Cooling fans inside Hub motors

[amberwolf]

Yes, I agree that a test is needed as coffee filters are normally meant to filter water, not air, so it is likely they may not let much through.

It's not really the kind of filter, as the kind of fan. Even muslin is probably too restrictive.


In my experience, "box" fans like those move lots of air when totally unimpeded, but with anything that causes an airflow limitation the partial vacuum behind the blades ends up causing the airflow to stop or nearly stop, and just recirculate around the fan blades.

Sometimes it even puts such a high load on the bearings that the fan overheats and fails.


I never used those particular fans, but I tried quite a few things back in my PC hotrodding days to cool the computers quietly but greatly, without sucking doghair and dust inside (both really big problems here), and fans with filters were my first option. The best filters were actually just large-hole windowscreen mesh, holes more than 1mm in size, on the inside part of an small box inlet that expanded after letting air inside the case thru a small hole. All fans were on the actual outlet, with no grillework or anything to impede airflow or cause turbulence.


I used slower fans, because there was less turbulence around the fan and so less noise and less wasted power (any noise is wasted power, usually from airflow turbulence), and the dust when not sucked in quick would settle on the filter or just before it, from the slowdown in the airflow from the expansion of hte air after coming in thru the smaller hole into the larger chamber.


I don't know that there's space to do all that inside the hubmotor, but if there was, it'd probably be the best way to work it.


If you could put the fans external to the hub, and have a nearly continous wide slot in the cover for the outlet and inlet, and a disc on the bike frame covering each slot as close to the cover as possible, then put the fan in a duct on the fork so it was far enough away from the cover to prevent turbulence at it's backside and the end of the duct was far enough away from the fans to prevent turbulence at it's front side, then you could put the screen (if needed) in the inlet duct, and the fan in hte outlet duct. You could also use a much larger diameter fan, and it'd be able to run slower with less turbulence and less noise and more fan blade/airflow area for the size of the fan (vs motor and frame area),


SOmething sort of similar in concept was used by Lowracer on his trike, but unfortunately his site, photos, and vids are gone and/or don't work anymore, and he hasn't been around in a long time and doesn't respond to contact attempts. There's a little descriptive info in his threads, but not much. IIRC his was super high speed fans ina duct that sucked or blew air thru a tube from next to the hub on a series of holes.



IN the PC's eventually I went to single slow-rotating but very large diameter fan with ducting and airflow redirectors to ensure flow past the right parts, and it worked way better than a bunch of fast little fans, and was also way quieter (necessary to do audio recording in the same neighborhood :lol. I ditched the filters entirely and just used a box chamber around the PC except at the exit hole, with relatively small inlet holes, and the pressure drop inside the inlets made most of the dust fall out of the air between the outer box and the PC case. I just had to pick up the box and vacuum it up every few days to prevent a miniature sahara diorama.

I think you would be surprised at how much air these little things can move though.

Oh, I know they move lots of air--when totally unimpeded. But if you even just put a protective grill (like on PC cases or power supplies) up against it's back, it gets a LOT noisier and airflow drops way down.

 

[fechter]

In considering a similar fan setup, I thought it may be possible to power the fans from the phase wires, eliminating the extra power wires through the axle. You'd use some fast-recovery diodes in a bridge setup to rectify the phase voltage and feed a capacitor. Even at very low throttle settings, the PWM goes all the way to pack voltage, so the cap will charge up to near pack voltage even at low speeds. May still need a voltage regulator between the cap and fan motor in my case.

 

[John Bozi]

My mate just uses a bit of old cloth from a t-shirt wrapped around the end of his fan modded controller.

Also to some degree the more you cake on that red paint the more heat you are keeping in too.

what kind of glue did you use there on the halls etc.? looks similar to epoxy I used - suppose to be high heat capable and hard as steel - but mine softened up, the halls floated up to the side covers and the vent holes ripped them.... so I have repairs ahead again


keep a the great work!

 

[Cowardlyduck]

Thanks all for the comments.

I just wish the magnetic field of the hub to not involve as negative factor to the magnetic field of the tiny fans.
Definably this project seems close to succeed at fighting the enemy of all hub motors.

This is the second time I'm doing this, but I did not notice any effect on the fans from the hub magnets previously. They are a good distance away from them, so I doubt the effect would be great if at all.

Amberwolf, I like your idea on using fly-screen. Should the coffee filters be too restrictive as suspected, I might try some fly-screen instead.

I realize the airflow will be restricted by doing so, however what I found when I did this mod previously, is that you do not need much air flow (compared to just straight vented) to make a large difference in cooling capacity. From page 2 of this thread:

Fetcher, the idea of using the phase wires to power the fans was previously suggested in this thread, however for simplicity sake, I've chosen to just run the extra wires. At least then I can turn the fans off when not needed saving ~35W and a lot of noise.

