Definitive Tests on the Heating and Cooling of Hub Motors

Get all your technical information about electric bikes here.
User avatar
richdeloup
100 W
100 W
Posts: 258
Joined: Mar 11, 2012 5:15 pm
Location: UK somewhere between the centre and the edge

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

Post by richdeloup » Nov 27, 2014 9:43 am

disc idea.jpg
9 ebikes, one pair of legs, small apartment, no wife

User avatar
richdeloup
100 W
100 W
Posts: 258
Joined: Mar 11, 2012 5:15 pm
Location: UK somewhere between the centre and the edge

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

Post by richdeloup » Nov 27, 2014 11:38 am

heat sink fan.jpg
heat sink fan.jpg (31.18 KiB) Viewed 2592 times
These are meant for small appliances, so not sure if of use for us lot?
The results show a good thermal performance and highlight the need to develop the heat sink and fan as an integrated thermal solution rather than in isolation as is the traditional methodology. An interesting finding is that the heat transfer scales are in line with turbulent rather than laminar correlations despite the low Reynolds number.


Sounds impressive even if not fully understood :oops:

I cant see anywhere possible to fit an electric fan but maybe well designed fins and a low profile heatsink as in the diagram may work?
9 ebikes, one pair of legs, small apartment, no wife

Punx0r
10 GW
10 GW
Posts: 4379
Joined: May 03, 2012 8:16 am
Location: England

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

Post by Punx0r » Nov 28, 2014 3:59 am

It means that turbulent (swirly) airflow cools better than laminar (smooth). The Reynolds number simply describes how turbulent the flow is (or not) and you'd want to be several thousands to ensure turbulent flow. It's not really relevant though. It's easy to calculate the Reynolds number in something like a pipe, but inside the complicated shape of a hubmotor interior would be more difficult. It's also much easier to get turbulent flow than laminar.

Just make sure the airflow is high ;)

User avatar
richdeloup
100 W
100 W
Posts: 258
Joined: Mar 11, 2012 5:15 pm
Location: UK somewhere between the centre and the edge

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

Post by richdeloup » Nov 28, 2014 5:20 am

Thanks that explains it!

I was really just looking for the best design on a simple lightweight CNC'd cover for use on an oil cooled motor.
9 ebikes, one pair of legs, small apartment, no wife

User avatar
richdeloup
100 W
100 W
Posts: 258
Joined: Mar 11, 2012 5:15 pm
Location: UK somewhere between the centre and the edge

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

Post by richdeloup » Dec 19, 2014 3:27 pm

trasfer through cover.jpg
trasfer through cover.jpg (85.36 KiB) Viewed 2489 times
I'm going to try this out, unless someone knows better and says that its a waste of time???

It's intended for oil cooled.
9 ebikes, one pair of legs, small apartment, no wife

User avatar
richdeloup
100 W
100 W
Posts: 258
Joined: Mar 11, 2012 5:15 pm
Location: UK somewhere between the centre and the edge

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

Post by richdeloup » Dec 19, 2014 5:41 pm

transfer through cover.jpg
transfer through cover.jpg (92.38 KiB) Viewed 2486 times
Edit; to keep with the title of the thread, ill test with standard motor, then oil cooled, then no oil but with above riveted on, then with both. I'm not tech enough to draw graphs but will take the same varied terrain route and video the temp gauge and timer to try to figure out whats going on.....
9 ebikes, one pair of legs, small apartment, no wife

User avatar
John Bozi
1 MW
1 MW
Posts: 1909
Joined: Apr 09, 2013 1:46 am
Location: Brisbane, Australia
Contact:

add more motor mass vs weight

Post by John Bozi » Dec 19, 2014 7:13 pm

are rivets leak proof? wouldn't it be easier to just epoxy on fins to the outside cover and then hope that one day in a crash that the spinning wheel doesn't rip through somebodies flesh?

What's the weight of the oil and and extra proposed metal added?

If it is significant then adding more weight is also adding more heat to get the same speed...

A light secondary motor to keep the bike moving might be a solution. The weight is a big thing I need to spend time considering before I decide to stay with dual motor set up or not. It's really pointless unless you are in the overheat zone quite a bit as I am up the mountains...


