stators are both thick aluminum spokes, which act as a heat sink to absorb temporary amp-spikes. If you desperately wanted to use a particular size of motor and add active cooling (rather than upgrade to a larger and more appropriate motor), you could bond shaped aluminum pieces just under the stator, so the water-misting would not be sprayed directly onto the coils (if that was a concern).

I believe that some airflow would be needed to make water-misting effective. Perhaps the outlets could be open, and the air inlets covered by a filter of some type (or a fine screen)?

I'm not sure where this trouble would be worthwhile. The thought of a race bike comes to mind, but that only makes sense where the physical size of the motor is restricted...(in the 1960's trans Am race class, Ford and Chevy both were restricted to running 302's, NASCAR engines grew every year until the displacement was capped at 7-liters (426-428 ci). Either way, it's a fun thought experiment.

If my only mandate on a paid commission was to incorporate liquid cooling on an e-bike motor, I'd use a non-hub inrunner, like the 4-1/2-inch dia, 101mm long Big-Block from the GNG family.

 

[999zip999]

I saw Luke take grinder to cut a hole in top of a hot motor put a hose in it. Then cut a hole in the bottom to let the water out, then take it on the track ?
It was hard to understand.

 

[liveforphysics]

I saw Luke take grinder to cut a hole in top of a hot motor put a hose in it. Then cut a hole in the bottom to let the water out, then take it on the track ?
It was hard to understand.

We only wanted to water fill that poor 10kW golden motor because it was making a concerning amount of bad smells while I raced.

The moment it was full of water, the Sevcon gave a fault for "phase winding short-circuit", and wouldn't run.

That was why we frantically drilled the hole to let the water it was filled with drain. After it finished dripping out, we were able to clear that crazy fault in the Sevcon and it ran again.

If I remember right, in the final event race I melted that motors windings down and it burned up in a shower of smoke, sparks, and plasma. It likely would have saved that motor (and perhaps won the raced as I was in 1st when she blew) if the Sevcon wouldn't have been so damn fussy as to detect that water.

 

[h0tr0d]

The moment it was full of water, the Sevcon gave a fault for "phase winding short-circuit", and wouldn't run.

For this reason and others, I'm loving more and more oil cooling...

 

[agniusm]

Bernoulli's principle.
I need advice on sidecover pattern. I have X5303 and I want to try forced air system. Here is a quick Note2 sketch:

The principal for the exhaust is this which I hope will work:
[youtube]UujAMPv3y-A[/youtube]

 

[cal3thousand]

Bernoulli's principle.

This is the same reason that I made the holes on opposite sides of different size and location. On the drive side, 9 holes 1.125" in diameter located near the circumference of the cover. One the brake side, 6 holes, 3 of them 1" and 3 of them 1.25", all located closely to the axle.

My hypothesis is that the speed of the air over the holes will differ, causing the pressure gradients to flow more air.

Now, I have a scientific principle that I can spout off as the reasoning if anyone asks

 

[John Bozi]

I agree with AW. You don't want to cut into the magnet backing ring so close the the magnets. It's much easier to make holes in the cover.

With effective ventilation you don't need any extra heat sink to cool the motor. As Justin proved earlier in the thread, holes or slots alone are relatively ineffective. I've always used some kind of interior blades, which explains the success I've had. More recently though I added some exterior blades to a hubbie with exhaust slots only on one side, and now the motor moves a tremendous amount of air. I started with a many slot approach to keep debris out, and my thought was the material between each slot would act like a centrifugal blade. It did work and flows some air, but I'm sure less than 5 or 10cfm, which can't take much heat out with it by my calculations. That changed with the blades and I can feel so much wind off of that side of the wheel that it has to be moving hundreds of cfm through and across the outside from left to right. I tested it with smoke and it sucked smoke both inside and across the outside quite strongly even at mid hundreds of rpm. My next iteration will be just 6 larger exhaust holes with a slightly angled blade in front of each instead of just straight radial like I did here. The slight angle will be so the blade covers the hole to some extent. With a more typical hubbie there's more room between the cover and the stator, so the narrow slot cut for the blade will allow a significant amount of blade on the inside to, which will create more turbulence at the stator for better cooling. Drilling bolt holes and tapping threads allows the blades to be fit perfectly from the outside with the motor assembled. I didn't even take the motor off the bike to install my blades. I used a piece of cardboard in the slots to get the interior amount of blade just right before cutting the metal ones. :
View attachment 2
View attachment 1

Thanks Justin and all the posters of stuff they have tried. Hard to say for sure but it looks like fins on the outside is the route to take. Images won't repeat so go here if you dont know what Im refering to...
http://endless-sphere.com/forums/viewtopic.php?f=2&t=48753&start=250#p777497
Only one guy here looks like he's tried it and would love to see Justin try something in that direction.

