Chalo
100 TW
Once you succeed at making e-bikes leaky and smelly, you can move on to figuring out how to make them emit toxic fumes and obnoxious noise.
Oil cooling a motor with transformer oil/dielectric fluid - the next frontier?
- Thread starter neptronix
- Start date
xfrankie said:
Just a note, silicone oils with viscosities below 5cSt are flammable, and a pain in the ass to ship.
Thus, 5cSt@25C dimethicone is probably the ideal silicone. Thermal conductivity is 0.0003g cal/cm·sec·°C = 0.1256W/(m*K).
Similar viscosity to Redline Likewater 4cSt@40C. Likewater is mostly hexadecene, thermal conductivity 0.1437W/(m*K) = 0.0003g cal/cm·sec·°C.
Silicone probably has better materials comparability. Likewater probably has marginally better performance and residual lubrication. Distilled water trounces both, but at the risk of decreased dielectric strength if contaminated.
No idea on it's properties, but a few years back I successfully cooled a small geared hub with chainsaw oil. :lol:
I only used it cause I had it on hand and didn't care if it killed the hub.
That stuff was super thick, but thinned out with heat and seemed to do a decent job. It didn't seem to negatively impact the wires or any part inside the hub like the nylon gears etc.
I used to monitor the heat with a thermal sticker on the shell as I didn't have a temp sensor inside, and I was never able to get it over 80C. This was a '250W' hub being run at around 800W.
Ironically I did end up killing the motor from heat after I incorrectly adjusted the drum brake so it was badly rubbing then towed my kids up a steep hill in a trailer. Pushing 800W through it was enough for me not to notice the increased drag and because the drum is attached to the hub it heated up way more than usual burning out the windings.
I've since also put ATF inside another geared hub on this same bike, which has been sitting for 5 months so far...and will continue to sit for at least another 7...will see if it kills itself, I hope not.
In terms of silicone, on several open hole hubs over the years I've sprayed a decent amount of generic 'silicone spray' lube in them periodically. Never seems to cause a problem, and many times it has helped stop dirt/scraping noises from build up of crap in there.
Cheers
I only used it cause I had it on hand and didn't care if it killed the hub.
That stuff was super thick, but thinned out with heat and seemed to do a decent job. It didn't seem to negatively impact the wires or any part inside the hub like the nylon gears etc.
I used to monitor the heat with a thermal sticker on the shell as I didn't have a temp sensor inside, and I was never able to get it over 80C. This was a '250W' hub being run at around 800W.
Ironically I did end up killing the motor from heat after I incorrectly adjusted the drum brake so it was badly rubbing then towed my kids up a steep hill in a trailer. Pushing 800W through it was enough for me not to notice the increased drag and because the drum is attached to the hub it heated up way more than usual burning out the windings.
I've since also put ATF inside another geared hub on this same bike, which has been sitting for 5 months so far...and will continue to sit for at least another 7...will see if it kills itself, I hope not.
In terms of silicone, on several open hole hubs over the years I've sprayed a decent amount of generic 'silicone spray' lube in them periodically. Never seems to cause a problem, and many times it has helped stop dirt/scraping noises from build up of crap in there.
Cheers
Chart on page 1 shows 0.1346W/(m*K) @ 40C, better than silicone but worse than hexadecene, but really close enough not to matter.neptronix said:
fatty said:
For silicone oil it depends on the viscosity:
I already posted some more infos about silicone oil in the ge10 thread, but thought i makes sense to post it here again:
Source https://ecllube.com/resources-for-engineers/
fatty said:
Current delta for idle full speed (~40 km/h, 305RPM) is about 0.2A @~41V,
comparing silicone oil 20cSt (1,1A) and 100cSt (1.3A) in a G310.
Unfortunately when measuring the 20cSt Oil I had smaller tires (schwalbe marathon supreme 622x32mm) with no/less profile vs (schwalbe g-one speed 622x35, tubeless). The weight of the tire doesn't matter for this measurement, but increased drag due to the tubeless milk and ajr resistance due to the changed tire profile could make a difference.
Based on this I would interpret the 8 Watt difference as an upper bound and assume that the increased viscosity has an even smaller effect on the resistance.
DogDipstick
100 kW
Chalo said:
OMG.
Lol I never leaked a drop chalo. You kmust dbe doing something wrong
SniperGaulois
10 kW
Cool thread guys! Very interested to know what products you are finding that have hi thermal conductivity and low-viscosity.
