Oil cooling your hub- NOT snake oil!
- Thread starter Spicerack
- Start date
GrayKard
10 kW
Thanks, I needed a good laugh. :lol:bigmoose said:
I'm considering oil cooling my next motor when I find the right one. I did an impulse buy off ebay recently of a hub motor but was pretty pissed off when I cracked it open and saw that the magnets are only 25mm wide in a 60mm wide magnet ring. Must be the low power version.
Gary
GCinDC
100 MW
seal on the drive side finally leaked, after 4 weeks?
i'll pop it open this weekend and take lots of pics.
i'll pop it open this weekend and take lots of pics.
zombiess
10 MW
If it weren't for all the leaking issues I'd love to give this a try, but I absolutely hate things that leak! If a hub motor was specifically designed to be liquid tight and contained wire which was oil safe I'd love to do this to smooth out the peak heat transients. I think my stepped drill bit is going to see some action again soon. I've been watching this thread hoping someone comes up with a great how to on sealing a motor that lasts longer than a few weeks.
GCinDC said:seal on the drive side finally leaked, after 4 weeks?
i'll pop it open this weekend and take lots of pics.
What do you think caused it? bearings moving around caused by shock? or some improper sealing? and what oil did you use, I've been reading about INOX MX3 and it seems to be the best despite costing 20 dollars for 300g
GCinDC
100 MW
big moose recommended dextron III and that was enough for me. i think it was actually an imitation, or replacement for dextron III.ian.mich said:
well, unfortunately, it's hard to tell, but i found oil inside side cover bearing housing, so i'm inclined to think you're right - leaking was due to shock, and obviously improper sealing. i'd suspected that the heat and shock would have caused my mountain of sealant covering the wires to peel away, but that was not the case as far as i could tell.ian.mich said:
there was NO leaking on either side when the wheel positioned horizontally in the vice. in fact, none came out of the tube! i've not measured how much came out, but i'd be surprised if it was more than 100cc (and i put 200cc in). the winding soak it in!
we've had temps here of 100F for a week and i regularly hit 110C (my buzzer limit) at mile 3.5 going uphill on the way home. and then nurse it. i suspect more oil would have helped cooling at lot. but cooler ambient temps made a big difference, i found.
anyway, some yummy photos of the examination:
Actually, this seems to have been moderately successful. I doubt that shock played a role - there is insufficient room for movement to cause meaningful shear forces from impact. This seems more likely to have been a slow degradation of the sealant bonding due to normal loading and movement.
There are four places this could leak:
The new bearing seals seem unlikely to have passed that much oil even though they are really dust seals not oil seals and lack garter springs, etc. This failure mode might be investigated further by cleaning the outside of a bearing and removing a seal to check for oil ingress.
Even with a fair fairly snug fit, (2) and/or (3) seem suspect in that the sealant was probably holding back the oil until normal shaft forces flexed, sheared, and worked the sealant loose. ATF has very low viscosity at room temp and is even thinner at the working temperatures you experienced. This coupled with its normal detergents and cleaners make it very difficult to seal by simple mechanical means.
If the source of the failure was indeed (2)/(3) then you might solve this by sealing the bearing races using a more aggressive ATF sealant as recommended in an earlier post. This should resist the ATF solvent action and bond more strongly to the metal. Sealing the joints with a fillet of sealant on the narrow faces of the races rather than on the load bearing surfaces might be more successful since such a fillet should experience little mechanical stress (problematic for the outer races because of assembly, but...). Here you should try to keep the bearing seals free of sealant so they retain their original flexibility and shape to properly seat the lips. (And, of course, be careful doing prep work with brake cleaner around bearings lest it leak in and dissolve the grease...)
Just some thoughts, FWIW
There are four places this could leak:
- the wire bundle/slot,
- between the outer race and bearing pocket,
- between the inner race and shaft, or
- through the bearing seals.
The new bearing seals seem unlikely to have passed that much oil even though they are really dust seals not oil seals and lack garter springs, etc. This failure mode might be investigated further by cleaning the outside of a bearing and removing a seal to check for oil ingress.
Even with a fair fairly snug fit, (2) and/or (3) seem suspect in that the sealant was probably holding back the oil until normal shaft forces flexed, sheared, and worked the sealant loose. ATF has very low viscosity at room temp and is even thinner at the working temperatures you experienced. This coupled with its normal detergents and cleaners make it very difficult to seal by simple mechanical means.
