Good! Someone finally oil cooled a hubbie!! HS3540

[whatever]

this is a bit off topic but.... if a hubmotor is getting hot and needs to be cooled, is it not better to rewind it to suit the power level being used? Would save alot of wasted energy

[Spicerack]

I need to go thrash it around and get a screenshot, but the air cooled doublepie with my inexperienced 200 pound self has exceeded 200wh/mile average on the socal racetrack, it did overheat but not to the point of destruction, The temp sensor did pass 190C (thank you 220C rated wire ). BUT about 1KW of the heat its making is iron losses due to a construction error, so when I get in two more pies ill have to do another test.

I know it's bit off topic...but... what was the construction error that led to the excess heat? Are the stator laminations shorted together or something?

[Miles]

this is a bit off topic but.... if a hubmotor is getting hot and needs to be cooled, is it not better to rewind it to suit the power level being used? Would save alot of wasted energy

Not sure what you mean? For a given torque output, copper losses will be the same, no matter how you wind it (assuming equal fill factor).

[Farfle]

I need to go thrash it around and get a screenshot, but the air cooled doublepie with my inexperienced 200 pound self has exceeded 200wh/mile average on the socal racetrack, it did overheat but not to the point of destruction, The temp sensor did pass 190C (thank you 220C rated wire ). BUT about 1KW of the heat its making is iron losses due to a construction error, so when I get in two more pies ill have to do another test.

I know it's bit off topic...but... what was the construction error that led to the excess heat? Are the stator laminations shorted together or something?

Yup, when the glue cured the two stators had rotated a very small amount relative to one another, so I couldn't fit the wire into the slot , so I took a skinny abrasive disk to the slot to open it up a small amount. which shorted all the iron together on the edges of the stator teeth.

[whatever]

Miles a hub motor is going to get hot if its overvolted, run within specs should be negligible heat.
Instead of cooling rewind for the amp/volts that are intended for it. Number of turns and thickness of wire are the main two variables, effect on torque I have no idea, but there should be some formulas out there to find out.

[hjns]

this is a bit off topic but.... if a hubmotor is getting hot and needs to be cooled, is it not better to rewind it to suit the power level being used? Would save alot of wasted energy

Not sure what you mean? For a given torque output, copper losses will be the same, no matter how you wind it (assuming equal fill factor).

Miles a hub motor is going to get hot if its overvolted, run within specs should be negligible heat.
Instead of cooling rewind for the amp/volts that are intended for it. Number of turns and thickness of wire are the main two variables, effect on torque I have no idea, but there should be some formulas out there to find out.

I think this is all completely true for motors that are being run within specs. If a motor is getting hot within specs, either the specs are wrong, or something else is wrong. And rewinding it would probably improve the specs, preventing the motor from getting hot.

However, feeding a HS motor >1kW can be considered out of specs. I think that there is a limit to the amount of rewinding that can be done due to the physical spatial limitations to fill the stator with more copper, regardless of amount of turns or size of copper wires. Therefore, cooling starts to become most important when you try to tease out the highest performance with a given stator and a given winding, where the variables are more "soft", e.g., battery voltage, battery current and phase current.

[Miles]

Miles a hub motor is going to get hot if its overvolted, run within specs should be negligible heat.
Instead of cooling rewind for the amp/volts that are intended for it. Number of turns and thickness of wire are the main two variables, effect on torque I have no idea, but there should be some formulas out there to find out.

Methinks you need to read up a bit more on motor theory......

Good resources to start with:
http://books.google.co.uk/books?id=gbID ... es&f=false

http://books.google.co.uk/books?id=lKs7 ... hi&f=false

[thepronghorn]

Miles a hub motor is going to get hot if its overvolted, run within specs should be negligible heat.
Instead of cooling rewind for the amp/volts that are intended for it. Number of turns and thickness of wire are the main two variables, effect on torque I have no idea, but there should be some formulas out there to find out.

Really? Really really? Miles knows exactly what he's talking about. He knows how hubmotors get hot, why they get hot, and how to prevent it. What we're talking about here is not running within specs, but running far beyond specs and still having the motor survive. The motors have already mostly maximized copper fill, so the only options for more power involve higher rpm/smaller wheels or better cooling.

[whatever]

As mentioned I have no idea how torque relates to rewinding a hub motor, my suggestion is not based on any thorough understanding of hub motor windings, you can take it more like a question if you like.
For example: lets say you take a x5304 ( older style) which I assume is wound for 48v 20-25amps to run within spec.
Then you overvolt and overamp it to get more speed and power ( torque ), would it not be possible ?? ( this is a question !)
to rewind it such that you could run it on say 100v and 60amps such that it wont overheat?????
Is this possible???

[Miles]

No, I'm afraid you can't gain anything by rewinding with a different turn count.

Terms like overvolt and overamp are a bit confusing and best avoided.

[whatever]

for noobs like me, found very nice site on hubmotor rewinds etc
http://www.avdweb.nl/solar-bike/hub-mot ... -variables

here is a very interesting idea for high effeciency at low rpm only winding each second stator
http://www.aerodesign.de/peter/2001/LRK ... o_eng.html

[parabellum]

here is a very interesting idea for high effeciency at low rpm only winding each second stator

Compromising a lot of things. Not very good for high power application and not usable in all motor designs.

[dfar]

posted the "plug/vent" system I'm using so far, working very well in combination with the ATF sealant and 100ml ATF fluid... no leaks yet!

viewtopic.php?f=2&t=37972&start=135#p572219

[hjns]

Hi dfar,

That is a huge plug. Doesn't it interfere with rotation of the motor?

