Definitive Tests on the Heating and Cooling of Hub Motors

justin_le said:
Punx0r said:
It's interesting to see the thermal conductivity plot flat-line or dip and then increase again with increasing motor speed.
Is it possible that the ferrofluid is being drawn back into the air gap at higher speeds due to a venturi effect? Or that the increased air turbulence is causing the ferrofluid to foam and re-enter the air gap?

Could the back emf- iron losses -flux pattern changes- or harmonics be changing how the stuff piles up at high speed. Something may cause some larger ripples to develop on the magnet surface that bridge the gap or mix fluid /heat a bit more. Just a thought. Hows that high speed internal camera setup coming. :p
 
instead of a loss in conductivity with the higher ml fills, I would propose that what you are seeing is the effects of drag, manifesting itself as less of a delta T. You think it is less heat transfer, but what you have is new heat being generated by the drag of fluid.
 
Not sure if it's been properly tested before, but is there any data available on trans oil ? I know it's an old method. FF has much better handling properties, ie not running up your phase wires and leaking out etc. but performance wise I'm guessing it's a little better based in the coverage it can achieve?

By the way Justin stoked your testing the hubsinks :)
 
Jestronix said:
Not sure if it's been properly tested before, but is there any data available on trans oil ?

Yes, right here on this thread and the results are in this post, have a read:
https://endless-sphere.com/forums/viewtopic.php?p=1106995#p1106995

Basically for a MXUS 45mm motor with Statorade I had 7 watts/degree core to shell, while 150mL of transmission fluid was a little better at 8 watts/degree. I was expecting a more substantial improvement than this. I didn't do much more extensive testing with ATF over a wide range of RPMs and fill levels etc. but it was enough of a result to tell me it's not interesting to pursue further given the mess associated with those tests.

By the way Justin stoked your testing the hubsinks :)

Yeah me too! Sadly I've been stung again by the hard realities of running a brick and mortar business during the height of summer season when everyone wants to go on vacation, so the experiments have been stagnant the last few weeks but when time and staffing allow me to spend time in the lab I'll be back at it. I did manage to run a full wind tunnel test the MAC geared motors which I need to post here, and also got the H motor results with hubsinks live on the trip simulator app, so you can see what the effect of the hubsinks are vs. just statorade.

http://www.ebikes.ca/tools/trip-simulator.html

The difference isn't nearly as big as going from raw hub to hub + statorade, but it in most situations it gives a ~10 degree cooler motor over not having the hubsink set.
 
I have followed " Emmetts " method of sealing my new 3540 using J and B black ultra RTV and used sealed Nachi bearings ! No leaks for wire side , some weeping from pressure relief hole ! Using 130 cc of synthetic transmission fluid ! The improvement in cooling is dramatic !! Climbing the 12 % hills with ambient temperature of 107 degrees no problem ! I also have the Grinn cooling fins which I believe help a lot ! Using fero fluid only I overheated in one mile up the 10 % street I live on , my bike was useless without oil cooling !!
 
burningwings said:
I have followed " Emmetts " method of sealing my new 3540 using J and B black ultra RTV and used sealed Nachi bearings ! No leaks for wire side , some weeping from pressure relief hole ! Using 130 cc of synthetic transmission fluid ! The improvement in cooling is dramatic !! Climbing the 12 % hills with ambient temperature of 107 degrees no problem ! I also have the Grinn cooling fins which I believe help a lot ! Using fero fluid only I overheated in one mile up the 10 % street I live on , my bike was useless without oil cooling !!

Impressive findings. No doubt the fins are helping the oil bath cool things better.
 
I noticed that the rear motor heats quite more than the front on my 2wd (2x Mxus V3 3T), when they receive exactly the same energy from two max e with the same settings. So the difference comes from air cooling. The rear wheel is trapped in the turbulence generated by the front wheel, the frame and my legs. So I made the cheapest venturi effect ever with duct tape and crate wood : tried to collect air from the sides to compress it in direction of the hub.

I gave a kind of crap wing shape in order to create a laminar effect on the external lower part and collect even more air.

Well believe me or not, the engine now runs 30°C cooler at 90km/h, without FF nor fins, which I'll ad later. I guess that this idea (maybe in a cleaner version !) will dramatically improve the fins' eficiency.

Edit : now the rear wheel is slightly cooler than the front, I suspect the moto tire and rim to have a noticeable bad aerodynamic effect

The aerodynamic drag of my experimentation is also noticeable, i've lost 5km/h (116 instead of 121 with OVS at 4) but I can cruise at 100 kph now with 90°C max. of course for lower speeds and uphill single track style, FF will be more appropriate

Thought it would be interesting for some of you !

