Cooling fans inside Hub motors

Electric Motors and Controllers
User avatar
madin88   1 GW

1 GW
Posts: 3056
Joined: May 27 2013 2:02am
Location: Austria

Re: 25mm PC cooling fans inside Hub motors

Post by madin88 » Sep 18 2014 2:10am

Offroader wrote: Are you saying you would need the holes drilled only on one side? How would you position an airduct exactly? Do you have any links or pictures of this solution?
think of how the air must flow. the airduct must be installed between inner holes (intake) and outer holes (outtake) with a very small gap to the sidevocer. maybe a part of a big plastic pipe would work for this. with this the air will be almost completely pushed through the stator teeth and windings.
- 10kW belt drive with Neumotor 8057: [url]viewtopic.php?
- KTM Hardtail Pedelec / eZee V2 6T / 27,5" / Kelly KBSX @ 1,5kW / Torque PAS / CA-V3 / 13s4p Sanyo GA in Shark case

User avatar
Allex   1 GW

1 GW
Posts: 3230
Joined: Dec 05 2011 8:46am
Location: Stockholm, Sweden

Re: 25mm PC cooling fans inside Hub motors

Post by Allex » Sep 18 2014 3:03am

Interesting, but I dont quite get it Madin, could you make a sketch?

User avatar
Cowardlyduck   100 MW

100 MW
Posts: 2736
Joined: Jun 26 2011 8:41pm
Location: Canberra, ACT, Australia

Re: 25mm PC cooling fans inside Hub motors

Post by Cowardlyduck » Sep 18 2014 7:47am

Peter Sternersson wrote:I´m kind of looking forward to what these little bastards can do for the cooling in my V2 Cromotor :D

They blow crazy amounts of air on 7V each (7,2V max), makes lots of noise also but hey...

3D-printed inlays for the pockets in the stator, and just some hot glue here and there.
Heeeeey, that is awesome. Definitely taking things to the next level. :) :D :mrgreen:
Did you buy those fans cause I posted about them, or found them yourself? That's great to know they work well...and I'm sure they could take even up to 10 volts ok if they are well built.
Peter Sternersson wrote:The windings are sprayed with a special paint made for electric motors "Ultimeg 2000/372" reads on the can. It has something like 30kV/mm resistance and withstands 150-160C.
This coating has been on there 200km with vented side covers.
Do you think it could potentially thermally insulate things also?
I don't have any paint to protect my windings, but I periodically spray liberal amounts of WD40 into to windings and fans themselves. I think doing this has saved my fans when riding through water.
Peter Sternersson wrote:The covers are milled as follows, and I intend to use fan blades on the exhaust side for extra suction.
I think you need either more, or bigger holes. I guess with better fans it is less necessary, but it won't hurt to add some more I think.
cal3thousand wrote:Good job guys!

The video and plot from Cowardlyduck are enough evidence for me. This mod certainly has it's benefits and I look forward to any developments you have.
Thanks for the kind words. Like Peter, I will definitely be using the Yeah Racing fans when I do this mod next time.
In case anyone wants the link again:
http://www.rcmart.com/rc-yeah-racing-ya ... 29521.html
madin88 wrote:think of how the air must flow. the airduct must be installed between inner holes (intake) and outer holes (outtake) with a very small gap to the sidevocer. maybe a part of a big plastic pipe would work for this. with this the air will be almost completely pushed through the stator teeth and windings.
Madin, I don't get this either.
What is wrong with the way I've done it? The air must flow right past the windings on both sides.
You don't want to make the holes too small (or narrow) as you will block the air flow.
Are you somehow wanting to use the Venturi effect. https://en.wikipedia.org/wiki/Venturi_effect?
The holes need to be big enough to allow air to easily flow, but small enough to direct it past (through?) the windings. They still also need to maintain the structural integrity of the hub cover...which round holes do best.

Cheers
My stuff:
Ebike Garage - My YouTube Ebike ramblings.
Cowardlyducks - Videos - Ride videos
BikeE recumbent commuter - 9C, 6Fet Grinfineon, internal 17AH, 52V, + on-board solar.

For Sale:
Custom made 18650/21700 battery packs - Australia only
High Power LiPo wiring harnesses - 4P - XT90, HXT4mm, 5.5mm. 200Amp+ capable. Global shipping.

Peter Sternersson   100 W

100 W
Posts: 145
Joined: Aug 14 2012 2:28am
Location: Ljungby, Sweden

Re: 25mm PC cooling fans inside Hub motors

Post by Peter Sternersson » Sep 18 2014 11:08am

I hade some scratches here and there inside the motor after pulling it apart the first time, so the paint was mostly extra protection. I do not think a 1/10th of a millimeter makes much difference in heat dissipation.