If the source of the failure was indeed (2)/(3) then you might solve this by sealing the bearing races using a more aggressive ATF sealant as recommended in an earlier post. This should resist the ATF solvent action and bond more strongly to the metal. Sealing the joints with a fillet of sealant on the narrow faces of the races rather than on the load bearing surfaces might be more successful since such a fillet should experience little mechanical stress (problematic for the outer races because of assembly, but...). Here you should try to keep the bearing seals free of sealant so they retain their original flexibility and shape to properly seat the lips. (And, of course, be careful doing prep work with brake cleaner around bearings lest it leak in and dissolve the grease...)
Just some thoughts, FWIW
I firmly believe that this oil-cooling will prove to be the ultimate for a race-hub. I'm glad to see several builders continuing to improve the oil-cooling. I cannot imagine a modern engineer designing an industrial motor without some type of air-cooling / oil-cooling / hybrid of both...which provides very real and measurable improvements.
GCinDC
100 MW
who are you!? 
:lol:
and had you noticed how much thick red stuff was left in the freewheel? the main leak must have been from the seals then!
i'm thinking that's ATF mixed with grease. (i presume the grease is not also red...)
i wasn't worried about the ATF getting on the seals cause i figured as soon as that part moved, the seal would break. i'm thinking of ways to prevent the oil from getting to the bearing. what i was trying to do was first seal the inner bearing race to the axle. (that seems to have worked fine. the bearing took some effort to remove and then looked dry.) then i figured with bearing seated on the axle up tight against the stator, i could fill in all around it with ATF sealant knowing that the bearing wouldn't stick when it had to move, and then it would be harder for the oil to get in there, and certainly wouldn't just splash down against the bearing. would it be bad if the bearing seal were fixed to the sealant? would it be better if in fact i wiped the bearing face with oil so the silicon wouldn't adhere, and then embed it in a mound of sealant anyway?
another relevant question may be, at what temp does the bearing grease melt and run? with ambient temps getting way up, the motor got hotter quicker, and i rode at 110C for a while, with occasional spikes to 120C. i wonder if the leak would have developed had i kept below say 100C...
btw, some clear jellyish stuff settled to the bottom of the used oil container. perhaps that's grease? or eroded bits of jbweld? or wire insulation? tempted to send it to the lab, but aw heck, i'll just do the old taste test.. :lol: just kidding.
and for the record:
meanwhile, a relaxing timelapse of taking apart the motor..
[youtube]dE13EtyOsAs[/youtube]
:lol: i used that awesome ATF sealant around the bearing races and side covers. i only used the clear, weaker ATF sealant for wire blob and slot. i just spent some time trying to peel away the clear stuff from the stator and it was definitely on tight!teklektik said:
well i'll be!teklektik said:
and had you noticed how much thick red stuff was left in the freewheel? the main leak must have been from the seals then!
i'm thinking that's ATF mixed with grease. (i presume the grease is not also red...)
i wasn't worried about the ATF getting on the seals cause i figured as soon as that part moved, the seal would break. i'm thinking of ways to prevent the oil from getting to the bearing. what i was trying to do was first seal the inner bearing race to the axle. (that seems to have worked fine. the bearing took some effort to remove and then looked dry.) then i figured with bearing seated on the axle up tight against the stator, i could fill in all around it with ATF sealant knowing that the bearing wouldn't stick when it had to move, and then it would be harder for the oil to get in there, and certainly wouldn't just splash down against the bearing. would it be bad if the bearing seal were fixed to the sealant? would it be better if in fact i wiped the bearing face with oil so the silicon wouldn't adhere, and then embed it in a mound of sealant anyway?
another relevant question may be, at what temp does the bearing grease melt and run? with ambient temps getting way up, the motor got hotter quicker, and i rode at 110C for a while, with occasional spikes to 120C. i wonder if the leak would have developed had i kept below say 100C...
btw, some clear jellyish stuff settled to the bottom of the used oil container. perhaps that's grease? or eroded bits of jbweld? or wire insulation? tempted to send it to the lab, but aw heck, i'll just do the old taste test.. :lol: just kidding.
and for the record:
meanwhile, a relaxing timelapse of taking apart the motor..
[youtube]dE13EtyOsAs[/youtube]
bigmoose
1 MW
Raaaats! It does look like it washed the grease out of the bearings. Now the oil in the bearings is as good as or better than grease for lubricity, so performance wise there is no problem with the oil getting in; if we can figure how to seal the bearing system to keep the oil from getting out. There is not much room on either side of the bearing either.
