For those who have actually put real mileage on your cars...

[TurboFource]

Yes Gavin, a set of opposed taper bearings. The .049 difference will move the housing out that amount but I will lengthen the lower control arm that amount, I can adjust the thickness of  the hub/drum spacer my son made to account for the outside difference. The housing depths could be machined to account for the difference but I don’t want to do that.

Kendo,
The plastic prints at 572 degrees so I will give it a try, if it does not handle it I will get aluminum ones made (I really need a lathe……)
Typical transmission and differentials only run in the 150-200 degree range and most of the heat comes from the helical bearing
surfaces "sliding" over each other as they mesh...so the hub bearings should run cooler than this...

[TurboFource]

I wasn’t going to post anything until I had it all complete but thought I would mention it since it came up ….
 If you search for “S2hubconversion” this is where I got the inspiration to try this…he says he has run the same set of bearings for 25 years…..

JB..thanks for posting the link.....above

[jbcollier]

https://www.lotuseuropa.org/S2HubConversion.htm

[SwiftDB4]

Has anyone tried a double row inner ball bearing ? 3006 2RS is 55 x 30 x 19 mm. I bought a pair, but never machined the axle shoulder for the extra width. Being that it's a double row bearing and 19mm wide vs 13mm on the standard bearing it should be appreciably stronger.

[BDA]

I was thinking about just that sort of solution. With the radius arm attachment to the upright, we're restricted in how large in diameter the bearing can be but there is room for longer bearings. I'll bet you could call up a bearing mfg.'s tech support line and get some guidance. To my knowledge, VWs and Porsches don't have a longevity problem with rear wheel bearings. Since they have about the same ID, I wonder how much bigger their OD is. MAYBE our upright could be adapted to use them?

I'm sure I'm not the first to think of any of this so I will commit to calling Richard (I'll try for tomorrow) and ask him what he can tell me. I'll report back

[Pfreen]

I just calculated the theoretical life (via online calculator) for one tcs rear bearing. 
I roughed it out.  1000 lb load and 100 mph.  The 90% reliability calculated as 1129 hours.

I think the average load would be significantly less.  The peak load during cornering may be close to that.  However, the axial load I think would be taken up by the two bearings.

Anyway, I don't think we will average 100mph over 1129 hours.

So, it seems to me, ideally the bearing size is fine.  Lubrication and tolerancing may be the cause of premature failure.  I know if the spacer was too short or the housing was machined incorrectly, the bearings would definitely fail prematurely. 

This discussion has been great and I hope the issue gets solved.

[TurboFource]

Tapered bearings can be set up to run with preload or clearance so I think they will be a lot more tolerant of assembly issues...

The front bearings on my '95 F150 have 232K miles on them with very minimal maintenance....half that should be doable on a car that weighs a fraction of what it does.....

[jbcollier]

Tapered bearings lubed with grease need a slight amount of free play, just a couple of thou.  Preloaded tapered bearings make heat and need at least splash oil lubrication to keep them lubed and cool.  Grease will just flow away.

[BDA]

I've forgotten most of what I ever knew about bearings from school but I suspect that a significant (but no idea how significant) factor might be shock loads when you hit a bump. I wonder if that is accounted for in their calculation.

Certainly there appears to be wide variations with the life of rear wheel bearings. Other than expected wear, I'm at a loss to explain how mine went out so quickly mileage wise. With a twin link, there is almost no lateral loading. I had sealed bearings and IIRC, I used good brand, I think SKF. I had a hardened spacer (though I believe hardened spacers are less important with a twin link). At most I do spirited street driving - no autocross or tracking. Maybe something will show itself when I finally get it all apart but otherwise, it seems like something I should expect going forward.

[TurboFource]

I didn't mean I would preload them in a car, (not intentionally anyway ) when I rebuilt industrial equipment, preload was set by running torque, of course these bearings had spray lubrication...just saying they will handle a wider range of assembly procedures and loads than a standard roller bearing.

[GavinT]

Thanks Turbo & JB,

I've seen that link before yonks ago but obviously didn't read it thoroughly - ooops.
I've usually been less enthusiastic re the 2 x tapered bearing idea because for a S2, that would mean adding metal to the upright and the associated re-machining - all that hassle.

What I didn't see was there's an option for 55mm inner bearing cups . . hmmm . . maybe that's worth revisiting.

[GavinT]

[...] To my knowledge, VWs and Porsches don't have a longevity problem with rear wheel bearings. Since they have about the same ID, I wonder how much bigger their OD is. MAYBE our upright could be adapted to use them?

I'm sure I'm not the first to think of any of this so I will commit to calling Richard (I'll try for tomorrow) and ask him what he can tell me. I'll report back

Yeah.
I've looked at the Datsun 1600/510 rear axles mainly because they have the wheel hub integral with the axle and the retaining nut on the inside end.

The Datsun uses 2 x 6206 bearings just like the TCS but with a more substantial spacer and in a steel housing (semi trailing arm). The spacer is within a poofteenth of the Lotus one in terms of length, so the bearing spacings are nearly identical. The Datsun, however, uses sliding half-shafts so the load paths are different. The torque spec for the axle retaining nut is 181 to 239 ft-lb. The Datsun manual recommends to replace the wheel bearing grease every 50,000km (30,000miles), so that tells us something.

Would I be right in thinking the TCS with it's larger inner bearing fairs better in terms of bearing life than a S2?

My impression (and that's all it is) is that shortish bearing life is perhaps associated with the relatively light (in terms of sturdiness) alloy upright as much as anything else. That, and the axle nut torque on a cheddar cheese axle being somewhat modest, which, even then, is probably more than the stock spacers can handle. Not to even mention the overhanging hub . . . so likely a combination of many things.

And, of course, we're latterly led to believe by the factory that Loctite on splines is the answer to a maiden's prayer. Seriously?
See . . . there's a question. How many similar axle arrangements (Datsun, VW, BMW etc.) specify Loctite for assembly?

I've sometimes toyed with the idea that pressing a 5mm wall thickness steel sleeve through the whole upright and building up alloy where necessary might just solve a lot of issues. Too drastic?

Pardon my rant . . thoughts?

[jbcollier]

Yes, the TCS is said to have a longer bearing life.

There are a number of factors to Europa rear axle issues:

- poor material choice for the axle
- mild steel spacers
- bearings with 2 rubber seals (less contact area for the spacers)
- outer bearing seating partially on the splines
- undersized inner bearing
- mild steel lock tabs
- poor repair practice (driving out the bearings crooked wearing and ovalizing the housing

Any one, two or three might be ok but all of them together is not good.

The housing itself is ok; a little light duty, true.  It’s poor repair techniques that take out the housings.

[TurboFource]

Swift,
That bearing looks nice, is there enough material at shaft shoulder or in the housing to account for a 6mm difference?

[SwiftDB4]

Yes there's enough space in the housing to fit that bearing. As far as removing 6mm of the shoulder end of the stub axle there is also enough material on the 'step'. I'm not an engineer, so I don't know if that might weaken the shaft. However looking at a TC axle compared to S1/2 there is less 'step' material on a TC.