My mate just uses a bit of old cloth from a t-shirt wrapped around the end of his fan modded controller.
Also to some degree the more you cake on that red paint the more heat you are keeping in too.
what kind of glue did you use there on the halls etc.? looks similar to epoxy I used - suppose to be high heat capable and hard as steel - but mine softened up, the halls floated up to the side covers and the vent holes ripped them.... so I have repairs ahead again
keep a the great work!

Thanks John.
The cotton/t-shirt cloth idea had crossed my mind and I might give it a try, but as Amberwolf points out, it's likely to be far more restrictive than helpful.
I agree, the red paint will impede some heat transfer, however it's worth it to protect the windings IMO.
Glue used was some high temp epoxy from SuperCheap Auto called DynaSteel. Claims to be able to handle up to 260C. I've used it before without any issues. Even if it doesn't like the high heat, I don't expect this motor to ever see north of 120C with the fans running.

Cheers

 

[Cowardlyduck]

If you run those tests by having your motor painted white, especially in the summer, you will note 15C lower. :wink:

LOL :lol: - Yeah, but that would totally ruin the look of my Black Stealth Fighter.
If I was going to go that far, why not build a aluminum foil dome around the motor to reflect all external heat away.

Cheers

 

[John Bozi]

.
Glue used was some high temp epoxy from SuperCheap Auto called DynaSteel. Claims to be able to handle up to 260C. I've used it before without any issues. Even if it doesn't like the high heat, I don't expect this motor to ever see north of 120C with the fans running.
Cheers

Yes exactly same stuff and place I bought mine. It's BS to have steal in the name when it's soft and flexible like plastic and even rubbery now that I removed some to play with in my hand.

I wish you the best with it. I only had about 2mm to play with from the side covers and a bit of an an accident while at high temps must have loosened it up enough to touch and rub and the walk home in shame ....

 

[Offroader]

If you run those tests by having your motor painted white, especially in the summer, you will note 15C lower. :wink:

LOL :lol: - Yeah, but that would totally ruin the look of my Black Stealth Fighter.
If I was going to go that far, why not build a aluminum foil dome around the motor to reflect all external heat away.

Cheers

When I got a high quality IR temperature gun 50:1, among my tests I did also an evaluation of the temperature of parked cars is the sidewalk.
Paint is the cheapest way to keep temperature down, and every outdoor installed motor is white.
When performance is the goal, aesthetics is not important.

I'm a bit confused here, I actually thought about painting my motor white once as that would look really cool on my white bike.

However, when researching paint color I learned that black is the best color to reflect heat out of the motor. This is the reason why hubmotors are actually painted black.

Even though white will reflect more heat from the sun, it will not shed as much heat as black paint, which is more important to shed heat because the motor is making heat internally.

Sorry about not using all the correct terminology here as I forgot what everything is called.

maybe I am misunderstanding something, are you saying white would be the best color to paint a hub motor to reduce heat?

 

[sn0wchyld]

........
LOL :lol: - Yeah, but that would totally ruin the look of my Black Stealth Fighter.
If I was going to go that far, why not build a aluminum foil dome around the motor to reflect all external heat away.

Cheers..............

When I got a high quality IR temperature gun 50:1, among my tests I did also an evaluation of the temperature of parked cars is the sidewalk.
Paint is the cheapest way to keep temperature down, and every outdoor installed motor is white.
When performance is the goal, aesthetics is not important.

I'm a bit confused here, I actually thought about painting my motor white once as that would look really cool on my white bike.

However, when researching paint color I learned that black is the best color to reflect heat out of the motor. This is the reason why hubmotors are actually painted black.

Even though white will reflect more heat from the sun, it will not shed as much heat as black paint, which is more important to shed heat because the motor is making heat internally.

Sorry about not using all the correct terminology here as I forgot what everything is called.

maybe I am misunderstanding something, are you saying white would be the best color to paint a hub motor to reduce heat?

Radiate, not reflect , but yea otherwise you're right... Dark will reflect less than light, but will consequently radiate more than light too. The absorbed energy from the sun will likely be significantly less than the waste generated inside the hub, so black should be better off even in direct sun - though i'd be interested to see a test.

put it this way, in the middle of the day in summer near the eqator, about 1.4kW/m^2 falls on the earth. A 30cm dia hub has a max surface area that can face the sun of about .25m^2 - this is lying it on its side... so worst case scenario, if we assume your black paint is 100% black - ie reflects no light, then you'll absorb an absolute peak of about .25 x 1.4 = 350w of heat from the sun. If you're putting in any more than about 1.2kw in to your hub then chances are you're generating nearly as much heat from your motor as this unrealistic 'worst case scenario', and a white hub (or rather, a 100% relfective mirror hub) will save you a absolute max of 350W of heat.

