Definitive Tests on the Heating and Cooling of Hub Motors

Get all your technical information about electric bikes here.
Jestronix
10 kW
10 kW
Posts: 505
Joined: Jul 18, 2015 4:59 am

Re: Definitive Tests on the Heating and Cooling of Hub Motors

Post by Jestronix » Dec 30, 2017 7:08 am

I'm sill amazed to this day that mxus etc don't make a from factory heatsinked hub! With a researched design for capturing maximum airflow. Makes me think most hubs are built for the China market?

I'm going to look at heat sinking the side covers, however just my simple mod has almost cured my main heat concerns. I only hit high 90s on my climbs today, however I'll by moving towards 10kw on a smaller diameter moto rim.

User avatar
amberwolf
100 GW
100 GW
Posts: 25055
Joined: Aug 17, 2009 6:43 am
Location: Phoenix, AZ, USA, Earth, Sol, Local Bubble, Orion Arm, Milky Way, Local Group
Contact:

Re: Definitive Tests on the Heating and Cooling of Hub Motors

Post by amberwolf » Dec 30, 2017 2:00 pm

Cowardlyduck wrote:
Dec 30, 2017 5:51 am
They apparently push 5.13CFM each, which is pretty impressive for a 25mm fan:
But I'd bet thats' only wiht no obstructions in front or behind. Put a dust filter on either end and I doubt you'd get much airflow (lot of noise though).

User avatar
DasDouble
100 kW
100 kW
Posts: 1097
Joined: Jul 12, 2015 1:35 pm
Location: GERMANIA :D

Re: Definitive Tests on the Heating and Cooling of Hub Motors

Post by DasDouble » Dec 30, 2017 3:12 pm

I have asked because that´s the only thing which keeps me away from drilling some easy holes inside the side cover as I don´t want the motor to get damaged by stuff like dirt etc... :roll:
"..the crazy kid exploring hipower ebike world ^^"
-Built own ebike: ✔
-Built own 3.2kWh LiIon battery (300Amps): ✔
-Overtake Porsche with own Ebike: ✔
-Take off with ebike: ✖

~14kW, 3.2kWh LiIon, QS 205 8ml ff, hubsink

Bullfrog
10 W
10 W
Posts: 65
Joined: Apr 08, 2017 10:37 am

Re: Definitive Tests on the Heating and Cooling of Hub Motors

Post by Bullfrog » Dec 30, 2017 3:42 pm

Assuming you are running a Direct Drive motor, I'd suggest you use ATF or Ferrofluid to conduct the heat to the case and then use a ducted fan (without a filter) to blow air over the motor. All of the things that work with a Direct Drive motor also work with an Internally Geared Hub Motor they just have to be applied appropriately.

Adding sinks (fins) and spraying a mist of water on the sinks when the motor gets hot will give you additional cooling.

If it were me, I'd first add fluid to the motor to thermally connect the motor to the housing, second I'd add sinks, third I'd add a fan and/or water misting.

Going to a smaller wheel will increase the motor efficiency and the way you ride can be a factor...try not to use full throttle at very slow speeds and never use your brake and your throttle at the same time. Your motor is most efficient at the rpm it attains when you pick the back wheel up and hold the throttle wide open. At rpms less than that it is less efficient and produces more heat. This is a broad general statement with a lot of caveats.

You can make the problem of heat rejection extremely difficult or very simple....keeping it towards the simple end should make it more reliable and less expensive.

A lot depends on the environment where you ride....I ride off road in the dirt and dust 100% of the time.

If you are going to ride on the pavement in a relatively nice clean environment, then cutting holes in the housing WILL help cooling. But, there is always a but...then you can't use ATF or Ferrofluid and you will always get some dust and dirt in your motor.

Only you can decide what is best for you and the way you want to ride.