Harold in CR
100 kW
Using a similar experience with under water metal detectors and leaks:
When the motor is cool and the oil is distributed around the shaft as dripping might occur after a ride, then, going on a long ride OR a stressful ride, heating the motor, it might be causing a little pressure rise inside the motor from hot air expanding, that will cause the oil to weep out of the seals ?? Maybe a double lip seal might help a little ?
Would it be possible to create a very slight vent for hot air to escape, without leaking ?
When the motor is cool and the oil is distributed around the shaft as dripping might occur after a ride, then, going on a long ride OR a stressful ride, heating the motor, it might be causing a little pressure rise inside the motor from hot air expanding, that will cause the oil to weep out of the seals ?? Maybe a double lip seal might help a little ?
Would it be possible to create a very slight vent for hot air to escape, without leaking ?
GCinDC
100 MW
It's got a vent/tube drilled near axle... i ididn't check to see if it was blocked but doubt it.Harold in CR said:
just drilled larger holes in backup motor while this project underway
Just another back-seat oiler!!!GCinDC said:who are you!?
Yep - Rats. This looked like it would eventually develop into an issue but it's too bad it happened right out of the gate and to such scale. As BigMoose noted, you're better off with oil instead of grease anyway- I was planning on switching to shrouded or open bearings in my BMCs since I have exterior oil seals...GCinDC said:
This strategy seems spot on. Apologies if I missed this detail in earlier posts. It takes (2) and (3) off the table.GCinDC said:
I couldn't say. We are trying to use a simple dust seal for liquids. It has a very small sealing surface and no way to apply sealing pressure except the resiliency of the seal itself (no garter spring and likely no sealing lips in the conventional oil seal sense). I have no special knowledge here but would think that anything that applied force to the seal might distort it, causing the narrow sealing surface to skew, and compromise this delicate situation - just a guess. I would think an unencumbered, free-floating seal would be best....GCinDC said:
You may be on to something there...GCinDC said:
Dunno - I think you would be surprised at how little grease there was in the bearings initially, so unless there is only a smidge, it's something else...GCinDC said:
This turn of events is unfortunate but the two big questions that spring to mind are: Others have done this successfully - what is different about your implementation? Why did it take so much time to begin pushing oil in quantity and at the same time on both sides?
Here's a line of thought that I'm just throwing out there for comment --- Perhaps the recent simultaneous leaks after weeks of dry running are due in part to the recent high temperatures that may have caused a rise in pressure within the bearing forcing out oil trapped in there (miniature version of the hub pressure situation). If so, this might be addressed by removing the interior seals so the bearings can benefit from the pressure equalization of the hub venting. Perhaps this combined with the sealing technique you mentioned above to keep the outer seals free might turn the trick.
GCinDC
100 MW
i think the secret still eludes us. :lol:
makes me wonder about sealed bearings. i guess the seal is relative. just looking a bit i see there are double sealed bearings? and triple sealed?
as you can see after i pulled the bearing away, i was non-committal about the amount of sealant to use against the inside of the bearing, fwiw:
i think i used much less on the other side where it really leaked.
makes me wonder about sealed bearings. i guess the seal is relative. just looking a bit i see there are double sealed bearings? and triple sealed?
as you can see after i pulled the bearing away, i was non-committal about the amount of sealant to use against the inside of the bearing, fwiw:
i think i used much less on the other side where it really leaked.
GCinDC
100 MW
duh, that's what i had. a seal on each side. also looks like 120C is max temp, so maybe that was it!GCinDC said:
i've been looking for better bearings, but not sure what to choose from. the size is 6203 2RS. nomencature explained here.
Just to be clear - a bearing seal is first and foremost to protect the bearing itself from dust and dry contaminants - not to keep liquids out of the device in which it is installed. We are sleazing here to try to press a bearing seal to provide service in an area for which it was not intended. That said - let's see what can be doneGCinDC said:
I'm out of my depth for bearings much beyond common automotive/shop tool stuff, but just winging it - here's an NSK doc that shows some of the more interesting seal types and which has some pics of the seals. It looks like the DDU seal is going to be the most successful - a full contact rubber seal with an embedded metal stiffener as with conventional oil seals to apply maximum pressure to the sealing lips.
Here's some snippets from other NSK docs that seem to confirm that the DDU may be your best bet...
View attachment 1
And finally - a snippet from an NSK doc with a bearing suffix crossref for other manufacturers. It seems the most desirable NSK DDU type is the RS type you already have from another manufacturer...
hmmmm.... since the seal style is already correct, I would think the that improvement might only be had by going with a better quality bearing like an NSK in hope of getting a better quality seal... again, just winging it.... :wink:
silverheels
10 mW
bigmoose said:
Compare this scavenge system:
HT/HS motor with bearing mod similar to Keynes bearing mod.