In reality though, we dont ride with our motors on their side - they rarely are in full sun, all the time, and no paint is even near 100% black, or else it would be impossible for our eyes to focus on... etc etc, so painting your hub white/mirrored could have a very small effect on how hot it gets in the sun, and (potentially, but I dont know how much) a notable reduction in heat dissipation due to reduced radation from the hub due to its lighter color.

Its not to say however that we should paint things like our batteries black - as they're far more heat sensitive, and (should) be generating far less heat internally when in use.


Edit
sorry, i did Pi D^2 not Pi R^2, so its more like 100W of energy absorbed from the sun on a 100% black motor.

 

[madin88]

Even though white will reflect more heat from the sun, it will not shed as much heat as black paint, which is more important to shed heat because the motor is making heat internally.

at our ambient temperature or motor temperature the colour of the surface area almost doesn't matter for radiation HEAT EMISSION because this mainly takes place in infrared region. hence black and white painted surfaces have similar emission there.
blank metal or glossy and reflective surfaces have poor emission. matt paint is best, but colour doesn't matter.

for HEAT ABSORPTION the color is important. while black absorbs almost all the radiation heat from the sun, white only soak up 20% or so..

 

[sn0wchyld]

though i'd be interested to see a test.

Get two eggs, paint the one black and let them all day at the sun. :lol:

If the black one will be cooked for free, you own me 5$.

never said it didn't absorb heat - i said it doesn't absorb a significant ammount more. like i said - the maximum difference between white and black is about 100W. Realistically more like 50W difference between a black and a white hub. if a hub at ~70 deg in partial/full shade emits 50W more for being black vs white then the benefits of a white hub are entirely negated. Even if its only a 25W difference, its probably still far better, as my example was calculating the maximum heat absorbtion, with pure black paint, and the hub's side perpendicular to the sunlight, in the middle of summer, 'near' the eqator - hardly a 'real world' situation.

THIS is where my desire to see any experimentation comes in - the difference in radation between black and white paint is likely quite small - as is the difference between its absorption. It may even be the same, ie inversely proportional. But the fact remains that the reduction in heat absorption from painting your hub white is likely very small - 50W would be the most, it could well be far less than that.

 

[sn0wchyld]

............

the difference in radation between black and white paint is likely quite small - as is the difference between its absorption. It may even be the same, ie inversely proportional. But the fact remains that the reduction in heat absorption from painting your hub white is likely very small - 50W would be the most, it could well be far less than that.

The DIFFERECENCE is know and is +15C for a black object compared to white one.

When direct sun light heating up an metal to 45C the black one will be at 60C.
What is crusial is the winding coating temperature to kept lower than it max spesification.
This is what cooling is needed for.

sorry mate, but you're really missing a few points here.

1.
you looked at cars and the like in broad daylight - with a comparatively huge surface area, likely absorbing an order of magnitude more light than a motor ever will, and with less surface open to the air to disipate this heat (back side of a bonnet/roof is not open to fresh air, and are also often steel, less thermally conductive than ally).

2.
Magnets, not wires, are the part of the motor with the greatest heat sensitivity, and are the part most likely to be affected by heat due to heating from sunlight. 80C is the usual accepted max for motor magnets, unless you buy one with specialist magnets. Wires are both separated by an air gap and also able to withstand 120C at a minimum, some enamels can handle over 200. The magnets will be dead and gone long before your wiring starts to suffer any damage at all.

3.
a difference in temperature on cars says nothing about the difference in color. cars may have had more sun prior, or been driven prior, or whatever else before your measurements.

I can guarantee I can put a plate of aluminium painted white, and another identical one painted black, and have it so the black one is cooler, just by having more air flowing over the black one, or having the white one with the back of the plate coated in insulation. Whats needed is an experiment that eliminates variables, not takes no account of them whatsoever. Moreover, it needs to compare the amount of heat RADIATED from black vs white, vs what is absorbed, not 'which gets hotter in the sun' - which we all know will be black, all other things being equal.

No one afik is disputing that black absorbs more light, and thus will be hotter when left in the sun. the debate is over whether black RADIATES more heat away from inside the hub, and thus provides better heat dissipation overall - regardless of the increase in temperature due to being left in the sun.

