Definitive Tests on the Heating and Cooling of Hub Motors

[sketchism]

Excited to help you test!

Me too haha, i literally already have your box addressed and ready to send as soon as i get the call from the depot that they're here.

We've only run a couple of bikes with them so far, but already it feels like riding without them is akin to running a car without a radiator or a cpu without a heatsink, but i'm biased so i think your findings will be great to see.

Did you do any study on heat effects on magnets since the FF bridges all the stator heat directly to the magnets this could cause some demagnetization. I fear that this solution that you have will cause people to dump ungodly amount of power into motors without consequences. This should be paired as a trifecta maybe venting FF and the ring rolled into an ultimate solution with a CA doing thermal limiting being a solution in itself. Keep up the good work.

This was a Good question asked on the ES facebook and i was thinking about including it in the FAQ but wasn't sure so ill include my thoughts here too.

My theory is that because the overall temp is so much lower across the spectrum. (In my normal commute during riding that would see my average at about 90c i was now sitting at 50C) perhaps this could actually be an overall solution to (or at least reduce) the increased risk of thermal demagnetisation from the FF coupling the stator to the magnets thermally, we were already doing that and getting them much hotter than before, now we are pulling the heat out from behind them.

I also found that running them cooler, also saw higher efficiency, lower power consumption and slightly longer range on the same ride at the exact same settings.

Granted people will now run more power, but i think the magnets will still be seeing the advantage of having a direct way to shed that heat instead of soaking.

During testing the hubs cooled down faster by a factor of 2 to 2.5 when parked straight after a ride at 100C (i plan to film this on the FLIR cam as one of the tests as well)

i do agree that i would 100% recommend running a CAv3 with thermal rollback set to avoid peak temps regardless though, but that would be the same with or without the sinks.

 

[Cowardlyduck]

I would have also jumped in to beta test these heat-sinks had I not already done my own thing with the little ones + epoxy. Any new motor's I get will be sure to have one of these new heat-sinks attached.

We still have to remember though, these heat-sinks don't make our hub motor's invincible.
I attached 4 sinks to a little 250w geared hubbie I'd previously oil filled. It worked great and allowed me to run up to 750w reliably for the last few months.

Well I'm yet to verify, but I'm pretty sure I completely cooked it yesterday after hitting 90c on the covers. Touching the heat-sinks nearly burned me so they were definitely trying they're hardest to dissipate heat.
Ironically, I don't think it was heat from running the motor at 750w, but heat from it's rubbing, misaligned drum brake attached to the side of the hub motor.

Cheers

 

[Cowardlyduck]

I'm sure the PM I got is probably indicative of other's also wondering where I got those temp sensitive stickers from, so I'll share here also.

I've been using these since 2012 and TBH I'm surprised I haven't seen anyone else on ES using them yet as I've posted about them several times.
30-60C http://au.rs-online.com/web/p/temperature-sensitive-labels/0286169/
60-90C http://au.rs-online.com/web/p/temperature-sensitive-labels/0286175/

Minimum order is a 10 pack so ~$20 AUD, but I've been through several packs now over the years as I used them on all my motors and I've also started putting them on things like oven doors, soldering irons, hot glue guns, lamps, etc, and they are very useful.

Cheers

 

[Samd]

Hang on a minute. Where did you put the heatsinks on a geared hub in relation to the heat flow path?


Sent from my iPhone using Tapatalk

 

[Rix]

I would have also jumped in to beta test these heat-sinks had I not already done my own thing with the little ones + epoxy. Any new motor's I get will be sure to have one of these new heat-sinks attached.

Well I'm yet to verify, but I'm pretty sure I completely cooked it yesterday after hitting 90c on the covers. Touching the heat-sinks nearly burned me so they were definitely trying they're hardest to dissipate heat.
Ironically, I don't think it was heat from running the motor at 750w, but heat from it's rubbing, misaligned drum brake attached to the side of the hub motor.

Cheers

I am sure you cooked your motor, If you went north of 90C, say 200F on the outside cover of a geared motor, your core would have had to seen 400F or more. Split the case and tell us what you go.

 

[liveforphysics]

Exciting stuff sketchism-- thanks for sharing all the juicy bits of data so we have more to drool over while we wait :lol:

Regarding FF, I may have missed a post which spells out why so many people are curious about testing evaporation specifically so I don't think evaporation is much of an issue for "disappearing" FF. Contamination and resulting loss of properties is the most likely cause, IMO. Please inform me (both here and via PM) if I've missed a part of the conversation which clarifies this definitively.

Regarding the magnets losing coating, this appears to be physical and not chemical in nature. The magnets are being scrubbed by a slurry of iron (and other particles if contaminated/vented).