The size and position of the holes might not be the best, I simply milled existing drilled holes a bit bigger, but not so big that I milled away the reinforcements in the casting on the covers. I´m a big bloke with a heavy bike, I need as much material to hold me up as I can get :D
With the holes at the windings on at least one side, the copper should succumb to at least some air flowing past it. And with a aluminum stator, I think the air exchange inside the motor, even if not directly on the windings will make a difference.


The fans was bought after seeing them in this thread. This thread even inspired me to try some forced aid cooling before trying to water cool the motor.

User avatar
Offroader   10 MW

10 MW
Posts: 2489
Joined: Sep 08 2013 9:03pm
Location: USA

Re: 25mm PC cooling fans inside Hub motors

Post by Offroader » Sep 18 2014 12:15pm

I am confused about how these fan blades will work. They will be blocking the air flow as the motor spins. Or is this where the air exits the motor?

Image

Peter Sternersson   100 W

100 W
Posts: 145
Joined: Aug 14 2012 2:28am
Location: Ljungby, Sweden

Re: 25mm PC cooling fans inside Hub motors

Post by Peter Sternersson » Sep 18 2014 12:48pm

They are on the air exit side.
The thought is that the wings will create a low pressure zone behind it, and draw air out of the motor, and it does work.

Me and a friend have experimented to a great extent this summer.

We got his Cromotor V1 to run the same temp or cooler than my Cromotor V2.

My bike, 55kg + 95kg person. 24" wheel, Cromotor V2 with aluminum stator and Kelly 250A controller.

His bike, 55kg + 100kg person, 26" wheel, Cromotor V1 with steel stator and Kelly 250A controller.

Without the wings, his motor ran 10-30C hotter than mine all the time.

User avatar
Offroader   10 MW

10 MW
Posts: 2489
Joined: Sep 08 2013 9:03pm
Location: USA

Re: 25mm PC cooling fans inside Hub motors

Post by Offroader » Sep 18 2014 1:18pm

Peter, very interesting. Yes makes sense now. Are you going to add vents to the intake side?

Have you done a test with and without the wings on your motor? I'm wondering how much of a difference you noticed by just adding those.

User avatar
cal3thousand   1.21 GW

1.21 GW
Posts: 3606
Joined: Mar 26 2012 4:47pm
Location: California

Re: 25mm PC cooling fans inside Hub motors

Post by cal3thousand » Sep 18 2014 5:09pm

Peter Sternersson wrote:They are on the air exit side.
The thought is that the wings will create a low pressure zone behind it, and draw air out of the motor, and it does work.

Me and a friend have experimented to a great extent this summer.

We got his Cromotor V1 to run the same temp or cooler than my Cromotor V2.

My bike, 55kg + 95kg person. 24" wheel, Cromotor V2 with aluminum stator and Kelly 250A controller.

His bike, 55kg + 100kg person, 26" wheel, Cromotor V1 with steel stator and Kelly 250A controller.

Without the wings, his motor ran 10-30C hotter than mine all the time.
But isn't your motor turning faster at the same speeds since you're in a 24" wheel? Not really an apples to apples comparison is my point
Get a Cycle Analyst and a Multimeter, you're still a noob if you don't have at least one of each.

Planning on posting questions or buying anything on this site? Put up your country (at minimum) on your profile. This is a worldwide forum and we haven't reached clairvoyance.

Peter Sternersson   100 W

100 W
Posts: 145
Joined: Aug 14 2012 2:28am
Location: Ljungby, Sweden

Re: 25mm PC cooling fans inside Hub motors

Post by Peter Sternersson » Sep 19 2014 9:44am

cal3thousand wrote:
Peter Sternersson wrote:They are on the air exit side.
The thought is that the wings will create a low pressure zone behind it, and draw air out of the motor, and it does work.

Me and a friend have experimented to a great extent this summer.

We got his Cromotor V1 to run the same temp or cooler than my Cromotor V2.

My bike, 55kg + 95kg person. 24" wheel, Cromotor V2 with aluminum stator and Kelly 250A controller.

His bike, 55kg + 100kg person, 26" wheel, Cromotor V1 with steel stator and Kelly 250A controller.

Without the wings, his motor ran 10-30C hotter than mine all the time.
But isn't your motor turning faster at the same speeds since you're in a 24" wheel? Not really an apples to apples comparison is my point
Exactly! My motor spins faster and produces more torque to the ground then my friends motor = mine should run cooler.
Also, the Cromotor V2 has greater efficiency than Cromotor V1, and finally, the aluminium stator makes the V2 cool down faster.

But despite all the above, my friends less efficient motor, running slower, ran colder.