2017-09-08_16.50.37[1].jpg
 
A bit of self moderation here. Since my tests above were done in a coastal landscape, my enthousiasm was biased by the 90 % flat roads. In that case my crap wing is enough to reduce the temperature from 145 to 115°C as soon as speed is above 70 kmh.

But back in Paris my double 3k 3t config happened to get warm way too fast, as soon as the speed is too low (aka under 70 km/h, climbing hills, trafic jam).

double heatsink.jpg

I've added FF and a pair of hubsinks to the rear motor : things are even better in "highway style" (speed 100km/h on long distances without overwarming) , with a stable 105°C at around 4.5kw continuous, and 11 kw peaks (21kw peak combining the 2 motors as you can see below). Compared to the 145° of my first trials it's a progress. But the hills are still a problem : no miracle, I have to reduce current under 3000w per motor to be able to climb without stopping.

21 kw.jpg

Going SERIOUSLY up hill (I mean climbing 2000m for example) will clearly not be possible with that config. So it's a testimonial to say that you'll really have to choose the good winding, in a motor having the appropriate nominal power according to your needs, before trying all sorts of time consuming modifications resulting in relatively low effects. I'm gonna have to go back to 2 x 3k 4t I'm afraid. Big mistake...
 
csc said:
I'm gonna have to go back to 2 x 3k 4t I'm afraid. Big mistake...

Instead of going back to a slower wound motor with all the work and time investment, i would first try to set a speed limit, and / or in combination with a lower phase amp limit. The result should be quite similar.

I wonder what your bike weights or do you pull a trailer with it? Otherwise i can't imagine why someone should have overheating issues with 2WD 3000W motors with FF + Hubsinks.
Man, your bike must need a hell of power to make it moving :)
 
madin88 said:
csc said:
I'm gonna have to go back to 2 x 3k 4t I'm afraid. Big mistake...

Instead of going back to a slower wound motor with all the work and time investment, i would first try to set a speed limit, and / or in combination with a lower phase amp limit. The result should be quite similar.

I wonder what your bike weights or do you pull a trailer with it? Otherwise i can't imagine why someone should have overheating issues with 2WD 3000W motors with FF + Hubsinks.
Man, your bike must need a hell of power to make it moving :)

I would check no-load on everything to make sure it's as low a possible. Something seems off to be making excessive heat in that rig.
 
madin88 said:
I wonder what your bike weights or do you pull a trailer with it?

108 kilos !!!! The motorcycle fork itself is around 15, but I won't go back to bicycle stuff for stiffness and braking reasons. Batteries are awfuly heavy too : it's a total of 20s 28p and it's going to be even worse soon with a total of around 20s 34p (super odd front battery box in progress). Yes I can put a trailer + somebody on the rear battery box (did 500 km like that with my girlfriend this summer). In that case it should still be able to climb, which is complicated if you want something also able to reach 100 kmh, when there's no passenger.

Rube : I made an air deflector to cool the rear motor as described above. I used Chanel crate wood and Hermès duct tape for obvious esthetic reasons !

Liveforphysics : No load is OK, I checked already. I slept on it and made a 55 km trip from Paris city center to the countryside this morning, after reducing batt and phase amps to 111 / 300. Did almost never WOT (just a few bursts less than 3 seconds to fit in the trafic). As soon as I keep nominal input around 4kw per motor, it's under 110° again. Just have to swap between boost and normal mode sometimes to push harder temporarily with the cooler front hub.

The only doubt I have is about my angle corr (-11°) and ind timind (590 us). Sorry, it's a question that should be in the Adaptto thread (already asked there BTW but no answer) : do you or anybody know if these are normal values for a 3k 3t ?

In fact it's more and more obvious to me that if I want a heavy duty fast cargo thing, rear 3k is just too small. Trying to cool a motor is OK within reasonable limits : at some point removing heat doesn't solve the problem of it being generated. I think everybody here is OK with the idea that the more heat generated - no matter it's dissipated or not - the less eficiency. So when I read people with a qs 205 claiming they have a 20 kw bike, please don't mind if I have doubts !!! They have a 20 kw roller toaster, as I do now with my 3t config.

My conclusion is that I need a 3k 4t front in a 21" motorcycle rim with OVS 4 + a 5k turbo 4t in a 19" moto rim, OVS at 6. My 100% empiric approach tells me that it will be more efficient in real world use and not slower. Do you agree with that ?
 
