Definitive Tests on the Heating and Cooling of Hub Motors

[Offroader]

Mac Ribs got you thinking Offroader, I like it, if you do this, please let us know the results. I haven't jumped on the FF bandwagon yet. My venting and riding with some mechanical sympathy has thus far worked on all my builds.

Yes, that is why I ask a lot of questions on this forum for idea generation.

Rix, you know you can easily seal up those holes and FF fluid and hub sink your hub motor. Especially considering you live out in the desert there and I know Las Vegas is dusty as hell.

 

[Rix]

Mac Ribs got you thinking Offroader, I like it, if you do this, please let us know the results. I haven't jumped on the FF bandwagon yet. My venting and riding with some mechanical sympathy has thus far worked on all my builds.

Yes, that is why I ask a lot of questions on this forum for idea generation.

Rix, you know you can easily seal up those holes and FF fluid and hub sink your hub motor. Especially considering you live out in the desert there and I know Las Vegas is dusty as hell.

True that. I was just down in Vegas the week before Christmas, amazed me how much warmer the weather is there than here. Yah I agree that asking questions gives us ideas we wouldn't have thought of otherwise also. If I go with FF, I will get new side covers, Yah, I can seal them up with a gas tungsten arc welder and some aluminum rod, but it would look like I sealed the holes up if you know what I mean. Anyway before I got this route, waiting to hear from you. All that said, I still may got with the heat sinks.

 

[Offroader]

For you guys who ordered the hub sinks in USA. Is there a USA distributor or did you have to order it from hubsinks.com Australia and pay $30 shipping?

Thanks.

 

[litespeed]

All of us beta testers got ours from Sketch.

Well worth the 30 clams shipping!

Tom

 

[Offroader]

I think I may try another idea so I don't have to lace a rim and buy spokes. When using forced air cooling ducted to take in fresh air and exhaust hot air, I noticed the magnet ring would get extremely hot even though I was taking in cool and exhausting the hot air. The air flow was heated up so quickly and hitting the magnet ring that it would get extremely hot.

If I close up the center of the hub and run the EDF in a sealed motor, it will rotate the air around the hub and through the windings and will heat the entire motor case up. It will basically be rotating hot air around the motor as the air heats up as it passes through the windings.

My current MXUS motor housing surprisingly does not get hot at all without FF fluid, and only peaks at 100 deg F / 38 deg C, after sitting for 15 minutes. I will know immediately if the EDF fan is heating up the motor housing much faster than when not using the fan.

The question is will circulating the air at high speed around the motor work as effective as FF, I don't know but I think it should or close enough. The air temp in the motor will quickly take the air of the windings, and then will be passing over the motor housing at very high speed. I don't know how quickly hot air can heat a surface up compared to hot oil. But what I do know is that if I hold my hand over metal that is 200 deg it won't burn my hand, but if I blow 200 deg hot air on my hand it will burn it very fast, so moving hot air has to be effective.

Here are some ideas:

1) The circulating air inside the motor will be heated close to the temperature of the windings.
2) The airflow will be very high because EDF fans blow a ridiculous amount of air. These are not low speed computer fans.
3) The magnet ring should get heated really fast because the circulating air will be at winding temperature and blowing over it very fast
4) Bonus here will be that the side covers will also get super hot, but they will not have heatsinks so won't do that much.
5) The air flow will be direct from windings to motor housing. When using FF fluid the heat still has to go through the stator, and then to the motor housing.

I'm not recommending using a fan inside the motor for the thermal transfer as FF fluid is easier, I'm just going to use this so that I don't need to fill up my motor with FF.

Please let me know if you think the EDF fan will not work as well as the FF and this will still not tell me if FF + hubsinks are effective enough.

 

[Punx0r]

The air temp in the motor will quickly take the air of the windings, and then will be passing over the motor housing at very high speed. I don't know how quickly hot air can heat a surface up compared to hot oil. But what I do know is that if I hold my hand over metal that is 200 deg it won't burn my hand, but if I blow 200 deg hot air on my hand it will burn it very fast, so moving hot air has to be effective.

I wouldn't like to speculate how a fan in a sealed motor would compare to ferro-fluid but will interject a bit of theory on heat transfer. Liquids are generally much, much better at transferring heat than gases. Speeding up the fluid (gas or liquid) will increase heat transfer as the mass-flow rate is increased and this along with the temperature difference (delta-T) between the cooling fluid and the object to be cooled is what determines how much heat energy is transferred.

Holding your hand over a piece of metal at 200°C is not the same as touching it (or touching liquid at 200°C). The air gap between you and the metal is insulating you, with just a small amount of heat being transferred by convective airflow and radiant heat.

