Differential Carrier Bearing Race corrosion - Replace?

[frankc]

I decided to replace all the seals in my 915 while it was out, and I'm glad I did as I discovered that the right axle flange seal had failed and allowed moisture to enter. I removed the side cover and the differential and noticed some tiny spots of rust on the Carrier Bearing Race. The bearing itself appears to be ok.

I used some Simichrome polish on the race and it is now perfectly smooth when I run my finger over it, but will this damage the carrier bearing if left as is? Should I simply replace the bearing and race as they are inexpensive from our host ($25). And can I replace just the right bearing, or should they be replaced as a pair?

Taking a quick glance at Peter Zimmerman's 915 Wiki, it appears there is more to replacing the Carrier Bearings than simply pressing them in, so if I could avoid this (or at least postpone it) that would be preferred as I've already taken way too many tangents on this project.

Before:



Close-up of after polishing:

[HawgRyder]

Definate replace...and make sure the new race sits tightly in the housing.
Bob

[Steve@Rennsport]

JMHO, but I replace bearing races when I see corrosion like that.

[Jcslocum]

That is a failure just waiting to happen. It's got to go.

[Nine9six]

Its your car...If you want to half-step something that needs addressing, that's entirely your perrogative; but the question has been asked and answered. The only question remaining is, would YOU replace it?
Rust/corrosion is metal cancer. That bearing has cancer.

[frankc]

Thanks for the responses. I guess this just verifies what I already suspected, but I just needed to hear it. Oh well, what's another couple of weeks tacked onto what has become a 5 year project .

So I have a few of follow-up questions:

1) Since I am replacing the right bearing, should (must) I replace the left (side cover) bearing as well?

1a) If the answer to (1) is Yes, then should I take this opportunity to move to the Aluminum side cover (915-301-311-06)? I have a Magnesium 915/61.

2) Do you know of a source for tool P 2646 to press on the Carrier bearing?

3) Looking at Part 5 of the 915 Tutorial, it appears I will need to set the pre-load when I replace the Carrier bearings, correct? Where do I get the special washer to set the preload? Do I also need to set the backlash between the pinion head and ring gear as part of this procedure?

The more I read the more involved this looks. Perhaps this is a job I pay somebody to do who has the right tools.

Here is the wiki I was reading:
How-To: Porsche 915 Transmission Repair Tutorial Part 5 - Porsche Wiki

[Steve@Rennsport]

Frank,

If the left side bearing & race are both in excellent condition, you don't necessarily need to replace them. The result of a close & detailed inspection will tell you the answer,..

IMHO, you should have this done by a pro who is skilled and equipped to rebuild 915's properly. You don't want to risk the R&P with incorrect backlash.

The stock Mg sidecover handles that 2.7RS engine without problems. The only folks who need the aluminum ones are ones that have a 3.4 litre or larger engine.

[frankc]

Steve,

Thanks for your responses. Now I'll have to find a local mechanic that I can trust to work on a 915.

Regarding the Al side cover, my next engine project is to go twin-plug and 10.5:1 CR. Is the HP generated by this setup (maybe 230-240?) still ok with Mg?

Also, I'm curious - what would be the progression to failure be due to a race with corrosion such as this? Does the bearing eventually fail (e.g.; rollers crack) due to the uneven race surface? Does the race fracture at the corrosion point?

[Nine9six]

Originally Posted by frankc

Steve,

Thanks for your responses. Now I'll have to find a local mechanic that I can trust to work on a 915.

Regarding the Al side cover, my next engine project is to go twin-plug and 10.5:1 CR. Is the HP generated by this setup (maybe 230-240?) still ok with Mg?

Also, I'm curious - what would be the progression to failure be due to a race with corrosion such as this? Does the bearing eventually fail (e.g.; rollers crack) due to the uneven race surface? Does the race fracture at the corrosion point?

frankc,
Potential metal fatigue or stress fractures as the corrosion progresses. Do I see the beginnings of a stress fracture in the before photo in the ID bore at approx. 6:30 position, or is that lint?

[Steve@Rennsport]

Generally speaking, bearing races are surface-hardened to protect them from roller wear and when that surface suffers from corrosion, tiny pits form which breaches the hardness layer.

This causes more pitting and brinnelling of the surface. Pretty soon, the rollers catch & drag on the surface imperfections which triggers bearing failure.

The consequences usually go beyond the race & bearing as metal gets distributed throughout the transmission in this example.

[frankc]

Originally Posted by Nine9six

Do I see the beginnings of a stress fracture in the before photo in the ID bore at approx. 6:30 position, or is that lint?

If you are referring to the area circled in the photo below, that is simply a scratch in the magnesium where the axle flange seal mounts (not the race). The scratch is from when I was using a dental pick to clean out the channel between the race and the seal area (there was a lot of hardened crud in there).

[Unobtainium]

"The more I read the more involved this looks. Perhaps this is a job I pay somebody to do who has the right tools"

Call Pat Williams Racing in Memphis. He will guide you on cost effective shipping, do the job right the first time, at a fair price, and get it back to you on time. Or it's a 650 mile drive from Austin.