 

[madin88]

THIS is where my desire to see any experimentation comes in - the difference in radation between black and white paint is likely quite small - as is the difference between its absorption. It may even be the same, ie inversely proportional. But

YES and NO

radiation is almost exact the same with any colour at operating temperature of such motors (or ambient temperature), but absorption is different! the white surface absorbs only about 1/5 of what the black surface.

Kiriakos GR has absolutely right and the 60°C black vs 45°C white sounds reasonable :wink:

I can guarantee I can put a plate of aluminium painted white, and another identical one painted black, and have it so the black one is cooler, just by having more air flowing over the black one, or having the white one with the back of the plate coated in insulation. Whats needed is an experiment that eliminates variables, not takes no account of them whatsoever. Moreover, it needs to compare the amount of heat RADIATED from black vs white, vs what is absorbed, not 'which gets hotter in the sun' - which we all know will be black, all other things being equal.

apples vs oranges

white emits almost exact the same heat as black, but absorption of white is only 1/5!!

 

[sn0wchyld]

THIS is where my desire to see any experimentation comes in - the difference in radation between black and white paint is likely quite small - as is the difference between its absorption. It may even be the same, ie inversely proportional. But

YES and NO

radiation is almost exact the same with any colour at operating temperature of such motors (or ambient temperature), but absorption is different! the white surface absorbs only about 1/5 of what the black surface.

Kiriakos GR has absolutely right and the 60°C black vs 45°C white sounds reasonable :wink:

I can guarantee I can put a plate of aluminium painted white, and another identical one painted black, and have it so the black one is cooler, just by having more air flowing over the black one, or having the white one with the back of the plate coated in insulation. Whats needed is an experiment that eliminates variables, not takes no account of them whatsoever. Moreover, it needs to compare the amount of heat RADIATED from black vs white, vs what is absorbed, not 'which gets hotter in the sun' - which we all know will be black, all other things being equal.

apples vs oranges

white emits almost exact the same heat as black, but absorption of white is only 1/5!!

so, white absorbs perhaps 10w, while black absorbs 50w, worst case scenario? And I was not disputing kriakos's claim of a 15 deg difference, merely disputing is relevance (or rather, completeness of this observation) when it comes to motors. I get that the difference in emmision is likely very small, but unless your hub is in full sun during a ride, then the absorption difference is also going to be quite small, particularly compared to waste heat from the motor. Like i said earlier, the maximum it could possibly be is 100w differnce, on a large hub - not insignificant at all in terms of heat, but by the same token, this was a un-realistic worst case scenario. Riding with the sun directly overhead will mean that only a small % of the hub is exposed to sun, and thus the difference could be just a few watts, or less.

http://littleshop.physics.colostate.edu/activities/atmos1/ColorAndCooling.pdf
http://www.hk-phy.org/iq/blades/blades_e.html

both these experiments show that there is a measurable difference in both absorption and radiation depending on color, at similar tempratures to what our motors operate at.

and once again, im not saying a light colored hub will be worse or better, merely that looking only at the difference in absorption and the resultant temperatures they reach is only looking at half the picture.

 

[madin88]

what about radiation heat from the environment (hot asphalt etc)? I believe white painting also has its benefits there and not only in direct sunlight..

 

[Offroader]

My take on the heat is that although white paint will not heat up as much, it will not emit (hope this is the correct word here) heat as rapidly.

I figure the heat that will be absorbed from the environment is insignificant compared to the heat made inside the motor. If you don't use your hub motor it will be basically cold to the touch, once you start using it the temperatures will hit 100c.

When the motor is 100c from heat made inside the motor it is more important to emit that heat and the best color is black. The heat absorbed is insignificant compared to the heat generated inside the motor.

I think the answer is probably that the color of the motor is insignificant as most heat is transferred by convection (I think this is properly used here when the heat is transferred from the motor by air). If the color makes a difference it is probably 1 or 2% difference and can't easily be measured.

What hubmotors really need is fins all along the sides and top as convection is probably the most important for cooling. With enough fins you could probably like double the surface area of the motor.

 

[madin88]

When the motor is 100c from heat made inside the motor it is more important to emit that heat and the best color is black.

black emits heat only a few percent better at the most so the benefits are negligible compared to white.

 

[Cowardlyduck]

So while this discussion on motor colour is intriguing, this thread is about forced air cooling using fans inside the hub motor. IMO much more effective than anything else.