The problem with the sealed hub in your drawing compared to the open version, where ambient air is drawn in, heated then exhausted, is that the recirculated air will quickly become hot. The delta-T will reduce and so will the ability to transfer heat from the windings. It will be more effective than a sealed hub with no fan, as more heat will be transferred (a better thermal pathway) to the normally cool side covers, which can then act as cooling surfaces.

Will it work better than ferro-fluid? Dunno. FF will transfer heat better, but it's limited to the relatively small area of the magnet ring, compared to the larger side covers to shed that heat to the outside world. That's where the hub-sinks come into play. Would a sealed motor with fans best a sealed hub with FF and hubsinks? Hard call!

 

[madin88]

It is really hard to say if FF or the EDF setup does a better job. Yes, liquids transport heat many times better than air, but the EDF will blow the air through the windings (where most of the heat is generated), while FF removes the heat "only" at one area, the top of the stator teeth.
hmm, how long can we hold our fingers in front of a heatgun set to 100°C until we get burned? It is only air, but at very high speed.

if one doesn't like to use use Ferrofluid or EDF Fan, try this:

https://endless-sphere.com/forums/viewtopic.php?f=2&t=48753&p=1111120&hilit=fan#p1111120

 

[macribs]

If I manage to get a tight seal on the motor and axle I will use plain engine oil inside the motor combined with heat sink. The engine oil should last forever if not leaking and should be easier to remove then FF. I've contacted in-seal to try to get a few of those seals but with the holidays and all I have not yet hear'd back from them. The in-seal pro should take care of the axle, and for the side covers I am thinking machining a groove for an O-ring should do the trick. Then adding pressure equalizer either just drill a hole in the cover or use some sort of breathing one way membrane like madin I think it was did. For me riding in harsh weather with salty air from the sea and tons & tons of salt sprayed on our roads for 4-6 months of the year corrosion is a big issue. And my thought is that an oil filled hub + heat sink will give me the best combo of cooling and corrosion protection.

 

[Offroader]

I agree in that it is hard to figure out how FF will compare to a high circulating air. It is really hard to beat liquids with heat transfer.

One thing you have to take into account is that liquids have a lot more thermal mass, so when you get a drop of hot oil on your skin it will burn a lot more because of the thermal mass. This thermal mass isn't all that important inside the hub.

Thermal conductivity of oil .15
Thermal conductivity of air .024
Thermal conductivity of steel ~40

I think it will come down to where the bottleneck is in the heat transfer. I found in my MXUS motor that the thermal transfer from the windings to stator and motor housing was very low, the motor housing took a very long time to get hot, and was mostly cool while riding, even though my windings were consistently at 100C. This was unlike the cromotor where the motor housing gets way hotter and hotter way faster.

I assume the windings thermal transfer to the stator on the MXUS is somehow limited. FF fluid probably won't help the heat transfer from the windings to the stator in an MXUS. But the QS205 which is more like the cromotor may be different.

If the windings to stator heat transfer is the bottleneck then forced air may actually be better than FF which only contacts the very ends of the stator. The bottleneck could also be on the thermal transfer through the magnet ring or possibly the limits of the hub sinks, remember they only give about 2 square feet of hub sink area as the fins are very thick, that is probably about the surface area of a large CPU cooler and those are only rated at 250 watts.

Maybe it could be possible just to run some low powered fans to get more than enough air flow to make it an alternative to messy and leaking FF.

I'll test it out anyway as I'm curious, it isn't all that hard to install the EDF and I have everything to do it. I can also adjust the air speed with my electronic speed controller, to determine just how much different air flows make.


If you look at that computer heat sink below (be quiet dark rock pro 3) it has as many individual fins as a hub sink, each of those fins is close to the same surface area as on a hub sink as that CPU cooler is very large. Those cpu coolers are rated at 250 watts, you can't expect the hubsinks to really shed all that much heat. Not bashing the hubsinks, just saying that the bottleneck may lie there.

 

[macribs]

Well the rated value might be 250watt but people that have been messing with o/c cpu knows that those first great heat sinks like katana et.al made all the difference in the world. Suddenly we could get away from the water cooling, noisy pumps and potentially leaking issues. I remember we clocked almost as high with the katana as we did with water cooling. And back in the days those cpu's got real hot not like cpu nowadays. Do a test to see if it will work for you, for your ambient air temperatures and your conditions. Only then will you know for certain. If it does not work put it in the for sale section and make a fellow ES member very happy

I mean the FF alone showed pretty good increase of cooling, according to Justin's tests. Then you can paint the inside non reflective black as heat radiant will become faster compared to the shiny stock surface. Worth doing if you already got your motor on the work bench. From memory I think the black paint alone increased cooling by 10-12 %. Then adding liquid and heat sink will also help. It might turn out riding steep climbs in 105 degrees weather will bring the FF/heat sink to its knees. Or it might work even for your conditions, in which case you got a closed system. No dust, debris etc will enter you motor.