If you grease the bearings with silicone teflon grease, that should make them a bit more resistance to ingress from light machine mineral oils.
Opticool and battery coolant are sounding good, I'll take any synthetic oil which can handle 150'C.
$20 a litre will go very long way, considering you only need 100_200ml hub motor, that's 2 or 4 dollars.
I recommend cheap CPU Grease to seal the hub plates, perhaps Colgate too. Expensive CPU grease can transfer 5 times more thermal energy, but the oil would probably be doing that anyhow...
Water can permeate into the hall sensors in a gas phase under pressure and it can cause slight corrosion to the iron parts including the bearing cans. It will we work though. You can make it very alkaline using some additives, calcium is an obvious one, enough to offset the effects of the ambient CO2 which turn it into an effective acid.
The water would evaporate as I go up mountains come I would have to refill it. It would work pretty well although I think that oil is durable and non-evaporative choice for me.
If you grease the bearings with silicone teflon grease, that should make them a bit more resistance to ingress from light machine mineral oils.
Opticool and battery coolant are sounding good, I'll take any synthetic oil which can handle 150'C.
$20 a litre will go very long way, considering you only need 100_200ml hub motor, that's 2 or 4 dollars.
I recommend cheap CPU Grease to seal the hub plates, perhaps Colgate too. Expensive CPU grease can transfer 5 times more thermal energy, but the oil would probably be doing that anyhow...
Water can permeate into the hall sensors in a gas phase under pressure and it can cause slight corrosion to the iron parts including the bearing cans. It will we work though. You can make it very alkaline using some additives, calcium is an obvious one, enough to offset the effects of the ambient CO2 which turn it into an effective acid.
The water would evaporate as I go up mountains come I would have to refill it. It would work pretty well although I think that oil is durable and non-evaporative choice for me.
Ianhill
1 MW
- Joined
- Sep 25, 2015
- Messages
- 2,871
To me we have yet to see if the added frictional losses are exceed with the increased thermal managment.
Looking at the hub that failed at 800w i wouldn't say theres a definitive answer of any improvement from adding the fluid used in that case at least.
Looking at the hub that failed at 800w i wouldn't say theres a definitive answer of any improvement from adding the fluid used in that case at least.
DogDipstick
100 kW
fatty said:
I love sealing cases. I like the Kawa over the others cause you can split the case in a day sideway without a puller. So you can not care about it. I been in side the 700R4 through teh 4l85E. IN the 289 through the 6.0.
If the biggest hurdle to sealing your hub motor is: That whatever you put in leaks out: you are doing it wrong.
You need to learn how to seal cases in assy. Its not easy. Some try and get leaks. Yes. but... it is not rocket science.
Millions of vehicles not leaking. On the road daily. I cannot imagine anything in the hub motor that would be incompatible with Atf.
I do have quite the love for hydrodynamic and thermodynamic engineering and own quite a few thermal daggers and dataloggers. For what it is worth. I do have many data sources and can offer testing of hypothesis if you present one. However, sealing a hub from the elements present within environment is not ( should not ) be the thing that makes the decision. I hate it when people say it will leak out. Have some skill and demand in the field,.. for what it is worth.
Yes if you have a hole in the oil pan, oil will leak out. Yes, if you've a hole in the gas tank, gas will leak out. Yes, if you dont know how to properly seal a hub motor, what you put into it, may leak out.
I ont know how competent you guys are when it comes to thermal dynamics and the engineering surrounding. But its is all we do. Speculation is all I really see here. IDK. Speculation is nothing compared to one good test. Get those numbers.
I mean.. CPU grease to seal things? Colgate? lol no. There are things in the world of heat transfer made to do these things that I know of.
One love. Test your theory and give us enumeration. I love doing that stuff.
I mean, Mineral oil ws the #1 thing my engineer told me to use.
#2 was WD40. .... Any.. of yall... try WD40 in volume ( like a quart in a DDhub)? Have data? I could do that very easy, with all dataloggers and imagers running. I think that would be worthwhile to test for fun and concrete infos.... , a mineral oil vs a WD40. Calculate the viscosity, fluid friction losses, calculate the heat transfer, BTU, lost, made, ect... Datalog it.. compare...
WD40 vs mineral. Plus the control.
Hell we can even measure the drag easy.