User avatar
DasDouble
100 kW
100 kW
Posts: 1097
Joined: Jul 12, 2015 1:35 pm
Location: GERMANIA :D

Re: Definitive Tests on the Heating and Cooling of Hub Motors

Post by DasDouble » Dec 31, 2017 5:24 am

Yeah I see, I already have ff and the hubsinks. I could now theoretically add another pair of hub sinks, that could help too..

Another question: Do you guys think, or maybe someone already did that, that some more CPU fans, placed on the whole sideplates of the hubbie could help cooling the motor? Did someone already tried that?
"..the crazy kid exploring hipower ebike world ^^"
-Built own ebike: ✔
-Built own 3.2kWh LiIon battery (300Amps): ✔
-Overtake Porsche with own Ebike: ✔
-Take off with ebike: ✖

~14kW, 3.2kWh LiIon, QS 205 8ml ff, hubsink

Bullfrog
10 W
10 W
Posts: 65
Joined: Apr 08, 2017 10:37 am

Re: Definitive Tests on the Heating and Cooling of Hub Motors

Post by Bullfrog » Dec 31, 2017 8:11 am

More sinks will help.

One suggestion is to use an infrared thermometer to measure the surface temps of your motor and find the hottest area and direct the fan or fans toward that area. Also if you position the fans slightly forward and direct them aft, the velocity of the fan is added to the velocity of the bike...if you point them entirely left or right then you don't get the additive effect. In theory more fans would be better but I can't guarantee it...best way to find out is try it. Theory and reality don't always agree :D .

If anyone hasn't seen it, the Grin Tech web site has some good info on ferrofuluid, how much to use, and how much it helps. Their motor simulator is a good tool as well and it has the capability to model with and without ferrofluid.

Probably the best way we can all benefit is if everybody will post what they have done and the results. One experiment is more valuable than all of my hypothesizing.

User avatar
DasDouble
100 kW
100 kW
Posts: 1097
Joined: Jul 12, 2015 1:35 pm
Location: GERMANIA :D

Re: Definitive Tests on the Heating and Cooling of Hub Motors

Post by DasDouble » Dec 31, 2017 9:19 am

Well I have posted it someowhere before I think but here again my results:

Faster cooling while standing still with ff and hubsinks then without.
I have 8 ml ff inside. Some folks said 5-7 would be the best for the QS 205V3. More would add some resistance between hub and outher shell.
"..the crazy kid exploring hipower ebike world ^^"
-Built own ebike: ✔
-Built own 3.2kWh LiIon battery (300Amps): ✔
-Overtake Porsche with own Ebike: ✔
-Take off with ebike: ✖

~14kW, 3.2kWh LiIon, QS 205 8ml ff, hubsink

Bullfrog
10 W
10 W
Posts: 65
Joined: Apr 08, 2017 10:37 am

Re: Definitive Tests on the Heating and Cooling of Hub Motors

Post by Bullfrog » Dec 31, 2017 10:38 am

Good info, thanks.

Where did you buy your QS motor and the sinks and what is your opinion of the QS motor? This article makes the QS205/50H V3 sound like it is the best option: https://www.electricbike.com/large-hot-rod-hubmotors/

I may try one at some point...just FYI, EM3ev may start carrying them in the very near future. They have been good to deal with on my BBSHD and my MAC but I am having a little trouble getting a battery problem resolved.

I currently have one bike running a MAC IGH motor and I had to use 3.5 ounces (104 ml) to maximize my cooling. Any more than 3.5 ounces in my motor and I would start leaking...a lot smaller diameter than most DD motors and a bigger gap between the motor and the housing. Running a BBSHD on another bike. I ride off road 100% of the time so the weight of a big DD is tough to swallow at this point.

User avatar
DasDouble
100 kW
100 kW
Posts: 1097
Joined: Jul 12, 2015 1:35 pm
Location: GERMANIA :D

Re: Definitive Tests on the Heating and Cooling of Hub Motors

Post by DasDouble » Dec 31, 2017 1:44 pm

Got the ff from merlin, the QS from Artur and the hubsinks from Sketchism. Well I can´t tell a difference between the QS and other brands as I have only experiences with the QS.