It makes sense to have "scoops" to swipe the air in even more aggressively than just having fins internally creating a simple direction.

Is the logic of the scoop to pull in the air on one side and push it out on the other side and so why?

Wouldn't it be be logical to just have them on both sides as they will only ever be scooping air in the direction of the bike is going. So for example pulling it in at the bottom of the cycle and spitting it out at the top of the cycle.

Instead of just having cuts across the whole hub just have them under the scoops.

 

[John in CR]

It makes sense to have "scoops" to swipe the air in even more aggressively than just having fins internally creating a simple direction.

Is the logic of the scoop to pull in the air on one side and push it out on the other side and so why?

Wouldn't it be be logical to just have them on both sides as they will only ever be scooping air in the direction of the bike is going. So for example pulling it in at the bottom of the cycle and spitting it out at the top of the cycle.

Instead of just having cuts across the whole hub just have them under the scoops.

Waste of time. There's no such thing as a centrifugal fan running in reverse, ie scooping air into the center. I used to think about scoops too until Jeremy Harris ran some quick calcs demonstrating how little air they could move. They'd just end up being deflectors for sand to hit and bounce into the motor.

Hillsofvalp came up with a nice scoop related idea though. He took one of those dish shaped metal shields that goes under the elements on electric stove tops and cut back about half of it to make a single stationary "scoop". I'd call it more of an air dam for the intake side of the motor. Without something like that the outside air flow is moving perpendicular to the direction of intake, which no doubt greatly decreases air flow through the system compared to if the motor could freely intake air like common centrifugal fans. HOV's air dam slows the velocity of that flow right in front of the intake holes, which not only makes it easier to turn and enter the motor, but it also increases the air pressure there, which makes it even easier to flow in to replace the low pressure inside the motor created by the centrifugal fan.

My understanding of hubmotor cooling has evolved greatly, so anything I posted on the topic prior to July 2013 is obsolete and superseded by this bladed approach with single sided intake. I haven't come anywhere close to optimizing it, so there is definitely room for improvement on subsequent iterations, but I've been running the my motor at 27kw peak input for about 10 months with no heat issues at all. I turn on the BBQ thermometer that I use as my temp sensor only very occasionally, and only out of curiosity. All heat anxiety is completely gone.

 

[richdeloup]

Some while back I had a 16" fully cast aluminium motor/wheel, which I seemed to be able to put crazy power through. Obviously being small, it was less prone to overheat, however just wondering if the fully cast wheel maybe acted as some sort of heat sink?

 

[John Bozi]

It would have been good for this thread if I had made my changes one at a time, but nobody enjoys pulling a hub apart for fun, so I made many changes at the same time.

I wasn't sold on the little holes big holes combinations, just went big holes at the windings and a 19" bicycle rim 20" tyre.

There is no doubt what so ever how much longer and higher power levels my motor can handle, before over heating up slow steep mountains. I would estimate more than half the heat has gone.

What interests me now is testing the stop and take a break vs keep going at minimum throttle as a way of cooling the motor down. Keeping that breeze flowing seems to be more advantagous even at a simple pedal cadence and if needed with a little bit of power than just stopping and waiting.

Did a few runs like this and it is more fun, full power mountains and then pedal for a bit repeat repeat repeat - I dont like the ca doing it for me but its great having it there to stop melt down...

recently i was given a dvd of the australian electric superbikes series - was very surprised to see one team using a 20kw plus hub with almost identrical hole pattern to mine - having said that they were in the process of trying to make water cooling work without leaks....

http://youtu.be/X0t_m-Tnft4?t=1m6s

 

[justin_le]

What interests me now is testing the stop and take a break vs keep going at minimum throttle as a way of cooling the motor down. Keeping that breeze flowing seems to be more advantagous even at a simple pedal cadence and if needed with a little bit of power than just stopping and waiting.

Hi John, the results here might not be what you think. When I was running the tests on the dyno bench last year I was at first leaving the motor spinning after it was done in order to "assist" the cooling down, but in fact the motors would never get close to reaching room temperature this way since the internal iron losses of the hub were generating a fair bit of internal heat and really reducing the rate of cool down. If the core is very hot (100+ oC), then for sure the gains you get by extra convection probably offset the additional heat you are putting in from spinning the motor again, but at some point it reaches a crossover where you'd be better off with the motor still, less convection but no heat source inside the iron either.