Chalo
100 TW
The seals on cartridge bearings are for dust, not oil. That's why they're greased instead of oil lubricated. Please explain how you seal those against oil leakage while allowing them to turn.
DogDipstick said:
Here you can find some pictures of an ATF flooded Motor after 2 years
https://www.epowerforum.de/forum/threads/oelkuehlung-vor-und-nachteile.49/post-6125
Translation, what was damaged:
"The paint on the stator and also the insulation of the cables, as well as the cord with which the windings are held together."
I dont say all ATF oils will damage parts in all Motors, but some combinations don't work.
Although i would welcome you doing more tests, were are only really speculating about long term compatibility. There are plenty of test regarding the cooling function.DogDipstick said:
https://endless-sphere.com/forums/viewtopic.php?f=2&t=48753
https://endless-sphere.com/forums/viewtopic.php?f=2&t=92124&start=50#p1425130
Regarding drag, my measurement represents an order of magnitude (inline with the measurements from Justin in the g310 thread).
Again more testing is appreciated, but its not like we are only speculating.
DogDipstick said:
WD 40 is thin liquid but no lubricant! So its not a good choice for geared motors ir any kind of bearings. For other motor types ferro fluid makes more sense anyway.
WD40 can also attack plastics, rubber and plasticizers.
In my G311 electrical parts including the phase and hall wires, windings, boards. string etc have ZERO exposure to ATF used to lubricate and cool. It cost less than 20ml of ER2225 high temperature and chemically compatible epoxy and was stupidly easy to do. And the entire materials compatibility and largely ignored lubrication discussion is made redundant. For mine at least.
It'd be even easier to isolate materials inside a DD motor with say, an $8 rattle pack of EL601. Then you can fill it with whatever makes you feel warm and fuzzy inside.
It'd be even easier to isolate materials inside a DD motor with say, an $8 rattle pack of EL601. Then you can fill it with whatever makes you feel warm and fuzzy inside.
DogDipstick
100 kW
Chalo said:
I'll get the part number off one of the ( three ) hub motors I have here and I;ll look it up to find its intended application and design.
I do contest what you say. My qs motor(s) had a labyrinth seal(s) like a SKF just like the tailshaft of ta transmission or the crank pulley seal in a block. Garter spring, labyrinth, and two sealing surfaces. Just like most cars. THey are holding quite a large ( and draggy) volume of ATF for some time now. Probally 2/3 of a quart by now. I do nto know why you would think they are designed for dust only. They are pressure seals. At least in this manufacrurers. I am pulling a 205 apart soon to drill, fill and seal, for exposure to 400+ amps. Revving up 350A on the 273 made me want to do it. I am going to ride both motors on 400+A soon. Compare them myself.
Thankyou for your consensus thought on the damage and the record. I do not ( in my ways, ) think it is detrimental based on my schooling an knowledge. I will continue to do it. One good test is worth.... a thousan.. Oh nevermind.
I am absolutely willing to try the greasy WD in a hub from the start.. to see what happens.. if no one has done it yet.
DogDipstick said:
Bafang motors also have spring single lip seals protecting the bearings. It'd be easy to swap them for double lip. Leaks haven't been a problem so far.
Regarding WD40... I put it in electric and vacuum motorcycle chain oiler systems many years ago. It was an experiment to successfully prove a point. The oilers leaked like sieves past all of the seals.
It is naptha based and I'd be wary of aggressive solvents like xylene in an unprotected hub motor.
All the obvious ones have been mentioned in this thread, except 3M Fluorinert and Novec.zzoing said:
Knowing that silicone oil has better compatibility, you'd want to do the opposite: use hydrocarbon grease in the bearings, and silicone oil as coolant.zzoing said:
$100 + $61 shipping (for me, at least) for 1 gallonzzoing said:
Thermal grease is often silicone-based, which would be dissolved by silicone oil as the compatibility chart above. And since the hub plates aren't intended to seal liquids, they likely won't be machined well enough to seal without a curing bulk gap sealant.zzoing said:
And yeah, since the coolant medium is in contact with both parts, they would already be at thermal equilibrium.
I don't think you'd be able to get a watercooled motor past 100C. You'd have to be so deep into saturation that you're dumping amps purely into heat.zzoing said:
Interesting idea of correcting pH with calcium (carbonate?). But pH only settles to a mild 5.8, and adding calcium would result in conductive calcium ions -- much worse than slightly acidic pH.