User avatar
richdeloup
100 W
100 W
Posts: 258
Joined: Mar 11, 2012 5:15 pm
Location: UK somewhere between the centre and the edge

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

Post by richdeloup » Dec 19, 2014 7:49 pm

are rivets leak proof? wouldn't it be easier to just epoxy on fins to the outside cover and then hope that one day in a crash that the spinning wheel doesn't rip through somebodies flesh?
Good point, I'm using rivets at the moment and of the 8 on there only two are leaking (part sealed with high temp silicone). There are open ended rivets which I'm using at the moment but also closed ended which ill try out next... Also with a load of copper rivets protruding, im figuring that braising fins onto them may be easier than aluminium welding and that they may provide the thermal flow to the exterior?. ye safety wise, well we can figure that out later:)
What's the weight of the oil and and extra proposed metal added?
100ml seems to be the optimum and i'm going for 2mm thick copper discs, more about surface area than density; ill weigh it all once made up.

wine=edits
Someone in the UK must have brought in a ship load of 9c copies which they are trying to get rid of, as they are super cheap on ebay. So a good opportunity to get them in the overheat zone to analyse what works and what doesnt:)
They actually have 250w stamped on them!

I think I get what you're saying about if adding extra weight, then why not just get a more capable motor of the same weight and stay out of the danger zone, but where do you stop maxing out? im thinking that we are still a long way off the 600miles a day at full speed motorbike world, which one day i hope we'll get to.
Last edited by richdeloup on Dec 19, 2014 8:57 pm, edited 3 times in total.
9 ebikes, one pair of legs, small apartment, no wife

User avatar
richdeloup
100 W
100 W
Posts: 258
Joined: Mar 11, 2012 5:15 pm
Location: UK somewhere between the centre and the edge

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

Post by richdeloup » Dec 19, 2014 7:58 pm

Twin drive vid, cool, i've been thinking about that, but there is always someone whose already done it on here; so when one motor overheats you switch?
9 ebikes, one pair of legs, small apartment, no wife

User avatar
John Bozi
1 MW
1 MW
Posts: 1909
Joined: Apr 09, 2013 1:46 am
Location: Brisbane, Australia
Contact:

snakes eat tails

Post by John Bozi » Dec 19, 2014 8:51 pm

richdeloup wrote:Twin drive vid, cool, i've been thinking about that, but there is always someone whose already done it on here; so when one motor overheats you switch?
In an ideal world the two motors at the same power level would be great, but realistically a mid drive cant run the same power through the hub drive train. So the main hub stays the same but has the much smaller motor there to pull it in granny gear for a cool off period, before or after a steep overheating section.

sometimes used at the same time if incredibly steep where the hub bogs to 10kmh.

I also use the mid for all 0-10kmh stuff, but its not perfect, you do need to put some weight on the pedal so that the mid doesn't clack into tension.

Only time will tell, for me now after only one propper long off road ride, it feels too heavy. The extra weight of the motor wasting more power and creating more heat that it then compensates for.....

the snake eating its tail.

John in CR
100 GW
100 GW
Posts: 13910
Joined: May 20, 2008 12:58 am
Location: Paradise

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

Post by John in CR » Dec 19, 2014 11:15 pm

richdeloup wrote:Twin drive vid, cool, i've been thinking about that, but there is always someone whose already done it on here; so when one motor overheats you switch?
It's more twice as good to run both at the same time. Assuming they're equal it's half the current to each, which creates half the heat to begin with, and then you have twice the heat dissipation. Half the heat and twice the cooling means you can significantly increase performance and still come out way way ahead.

User avatar
John Bozi
1 MW
1 MW
Posts: 1909
Joined: Apr 09, 2013 1:46 am
Location: Brisbane, Australia
Contact:

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

Post by John Bozi » Dec 19, 2014 11:57 pm

John in CR wrote:
richdeloup wrote:Twin drive vid, cool, i've been thinking about that, but there is always someone whose already done it on here; so when one motor overheats you switch?
It's more twice as good to run both at the same time. Assuming they're equal it's half the current to each, which creates half the heat to begin with, and then you have twice the heat dissipation. Half the heat and twice the cooling means you can significantly increase performance and still come out way way ahead.
It just isn't going to happen unless I dunb down the hub to run at the mid drive power levels. The problem is that you can't shift gears anywhere near the way say the hub ramps from 0-50kmh.

sure on a long section where the road stays the same steepness, but not on complex up and down sections I see on my firetrails. Having to power off to change gear is a big pain too.