User avatar
cal3thousand   1.21 GW

1.21 GW
Posts: 3606
Joined: Mar 26 2012 4:47pm
Location: California

Re: 25mm PC cooling fans inside Hub motors

Post by cal3thousand » Sep 19 2014 10:26am

Peter Sternersson wrote:
cal3thousand wrote:
Peter Sternersson wrote:They are on the air exit side.
The thought is that the wings will create a low pressure zone behind it, and draw air out of the motor, and it does work.

Me and a friend have experimented to a great extent this summer.

We got his Cromotor V1 to run the same temp or cooler than my Cromotor V2.

My bike, 55kg + 95kg person. 24" wheel, Cromotor V2 with aluminum stator and Kelly 250A controller.

His bike, 55kg + 100kg person, 26" wheel, Cromotor V1 with steel stator and Kelly 250A controller.

Without the wings, his motor ran 10-30C hotter than mine all the time.
But isn't your motor turning faster at the same speeds since you're in a 24" wheel? Not really an apples to apples comparison is my point
Exactly! My motor spins faster and produces more torque to the ground then my friends motor = mine should run cooler.
Also, the Cromotor V2 has greater efficiency than Cromotor V1, and finally, the aluminium stator makes the V2 cool down faster.

But despite all the above, my friends less efficient motor, running slower, ran colder.
Ah, ok. That seems reasonable.

I have varying hole patterns and sizes between left and right, trying to get the same effect. But I don't have a way to test the effectiveness. Maybe a smoke test is in order. Have you ever thought about trying that?
Get a Cycle Analyst and a Multimeter, you're still a noob if you don't have at least one of each.

Planning on posting questions or buying anything on this site? Put up your country (at minimum) on your profile. This is a worldwide forum and we haven't reached clairvoyance.

Peter Sternersson   100 W

100 W
Posts: 145
Joined: Aug 14 2012 2:28am
Location: Ljungby, Sweden

Re: 25mm PC cooling fans inside Hub motors

Post by Peter Sternersson » Sep 19 2014 2:10pm

I have tried smoke in various forms, but I do not have a reliable source of smoke to make a serious judgement.

However, with only holes without any form of wings, the flow through the motor is more or less zero...
And I mean, why should the air that rushes by the covers in 20-50km/h go inside the motor? Just because there are some holes there, naaah...

User avatar
madin88   1 GW

1 GW
Posts: 3056
Joined: May 27 2013 2:02am
Location: Austria

Re: 25mm PC cooling fans inside Hub motors

Post by madin88 » Sep 19 2014 2:33pm

Allex wrote:Interesting, but I dont quite get it Madin, could you make a sketch?
Image
Last edited by madin88 on Sep 22 2014 1:53pm, edited 1 time in total.
- 10kW belt drive with Neumotor 8057: [url]viewtopic.php?
- KTM Hardtail Pedelec / eZee V2 6T / 27,5" / Kelly KBSX @ 1,5kW / Torque PAS / CA-V3 / 13s4p Sanyo GA in Shark case

User avatar
Kent   1 kW

1 kW
Posts: 382
Joined: Jun 04 2011 3:14pm
Location: Vermont, USA

Re: 25mm PC cooling fans inside Hub motors

Post by Kent » Sep 19 2014 4:55pm

Peter Sternersson wrote:I have tried smoke in various forms, but I do not have a reliable source of smoke to make a serious judgement.
Love it. :D
72 V, (2x36v, 11.5 ah in series) Cellman, HS3540, 18fet Lyen, Jamis Dirtjumper Hardtail. Its streetlegal w license plate, inspection sticker.

Only by setting an example of good road manners and public courtesy can we convince the general public to take the electric bike seriously as a transportation option. Obey traffic laws!

My build: http://endless-sphere.com/forums/viewto ... =6&t=45706
Going legal: http://endless-sphere.com/forums/viewto ... =3&t=49390

User avatar
Cowardlyduck   100 MW

100 MW
Posts: 2736
Joined: Jun 26 2011 8:41pm
Location: Canberra, ACT, Australia

Re: 25mm PC cooling fans inside Hub motors

Post by Cowardlyduck » Sep 19 2014 6:29pm

madin88 wrote:Image
Thanks for that. It makes so much more sense now. :lol:

I think this could work actually...the tricky part would be blocking the flow without having the block touch the side covers. I could see using some plastic here as the best option, so it doesn't matter if it does touch the side covers slightly.

I might well try this on my next motor as suggested. :D

Cheers
My stuff:
Ebike Garage - My YouTube Ebike ramblings.
Cowardlyducks - Videos - Ride videos
BikeE recumbent commuter - 9C, 6Fet Grinfineon, internal 17AH, 52V, + on-board solar.

For Sale:
Custom made 18650/21700 battery packs - Australia only
High Power LiPo wiring harnesses - 4P - XT90, HXT4mm, 5.5mm. 200Amp+ capable. Global shipping.