So if you have a CA and logger and can clock some data that would be easy to do and the results would be interesting to look at. Running the motor at no load generates almost no copper loss (I^2R is negligible when I is like an amp or two), but results in a fair bit of core loss, so a low RPM will be much better than a fast RPM in this regards. It sounds like what you want to do is cruise around riding the bike rather than prop the wheel up and free spin it?

-Justin

 

[John Bozi]

Hi Justin,

I agree with the starting temperature making a difference, no definitive test but just what I notice is that if I have reached 120 temps drop faster than from say from 90 degrees Celsius (bigger difference to ambient). I often watch temps down a mountain with and without regen dropping and would be a nice video record too.. as it didn't seem to heat the motor as much as not let cool more. (guessing that's because there's a big difference in the power in and out.)

I don't have a logger, enough things on the bike already but am uploading a video ...

that should give some better data.

Here's a picture from the video although most of it is not that clear with a lot of glare:

ps It will be done with the blue wire - throttle wire, out of the CA/controller connection I pulled out yesterday because I haven't sorted out why I get cuts on bumps/jumps , hard acceleration and also a creep in power from power up. Probably a combination of yanking wires and settings I currently don't need to deal with as the controller throttle set up does everything I want correctly.

 

[Cowardlyduck]

Justin/John, I did some mods with 25mm fans to my HS4065 recently that have been working well so far. I do a lot of slow off-road riding, so for these conditions the little fans really help.

For details see here: http://endless-sphere.com/forums/viewtopic.php?f=30&t=56965&start=25#p946731
I plan on doing more tests with longer real world rides and plotting the temps, but for now I managed to get this graph down in mostly controlled conditions.

For anyone else looking to do slower riding where air flow over the hub shell is not as significant (say <40kph), I would highly recommend doing a similar mod.

Cheers

 

[Arlo1]

Justin/John, I did some mods with 25mm fans to my HS4065 recently that have been working well so far. I do a lot of slow off-road riding, so for these conditions the little fans really help.

For details see here: http://endless-sphere.com/forums/viewtopic.php?f=30&t=56965&start=25#p946731
I plan on doing more tests with longer real world rides and plotting the temps, but for now I managed to get this graph down in mostly controlled conditions.

For anyone else looking to do slower riding where air flow over the hub shell is not as significant (say <40kph), I would highly recommend doing a similar mod.

Cheers

Good job. I use the same thing in my X5 although I managed to get mine pointer out wards so the air will blow from the center (cold air inlet) to the windings then out the holes around the outer diameter. NON the less its amazing how well the fans work.

 

[Arlo1]

BTY I got 3 year warranty with my 12v fans then I run 25v (fully charged) to them I just run 4 in series and hooked pack voltage to them

 

[icecube57]

Those are really dark windings you probably exceeded 150C.... Many times. My last build I threw 5kw at a Magic Pie and had the thermal limiting activated on the CA and set to about 140 and after a year of abuse the windings were shiny like they came out the factory. The winding tie downs were intact and not discolored. The plastic that insulates the laminations was intact and not melted. It was a brand new motor. You my friend dont have many more lives left on this motor. If the fans werent already installed I would suggest some sprayon varnish to reinforce the enamel on the windings.

 

[alsmith]

Justin/John, I did some mods with 25mm fans to my HS4065 recently that have been working well so far. I do a lot of slow off-road riding, so for these conditions the little fans really help.

....

For anyone else looking to do slower riding where air flow over the hub shell is not as significant (say <40kph), I would highly recommend doing a similar mod.

Cheers

Good idea!

 

[John Bozi]

HEy wow! first time seeing this implemented and great graphs thanks..

Many questions:

1. does no fans mean a sealed motor or the holes are open with no fans in the way or fans are covering the holes not turned on?
It would make a big difference to the data.

2. If you pump 5000 watts up a mountain riding slowly are you showing us that you don't get over 50 degreesC?
What temperature are the fans rated at? I can only imagine they would melt at 120Celsius....

We don't need fans at 50 degrees celcius so the graph and usefullness of them would be amazing at the 110-80 mark....
please!

3. If you do both slow riding and fast riding wouldn't fans just get in the way of higher convection rate of going 50kmh?

How much battery do they take to run on a 30k trip?


oops I should visit the link.....

 

[John Bozi]

I was under an intuitive impression that light pedalling with a bit of power might increase the cooling of the motor more than just not using power at all. It actually takes 1 minute to cool from 90 to 80 degrees with no power used, or three times longer when using a few amps at pedalling speed. Justin was right and Regen also heats up the motor especially when on strong. On lighter regen, the convection seems to cancel out the heating and the temperature stays the same.

