105 rear wheel bearings

[Divano Veloce]

Theres a bit of play in my rear wheel bearings so its time to replace them.

I would like to do the double row 4207-2RS mod as designed by Colin, Paul and Bruno however the 6mm increase in track might be a bit of a problem with the berlina rear arches....

After reading many forum threads and opinions i have a couple of questions regarding the 105 lsd axle bearing arrangement....

is the rapid wear and play at the outer bearing due to shaft flex or is it due to the inner end of the shaft not being particularly well located radially at the diff center?

is this problem worse on LSD axles?

how much better does the double row bearing hold the rear wheels straight?

how much longer does the double row bearing last?

is it possible to use the double row without adding track (ie, machining the axle tubes instead of the brake backing plate?

Would running the LSD with preload on the clutch packs assist to locate the inner end of the half shafts?

thanks!

[LaStregaNera]

You can skim 3mm off the outside face of the brake rotors to remove the additional track.

[Divano Veloce]

Or properly lip the rear arches...

[Divano Veloce]

The double row bearings too have durability issues albeit under extreme conditions. The issue seems to be the poor radial location of the shafts at the diff center allow misalignment of the shaft/inner race of the wheel bearing. If so then a double row wheel bearing isn't going to do much (other than last a bit longer). What are the experiences of those who have club sprinted with this mod?

[Paul Byrne]

G'day Matt,
The double row mod certainly reduces the play and they last a long time, and the extra 3mm of track per side is probably not going to be a problem in a Berlina, but as you say they don't completely fix the issue. I don't think there is any way to deal with it at the diff end as the clearances in the splines/gears allow substantially more movement than do the outer ball bearings. Also it is certainly not axle flex which causes the play!

I set up a jig and measured the play with a single row bearing, double row, double row angular contact and four point contact single row. All have their benefits and problems, eg width etc. Results attached.

Pad knock back is still an issue with the double row, and in Targa conditions it is really desirable to have a very reliable brake pedal position, so Colin and I investigated adding a second bearing inboard of the single or double row outboard bearing. The photo shows the adaptor/bearing holder welded to the cut off end of the axle tube.After fitting the adapter, the end is welded back on to the axle tube and the new inboard bearing pressed lightly in. Once the mod is completed axle removal/replacement, oil seal etc is completely standard.

The bearing is a needle roller and the axle shaft is ground very slightly to suit it. The shaft is quite hard enough to run inside the rollers, and of course lubrication is not an issue.

The mod is highly successful and results in no play and hence no knockback. No perceptible wear after a 5 day targa tasmainia and several circuit events.

It's not a trivial mod but it certainly does the job.
ps there is no advantage in running the double row bearings with this mod in place.

[Divano Veloce]

Thanks Paul, you have all the answers!

So the simplest way to improve the bearing life and pedal feel for a roadgoing/sprinted 105 seems to be the double row bearing mod.

Is there enough meat in the axle tube/backing plate to machine room for the double row angular contact?

Is the challenge with the 4 point contact single row clamping the inner races together?

I did a similar test where I measured pedal travel, then with the rear end raised i rocked each wheel in line with the caliper (so to knock the pads back) and measured pedal travel again. There was about 25mm of difference. I reckon they'd get knocked back a bit more when driving.

The Berlina brakes are excellent and the car tracks ok but the pedal is terrible, not something i recall being an issue with my open diff GTV.

[LaStregaNera]

There was a known bad batch of bearings getting around for a while - mine flogged out in maybe 10-15,000km.

[Colin Byrne]

So the simplest way to improve the bearing life and pedal feel for a roadgoing/sprinted 105 seems to be the double row bearing mod.

yep, for sure, they will last along time and are a great improment with that sort of work out, but i managed to degrade a set within 2 days of targa tas to the point were high speed stability was effected

Is there enough meat in the axle tube/backing plate to machine room for the double row angular contact?

we already machine the backing plate 3mm deeper to accept the bearing, you'll be hard pressed to find another 3mm anywhere else.  one option (and something i did originally for clearance) is to cut off the fitting at the end of the axle tube, shorten the tube by 3mm and weld it back on, fairly big job and if you were to do that you might as well add the ultimate fix with the needle roller bearing

Is the challenge with the 4 point contact single row clamping the inner races together?

sorry don't follow this question??

The Berlina brakes are excellent and the car tracks ok but the pedal is terrible, not something i recall being an issue with my open diff GTV.

We got hung up for a while assuming all the knock back was coming from the rear end, and a lot of it was, but after we did the full needle roller mod and the brake pedal still wasn't perfect we started looking at the front and sure enough the front stub axle was flexing enough to cause knock back, but that's another story....

[Divano Veloce]

Thanks Colin, i just need to source a boring bar small enough to machine up the front wheel bearing fix.

[Divano Veloce]

I have double row bearings and machined backing plates in hand.... from my measurements this mod moves the backing plate outboard by 3mm and the axle and hence wheels outboard by 2mm.... Am I right? I assume that this is not a problem for hand brake or disc brake alignment....

The issue I have is the concentricity of the machining of one of the backing plates.. The bearing will go in the axle tube or in the backing plate but wont go into the axle tube and the full depth of the backing plate.... I should have done a trial assembly before I fitted the bearings onto the axles...

[Paul Byrne]

G'day Matt,
The offsets are not a problem at all.
Re concentricity; since the back plate just holds the axle/bearing in place it does not matter if its bore is a little larger than standard, or a bit eccentric. I would just dremel/file it until it fits -that won't affect anything important.
Cheers
PB

[Divano Veloce]

Thanks Paul,

This might be a silly question but is there any reason why not to try this axle on the other side???

[Paul Byrne]

The path less travelled! Go for it!

[Divano Veloce]

Tried axle on the other side, same problem...

So i did a trial assembly of the bearing and backing plate on the passenger side axle tube, all fitted together nicely. So I transferred the handbrake mechanism (you can do this by just removing the retaining pins and moving the shoes as an assembly) to the new backing plate and installed the bearing and retainer on the axle. Now handbrake actuator fouls on the wheel stud!

So i pulled the springs and shoes etc apart and measured the offset from the backing plate to the wheel flange on both axles... it was the same. I then assembled the drivers side handbrake parts onto the passenger axle and viola! no interference! Turns out there's more than one style of handbrake actuator mechanism, and more than one length of shoe retaining pin.... Given that the difference in offset with the new bearing is only 1mm I'm not sure how the handbrake parts would have cleared the studs with the original bearings...

So i have one side ready to go back in, the other side requires the bearing/retainer to be removed, backing plate fettled and a replacement handbrake actuator lever assembly and maybe some shoe retaining pins.....

Theres still a chance for Broadford!

[Divano Veloce]

Machining non concentricity. I hope its at least square