User avatar
Offroader   10 MW

10 MW
Posts: 2489
Joined: Sep 08 2013 9:03pm
Location: USA

Re: 25mm PC cooling fans inside Hub motors

Post by Offroader » Sep 19 2014 6:49pm

Cowardlyduck wrote:
madin88 wrote:Image
Thanks for that. It makes so much more sense now. :lol:

I think this could work actually...the tricky part would be blocking the flow without having the block touch the side covers. I could see using some plastic here as the best option, so it doesn't matter if it does touch the side covers slightly.

I might well try this on my next motor as suggested. :D

Cheers
Please do try it!, I need to figure out what I am going to do for my 2nd cromotor. Can't make up my mind if I should go oil or air cooling.

John in CR   100 GW

100 GW
Posts: 14255
Joined: May 20 2008 12:58am
Location: Paradise

More on John's approach to hubmotor cooling

Post by John in CR » Sep 19 2014 11:35pm

If you're going to take the risk of putting oil in your motor, then be sure to add lots of surface area to the outside of the covers. Then it will truly accomplish a lot. More surface area means proportionately more heat dissipation at the same temp. If that new surface area is done in the form of blades, then it can also create a greater and more turbulent flow over the surface, which will also significantly increase heat dissipation at the same temp.

Without adding motor shell surface area the effect of adding oil is mostly just to delay the eventual heat up as GCinDC documented quite well. It helps increase cooling in intermittent operation but to be the effect is unimpressive. While it conducts heat more quickly to the covers, you're still stuck with the same surface area, which means to dissipate more heat the covers must be hotter. Guess what, that means your magnets must get hotter too.

Oil fill is definitely for moderate power builds unless you go the extra step and add something like double the surface area of the outside shell.


Fans sure, but people continue to use axial fans instead of radial fans (blowers), which will be far more effective because they're better able to create pressure in tight turbulent spaces where axial fan flow disintegrates. Plus with the blowers you direct the flow at the stator instead of the other side cover, so a double barrel benefit over axial fans. At low speed and forcing air flow while stopped is where interior fans pay the biggest dividends.

Holes in side covers help, but really only to the extent they let hot air leak out a bit more freely, especially when stopped, and maybe create a bit more turbulence at the windings. The most copied approach does very little compared to the risk of something getting into the motor with no way out. How many fans have you seen that consist of a relatively flat spinning plate with circular holes? Answer exactly NONE. I've tried hole shapes that I thought should create the right pressure differentials and a number of hole placements, and while all worked with each newer adaption working better with consideration always to minimize the chance of ingesting something problematic, NONE FLOWED MUCH AIR. Interior blades helped, but still without significant and meaningful air flow through the motor, it always remained possible to to overheat the motor at high power without getting abusive. I can overheat any motor in the right conditions and everyone should learn that limit for their own system and conditions.

Do yourselves a favor and do some research regardless of the cooling approach. With air cooling the numbers will quickly tell you that you need substantial fresh air flowing through the motor to carry meaningful heat out with it. I run my hubmotor at 27kw peak input, and while it's quite a bit more efficient than the common hubbies used around here, going from 16kw peak input with a stock motor and having to take heat generation into account while riding demanded serious cooling to add the additional 11kw via a 50% voltage increase and a 10% increase in current. Even riding at the same speeds as before required more cooling, because the higher voltage shifted the lower efficiency range of operation to higher rpm.

First I put about 100 narrow slots at the motor's perimeter on the exhaust side, and opened up the intake side as much as I dared. My hope was that these slots at the corner would act as small centrifugal blades. They did and it did flow enough air to feel with my hand very close and wheel off the ground, but we're talking about 10's of cfm not the hundreds needed. I was prepared for that result, and while I had the motor apart I drilled and threaded bolt holes near 6 extra long slots I cut in the shell, so I could experiment with exterior blades that also extended inward to also get the interior blade benefits I've obtained before.

The end result speaks for itself. Me and my bike together weigh about 190kg. The shell of my motor has less surface area than most DD hubmotors used here, yet despite running higher power through a hubmotor than anyone around, and pushing a much heavier load than most, I rarely even turn my temp sensor and alarm on. When I do it's only out of curiosity. A friend has a dense smoke machine for finding air system leaks in cars, and it showed the motor both sucking significant air into the motor, as well as a surprising by-product of sucking a lot of air across the motor through the rim holes. That meant it was providing extra cooling of the magnet backing rim from the outside with extra air flow from left to right. I didn't video it primarily because when I moved the hose further than about 1cm from the intake side, the turbulence of the spinning tire made the smoke mix with the air too much to really see it being sucked in, but the generally smoky air on the other side told me it was still drawing in. When I ventilate my next one I'll make a collar using a large diameter of cardboard to move the intake well away from the wheel turbulence and really show how strong the air flow is.