Agreed it had a physical component, but my hunch is that it would not have had that physical contact had the base of the oil not softened the red varnish. It's not at all uncommon to find various specialty oils that soften or dissolve various plastics/glues/paints etc. Many types of brake fluid and loc-tite soften and consume many plastics and paints as examples.

 

[Cowardlyduck]

Hang on a minute. Where did you put the heatsinks on a geared hub in relation to the heat flow path?

They are on the outer shell between the spoke flanges as you can see.
Yeah normally there is no heat path to that point, but I added oil to this hub to bridge that gap.

I am sure you cooked your motor, If you went north of 90C, say 200F on the outside cover of a geared motor, your core would have had to seen 400F or more. Split the case and tell us what you go.

Yeah, but as above, it has oil in it, so the difference between stator and shell would not be much.
Will post the results when I get around to opening it up.

Cheers

 

[Rix]

DURRRRR! I missed that CD, yah, the oil cooling, it may not be cooked. Either way let me know.

 

[sketchism]

They're here!

Just want to do a little more testing on the thermal bridging methods, Paste/grease, Epoxy or Thermal tape before we go to full beta to everyone interested.

ill probably send them out with a few different methods to ensure we find the best all-round option but it might work out that different methods suit everyone.

thermal tape is the neatest and fastest method i've found so far
arctic silver thermal epoxy has better thermal conductivity but is super permanent & very expensive (50% of the cost of the sinks themselves)
thermal paste has the best conductivity but i want to ensure it's easy to assemble

 

[litespeed]

Those look just excellent.

Tom

 

[fechter]

They're here!

Just want to do a little more testing on the thermal bridging methods, Paste/grease, Epoxy or Thermal tape before we go to full beta to everyone interested.

ill probably send them out with a few different methods to ensure we find the best all-round option but it might work out that different methods suit everyone.

thermal tape is the neatest and fastest method i've found so far
arctic silver thermal epoxy has better thermal conductivity but is super permanent & very expensive (50% of the cost of the sinks themselves)
thermal paste has the best conductivity but i want to ensure it's easy to assemble

Silicone glue has good thermal conductivity and can be pulled of later if needed. If the gap is small, the conductivity of the bridging material is not so important.

 

[sketchism]

Looks like our best options are

Regular ebay silicone heatsink glue/plaster @ 0.67W/m-k (very cheap and feels it lol)

For ease of install - 3m Thermal tape 0.2mm @ 2W/m-k

AS1802 silicone adhesive @ 2.30 W/m-k (but i cant find it any smaller than enormous 310mm gun tubes)

Hy880 Thermal paste with Thermal Conductivity: >5.15w/m-k (standard thermal CPU paste is about 1Wm-k to 1.5W/m-k) this is probably the best non permanent solution

or Arctic silver permanent 2 part epoxy as mentioned before at 7.5 W/m-K (this has to be broken off to change spokes etc)

 

[macribs]

Nice heatsinks. I am thinking regular thermal paste will work wonders. Those are great performers and can are commonly available at fair prices.
Thermal tapes or pads seems very easy so will be nice to see how they performs.

 

[Allex]

Nice set sketch!
I will take one 30mm in black

 

[Cowardlyduck]

Looks like our best options are

Regular ebay silicone heatsink glue/plaster @ 0.67W/m-k (very cheap and feels it lol)

For ease of install - 3m Thermal tape 0.2mm @ 2W/m-k

AS1802 silicone adhesive @ 2.30 W/m-k (but i cant find it any smaller than enormous 310mm gun tubes)

Hy880 Thermal paste with Thermal Conductivity: >5.15w/m-k (standard thermal CPU paste is about 1Wm-k to 1.5W/m-k) this is probably the best non permanent solution

or Arctic silver permanent 2 part epoxy as mentioned before at 7.5 W/m-K (this has to be broken off to change spokes etc)

I initially used thermal pads with my heat-sinks, but only held them on with tightened string, so not much pressure. It didn't perform too great, but still did work ok.

I then went straight to Arctic Silver epoxy. It works great for my setup with the sinks spaced out around the hub. Thermal conductivity is great and gets better as the hub gets hotter. That being said, as mentioned, it is most definitely permanent and would be super difficult to get off. Even more so with spokes in the way.

I've also used Hy880 for my side covers and a few other projects and it works great. Can be found very cheap if you look around also, so that's my pick of the options.

Cheers

 

[eyebyesickle]

I love what all you are doing here... So much to learn from on this thread... The progress is gaining momentum for EVs... I mean with the ferrofluid and the heatsinks... whats next?

What about adding a very slight turbine along the outer edge of the motor shell... not even all the way through for ventilation, but just a very thin portion, that directs air straight towards the heat sinks anyway...

I mean, you could even have one around the axle, like the controller turbine covers for magic pie... With an overlapping design, to better protect from the elements at a standstill, but I am just talking here.