One of the key problems I experience when I did this mod last time was the build up of dirt and dust inside the hub over time.
I previously posted some 25mm drain hole covers I was thinking of using to cover the holes, however I didn't really like the idea of drilling 25mm holes in the side covers.
I have drilled 25mm holes before (on a non fan cooled HS3548), and they were wider than the width of the windings allowing air to flow easily underneath the windings without going over/past them likely removing less heat in the process.
My previous fan cooled motor from this thread had 20mm holes, and these lined up perfectly with the width of the windings forcing most of the air to flow over the copper before it gets a chance to go anywhere else.
So with that in mind, I just ordered 40 of these:
http://www.ebay.com.au/itm/301250153543

These will obviously need some work to allow air flow, not block it, but at least they fit, and I can customize the hole pattern how I want...maybe with a smile face.
Or I might just drill out the entire middle, and fit some fly-screen as previously suggested.

The only issue I may now face is reducing the depth without destroying the locking mechanism. I may just have to cut it off entirely though to avoid risking it touching the windings or hall sensor wires.
Without the locking mechanism, I will have to either jam them in with a filter/fly-screen taking up the slack, or epoxy them in permanently. Either viable options IMO.

Cheers

 

[flathill]

Paint the inside of your covers flat black. It will make a huge difference in a sealed motor. The worst thing you could do is have shiny aluminum covers in a sealed motor. The difference is larger than you imagine as the thermal radiation bounces back and forth. You can see a 20+ C drop doing this simple mod on a sealed motor.

Don't paint the magnets. You don't want them absorbing heat. Shiny nickel plating is good here.

Don't paint the winding. Pot them in a material with high thermal conductivity to prevent hot spots such as SC-320

http://chargedevs.com/features/impregnating-motors-with-thermal-management-materials-to-reduce-hot-spots/

With air cooling, it doesn't matter so much if painted or not but it still helps to paint your side covers flat black.

 

[Offroader]

I agree with painting the inside of the side covers black. I was thinking about this as the inside of my cromotor covers are bare aluminum and this is a very bad absorber.

I'll have to paint mine.

 

[John Bozi]

Paint the inside of your covers flat black. It will make a huge difference in a sealed motor. The worst thing you could do is have shiny aluminum covers in a sealed motor. The difference is larger than you imagine as the thermal radiation bounces back and forth. You can see a 20+ C drop doing this simple mod on a sealed motor.

Don't paint the magnets. You don't want them absorbing heat. Shiny nickel plating is good here.

Don't paint the winding. Pot them in a material with high thermal conductivity to prevent hot spots such as SC-320

http://chargedevs.com/features/impregnating-motors-with-thermal-management-materials-to-reduce-hot-spots/

With air cooling, it doesn't matter so much if painted or not but it still helps to paint your side covers flat black.

Can you explain hot spots? I noticed I have one about 2 cms wide on one side, and I have never taken my temp sensor over 120 degrees celcius and mostly stay under 100 easily.

I don't know if it is the venting and running two motors because now I can approach a massive steep mountain using my bafang and then just blast from pretty cold a 30 degree spike up to 100 degrees. The readings of the thermo would say Im fine but I am guessing that without the usual ambient hot air heating evenly I am thumping heat into one point... does this sound kind of right?

Painting the inside of the covers is a new idea for me ... thanks I dont quite understand how that works as I thought black was to do with light being absorbed and if there is no light then it makes no difference the colour. If you paint the outside of your covers white or black it makes no difference if you ride at night I would presume...

 

[amberwolf]

You may want to do some googling about it, perhaps some wikipedia reading at least, but light is a spectrum from way below the red up to way past the purple.

Objects do not *only* absorb or reflect the light your eye can see.

 

[Cowardlyduck]

Well progress is continuing, and look what colour I painted the inside of my side covers.

This is what was left...from just one side cover.

Probably the total added weight of the fans there.

Cheers

 

[Offroader]

I remember this post by someone who had failure from too many holes, or ones placed in bad location.

http://endless-sphere.com/forums/viewtopic.php?f=2&t=62803

Cowardly Duck, do you think that you drilled too many and it will compromise the integrity of the hubmotor? You may be alright because the holes are drilled more on the outside.

 

[Cowardlyduck]

Well this was my previous side cover...from page 2 of this very same thread your reading now:
http://endless-sphere.com/forums/viewtopic.php?f=30&t=56965&start=25#p907537

No issues there.

And before that:
http://endless-sphere.com/forums/viewtopic.php?f=9&t=50568#p748271

Although this one was a little overkill, and I wouldn't risk drilling another motor out that much.

So since this cover has less, and smaller holes than both my previous attempts of which neither failed riding heavy off-road conditions including jumps, stairs etc, I don't think I'll have any issues, but thanks for your concern anyway.

Gosh, the nay sayers in the place sometimes.

Cheers

 

[fechter]

Keeping all the holes toward the outer diameter will help. The stress around the axle is much higher.