I am interested in ebikes.ca new splined hub motorm no idea of it's strength to run a mid through but it would be my choice so that I could have a choice of modern cassettes which should shift just as well as a mountain bike does. These 7 speed freewheels never seem to be spaced properly or just I am lousy at getting them to shift properly.... I only have a bout 3 speeds that are clean usable...

User avatar
richdeloup
100 W
100 W
Posts: 258
Joined: Mar 11, 2012 5:15 pm
Location: UK somewhere between the centre and the edge

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

Post by richdeloup » Jan 13, 2015 6:09 am

Ive done a few tests over the Christmas period without the heat transfer idea that I want to try out. I'm sure this oil cooled bit has been covered before in more detail, but running at 100v 60A on a 9c clone.
Standard motor not modded Overheat (90 degrees) 8.12 minutes
Oil cooled with 100ml ATF Overheat (90 degrees) 10.14 minutes
Oil cooled with 200ml ATF Overheat (90 degrees) 19.34 minutes

Same 1 mile route, varied terrain on road/ off road.

I thought that 100ml ATF was thought to be the optimum amount? I suppose any extra AFT takes longer to heat up, anyhow now for some copper rivets along the lines below, to try to get the heat out. Thinking more rivets may add more surface area than a copper disc but will try both ideas....
Copper Pin Fin Heat Sink-web.jpg
Copper Pin Fin Heat Sink-web.jpg (108.09 KiB) Viewed 2329 times
9 ebikes, one pair of legs, small apartment, no wife

User avatar
izeman
10 GW
10 GW
Posts: 4561
Joined: Jun 21, 2011 8:25 am
Location: vienna, austria
Contact:

Definitive Tests on the Heating and Cooling of Hub Motors

Post by izeman » Jan 13, 2015 8:04 am

The extra oil will not delay heat up because of it's mass/volume but because it can cover the windings better and transport more heat away from them to the covers.

User avatar
richdeloup
100 W
100 W
Posts: 258
Joined: Mar 11, 2012 5:15 pm
Location: UK somewhere between the centre and the edge

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

Post by richdeloup » Jan 13, 2015 8:32 am

Great thanks, I was wondering if this was the case but as I read that 100ml was a good amount to use, I though that this would be enough to cover the winding. Each motor is obviously a different size;
the motor I'm using is pretty small with just a 30mm wide stator but obviously needs more than 100ml.
9 ebikes, one pair of legs, small apartment, no wife

itchynackers
100 kW
100 kW
Posts: 1211
Joined: Dec 27, 2009 11:11 am
Location: Janesville, WI

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

Post by itchynackers » Jan 13, 2015 10:08 am

No one ever said 100mL was optimal. Its just what people were comfortable using (not sure why). It somehow morphed into gospel. FWIW, I was using 600mL running up Pikes. This is half the total capacity of my 9c motor. The more you use, the more risk of leakage out the axle (per my empirical data).
9c 9x7f w/ 12awg phases
9c 8x8r w/ 10awg phases, temp sensor, vented
9c 6x10r w/ 10awg phases, temp sensor, liquid cooled, 3rd Place Pikes Peak 2012!

"Don't argue with the galactically stupid...they drag you down to their level and beat you with experience"

User avatar
izeman
10 GW
10 GW
Posts: 4561
Joined: Jun 21, 2011 8:25 am
Location: vienna, austria
Contact:

Definitive Tests on the Heating and Cooling of Hub Motors

Post by izeman » Jan 13, 2015 11:33 am

ppl added oil and tested no load current. there was a point where adding more oil leaded to higher amps. that's where they stopped saying it was the optimal fill. not saying that more oil can/can't cool the motor better. and the limit is the height of the breather hole. if you fill the motor beyond that point it will leak all the time. ;)

User avatar
richdeloup
100 W
100 W
Posts: 258
Joined: Mar 11, 2012 5:15 pm
Location: UK somewhere between the centre and the edge