So you mean you have a liquid cooled motor? Awesome :D
"..the crazy kid exploring hipower ebike world ^^"
-Built own ebike: ✔
-Built own 3.2kWh LiIon battery (300Amps): ✔
-Overtake Porsche with own Ebike: ✔
-Take off with ebike: ✖

~14kW, 3.2kWh LiIon, QS 205 8ml ff, hubsink

Bullfrog
10 W
10 W
Posts: 65
Joined: Apr 08, 2017 10:37 am

Re: Definitive Tests on the Heating and Cooling of Hub Motors

Post by Bullfrog » Dec 31, 2017 4:54 pm

Well I guess the motor is liquid cooled :) but the case is still air cooled :( .

Haven't had a chance to run it much due to a battery problem that is currently being worked between me and EM3ev but my initial testing showed that the ATF helped a lot. My MAC came prewired with temp sensors and I am running a Cycle Analyst to display the temps...the section that used to get my motor to 105 C now gets it to 83 C so the ATF works great for my particular application.

Jestronix
10 kW
10 kW
Posts: 505
Joined: Jul 18, 2015 4:59 am

Re: Definitive Tests on the Heating and Cooling of Hub Motors

Post by Jestronix » Jan 02, 2018 6:37 am

So impressed with what a simple fan mod has done, I moved to a smaller fan and got the same/ similar results and reduced weight so slightly :) I also found running the fan all the time gave the best results, rather than when things get stinking hot, eg 70c up.

Be great to see some results in Justin's test rig :) with hub sinks too :)!

Bullfrog
10 W
10 W
Posts: 65
Joined: Apr 08, 2017 10:37 am

Re: Definitive Tests on the Heating and Cooling of Hub Motors

Post by Bullfrog » Jan 02, 2018 4:15 pm

Everybody needs to remember that Jestronix is running a fluid (can't remember if it is ATF or FF) which thermally connects the motor to the housing.

Without the fluid, I don't think the fan would help nearly as much.

User avatar
DasDouble
100 kW
100 kW
Posts: 1097
Joined: Jul 12, 2015 1:35 pm
Location: GERMANIA :D

Re: Definitive Tests on the Heating and Cooling of Hub Motors

Post by DasDouble » Jan 02, 2018 4:43 pm

So thats the next step further from now on? First ff, then hubsinks and then active cooling by a fan? Awesome, I like progress :)
"..the crazy kid exploring hipower ebike world ^^"
-Built own ebike: ✔
-Built own 3.2kWh LiIon battery (300Amps): ✔
-Overtake Porsche with own Ebike: ✔
-Take off with ebike: ✖

~14kW, 3.2kWh LiIon, QS 205 8ml ff, hubsink

markz
100 GW
100 GW
Posts: 5244
Joined: Jan 09, 2014 11:38 pm
Location: Alberta Canada

Re: Definitive Tests on the Heating and Cooling of Hub Motors

Post by markz » Jan 02, 2018 6:22 pm

More aluminum mass on the inside, heat sinks between the flanges, ferrofluid, pc fan then the last resort would be drilling holes.

You could add the alum mass yourself, balancing it all out.

User avatar
DasDouble
100 kW
100 kW
Posts: 1097
Joined: Jul 12, 2015 1:35 pm
Location: GERMANIA :D

Re: Definitive Tests on the Heating and Cooling of Hub Motors

Post by DasDouble » Jan 03, 2018 5:16 pm

markz wrote:
Jan 02, 2018 6:22 pm
More aluminum mass on the inside, heat sinks between the flanges, ferrofluid, pc fan then the last resort would be drilling holes.