The changes I make will be to make only as many exhaust holes in the motor as I have blades, probably 8 next time. I'll cut the slots to secure the blades at an angle instead of straight radial, so the blade protects the hole from debris (I've never gotten any debris in any of my ventilated motors, and I go small with holes and/or slots to keep it that way). To counter the reduce flow effect of the angled blades, I'll put a forward curve near the blade end to make the tangent at the tip parallel with a straight radial. This will help the blades fling more air away at the blade tips, which is how centrifugal fans work...ie flinging air away from the blade edge creating low pressure behind the blade.

Here's a picture of one of the blades made from rust resistant 1"x1" thin angle steel used in home roofing. I cut them simply with tin snips. The silver part is outside of the motor. The blade edge is at the bottom in the pic. The green is the thickness of the motor shell and the black and yellow protrude inside almost touching the motor end windings.
HubMonster blade.JPG
HubMonster blade.JPG (34.78 KiB) Viewed 3509 times
Here's what they look like mounted on the motor. Note that the angle steel give it good rigidity, and the side facing portion at the perimeter helps prevent air from rushing in to fill the low pressure area behind the blade. I did bend those sides slightly away from the motor thinking that would open up more of the slot behind the blade as well as push air away from that side.
Hubmonster slots and blades.JPG
Hubmonster slots and blades.JPG (26.92 KiB) Viewed 3509 times
I apologize for so little documentation, but this started as a test, and it's worked so well I haven't done anything further. The best examples of how well it works are:
1. With the motor already at operating temperature of 50-60°C, I did a high speed climb with the first few km averaging 5% grade, and the last 2.5km a solid continuous 20% very curvy and accelerating hard from very slow in each curve to up to above 80kph when there was enough straight to allow it. It was my first time up the road so I was extremely slow in each curve. The stator only got to 105°C, the hottest I've ever had it.
2. On a 15km ride mostly highway including one 2km 7% climb and back down the other side to a long slight grade into a headwind ending at a 200m net increase in grade, and speeds never below 90kpm after entering the highway and showing off for cars a couple of times up to about 120kph, stator temp very gradually climbed to 68°C when I parked. I haven't worried about motor temps since.

If I used a hubmotor for an offroad bike sure I'd use fans in the motor due to the low speeds, but they'd definitely be radial fans, not axial fans like everyone mistakenly uses. For the motor to act as a centrifugal fan itself, the higher the rpm the better. However, an offroad bike for me would see Pacific beach use too, so I'd go mid-drive using similar ventilation, but put a shell over the motor to provide a protected intake and an exhaust duct common of centrifugal blowers all in an effort to keep salt and sand out. With a hubmotor I'd get someone to build me a Lebowski controller or two, so I could forget about the halls, and epoxy up everything very well, add lots of outside surface area by welding AL blades to the covers, and fill it to an inch or two below the axle with ATF.

One more word about blades. People often mention exterior blades in the form of scoops or whatever. They're all just axial blades, and a waste of time. It's totally impractical to get sufficient blade size for the operating rpm, and even if you did the intake side would deflect sand, dust, and debris directly at the intake side of your motor. OTOH at the motor diameter and rpms we operate a well designed centrifugal fan can create enough pressure to move a lot of air. I can feel the exhaust side of mine throw off more air than the intake even from a couple of feet away from the wheel when spinning it up no load, so I know I moving air. While it's nowhere close to optimized, I quit because it works well enough and I can barely detect any increase in no-load current, so even at high speed the power cost has to be less than the reduced resistance of a cooler motor.

I'll do a good thread with video and lots of pics someday, but I'm really slow with that kind of stuff.

John

PS- Please ignore all of my posts related to air cooling prior to 2013 when I implemented this exterior blade approach. Much of it was accurate, especially the use of interior blades and putting exhaust holes at the extreme perimeter of the motor along with one sided intake and exhaust to get air flow through the magnetic gap where it is badly needed. The bladed approach is so each and effective and only needs a fairly small number of exhaust holes to function far better than anything short of putting a noisy high power blower pushing massive air at and through the motor like Toolman2 did. He shared even less detail than I have other than to measure the effect of 1000W reduced resistance in the windings of his extremely stress RC motor.
Last edited by John in CR on Oct 01 2014 11:34am, edited 1 time in total.