How about a **thermally conducting** (& obviously dielectric insulating) "potting" sealant that starts out in liquid form?
Use that to coat all the (possibly reactive) components, seal up any (possibly leak prone) gaps
let it thoroughly dry & cure.
Then (if even needed) your permanently-liquid heat transfer fluid only needs to be non-reactive with the potting compound and those few materials exposed at the interface of moving parts where the latter can't adhere a thick enough layer.
Use that to coat all the (possibly reactive) components, seal up any (possibly leak prone) gaps
let it thoroughly dry & cure.
Then (if even needed) your permanently-liquid heat transfer fluid only needs to be non-reactive with the potting compound and those few materials exposed at the interface of moving parts where the latter can't adhere a thick enough layer.
SniperGaulois
10 kW
ampcool is a bit too viscous? https://www.engineeredfluids.com/ampcool
Usually hub disks which hold the cartridge bearings are reinforced so that the aluminium is higher near to the cartridge forming a dam when the hub is on it's side.Chalo said:
SniperGaulois
10 kW
Here's a list of light machine oils which can easily be found on ebay/amazon
Shell tellus s2 vx 15 flashpoint 200 cst 15 to 3.8... ISO VG 15 it's hydraulic oil
CASTROL Hyspin Spindle Oil ZZ 5/ E5/ E10, flash point 115 / 150, ISO VG 5/10
mobil Velocite No 6, flash point 180 C ISO VG 10 specifications here: https://www.mobil.com/en/lubricants/for-businesses/industrial/lubricants/products/products/mobil-velocite-oil-no-10
Total Azolla Zs10, 165 flash point, ISO 10 hydraulic
Lubrilam C10 by Total , flash point 134, cst 2.7 at 40'C , laminating oil
Shell Morlina S2 BL 5, flash point 120, iso vg 5, BL 10, ISO vg 10
No promises about the cancerigenic properties of the above products!
Shell tellus s2 vx 15 flashpoint 200 cst 15 to 3.8... ISO VG 15 it's hydraulic oil
CASTROL Hyspin Spindle Oil ZZ 5/ E5/ E10, flash point 115 / 150, ISO VG 5/10
mobil Velocite No 6, flash point 180 C ISO VG 10 specifications here: https://www.mobil.com/en/lubricants/for-businesses/industrial/lubricants/products/products/mobil-velocite-oil-no-10
Total Azolla Zs10, 165 flash point, ISO 10 hydraulic
Lubrilam C10 by Total , flash point 134, cst 2.7 at 40'C , laminating oil
Shell Morlina S2 BL 5, flash point 120, iso vg 5, BL 10, ISO vg 10
No promises about the cancerigenic properties of the above products!
Ianhill
1 MW
- Joined
- Sep 25, 2015
- Messages
- 2,871
Cooling happens best with a larger temp differential put your warm hand iced water if you don't believe me see how long you last.
Same for heating something large temp swings boil a kettle fast for example.
But here theres just a mass if liquid trying to be a heat well as there's no source to dump the heat other than the case, the temp difference is very little so it doesnt suck heat from the core thats why its not in production chaps.
Engineers have actually studied thermodynamics and have a good solution in place for cooling its not this thats for sure you need a flow of cold water into the hub.
Imagine your car when the water pump breaks your left with a system that's simular to your own but theres a rad with fins for convection cooling to occur so an engine with a failed water pump still outperforms your type of closed setup and even that will melt.
Drag cars fill the water galleries and run it as a thermal block of material heatsoak the material rather than cool it becuase it saves weight but they have a 10 second limit as a result ??? Sound familiar
electric motors mmmmmmmmm
Same for heating something large temp swings boil a kettle fast for example.
But here theres just a mass if liquid trying to be a heat well as there's no source to dump the heat other than the case, the temp difference is very little so it doesnt suck heat from the core thats why its not in production chaps.
Engineers have actually studied thermodynamics and have a good solution in place for cooling its not this thats for sure you need a flow of cold water into the hub.
Imagine your car when the water pump breaks your left with a system that's simular to your own but theres a rad with fins for convection cooling to occur so an engine with a failed water pump still outperforms your type of closed setup and even that will melt.
Drag cars fill the water galleries and run it as a thermal block of material heatsoak the material rather than cool it becuase it saves weight but they have a 10 second limit as a result ??? Sound familiar
electric motors mmmmmmmmm