The video is boring and a pain to watch because of glare for one thing and only proved what everybody knows, at least it proved it for me

If you want to see the relevant parts skip to watch from 4:00 to see the no power 1 minute to cool 10 degrees C or from 13:00 to see it take 3 minutes to cool 10 degrees with low power pedalling.
[youtube]LP1AqYHuhAs[/youtube]

 

[fechter]

I was planning to do something similar to my A2B motor. The idea is the fans force air around the windings in the slots and tranfer the heat to the inside of the casing. The casing normally has good air flow around the outside, so outside temp stays low. The rate at which heat can be tranferred from the copper to the outside of the case is greatly increased by using the fans.

Axial fans may not be the best choice for pushing air through a thin passage. The back pressure really kills the flow. Centrifugal blowers will push more air through a small opening. Stacking axial fans might improve the pressure performance.

Another approach might be to use axial fans and have return holes in the stator. This will increase total air flow, but it won't be directed through the slots as much. Heat from the windings can tranfer to the stator and get picked up there.

 

[Cowardlyduck]

Thanks for the kind comments guy's. This mod was a first attempt for me. It was very cheap and easy comparatively to other modes like water cooling and oil cooling, but still better than the straight venting approach I was using before.

BTY I got 3 year warranty with my 12v fans then I run 25v (fully charged) to them I just run 4 in series and hooked pack voltage to them

That's a good idea actually. I run 50v fully charged, so my 6 fans would run at about 8v each if I did that. That's a little slow, but still acceptable if I needed to. For now I use a small DC-DC converter which I can adjust the voltage on when needed. This way, when cruising around town I can turn the fans down for less noise, but when flogging it off-road, I can crank them up to help cool better.

This is mounted just beneath the frame.

Those are really dark windings you probably exceeded 150C.... Many times. My last build I threw 5kw at a Magic Pie and had the thermal limiting activated on the CA and set to about 140 and after a year of abuse the windings were shiny like they came out the factory. The winding tie downs were intact and not discolored. The plastic that insulates the laminations was intact and not melted. It was a brand new motor. You my friend dont have many more lives left on this motor. If the fans werent already installed I would suggest some sprayon varnish to reinforce the enamel on the windings.

My windings are not dark from over power I think. I previously sprayed a silicone based black paint all over the windings to protect them. That shot above was after attempting to wipe most of it off. I've previously put 6+KW through this motor, and shorted 2 strands on one phase to the stator. I then removed those strands, and the motor still works. See the end of this thread for details: http://endless-sphere.com/forums/viewtopic.php?f=2&t=55486
Since adding the fans, I haven't exceeded 120C guaranteed.

1. does no fans mean a sealed motor or the holes are open with no fans in the way or fans are covering the holes not turned on?
It would make a big difference to the data.

No fans in that test means just turning them off. The covers are still vented, but just at the perimeter near the windings. The point I'm making is straight venting can only help so much...add some fans and it's that much better.

2. If you pump 5000 watts up a mountain riding slowly are you showing us that you don't get over 50 degreesC?
What temperature are the fans rated at? I can only imagine they would melt at 120Celsius....

No, that test was about keeping things consistent. I've had the fans running with the motor up to over 110C a number of times up steep hills. My temp gauge maxes out at 110C. It cools off very quickly, even still pushing 1+KW through it as the hill tapers off.
These little fans are sucking in fresh air, so they don't see the same temp the windings themselves see. I would guess I would need to hit almost 200C for the fans to start melting.

We don't need fans at 50 degrees celcius so the graph and usefullness of them would be amazing at the 110-80 mark....
please!

Agreed! As mentioned, more tests planned. The point of the first test was to prove the concept.

3. If you do both slow riding and fast riding wouldn't fans just get in the way of higher convection rate of going 50kmh?

Negative. Once I hit 40kph+ the normal air flow over the motor seems to cool better than the fans, or at least faster. The fans don't make a difference to this as they would only prevent a cross wind from flowing through the motor in one direction.

How much battery do they take to run on a 30k trip?

Not much. The DC-DC converter I use maxes out at 15W. I'm running 6X 12v fans. Previous tests showed about 0.5A at 15V, so 7.5W. If anything the lower motor temps mean it is more efficient, so I use less power overall than without the fans.

Another approach might be to use axial fans and have return holes in the stator. This will increase total air flow, but it won't be directed through the slots as much. Heat from the windings can tranfer to the stator and get picked up there.

Fechter, check out the other thread. Some similar suggestions have been made already.
http://endless-sphere.com/forums/viewtopic.php?f=30&t=56965&start=75

Thanks all for the kind comments...inspires me to to more of the same, and a better job next time...I have another motor (HS4080) I also plan on modding this way, but with better fans, pointing them the other direction, and adding radial veins/flaps to assist the flow out on one side.

Cheers