User avatar
madin88   1 GW

1 GW
Posts: 3056
Joined: May 27 2013 2:02am
Location: Austria

Re: 25mm PC cooling fans inside Hub motors

Post by madin88 » Sep 20 2014 4:46am

Offroader wrote: Please do try it!, I need to figure out what I am going to do for my 2nd cromotor. Can't make up my mind if I should go oil or air cooling.
i guess i will do, but certainly not with my cromotor. Its because i want to be able to ride with my streetbike in rain. For my next high powered DD ebike build i will take the improved MXUS 3000 from e4bike.ru
the russians claim the performance will be close to cromotor but about 3kg less heavy (less unsprung weight is always good - especially with small wheels and bicycle frame / suspension).
I'm a bit worried if there is room for additional wires for the fans inside the small axle the MXUS has, but maybe the russians will also improve this part of the motor.

Oil cooling will transport heat very fast to the motor case, but as John in CR mentioned i also guess the surface area of ususal hub sidecovers is to small to shed the heat quick enough.

improved sketch:

Image

maybe this type of RC Fan is a good choice:

Image

http://www.gws.com.tw/english/product/p ... /edf30.htm
Last edited by madin88 on Sep 22 2014 1:54pm, edited 1 time in total.
- 10kW belt drive with Neumotor 8057: [url]viewtopic.php?
- KTM Hardtail Pedelec / eZee V2 6T / 27,5" / Kelly KBSX @ 1,5kW / Torque PAS / CA-V3 / 13s4p Sanyo GA in Shark case

User avatar
madin88   1 GW

1 GW
Posts: 3056
Joined: May 27 2013 2:02am
Location: Austria

Re: 25mm PC cooling fans inside Hub motors

Post by madin88 » Sep 20 2014 5:04am

John in CR wrote: I apologize for so little documentation, but this started as a test, and it's worked so well I haven't done anything further. The best examples of how well it works are:
1. With the motor already at operating temperature of 50-60°C, I did a high speed climb with the first few km averaging 5% grade, and the last 2.5km a solid continuous 20% very curvy and accelerating hard from very slow in each curve to up to above 80kph when there was enough straight to allow it. It was my first time up the road so I was extremely slow in each curve. The stator only got to 105°C, the hottest I've ever had it.
2. On a 15km ride mostly highway including one 2km 7% climb and back down the other side to a long slight grade into a headwind ending at a 200m net increase in grade, and speeds never below 90kpm after entering the highway and showing off for cars a couple of times up to about 120kph, stator temp very gradually climbed to 68°C when I parked. I haven't worried about motor temps since.
Tanks for the pics. Good Job! I can imagine those big blades will move lots of air through the motor when it spins.
I guess the inlet holes are on the other sidecover?
- 10kW belt drive with Neumotor 8057: [url]viewtopic.php?
- KTM Hardtail Pedelec / eZee V2 6T / 27,5" / Kelly KBSX @ 1,5kW / Torque PAS / CA-V3 / 13s4p Sanyo GA in Shark case

User avatar
madin88   1 GW

1 GW
Posts: 3056
Joined: May 27 2013 2:02am
Location: Austria

Re: 25mm PC cooling fans inside Hub motors

Post by madin88 » Sep 20 2014 10:12am

for the airduct the lids or sealing covers used for larger plastic tubes would work.

Image

Image

http://www.kunststoffrohre-einecke.de/E ... enstopfen/
http://www.kunststoffrohre-einecke.de/E ... /KG-Kappe/

maybe this can be used as cooling fins:

Image

http://alutronic.com/products/heat-sink ... 247?c=1360
Last edited by madin88 on Sep 22 2014 1:55pm, edited 1 time in total.
- 10kW belt drive with Neumotor 8057: [url]viewtopic.php?
- KTM Hardtail Pedelec / eZee V2 6T / 27,5" / Kelly KBSX @ 1,5kW / Torque PAS / CA-V3 / 13s4p Sanyo GA in Shark case

John in CR   100 GW

100 GW
Posts: 14255
Joined: May 20 2008 12:58am
Location: Paradise

Re: 25mm PC cooling fans inside Hub motors

Post by John in CR » Sep 20 2014 11:20am

madin88 wrote:Oil cooling will transport heat very fast to the motor case, but as John in CR mentioned i also guess the surface area of ususal hub sidecovers is to small to shed the heat quick enough.
It's not so much that it's too small. It's that it's exactly the same surface area as a stock motor, so at the same temperature of that surface it will dissipate exactly the same amount of heat. Yes it heats up the side covers more quickly, but in stock form covers get quite hot too, so as GCinDC's graphs showed the motor is much cooler than stock early on, but the bottleneck rejecting the heat off of the covers shows up as a steadily climbing stator temp. It probably does end up somewhat hotter at the surface than stock, but that means only somewhat more heat rejection, not multiples of it.

Another example is Itchynackers climb up Pikes Peak. He was only running a few Kw, and with oil in the hub he still had to back off and pedal harder with the temperature pushing to 130° or so if I recall correctly. That was in near freezing temperatures. A relatively easy doubling of the surface area would have guaranteed a far cooler motor and no heat concerns at the power he ran.