 

[liveforphysics]

The Rth values for static air and static liquids are just about irrelevant in a turbulent application.

 

[Offroader]

In my MXUS turbo, which seems to have very bad thermal transfer because the motor housing does not get very hot, comparing this to a cromotor which gets very hot.

When looking at the MXUS windings, they are encased in a plastic when going through the stator, which also suspends the windings somewhat in air, you can see all the space around them.

The windings on the outside are then separated from the iron stator by that yellow lamination on the outside, which may not have a very good thermal transfer when compared to the steel laminations.

This is why I question FF fluid in that it doesn't do anything for the thermal transfer from the windings to the stator. This may very well be a huge bottleneck.

That being all said, Justin did pot a stator and it didn't show any benefits when used with FF. So I just don't know, but that was a different hub motor. The MXUS seems particularly bad because the motor case would not get hot, while the cromotor motor case would get super hot when not using FF.

I'm hooking up my EDF now so I'll know just how much of a difference the hub temp gets. Right now my MXUS turbo without FF stays cool while riding, takes about 20 minutes and will finally peak at about 38C. My Cromotor would peak at about 71C after sitting with no FF and sealed motor. The cromotor would also get hot quickly, like 20-30 seconds after riding would feel hot to the touch.

I remember warming up my fingers in the winter on a sealed cromotor right after I stopped riding, that is totally out of the question with the MXUS, it just doesn't transfer the heat to the motor housing. You can see the reason is probably the windings being encased in plastic with a lot of air around them also.

 

[madin88]

It is highly possible that those fiber glass sheets and the insulation in the teeth are the main reason for the lower thermal flow to the case.
The cromotor has, iirc also insualtion betweeen the teeth, but it has dark blue paint on the stator and i also have seen one with dark colored magnets. that also would improve thermal flow..

 

[macribs]

Offroader if your thesis is correct FF will not work optimal in your motor, maybe use old fashioned oil filled hub? Surely that will allow for steep climbs even in the harshest condition. Add a heatsink for good measures and forget about overheating, stopping to cool down or avoid riding on the hottest days. It does come with a price, oil drippings.

Cardboard underneath where you park your bike and half a quart of motor oil might work.

 

[liveforphysics]

If you have a flir and a high current power supply, it can reveal a lot to just pass steady current through a couple phase leads and observe the deltaT ramp between windings and stator iron.

 

[Offroader]

Offroader if your thesis is correct FF will not work optimal in your motor, maybe use old fashioned oil filled hub? Surely that will allow for steep climbs even in the harshest condition. Add a heatsink for good measures and forget about overheating, stopping to cool down or avoid riding on the hottest days. It does come with a price, oil drippings.

Cardboard underneath where you park your bike and half a quart of motor oil might work.

macribs, well, read this post here by justin when he compared FF against filled with ATF oil, while the ATF oil did better, it didn't change the cooling results all that much.

Then look at the leakage of the ATF oil, this is why I won't go oil filled. Everyone who oil filled had serious issues no matter what they tried.


https://endless-sphere.com/forums/viewtopic.php?p=1106995#p1106995

Here is what gets interesting. Here are the posts of Justin potting the MXUS motor. The last thread below he found that potting didn't do all that much.

https://endless-sphere.com/forums/viewtopic.php?p=1107953#p1107953

https://endless-sphere.com/forums/viewtopic.php?p=1108283#p1108283

https://endless-sphere.com/forums/viewtopic.php?p=1116645#p1116645

Justin found that ATF oil didn't do all that much better than FF. Potting didn't do all that much better also. It would seem then it probably isn't worth the trouble to increase the thermal transfer of the windings to stator.

Without a doubt if you are running peak watts and climbing a hill, the transfer would be more important from windings to stator. When my MXUS motor was not setup correctly in my MAX-E, I would hit peak temps of 150C just riding up the street, the heat just obviously got stuck in the windings and could not transfer out because the motor case was cold to the touch. But I guess for most of the riding we do, the delay of the heat transfer isn't that big of a deal. Most of us wouldn't climb that steep hill for such a long time at such a high throttle for this to matter.

What I think is that there is a thermal bottleneck from the heat escaping through the motor case and you are mostly limited on what you can do with a sealed hub motor. The heat takes time to conduct through the metal case, and then it is limited in how fast it can also shed it on the outside.

 

[fechter]

Most of the heat transfer is going to be in the slots. Filling the air spaces will help. Motor varnish or high temp runny epoxy might be good. Anything is better than air.
I have a similar issue with my A2B motor.