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

Post by richdeloup » Jan 13, 2015 11:49 am

As said, I suppose a lot of people were talking about the optimal balance level between cooling, higher amp draw and every day hassle factor, good to know:)
Think ill stay at 200ml as I'm more interested in trying to get the heat out of the oil once it's max'd out....
9 ebikes, one pair of legs, small apartment, no wife

itchynackers
100 kW
100 kW
Posts: 1211
Joined: Dec 27, 2009 11:11 am
Location: Janesville, WI

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

Post by itchynackers » Jan 13, 2015 6:21 pm

izeman wrote:ppl added oil and tested no load current. there was a point where adding more oil leaded to higher amps. that's where they stopped saying it was the optimal fill. not saying that more oil can/can't cool the motor better. and the limit is the height of the breather hole. if you fill the motor beyond that point it will leak all the time. ;)
This is somewhat very misleading, and part of reason ppl are stuck on the 100 ml. In fact, adding the first 100ml will add the MOST resistance (no load). The fluid added beyond that first 100 ml does very little to increase no load current. You'd probably be surprised how little the no load increases with lots more coolant.
Last edited by itchynackers on Jan 13, 2015 6:30 pm, edited 1 time in total.
9c 9x7f w/ 12awg phases
9c 8x8r w/ 10awg phases, temp sensor, vented
9c 6x10r w/ 10awg phases, temp sensor, liquid cooled, 3rd Place Pikes Peak 2012!

"Don't argue with the galactically stupid...they drag you down to their level and beat you with experience"

itchynackers
100 kW
100 kW
Posts: 1211
Joined: Dec 27, 2009 11:11 am
Location: Janesville, WI

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

Post by itchynackers » Jan 13, 2015 6:28 pm

Just to add some real data rather than talk, here is some testing I did a while back from this thread...

https://endless-sphere.com/forums/viewto ... d&start=25
itchynackers wrote:After being the first bike up the Pikes Peak race this year, I can attest that liquid cooling does work. I just completed my winter teardown to make a few improvements. Found out I had a detached capacitor. Who knows how long I was riding like this.

Controller:
Soldered the cap back on after hearing it rattle inside the controller. Not too bad.
Siliconed each cap for vibration resistance.
Snugged the fet screws. They were ALL loose!
Spread more thermal compound on backside of aluminum fet bar.
Installed 12v cooling fan in end cap of controller.

Motor:
Installed new Japanese bearings with thin layer of silicone sealant
Re-sealed interior axle/wire slot with silicone sealant.
Tapped hole in side cover for one way pressure valve like these...
http://www.ebay.com/itm/221106438184?ss ... 1497.l2649
Hopefully this will work a bit better than the "tiny hole" method. I will have to stop and bleed the air once I reach maximum temperature. Hopefully the motor will be in a vacuum state after cooling, which should help avoid leaking fluid.
I bolted everything back up and I'm functional, so I didn't botch the controller, woot!

After charging my 25s 12ah battery to 4.1v/cell, I get the following results:
Edit: Working on filling in the data...

ATF Fluid level Resting V Sagging V No-Load Current Wheel Speed RPM
___________________________________________________________________________
0 ml:................102.5.....102.3............1.29...........46.4.........609.5
50ml:...............102.5.....102.2............1.37...........46.4..........609.5
100ml:..............102.5.....102.1............2.16...........46.0..........604.2
150ml:..............102.5.....102.1............2.24...........45.9..........602.9
200ml:..............102.5.....102.1............2.24...........45.9..........602.9
250ml:..............102.5.....102.1............2.24...........45.9..........602.9
300ml:..............102.5.....102.1............2.24...........45.9..........602.9
350ml:..............102.5.....102.1............2.24...........45.9..........602.9
400ml:..............102.5.....102.0............2.24...........45.9..........602.9
450ml:..............102.5.....102.0............2.24...........45.9..........602.9
500ml:..............102.5.....102.0............2.24...........45.9..........602.9
550ml:..............102.5.....102.0............2.24...........45.9..........602.9
650ml:..............102.5.....102.0............2.25...........45.9..........602.9 (this is exactly half full)
700ml:..............102.5.....102.0............2.31...........45.9..........602.9
Ran out of fluid...
750ml:
800ml:
850ml:
900ml:
950ml:
1000ml:
1050ml:
1100ml:
1150ml:
1200ml:
1250ml:

I measured the total full capacity of the motor at 1300ml using water. I calculated the thermal expansion of atf3 at 0.3% per degree C, so it really doesn't expand too much even with a delta T of 100C. Its the hot air I worry about. If I don't bleed that valve properly, or didn't seal the motor properly, I may have created a pressurized, hot oil filled, spinning electric device that I'm riding at 40mph. What could happen?
9c 9x7f w/ 12awg phases
9c 8x8r w/ 10awg phases, temp sensor, vented
9c 6x10r w/ 10awg phases, temp sensor, liquid cooled, 3rd Place Pikes Peak 2012!

"Don't argue with the galactically stupid...they drag you down to their level and beat you with experience"

User avatar
richdeloup
100 W
100 W
Posts: 258
Joined: Mar 11, 2012 5:15 pm
Location: UK somewhere between the centre and the edge

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

Post by richdeloup » Jan 14, 2015 10:10 am

Great information, best buy meself another can of ATF:)
9 ebikes, one pair of legs, small apartment, no wife

Emmett
1 kW
1 kW
Posts: 331
Joined: Aug 27, 2014 9:45 am
Location: NSW Australia

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

Post by Emmett » Jan 17, 2015 10:37 pm

Castor oil is better. No foul smell. Easy to clean if it weeps out the breather or axle slot. Way better IMO. 120ml in a H4065.

User avatar
richdeloup
100 W
100 W
Posts: 258
Joined: Mar 11, 2012 5:15 pm
Location: UK somewhere between the centre and the edge

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

Post by richdeloup » Jan 20, 2015 11:30 am

Pretty crude mod so far but got bored of riveting. Pretty interesting results which looks like something is working. I ran out of battery so need to retest to see if its a real solution for longer trips or if the extra mass is just taking longer to soak up the heat??
results so far.jpg
9 ebikes, one pair of legs, small apartment, no wife

User avatar
richdeloup
100 W
100 W
Posts: 258
Joined: Mar 11, 2012 5:15 pm
Location: UK somewhere between the centre and the edge

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

Post by richdeloup » Jan 20, 2015 11:55 am

What I cant figure out is how exactly this is working, any suggestions welcome?

So the internal copper discs runs very close to the windings and should be taking some of the heat either directly or helped by being covered in oil.
Some of the heat should run to the external discs through the copper rivets.
Some of the oil/heat should get traped between the cover and external disc. Maybe because the oil is trapped, it has more time in contact with the cover to transfer the heat through, rather than constantly being on the move? im not really sure, a micro camera would be good to have...
9 ebikes, one pair of legs, small apartment, no wife

User avatar
justin_le
1 MW
1 MW
Posts: 1999
Joined: May 28, 2007 3:27 am
Location: Vancouver
Contact:

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

Post by justin_le » Mar 18, 2015 6:27 pm

Hi guys, 2 years after starting this thread we're now aiming to wrap up part of the original purpose and intent, which is to have an empirically verified thermal model for not just one specific hub motor, but a general case that we can apply to all the motors on our motor simulator page. I've now got 57 motor types characterized and modeled, or 33 unique motor families if you don't count winding variations of the same hub:
Simulator Motor List.jpg
Simulator Motor List.jpg (55.26 KiB) Viewed 1933 times
Considering that it generally takes a couple days of setup and experimental tests to measure and exactly fit the motor's thermal characteristics to a model over the full range of RPMs and wind speeds, as was done with a 9C hub in this thread (see https://endless-sphere.com/forums/viewto ... 24#p718924 and https://endless-sphere.com/forums/viewto ... 48#p722948, there is no way to realistically repeat that for each of these motor. So what I want to do instead is generate a good ballpark set of thermal parameters for a hub based only on easily measured values. Ideally you could feed in the mass of the motor core, the mass of the hub shell, the diameter and width of the casing, and out comes a decent first order guess for the heat conduction coefficients from the core to the shell, and the shell to ambient, as well as the heat capacities both of the Stator core and the shell.

It won't be quite as exact as measuring every single hub individually, but it would be a lot more versatile an approach, especially if as we want to keep adding motors to the simulator database without needing days of testing for each one.
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

Post Reply