I mean the next step, would be to implement this stuff all, including heat sinks, directly into the motor shell/housing... correct?

I just want to learn from the experts, then do something. Someone point me!

 

[sketchism]

If anyone is considering the HubSinks or just wants to be involved in general as the beta versions are getting sent out, i'm collecting a database of actual outer diameters/circumferences of the magnet rings on various hubs in real life, i've found a few cases of the tech drawings being a couple of mm off the practical measurements even within identical models, so i'm making sure i build the tolerance into final design.

if you have a second to measure the real life circumference of the magnet ring of your hub, i'd be certainly interested in everyones results!

 

[eyebyesickle]

I would have also jumped in to beta test these heat-sinks had I not already done my own thing with the little ones + epoxy. Any new motor's I get will be sure to have one of these new heat-sinks attached.

We still have to remember though, these heat-sinks don't make our hub motor's invincible.
I attached 4 sinks to a little 250w geared hubbie I'd previously oil filled. It worked great and allowed me to run up to 750w reliably for the last few months.

Well I'm yet to verify, but I'm pretty sure I completely cooked it yesterday after hitting 90c on the covers. Touching the heat-sinks nearly burned me so they were definitely trying they're hardest to dissipate heat.
Ironically, I don't think it was heat from running the motor at 750w, but heat from it's rubbing, misaligned drum brake attached to the side of the hub motor.

Cheers

Nice. I like the low profile heatsinks, as very large versions might not suit everyone's taste, and are obviously not for smaller hubs - or low key conversions.

A few questions, if you don't mind:

1) - Generic Heat Sinks - So obviously there is a wide variety available, and without getting too scientific (or be as detailed as you want, not trying to hassle), want to summarize any observations/findings? I am guessing you tried a variety of them... (I see you briefed us on the adhesive, seems spot on)

2) - What type of Oil have you tried/do you prefer
2a) - Any sealant methods/tips ~ Are you retaining most of the oil or is there leakage?

3) Is that a 26" rim it's laced to? What top speed are you getting @ 750w?

The legal limit here in Nashville is ridiculous, totally messed up my high power build plans with the July 1st legislation. So next to that, I really like low key/stealth conversions, which honestly, the geared hubs are pretty much your only option for. I already get geared hubs with high temperature magnets, and 180C copper, but I am thinking (well, planning) about introducing some oil or what not into the mix, and of course implementing heat sinks as well. That should get these things kickin...

Also, any update on the failure/cause?

THANKS!

 

[Cowardlyduck]

Nice. I like the low profile heatsinks, as very large versions might not suit everyone's taste, and are obviously not for smaller hubs - or low key conversions

A few questions, if you don't mind:

Thanks! and no worries.

1) - Generic Heat Sinks - So obviously there is a wide variety available, and without getting too scientific (or be as detailed as you want, not trying to hassle), want to summarize any observations/findings? I am guessing you tried a variety of them... (I see you briefed us on the adhesive, seems spot on)

Nah, I don't have the resources to test a bunch...I just thought about it a bunch first and ordered what I thought would work well. For a standard square or rectangle heat-sink with a flat base, you need a smaller one to make up for it's lack of conforming to the round magnet ring. That is unless you can bend it to fit, but then I would just get one of Sketch's properly made heat-sinks.
These are the heat-sinks I bought:http://www.ebay.com.au/itm/221725657473

As for thermal performance, they definitely do help a lot. No idea on thermal capacity, but they get real hot when the hub is hot and seem to match the temp of the hub shell pretty closely. The more you add, the more effective they will be also. In the case of my Leaf motor, I added about 12 of them.

2) - What type of Oil have you tried/do you prefer

This was my first attempt at doing oil cooling and I didn't want to go buy new oil to try it, so I just used what I had on hand which happened to be chainsaw oil. It's sticky, thick oil, but it actually seemed to work quite well as it became far less viscous as it heated up and also helped lube the planetary gears quite well.

2a) - Any sealant methods/tips ~ Are you retaining most of the oil or is there leakage?

Nope! I failed in this regard. I tried adding plenty of silicone sealant and added a very small vent hole near the axle, but oil kept seeping out at a decent rate, especially after the motor got hot. I would come back to the bike after it had been parked all day at work to find a wet patch of oil about the size of a small melon under the rear motor. I honestly don't know how much is left now.

3) Is that a 26" rim it's laced to? What top speed are you getting @ 750w?

Yup, 26" rim. On 12S (45V nominal) I'm getting 30-35kph, when I bump the throttle into 120% mode I can get 40-45kph, but that chews the battery.

The legal limit here in Nashville is ridiculous, totally messed up my high power build plans with the July 1st legislation. So next to that, I really like low key/stealth conversions, which honestly, the geared hubs are pretty much your only option for. I already get geared hubs with high temperature magnets, and 180C copper, but I am thinking (well, planning) about introducing some oil or what not into the mix, and of course implementing heat sinks as well. That should get these things kickin...