[dfar]

Hi dfar,

That is a huge plug. Doesn't it interfere with rotation of the motor?

No it does not. It is placed on the disc brake side and drilled low enough that it misses the break caliper and wiring.

Good point though, one has to be certain that they have enough clearance with the brake calipers they are using. On my other motor I run regen braking so I don't use a back brake caliper so no worries about clearance there.

[RichiePA]

Good thread!

I've been resisting the idea of drilling holes in my sidecovers for better cooling as I can imagine the motor getting mud and sand and suchlike in there so sealed with oil seems like a good idea and I guess if I find it to be too leaky or whatever it should be simple enough to reverse the changes since it sounds like it would be just one filler/breather hole that would need to be plugged.

At the moment I've been pretty restrained in my riding style since I only run 72V 5A/hrs pack and the motor generally doesn't get very hot or if it does I expect the lipo sag is having a greater affect. once my new 72V 15A/hr pack is complete I expect I'll be giving it the beans a bit more so if I think the motor is getting hot then this is defiantly going to be the next mod on the to do list.

Keep up the good work guys.

[megacycle]

Read a lot about drilling holes in the covers, no reports of any damage due to water or any other penetration exception of one pebble incident someone was saying they even wash theirs out after salty conditions, forgive me doc if it was'nt you, i think that they have been coating the stator inhibiting corrosion.

Newbescence probably glaringly obvious, but is it possible for the covers to be cut down to a 3 spoke arrangement like a mag wheel shape with just 3 wide spokes and cover the inside with fine mesh and say gortex or similar, thinking something like this for a 5404 build in progress with some series air cooling, besides hating oil,it loves to soak into my lily white dry skin.

[amberwolf]

You're probably thinking of my post suggesting you can wash out salt after a ride. We don't get ice on the roads here in Phoenix, so we don't have salt, but we have all sorts of dirt, sand, construction debris, nails, metal and plastic from car crashes, etc., and none of it has gotten into my air-cooled 9C.


Back to the topic of oil-cooling...I just started testing this today with my 36V-Fusin-@48V, which has had problems at low speeds and medium loads where it gets so hot I can't touch it, and the halls stop responding. This is a geared hub, so the heatsoak problem is much worse than a DD. I'm using regular ATF F-type, cuz that's what I had lying around.

viewtopic.php?f=2&t=15570&p=576368#p576368

So far, the only real test results I have are:
--the hub still works
--the ATF is splashing around inside pretty good, then running down the wiring and stuff into the axle wiring hole, and out the end, despite attempting to seal it with rubber cement (which I didn't expect to work, but the only other stuff I had was automotive silicone that had hardened in the tube and was useless for this purpose).

Until I solve that problem, I can't take it on a road test, as it gets on the rim braking surface, and I only *have* that wheel to brake with right now--the rear has no brakes yet.

I have done no real temperature testing, as it has only been off-ground so far. But at the ~200W it uses at full throttle no load as an overvolted-36V@48V(58V, really) motor, it gets warmish on the outer covers in only a minute or two, while still running the test, whereas before this as aircooled only, it would not seem to be warm at all, until long after the test run was over. Only the axle would seem to be warm (sometimes even hot).

So the ATF is doing *something* for the heat transfer; I just don't know how much. It *should* help keep the motor windings and especially the magnets (on teh spinning bell) cooler, by moving the heat in them to the outer cover much faster than with air only. I think it probably is doing that.

Now I need to find a temperature sensor, like an old BBQ one, and install that when I seal up the axle, and retest without the ATF first, and then with it.

[John in CR]

AW,
I hope I can count on you for the needed test info pervasively lacking, which is with and without oil no-load current and no-load speed. It will be important for determining optimum fill levels. I think the warmish covers is good in your test so far, but that's as long as much of the heat isn't new.

[hjns]

HI John in CR,

I am waiting for my 10k thermistors ntc to come in. Then I will give everyone just that, for both motor and controller (if successful). If in the meantime somebody else does it first, it means we will have an independent control from my results - even better!

[hjns]

On my other motor I run regen braking so I don't use a back brake caliper so no worries about clearance there.

Hmm. I use regen AND a disc brake. I find that my regen braking just does not cut it in case of an emergency brake. In such a case, I want both my front and rear hydraulic calipers to be fully engaged in providing the maximum braking without blocking.

[amberwolf]

AW,
I hope I can count on you for the needed test info pervasively lacking, which is with and without oil no-load current and no-load speed. It will be important for determining optimum fill levels. I think the warmish covers is good in your test so far, but that's as long as much of the heat isn't new.

That's part of what I was going to start doing last night, but all the little issues ate up my time and I didn't get teh chance. So these will be done next chance I get to work on this project, when I have the sealant for the axle and the temperature sensor to install.

I can say that no load current appeared normal, from what I can remember of previous testing, but I need actual numbers at specific speeds to be sure, and to see any small differences that come up, which will affect efficiency (and thus heat).

[amberwolf]

FWIW, I did a little test overnight: I took some of the dried-up rubber cement off the outside of the lid and bottle, and left it soaking in a capful of ATF. Today before I left for work, about 13 hours after doing that, there was only liquid in the cap. ATF most definitely completely dissolves rubber cement, even if it's been dry for months!

So don't use it to seal an ATF-filled motor.

[Xanda2260]

Now that's cool. Eagerly awaiting further results AW,

Cheers

[dfar]

On my other motor I run regen braking so I don't use a back brake caliper so no worries about clearance there.

Hmm. I use regen AND a disc brake. I find that my regen braking just does not cut it in case of an emergency brake. In such a case, I want both my front and rear hydraulic calipers to be fully engaged in providing the maximum braking without blocking.

kids this is good advice,