 

[liveforphysics]

Most of the heat transfer is going to be in the slots. Filling the air spaces will help. Motor varnish or high temp runny epoxy might be good. Anything is better than air.
I have a similar issue with my A2B motor.

That's a common misconception. It's true almost anything is better than static air, but air that is turbulently and rapidly flowing around easily gets better than varnish or thermal epoxies and lots of other materials that on paper look many orders of magnitude better than static air.

 

[madin88]

That's a common misconception. It's true almost anything is better than static air, but air that is turbulently and rapidly flowing around easily gets better than varnish or thermal epoxies and lots of other materials that on paper look many orders of magnitude better than static air.

exactly!
most hairdryers have 1500-2000Watts and this heat is removed by one tiny fan.
or a heat gun can be used as barbecue lighter

 

[fechter]

OK, if you have forced air, I could see that working. In a sealed hub motor, the air is nearly static in those slots.

 

[Cowardlyduck]

All this discussion reminds me of the guy that tried attaching fan blades to the inside of the side covers....can't remember who, but I do remember it ended badly when they ripped off during use.

If moving the air around is all that is needed (and I already know this from my own fan experiments) then attaching fan blades to the inside of the covers has as much merit as and EDF, IMO. While you will get much better performance with a vented motor, if that's not an option then I would try the blades inside the covers....just make sure you tuck all the wires away securely first.

On another but similar note, I accidentally passed 18S through the fans mounted in my Leaf Motor and fried them. I forgot that I only had 7 fans mounted and they are max 7.4V fans. I've already got a bunch of replacements so will be mounting 12 soon. This is the other motor I also put FF in together with the fans so I will be interesting to see if it also looks like my HS4080 did when I cracked it open.

Cheers

 

[Offroader]

Working on installing my EDF today. I'm using a 35 mm EDF, with a 6000 KV motor. Comparing this to my old 30mm EDF with a 13,000 KV, I can run this at 12 volts instead of 8 and use much less amps. This 35mm EDF will easily blow double the air as the old 30mm 13,000 KV one, and the old one blew a lot of air.

Here are some pictures for those who are interested in how I set this up.

I use this sticky aluminum tape I think is used for duct AC or heating. I put it on both sides and it simply works well, is light weight, easy to install, and may offer some thermal advantages compared to plastic.

You can see the enlarged hole for the 35 mm EDF





The MXUS Turbo has two sets of halls and temp gauges. Since I only need one side they will give me a free 6 wires to easily connect the 3 phase EDF, I will use 2 wires per gauge. In my old setup I ran a single 24 AWG wire to each phase and that was not enough as the wires got hot.

I certainly don't want to run new wires and have to deal with getting them through the hub and replacing the stock setup.

If you look closely you will see that MXUS temp gauge wire actually broke off the temp probe by just lightly pushing on it, MXUS is known for their bad quality control.



Here I use silicone to glue the EDF in place because it holds it sturdy enough and I can easily remove the EDF and remove the silicone if needed.

 

[Cowardlyduck]

Great stuff Offroader. Looking forward to the results.

Perhaps we should shift this stuff over to the cooling fans inside hub motors thread and keep this thread as it's title states to be the definitive tests side of things.

Cheers

 

[madin88]

WOW, i'm impressed with your work!

About the hall wires and using them for powering the EDF is a great idea. This way you can install the ESC in the frame
However, if the hall wires are not shielded, i would change the wiring and split them into two cables outside the motor. I can understand that you not planning to touch this since swapping motor wire through the axle is a time consuming job, but it would improve the signal quality from the sensors to the controller.
The phase wires will create a BIG magnetic field and this field creates lots of noise on those wires.

When i was abusing one of my MXUS with max phase amps on a Max-E, i sometimes have seen HALLS! error which must have been because of the noise.
Adding a small capacitor between the hall supply lines inside the motor (ceramic SMD type) would also improve the signal quality

 

[Offroader]

Yes the bikes phase wires do put out a lot of noise as I have discovered when using a servo tester to control the ESC / EDF. I wanted to put the servo tester on the handlebar but every time I gave my bike throttle the EDF would shut off, so I had to keep the servo tester right next to the ESC. The long servo wires run to the servo tester were picking up the noise from the bikes phase wires.

It really sucked not being able to control the fan speed at my handlebars so I looked into a way to get it to work. I found this thread here which goes into a lot of detail about how to eliminate the noise, this is from guys flying those RC airplanes around at long distances.

http://fpvlab.com/forums/showthread.php?3641-The-Most-Amazing-Invention-Ever-For-FPV-%28CAT-6%29

Basically, I need to use shielded CAT 6 cable or something similar, and then I need to make sure the shield in the wire is grounded properly.

I don't have trouble with my halls picking up noise, I did once have to run a separate ground wire for my temp gauge.