You could add the alum mass yourself, balancing it all out.
Well my friend, you have just brought some crazy good idea right there...! More aluminium mass on the inside.. Why did I not came to the same idea some earlier? Not only on the outside, but the inside also! Crazy good idea, crazy good.. :!:
Cheers
"..the crazy kid exploring hipower ebike world ^^"
-Built own ebike: ✔
-Built own 3.2kWh LiIon battery (300Amps): ✔
-Overtake Porsche with own Ebike: ✔
-Take off with ebike: ✖

~14kW, 3.2kWh LiIon, QS 205 8ml ff, hubsink

User avatar
mrbill
10 kW
10 kW
Posts: 516
Joined: Jun 10, 2008 6:47 pm
Location: Silicon Valley, California
Contact:

Re: Definitive Tests on the Heating and Cooling of Hub Motors

Post by mrbill » Jan 23, 2018 1:36 pm

I recently built a 406 wheel around an Edge1500 motor (35mm stator). I started with about 4.5 mL of FF, but I found that the rolling resistance felt significantly greater than with the same amount of FF in a motor build into a 559 wheel. The rolling resistance was also noticeable as reduced measured efficiency.

I suspect that one can get away with less FF in the motor, making a smaller thermal bridge across the gap that sweeps around the stator at a higher rate. Moreover, the smaller wheel effectively gears-down the motor, letting it operate more of the time with greater efficiency and to produce less heat. I get higher peak regeneration current for a given speed with the smaller wheel, too: at 32 kph, 1.5kw-1.7kw in a 406 wheel vs. 800-1000 watts in the 559 wheel.

I'm currently running with about 3.5 mL, and I notice the reduced rolling resistance--easier to pedal when the motor is idle. Thermal performance appears to be about the same as with 4.5mL, but I haven't been able to test thermal performance in hot weather yet. Maximum observed temperature inside the motor has been just over 50C.

Has anyone discovered how much less ferro-fluid (FF) one can use in a 406 wheel and still achieve similar thermal characteristics to that of a hub motor built into a standard 559 or 622 wheel? I'm trying to find a balance between sufficient heat-shedding and overall efficiency.

Thanks.
Bill Bushnell
http://mrbill.homeip.net/
Bike build thread can be seen here.

User avatar
amberwolf
100 GW
100 GW
Posts: 25055
Joined: Aug 17, 2009 6:43 am
Location: Phoenix, AZ, USA, Earth, Sol, Local Bubble, Orion Arm, Milky Way, Local Group
Contact:

Re: Definitive Tests on the Heating and Cooling of Hub Motors

Post by amberwolf » Jan 23, 2018 1:54 pm

THe drag vs cooling of the FF has to do with the airgap space in the motor and the magnet/stator width (and spacing between magnets if there is any) rather than the diameter of the wheel, AFAIK. Once there's enough FF to...splash? from the passage of the magnets past the stator teeth, that's where the drag starts to get worse, IIRC from Justin's tests.

His tests were probably done without a wheel at all, just a bare motor, but I dont' remember for sure--you'd have to go further back in the thread for that (or the article on Grin's web page).


Regarding wheel diameter vs rolling resistance, doesn't a smaller wheel offer more RR anyway, vs a larger, all other conditions identical?

User avatar
Rix
100 GW
100 GW
Posts: 6302
Joined: Mar 29, 2012 11:26 am
Location: Fallon NV USA

Re: Definitive Tests on the Heating and Cooling of Hub Motors

Post by Rix » Jan 23, 2018 2:57 pm

Regarding wheel diameter vs rolling resistance, doesn't a smaller wheel offer more RR anyway, vs a larger, all other conditions identical?
Yes, because of reverse mechanical advantages and less lever affect against eddy resistance.

User avatar
Cowardlyduck
10 MW
10 MW
Posts: 2383
Joined: Jun 26, 2011 8:41 pm
Location: Canberra, ACT, Australia

Re: Definitive Tests on the Heating and Cooling of Hub Motors

Post by Cowardlyduck » Jan 23, 2018 4:46 pm

Yeah I have a 406 (20") wheel on my recumbent, and I definitely notice the drag from the ~4ml of FF I added.

