Hi again,
Well, you can't argue with that picture, it looks very worn. Mine are nothing like and must have at least the same sort of mileage on them. My hubs are a loose sliding fit on the splines so I think that's normal production tolerances, I don't think the splines are worn because you can still see machining marks present on mine. (or could the last time I did the job, but admittedly I can't check the bores of the hubs themselves for wear)
I agree that the wear looks to come from the spacer, but in order to wear surely there must be movement ? Following logic (dangerous, as proven above
) I'm thinking that it can only happen when the compression on the whole arrangement reduces either by the hub nut loosening, the inner bearing spacer compressing, wear in either the bearings or UJs, the sub axle stretching under tension as you say, or finally the gearbox output shaft yoke not being shimmed correctly.
Personally I'd rule out either steel compressing or stretching under tension because you'd need much higher loads than you'd get from cornering and my thoughts go to either incorrect assembly or cumulative wear in the bearings/ujs in the upper link. It always surprised me that checking the end-float on the rear wheels/transaxle output was supposed to be done at the 5k & 10k services, and I'd be willing to bet that once the cars passed to second owners or DIY servicing, that job didn't happen.
So lots of places for potential wear & subsequent movement and my guess would be that's what leads to wear by that collar, and the subsequently fatigue failures on the threads at the end of the stub axle.
I can see a failure at the threads due to sideways loading/fatigue but not to power transmission because if the splines are doing their job, 99% of the loading will go through the stub axle splines which are before the thread failure points ? (a question for debate, not a definitive statement
)
The more I think about this problem the more I come round to thinking it's not the stub axles at fault, it's the maintenance or if you like the potential failure mechanisms due to the OEM design. Perhaps your twin link suspension is a way around this issue on well used cars ? I'm so intrigued now that I think I'll post a separate thread about the topic, please chip in with pictures !
brian