John in CR   100 GW

100 GW
Posts: 14255
Joined: May 20 2008 12:58am
Location: Paradise

Re: 25mm PC cooling fans inside Hub motors

Post by John in CR » Sep 20 2014 12:11pm

Ducted fans push pressure well, though I couldn't live with the noise. They do rely on a free air intake, which is the reason for the smooth taper on the intake, so do some testing to make sure they're don't end up starved of air, which will drastically decrease the flow rate. I'd also put a 90° elbow on the output to direct the wind toward the windings and at an angle to compliment the natural flow in the direction rotation.

It also won't end up with flow over the windings on the side with holes, so at a minimum add blades to the inside cover on the intake/exhaust cover to ensure turbulence at the end windings on that side.

Overall though, why bother with all that? What current and voltage do you plan? I've run a similar 40mm wide stator motor (not Xlyte overpriced garbage) with interior blades on both side covers, intake via a bunch of small holes on the wire side, exhaust on the other side with a number of 10mm hole AT the perimeter, but prior to my discovery of highly effective exterior blades. That motor has been running for 3 years at 140A peak battery current on 74V nominal, and has never gotten really hot, so I never bothered with a temp sensor. That's running a 16rpm/v winding, so less efficiency during acceleration, and many mountain climbs. Yes the 20" wheel is an advantage, but the load of mountain roads more than offsets that. Exterior centrifugal blades would make it capable of bigger loads and power even in a larger wheel. No new wires to run or fans to install. The "fan" can't break and leave me heat stranded, and best of all the only added noise is the little bit from having the motor open.

Low rpm + high load + high power is where I believe fans become an advantage.

Regarding your rain worry, to my knowledge no one has had such a failure other than those with stock seal motors that end up with water trapped inside when ridden to often commuting in the rain. Cold water hitting a hot motor cause the air inside to contract and suck water in. Zappy had some fans short out riding his fan driven vented motor through small streams in muddy offroad riding. Salt on snowy/slushy roads or riding at the beach would be my only concern with a vented motor, and a high quality coating on exposed metal with exhaust AT the perimeter to allow a rinse with hose through the intake holes could eliminate those concerns, though I would still avoid submersing in salty water like happens to me at the beach. I wouldn't beg for failure with the big holes so commonly used, especially with none at the perimeter to allow things that get in to flow right back out. I too used to have similar concerns as you regarding opening a motor up, but the lack of failures combined with buying a used sealed motor that upon arrival was seized from so much rust filling the gap between the stator lams and the magnet backing ring....Be careful buying a used motor from a year-round northern commuter.
madin88 wrote:.....
improved sketch:

Image

maybe this type of RC Fan is a good choice:

Image

http://www.gws.com.tw/english/product/p ... /edf30.htm

User avatar
madin88   1 GW

1 GW
Posts: 3056
Joined: May 27 2013 2:02am
Location: Austria

Re: 25mm PC cooling fans inside Hub motors

Post by madin88 » Sep 20 2014 1:02pm

John in CR wrote:Ducted fans push pressure well, though I couldn't live with the noise. They do rely on a free air intake, which is the reason for the smooth taper on the intake, so do some testing to make sure they're don't end up starved of air, which will drastically decrease the flow rate. I'd also put a 90° elbow on the output to direct the wind toward the windings and at an angle to compliment the natural flow in the direction rotation.
sure the noise is bad, but i only would turn them on if i want to go crazy with the bike or when climbing steep hills. During normal riding or my commute it is not necessary for me.
The biggest advantage of having fans inside the motor is the possibility to cool down the motor pretty quick after hard abuse, or when standing still during lots of stop and go traffic. without, it takes half an hour or more..
Thanks Cowardlyduck for the demonstration.
It also won't end up with flow over the windings on the side with holes, so at a minimum add blades to the inside cover on the intake/exhaust cover to ensure turbulence at the end windings on that side.
thats a good idea. the blades (i guess you mean on the inside of the cover?) will move air through the motor and over the end turns if the fans are off and give assist when they are on.
- 10kW belt drive with Neumotor 8057: [url]viewtopic.php?
- KTM Hardtail Pedelec / eZee V2 6T / 27,5" / Kelly KBSX @ 1,5kW / Torque PAS / CA-V3 / 13s4p Sanyo GA in Shark case

John in CR   100 GW

100 GW
Posts: 14255
Joined: May 20 2008 12:58am
Location: Paradise

Re: 25mm PC cooling fans inside Hub motors

Post by John in CR » Sep 20 2014 2:32pm

The way I look at interior blades is that air flow takes the path of least resistance, so it will tend to hug the smoother side covers. I use the blade to ensure the air is spinning at motor speed, and the blades force a restriction right at the stator. It pushes the spinning flow away from the side covers toward the end windings, and because the space between the blade tips and the stator is much smaller than the space between the stator and side cover, it is accelerated and more turbulent. That absolutely and significantly increases the coefficient of convective heat transfer. Think of the difference between blowing on your cup of hot coffee from 2 feet away vs from 2 inches away, or better yet the difference between drinking coffee that's too hot or cooling it with extra airflow right as it's going in your mouth by sipping it. It's not a subtle difference.