The thing is when the motor is cold, the drag is not that bad....once it gets hot however the drag from the FF definitely goes up substantially. I can even hear it rubbing inside the motor at low speeds once hot.

Justin if your still reading, I would really appreciate another test run on the dyno to compare the FF resistance hot vs cold.

Cheers
High Power LiPo wiring harnesses - 4P - XT90, HXT4mm, 5.5mm. 200Amp+ capable. Global shipping.

Modified Stealth Fighter - Force air cooled Leaf motor @ 6KW, Heat-sinked Adaptto Mini-E. Battery = 49AH, 14S (51.1V) Sanyo 18650GA - 2.5Kwhr = 200km range!
Cowardlyducks - Stealth Fighter Videos

BikeE recumbent - Golden motor, 6Fet Grinfineon, internal 17AH, 14S Sanyo 18650GA, + on-board solar.

The stuff I make and modify.

Bullfrog
10 W
10 W
Posts: 65
Joined: Apr 08, 2017 10:37 am

Re: Definitive Tests on the Heating and Cooling of Hub Motors

Post by Bullfrog » Jan 23, 2018 5:00 pm

Check out this site: https://ferrofluid.ferrotec.com/technology/

Might be worth contacting them to discuss your application...I believe you can tailor the FF to manipulate the properties to some extent.

If I remember correctly, Grin did not use much FF to achieve their best results...their results and experiment are posted on their website somewhere, just been a long time since I saw it.

User avatar
Rix
100 GW
100 GW
Posts: 6302
Joined: Mar 29, 2012 11:26 am
Location: Fallon NV USA

Re: Definitive Tests on the Heating and Cooling of Hub Motors

Post by Rix » Jan 23, 2018 9:35 pm

Cowardlyduck wrote:
Jan 23, 2018 4:46 pm
Yeah I have a 406 (20") wheel on my recumbent, and I definitely notice the drag from the ~4ml of FF I added.

The thing is when the motor is cold, the drag is not that bad....once it gets hot however the drag from the FF definitely goes up substantially. I can even hear it rubbing inside the motor at low speeds once hot.

Justin if your still reading, I would really appreciate another test run on the dyno to compare the FF resistance hot vs cold.

Cheers
I wonder if the oil carrier for the ferrofluid is a multi weight oil? Like 5w-30w, it acts like 5W when cold, and 30w when warm/hot.

User avatar
justin_le
10 MW
10 MW
Posts: 2095
Joined: May 28, 2007 3:27 am
Location: Vancouver
Contact:

Re: Definitive Tests on the Heating and Cooling of Hub Motors

Post by justin_le » Feb 07, 2018 11:35 pm

Cowardlyduck wrote:
Jan 23, 2018 4:46 pm
Justin if your still reading, I would really appreciate another test run on the dyno to compare the FF resistance hot vs cold.
Cheers
Oh man, I have SOOO many interesting new tests on Statorade from the last 3-4 months of lab work that I need to share here. But none of them revisiting the drag vs. temperature data, since that was pretty definitive match of reduced viscosity and drag with increasing temperatures.
https://endless-sphere.com/forums/viewt ... 8#p1107308

If you are seeing the opposite effect, more drag when the motor is hot, then there must be something else at play in your system. Have you quantified the no load current draw of your motor when it is hot vs cool?

The fact that you also report hearing sounds from the motor when it is warm further hints to a purely mechanical effect and not one from fluid viscosity.
Big Dummy Frame (yes This One, thanks ES!) with prototype 26" Grin all-axle front hub, Phaserunner controller, and 52V 11Ah Cellman triangle pack
My website: http://www.ebikes.ca
Please contact via email, info@ebikes.ca, rather than PMs, which are disabled

User avatar
Cowardlyduck
10 MW
10 MW
Posts: 2383
Joined: Jun 26, 2011 8:41 pm
Location: Canberra, ACT, Australia

