Coupling of EV Motor to Reduction Gear Pinion

[KiwiME]

Hello,

I'm noting several different approaches to the above detail used on production EVs. The context of my question is that a handful of near-new e-Niro and Kona Electric examples have developed clicking noises reportedly at approx motor shaft frequency. The manufacturers don't seem to have a field solution to this yet, as best as I know, despite replacing both motors and gearboxes. I'm wondering if there is a fundamental flaw in their design related to coupling two fully-constrained rotating shafts which may have small parallel offset errors due to production and mounting tolerances. I'm also wondering how splines outside of oil-wet areas would be lubricated, whether grease is sufficient.

Here's what I have found on several EV designs.

GM EV1: The rotor (induction motor) and pinion are mounted on a single shaft with a bearing at each end. Between them is only an oil seal.

GM Bolt: The pinion is held by a large nut onto the end of the (PM) motor shaft. There is no outrigger bearing.

Nissan Leaf: The motor (PM) has a splined shaft output. The reduction pinion (and parking pawl wheel) is supported at both ends by bearings. The detail of the motor's male and pinion's female splines is not clear.

Hyundai Kona, Kia e-Niro: Not much info is available but the (PM) motor has a splined output shaft like the Leaf. The design detail of that spline and design of the reduction gearbox internals are all unknown. These two vehicles reportedly use the same motor and reduction gearbox but the overall gear ratio is slightly different.


Any insight is appreciated.

thanks,
Paul

 

[JayMaechtlen]

Fully constrained shafts - seems like any misalignment would cause problems. For sufficiently close alignment I'd expect no sliding within the coupling except for axial motion due to thermal changes.
For very slight misalignment you might expect a bit of - rocking? along the contact lines of the splines.
For more than slight misalignment, I'd expect issues. Depending on how the two units are mated, there might be a chance of field alignment (or misalignment) when installing.
Seems like this calls for really close inspection of a noisy unit, especially the splines.
And really close inspection /measurement of the mounting arrangements and dimensional variations.

Jay Maechtlen

http://www.laserpubs.com/techcomm

 

[TugboatEng]

We operate a fleet of hydraulic pumps with SAE744a mountings, splined couplings, and a very poor design with regards to lubricant retention. We have many problems with fretting wear of the coupling. I will also confess that if one forgets the grease a new coupling during installation it fails catastrophically at 1800 hours. Lubrication is very important to spline life as small amounts of movement will cause then to fret into oblivion without it.

 

[KiwiME]

Thanks for the comments so far. The issues with the hydraulic pump are exactly what I would expect unless tolerances and fit were extremely good.
I found a video showing the relevant details inside the Tesla Model S but the design is not exactly clear. On first glance it appears to have three bearings but it's not certain that the pinion is part of the motor shaft or splined to it.
Found a pic of the Model 3 where the pinion is clearly integrated. It may the same design as the S.

 

[KiwiME]

Just to warm this thread up, here are images of the parts in question, the gear reducer and the motor output shaft. Noting that the first stage pinion shaft has bearings at both ends, there must be an over-constrained condition when the splined coupling is under torque. The last reported incident was of catastrophic gearbox failure after a few days of a ticking noise at about motor shaft frequency. The motor spline was apparently in serviceable condition and only the reducer was replaced. I'm guessing that the gearbox bearing closest to the motor was overloaded and perhaps the cage broke apart.

 

[GregLocock]

The system is only overconstrained axially if the plate bending stiffness of the casing is ignored.

Cheers

Greg Locock


New here? Try reading these, they might help FAQ731-376

http://eng-tips.com/market.cfm?

 

[KiwiME]

Any thoughts on the radial situation, Greg?

 

[GregLocock]

Looks pretty standard to me. The two halves of the housing are probably dowelled if that is an issue, or of course they may hvave used a self aligning bearing.

Cheers

Greg Locock


New here? Try reading these, they might help FAQ731-376

http://eng-tips.com/market.cfm?

 

[gruntguru]

Further to that, the motor shaft and the gearbox input shaft should be concentric - the motor housing is located in a concentric recess in the gearbox.

je suis charlie

 

[KiwiME]

Yes, the photo of the motor in my post just above shows the two hollow dowels. The flange around the motor shaft appears to have an O-ring and may or (more likely) may not provide any further concentric locating between the motor and reducer shafts.

The Tesla Model 3 and Model Y teardown videos by Sandy Monroe indicate that the motor/pinion arrangement is similar with the exception that the gearbox bearing nearest the motor is not present.
Adjacent to the spline coupling in all cases are locating diameters to sensibly center the splined connection when torque is relaxed. In the Teslas, this locating diameter alone appears to support that side of the pinion, effectively utilising the motor's nose bearing. Another interesting feature of the Tesla design is that about the first 1/3 of the motor housing is provided by the reducer housing. So the machining of the motor nose bearing is on the reducer casting. No doubt that's one advantage of manufacturing both units in-house.

My argument is that the Korean design is over-constrained and therefore unnecessarily sensitive to the tolerance stackup between the two shafts. Numerous owners of these EVs are having either or both the reducer and/or motor replaced under warranty due to a ticking noise and one case recently of catastrophic gearbox failure. In a few cases the problem has returned.

 

[gruntguru]

With reference to the drawing and "This bearing missing in Tesla 3, Y"). It would be unusual to use the spline itself to locate the pinion shaft. You would normally see a spigot bearing, spherical bush or similar to locate the pinion shaft to the motor shaft.

je suis charlie

 

[KiwiME]

Yes, certainly, that's what I mean by "spline locating dia" on my sketch and the text "in the Teslas, this locating diameter alone appears to support that side of the pinion". Here's a photo of the Tesla motor without the pinion fitted, noting the oil grooves.

 

[mfgenggear]

Just my observations.
a) misalignment destroying the spline , should be crown spline at the external spine. if free float major diameter fit.
b) not enough lubrication, causing scuffing & fretting, needs oil holes,
and one of the members needs to be coated, eg dry lube, phosphorus, or silver for break in.
c) adding a pilot diameter will also correct misalignment. (this would be a better design)

however if there is not enough lubrication grease will not be enough
it will be only temporary.. it will be pushed out then metal to metal contact.
this has been a hot topic on gear & pulley forum, (Gearcutter)& others
Please search forum on spline being lubed only with grease , high failure rate.