Having a 2wd is useful for that kind of questionning. When I throw more than 4 KW nominal in the front 3k 3t, while driving around 50 km/h in the city, stopping, accelerating and so on, temperature rises too. In fact both motors behave exactly the same : no halls issue here. Just an unapproprite belief in the fact that cooling the motor increases a lot the power. It just improves it a little by helping not to go over 120°.

Starting now my plan is to stay under 80° when going uphill with a 30 kilos payload, and under 120° with a passenger, while keeping a 100kmh long range top speed;

Are there hubsinks available for the 5k ?
 
csc said:
My conclusion is that I need a 3k 4t front in a 21" motorcycle rim with OVS 4 + a 5k turbo 4t in a 19" moto rim, OVS at 6. My 100% empiric approach tells me that it will be more efficient in real world use and not slower. Do you agree with that ?

If I were you I would try to reduce OVS as low as possible. OVS=0 is my best friend to reduce motor heating and I've been riding a QS205 at about 7-10kW from a couple of years.
 
You're half right, I think. Ovs is 2 on front wheel and 1 on rear, which is quite reasonable in theory but here, with the 2x3t, increasing it just makes more heat, no more speed.

My previous mxus v2 4t could go till ovs 6, and turn field weakening into real speed without warming that much...
 
How fast does the bike go with out feild weakening? Doesn't make sense to run it all the time to me. I have a MaxE/QS205 3.5T and have never used it since mine will hit 60 mph with out it. Luckily I don't ever need that kind of speed....I'll occasionally get in the 50's just for some excitement. As much info about feild weakening that has been posted on here would tell you to only use it when need that extra speed, not as a constant. Maybe try some different profiles that you can switch between?

Tom
 
csc said:
You're half right, I think. Ovs is 2 on front wheel and 1 on rear, which is quite reasonable in theory but here, with the 2x3t, increasing it just makes more heat, no more speed.

My previous mxus v2 4t could go till ovs 6, and turn field weakening into real speed without warming that much...

if you had the same battery voltage and wheel size on both the 4T and 3T setup, it would be clear that 4T has MUCH more room above and response to OVS / field weakening.
or did you adjust the voltage and current :?:

as example: 100V 100A on the 4T

would mean 3/4 the volt and 4/3 x the current on the 3T

so 75V 133A

doing so, both setups should perform much the same.
 
Litespeed : 102 kmh at OVS 0 but... ind timing 590 us and angle corr -11°. I'll be testing this afternoon a frankly lower ind + frankly bigger ovs, just to see what happens.

What are your Ind timing and angle corr values with the 3.5t ? To the others : sorry for not being in the adaptto thread but I would'nt like to miss the answer

Madin 88 : ok thanks ! I stayed in 72v (so 84.2 full) and kept a 26 wheel size (though front is a moto 21", it's the same diameter).

Question : will a 3 t 75v 133A and a 4t 100v 100A perform the same when going uphill in terms of heat ?
 
I still think 1 of the best ways to reduce hubmotor heat, is cooling holes . Letting the hot air easily escape instead of allowing it to be contained in the hub motor and trying to dissipate it thru heat sinks or ferro fluid or air scoops, etc.

Of course the cooling holes allow humidity / contaminents a easier entry point into the motor, so that is the big negative, especially if you are riding in dusty/ wet conditions. For just road use, cooling holes are still my favorite cooling method.

I had a Lyen controller which was getting very warm when I was hitting 60 + amps on a 72 volt setup. The easiest solution for me was to purchase a square vinyl 90 degree downspout
and mount it on my bike frame so it would scoop air over the controller. This has worked fabulous, and it was simple and inexpensive to do it.

Since I use this ebike for riding on dusty roads/ dirt trails, I was thinking of adding 2 of these vinyl 90 degree scoops to the rear wheel and attaching them to the spokes so it diverts the air directly across the hub motor { in the middle inbetween where the spokes attach and the inner magnets are mounted} .

I have tried fluid cooling of hub motors, and was not happy with the results and I never liked the idea of adding external small fans , because it just adds another electrical element that can get damaged.
 

Attachments

  • wind.jpg
    wind.jpg
    122.4 KB · Views: 1,871
  • wind2.jpg
    wind2.jpg
    166.9 KB · Views: 1,871
I never liked the idea of adding external small fans , because it just adds another electrical element that can get damaged.

Can you elaborate, I am not understanding what you mean with adding "another electrical element that can get damaged". Seems to me the fins, whether they got damaged or not, wouldn't affect the functionality of the motor.
 
Back
Top