Holes for 4 x 6mm OD (1/4") pipes (in addition to the phase and temp sensor wires - no halls on this motor).
Two pipes for oil input spray near the top of the motor on each side of the stator.
Two pipes for scavenge (one each side of the stator).
One scavenge pump
One input pump.
One Radiator
One de-frothing tank (approx 1 pint, with input and suction on the bottom of the tank and a breather pipe).
To Keyne's amazing cooling system (dedicated to cooling the inside face of the stator ring)
How do you think the two would compare (in terms of their ability to cool the motor)?
John in CR
100 TW
You might try looking at preventing the oil from heading toward the axles. Operation won't cause it, and neither will sitting stationary. How are the wires oriented in the motor. If they go up after entering the axle, then just rotate the axle 180° so the wires themselves don't channel falling oil directly to the exit.
That leaves oil running down the side covers, and maybe some kind of circular shield will do the trick.
Another alternative is to contact MWKeefer who has been doing oil in geared hubbies for years.
Last, if you're hitting 110-120° C the approach obviously isn't effective enough for the settings you're running. Try some different current settings, specifically a much lower phase/battery limit ratio. Block Time may be a valuable tool in getting the excess heat under control by allowing extra current during acceleration, without the sky high phase currents on hills. MWKeefer also has some interesting discussion about controller tuning in a thread about Block Time. As far as ventilation goes, the holes in a pizza pan approach is marginally effective, because it doesn't direct what little flow you get to where it's needed the most, and most of the flow just follows the side cover and out without flowing over and cooling the end windings and stator.
John
That leaves oil running down the side covers, and maybe some kind of circular shield will do the trick.
Another alternative is to contact MWKeefer who has been doing oil in geared hubbies for years.
Last, if you're hitting 110-120° C the approach obviously isn't effective enough for the settings you're running. Try some different current settings, specifically a much lower phase/battery limit ratio. Block Time may be a valuable tool in getting the excess heat under control by allowing extra current during acceleration, without the sky high phase currents on hills. MWKeefer also has some interesting discussion about controller tuning in a thread about Block Time. As far as ventilation goes, the holes in a pizza pan approach is marginally effective, because it doesn't direct what little flow you get to where it's needed the most, and most of the flow just follows the side cover and out without flowing over and cooling the end windings and stator.
John
FastDemise
100 W
I have been running mineral oil for a couple months now. My setup is in my signature and recently went to 20s at 60amps battery and 150phase. That's roughly 4.2kW peak and 2-2.2kW continuous. Running errands ~10miles with mid 80 degree weather both the controller are just warm to touch. I prefer mineral oil due to my understanding that it's perfect for electronics and it has no detergents or cleaners added that will do any long term harm to anything in my motor. I just filled my motor with a modest 50-60mL and been happy with the results ever since. I strongly strongly recommend mineral oil (preferably the laxitive kind) cuz it's easy to get a hold of and cheap.
FastDemise said:I have been running mineral oil for a couple months now. My setup is in my signature and recently went to 20s at 60amps battery and 150phase. That's roughly 4.2kW peak and 2-2.2kW continuous. Running errands ~10miles with mid 80 degree weather both the controller are just warm to touch. I prefer mineral oil due to my understanding that it's perfect for electronics and it has no detergents or cleaners added that will do any long term harm to anything in my motor. I just filled my motor with a modest 50-60mL and been happy with the results ever since. I strongly strongly recommend mineral oil (preferably the laxitive kind) cuz it's easy to get a hold of and cheap.
But how did you seal it?
I wonder if there might be a way to maintin a slight vacuum inside the hub? Something like a miniature air pump with an air/oil separator. Seems like technological overkill, but would certainly keep stuff from leaking out.
Kingfish
100 MW
That could work via the axle, and there would also need to be an oil-scrubber or drip-line feedback:fechter said:
Several industrial systems work that way to solve similar problems. When you think about it, ICEs pull mild vacuum on the crankcase via the intake manifold.
The problem with a hub motor though is reversed because it wants to pull vacuum through the axle like a centrifugal pump. Perhaps a capillary tube could pull from the hub to a small remote accumulator located on the fork above. A drain from the accumulator could connect to a docking collar encircling the axle where all things exit, thus provide an imperfect route back into the system. The accumulator is periodically recharged (depressurized) by a small non-positive displacement pump; a Piezoelectric Micro Pump would likely do the job quite well.
More coffee <slurp!> KF
Similar threads