It makes little difference if you have little or no fresh air flow, because the air trapped inside can't hold much heat at all due to the low mass and low heat capacity, so the air gets up to max temp either way, but when you have flow then it helps get the air hotter and takes it more directly from the source instead of where it spreads.

I definitely like the ducted fan over centrifugal blowers for generally OFF usage...assuming the non-free intake doesn't harm the air flow rate. That's because they will allow air to flow relatively freely between the blades when off.

I'd suggest the following change to your flow path and wait to bother even installing the ducted fans:
Put your perimeter exhaust on the opposite side as intake. Intake via many small holes in the intake side located from near the center out to even with the interior perimeter of the stator steel. Some will be outside of your cylindrical duct, and those would intake air to flow over the end windings on the intake side. Bolt on some interior blades on both covers. If you're resistant to the look of exterior blades, then try it first with only the interior blades. Evaluate results. Add exterior blades if you want more cooling. Evaluate results. If you have enough high load low rpm demand that heat is still an issue, then add the fans and intake duct. Evaluate, especially checking the intake side cover near the perimeter and between the spoke flanges, which if hot would indicate that the fans are stuffing up the flow through the magnetic gap, in which case simply add some exhaust holes at the intake side perimeter. Don't put exterior blades on that side because it would reduce flow through the gap created by the exterior chain side blades when the fans are off.

The hard part is changing the wiring harness anyway, which is warranted the first time you open the motor, so add your fan wires then. You could go ahead and go whole hog and just do all of the steps I outlined all on the first pass. I like to open motors either never, or only once, but I don't like to take the motor off the bike once installed, so it's do everything all at once for me. Actually I get motors with as much wire as will fit in the axle, so I'd just do the centrifugal flow approach, and if the motor demanded active fans, then I do that on the next motor and do some more things to improve flow on the current motor without removing the motor. That would be doing an air dam on the intake side to increase pressure outside of the intake. Doing so would also prevent the natural air flow at bike speed that is flowing at 90° to the intake direction that hurts flow without a doubt. Hillsofvalp did that using a bowl shaped metal piece fount under spiral electric stovetop burners. He just cut nearly half of one side off to face forward and allow air in. I'd probably go bigger and not use something with a curve that could deflect sand and rocks at the intake side cover. I haven't needed it, so I haven't put much thought into it, but I know an air dam would increase the air flow through the motor.

User avatar
Cowardlyduck   100 MW

100 MW
Posts: 2736
Joined: Jun 26 2011 8:41pm
Location: Canberra, ACT, Australia

Re: 25mm PC cooling fans inside Hub motors

Post by Cowardlyduck » Sep 20 2014 6:10pm

Yikes, just catching up on this thread now...John take it easy, we all have our own approaches to some degree.
I don't disagree with most of your points, however you do need to consider the type of riding we each do.

Most of the riding I do that risks overheating the motor is slow off-road riding often followed by stopping. A typical ride like this involves pulling 50-80+ Amps for 5-10min to get up a steep hill, then stopping at the top, or going down the other side with regen.
For this type of riding, adding blades like you have will only help so much, and I think running fans like I have is the better approach. Both together would be even better. :D

I suggest if you really want to demonstrate the effectiveness of your approach (which I do believe can be effective), do some tests and graph the results similar to what I have on the previous page. Only then can you definitively prove your approach without anyone saying otherwise. As the saying goes, a picture (graph) paints a thousand words. :D

Cheers
My stuff:
Ebike Garage - My YouTube Ebike ramblings.
Cowardlyducks - Videos - Ride videos
BikeE recumbent commuter - 9C, 6Fet Grinfineon, internal 17AH, 52V, + on-board solar.

For Sale:
Custom made 18650/21700 battery packs - Australia only
High Power LiPo wiring harnesses - 4P - XT90, HXT4mm, 5.5mm. 200Amp+ capable. Global shipping.

User avatar
RLT   10 kW

10 kW
Posts: 683
Joined: Oct 02 2007 6:47pm
Location: Ruidoso, NM; USA
Contact:

Re: 25mm PC cooling fans inside Hub motors

Post by RLT » Sep 20 2014 10:54pm

Cowardlyduck wrote:......Anyway, I won't ramble on too much, and try and let the video/graph speak for themselves........
Nicely done test. Thank you

Post Reply