I would be careful about marketing something like this as a reliable product. It doesn't take much to cook these things even with all the cooling mods in place. I managed to cook a completely stock Stealth Fighter motor back in 2011 just from riding on sand on a 40C day.

Also, any update on the failure/cause?

Not yet....flat out with work even on the weekends atm....hopefully soon.

Cheers

 

[eyebyesickle]

Thanks. FYI not talking about marketing a dripping oil filled 'hot rodded' lil geared hub, just learning and tinkering for now! It would be hard to ever expect/rate 750w continuous from one of those long term...

But still, you're on to something!

I wonder if the final failing had anything to do with a lack of oil (due to leakage) to transfer heat to the outside...

Thanks for your reply!

 

[macribs]

CD did you ever try that one way "breathing valve" I saw in this thread? (can't remember who did it) It seems that valve will allow pressure to escape to the outside as the hub and oil heats up. And that should minimize the oil spill. Kudos for taking a life hack approach when choosing oil, I liked the use of chain saw oil very much.

 

[LI-ghtcycle]

If anyone is considering the HubSinks or just wants to be involved in general as the beta versions are getting sent out, i'm collecting a database of actual outer diameters/circumferences of the magnet rings on various hubs in real life, i've found a few cases of the tech drawings being a couple of mm off the practical measurements even within identical models, so i'm making sure i build the tolerance into final design.

if you have a second to measure the real life circumference of the magnet ring of your hub, i'd be certainly interested in everyones results!

I have a CroMotor (Hubzilla) V2 IIRC, (confirming this shortly) as mid-drive, so I can measure it for you pretty easy, will post measurements in the next few minutes.

***edit*** cannot seem to find my cloth tape, will post when I can.

Very excited to see this!

 

[GmagNeato]

If anyone is considering the HubSinks or just wants to be involved in general as the beta versions are getting sent out, i'm collecting a database of actual outer diameters/circumferences of the magnet rings on various hubs in real life, i've found a few cases of the tech drawings being a couple of mm off the practical measurements even within identical models, so i'm making sure i build the tolerance into final design.

if you have a second to measure the real life circumference of the magnet ring of your hub, i'd be certainly interested in everyones results!

I have a CroMotor (Hubzilla) V2 IIRC, (confirming this shortly) as mid-drive, so I can measure it for you pretty easy, will post measurements in the next few minutes.

***edit*** cannot seem to find my cloth tape, will post when I can.

Very excited to see this!

You could use a length of flexible wire if you have any (such as 20AWG silicone wire) and wrap it around the hub. Pre-mark one end with a sharpie and start with that end. Be careful to pull it taught and make sure it runs along the middle of the magnet ring all the way around (the wire inside will minimize any stretching so pull it tight). Then mark it where it meets the pre-marked end with a pen or sharpie. Then take the wire out and measure it from the pre-marked end to your second mark with a regular tape measure. If you're careful it should be at least as exact as a cloth tape would be.

 

[Cowardlyduck]

Thanks for your reply!

No worries. Happy to help/share my findings.
Yeah, I'm not saying it's impossible for a commercial product to be made from this, and I hope someone does, but it would require extensive testing the likes of which few of us could manage.
That being said, I hope some existing manufacturers are paying attention as we all want lighter, more powerful motor's and heat-sinks/oil/FF is one key way to doing just that.
I also forgot to mention, in my geared hub I also added a small amount of Ferro Fluid. I added about 3-4ml prior to adding the oil. My theory was that it would help move the heat to the internal magnet ring which the oil contacts more of. I'm not sure how the oil/FF interact together, but it didn't seem to cause any issues.

CD did you ever try that one way "breathing valve" I saw in this thread? (can't remember who did it) It seems that valve will allow pressure to escape to the outside as the hub and oil heats up. And that should minimize the oil spill. Kudos for taking a life hack approach when choosing oil, I liked the use of chain saw oil very much.

Thanks!
Given the amount of oil that was leaking, I don't think it was only coming out of the breather hole. I think it was also making it's way past the silicone seal on the side cover, and out the axle on both sides. I think that's part of the reason the performance of my hub brake was so poor. :lol:

My main lesson learned from oil cooling at this stage is that while it can work, it's not easy to seal it up so no oil leaks. I think we electric snobs could learn a thing or two from the decades of difficulties ICE engines have had with containing their oil.
Maybe someone needs to make a side cover oil sealing gasket?

Cheers

 

[eyebyesickle]

I think that with the valve, it is more about the overall pressure not building up, still resulting in less overall leakage...

Thanks again for the answers/motivation.

Im going to go fry a little geared hub now hahaha j/k (kinda)