Re: Definitive Tests on the Heating and Cooling of Hub Motors

Post by Cowardlyduck » Feb 08, 2018 5:58 am

justin_le wrote:
Feb 07, 2018 11:35 pm
Oh man, I have SOOO many interesting new tests on Statorade from the last 3-4 months of lab work that I need to share here. But none of them revisiting the drag vs. temperature data, since that was pretty definitive match of reduced viscosity and drag with increasing temperatures.
https://endless-sphere.com/forums/viewt ... 8#p1107308

If you are seeing the opposite effect, more drag when the motor is hot, then there must be something else at play in your system. Have you quantified the no load current draw of your motor when it is hot vs cool?

The fact that you also report hearing sounds from the motor when it is warm further hints to a purely mechanical effect and not one from fluid viscosity.
Thanks for the reply Justin!
And so awesome that this is still being tested by you!

Well, I do hope it's something up with my setup rather than an issue with FF causing increased drag with heat. The strange thing is though, that I see this same behaviour on 3 motors I'm using FF in.
That being said...the one thing these 3 motors have in common is that they are open (vented) to the elements. Two have forced air flow through them (Leaf '1500W / HS4080) and one just has a drilled side cover on one side ('1000W Golden motor, 9C clone).

I'm currently putting together a MXUS 3K for an Alpha build which also has FF added and will also have hubsinks. When I was 'carefully' adding the FF with the syringe I had already added about 6ml when the resistance on the plunger suddenly dropped and the whole 10ml went in. :shock: :lol:
Oh well...at least it will have the best possible thermal bridge which is just as well since I plan on pushing about 10KW peak through the thing.

All that to say....I will soon have 4 motors with FF. 3 vented and 1 not, so if it's somehow the venting causing the drag to increase, I should know soon enough.

Cheers
High Power LiPo wiring harnesses - 4P - XT90, HXT4mm, 5.5mm. 200Amp+ capable. Global shipping.

Modified Stealth Fighter - Force air cooled Leaf motor @ 6KW, Heat-sinked Adaptto Mini-E. Battery = 49AH, 14S (51.1V) Sanyo 18650GA - 2.5Kwhr = 200km range!
Cowardlyducks - Stealth Fighter Videos

BikeE recumbent - Golden motor, 6Fet Grinfineon, internal 17AH, 14S Sanyo 18650GA, + on-board solar.

The stuff I make and modify.

User avatar
fechter
100 GW
100 GW
Posts: 12429
Joined: Dec 31, 2006 3:23 pm
Location: California Bay Area, USA

Re: Definitive Tests on the Heating and Cooling of Hub Motors

Post by fechter » Feb 08, 2018 1:09 pm

As the motor heats up, it's possible the stator is expanding and making the gap between the stator and magnets smaller. Thinner gap would increase the drag.

I had an issue with one of my A2B motors where it would start making a rubbing noise when it warmed up. It turned out the spoke tension was very uneven causing the rotor to be out of round. Once I got the spokes evened out, the noise went away.
"One test is worth a thousand opinions"

John in CR
100 GW
100 GW
Posts: 13999
Joined: May 20, 2008 12:58 am
Location: Paradise

Re: Definitive Tests on the Heating and Cooling of Hub Motors

Post by John in CR » Feb 11, 2018 4:45 am

Of course a hotter motor has less performance. Plus by then your battery voltage is lower and voltage is sagging even more.

Regarding the earlier mention of adding more non-working weight to a motor by stuffing more aluminum on the stator isn't a good idea. That would actually slow heat transfer to the outside world early by storing more of the heat inside. Sure early in the ride there would be some benefit in the form of cooler windings, but it will quickly get to the same temperature as without since you haven't changed the circuit for heat transfer to the shell. FF helps because it conducts heat faster to the outer shell, and hubsinks transfer heat faster from the shell to the outside air.

Of course the easiest route to cooler temps (other than proper controller settings) remains the least used. That is directing more air flow toward the hubmotor shell.

Post Reply