[frankc]

Originally Posted by Steve@Rennsport

The consequences usually go beyond the race & bearing as metal gets distributed throughout the transmission in this example.

Yikes! Thanks for the thorough explanation, Steve.

I did some more reading today and just realized that to set the preload the pinion needs to be out, which means opening up the gearbox . This is becoming one expensive seal.

Speaking of the seal, take a look at the picture of the failed seal below. Either the seal was originally defective, or something wore the rubber coating off the outside of the seal, allowing moisture to enter (the severe damage at the 5 O'clock position is from when I removed it). Oddly, the gearbox did not leak gear lube. I am trying to figure when the gearbox would have been exposed to a sufficient amount of water to cause this. The car has only been driven in the rain twice in the past 27 years (the last time more than 15 years ago). But the corrosion looks relatively fresh, and in the pattern of the rollers so static. So either this damage has been there a very long (unlikely), or it happened while sitting in my garage for the past 5 years, even though I live in a somewhat dry area.

I spoke with the owner of one of the local shops today. He was very nice and spent 10-15 minutes on the phone with me explaining everything that needed to be done to do it right, and assured me that he has done this hundreds, if not thousands, of times on 915s and 901s. He stated that if nothing else needs to be done and everything goes right, it could be as little as 5 hours of labor. Worse case scenario: 10-15hrs (I could probably buy a used gearbox for less at that point).

Since the synchros were replaced only about 15K miles ago, I'm hoping there will be no surprises when he opens it up.

[Matt Monson]

Originally Posted by frankc

Worse case scenario: 10-15hrs (I could probably buy a used gearbox for less at that point).

Better the devil you know than the devil you don't...

[bazar01]

Originally Posted by frankc

Yikes! Thanks for the thorough explanation, Steve.

I did some more reading today and just realized that to set the preload the pinion needs to be out, which means opening up the gearbox . This is becoming one expensive seal.

Speaking of the seal, take a look at the picture of the failed seal below. Either the seal was originally defective, or something wore the rubber coating off the outside of the seal, allowing moisture to enter (the severe damage at the 5 O'clock position is from when I removed it). Oddly, the gearbox did not leak gear lube. I am trying to figure when the gearbox would have been exposed to a sufficient amount of water to cause this. The car has only been driven in the rain twice in the past 27 years (the last time more than 15 years ago). But the corrosion looks relatively fresh, and in the pattern of the rollers so static. So either this damage has been there a very long (unlikely), or it happened while sitting in my garage for the past 5 years, even though I live in a somewhat dry area.
.

Steve already suggested to replace just the bad bearing.

Why don't you try replacing just the bad bearing and oil seal?
If you buy the same brand tapered roller bearing, it should have the same factory dimensional specs.
There is a shim behind the outer race and the flange bore that was established by the rebuilder for the correct bearing pre-load. If you are careful enough and have an expanding collet type puller to pull the race out, you should be able to preserve and reuse the shims.
Since you are not replacing the left side carrier bearing, the diff should maintain the correct backlash from that side with the pinion.
With a new right side bearing and the same shims, just hope that there is no axial play on the diff. If there is axial play, the existing shim is not enough to preload the bearing and you may need special tools for measuring the shim thickness and output shaft rotational torque value.
That's when you need to ship it to a pro

[frankc]

Bazar,

Yes, it was suggested to replace only the bad right bearing if the other (left) is fine, but no where have I read (until your post) that one can replace only the right bearing and get away with not setting the preload & backlash.

What you say makes sense, and is what I was hoping to hear when I initially posted. I thought that using the same brand bearing and number of shims would ensure the ring was in the same axial position as before, so there would be no issue. But from what I have read, the bearings have enough variance that this is not likely. The mechanic I spoke with stated that even the pinion depth may need to be reset.

Have you ever tried this and the dimensions were ok?

[frankc]

Originally Posted by Matt Monson

Better the devil you know than the devil you don't...

Yes, I agree and would not gamble on a used 915 in this case since my internals are known to be good. But I was just trying to make a point that 15 hours of labor would be expensive comparatively.

[Steve@Rennsport]

Originally Posted by frankc

But from what I have read, the bearings have enough variance that this is not likely. The mechanic I spoke with stated that even the pinion depth may need to be reset.

This is true for the most part and one really needs to measure ring gear backlash rather than depend on "Kentucky Windage". Bearing QC isn't what it used to be and its both prudent and wise to never assume anything in this regard.

Its most like that any corrections needed can be done with bearing shims; its rare to require pinion depth adjustment.

If you have a choice, buy the same brand of bearing that you removed.

[frankc]

Steve,
Thanks again for your wisdom.

To set the preload, it appears to me that the pinion can be disengaged from the ring gear by simply unbolting the diff housing from the center housing - is this correct? That is, there should be no need to tear into the internals of center housing. And then to check the backlash, bolt the two (diff & center housing) back together? Or am I being too optimistic here?

[bazar01]

Originally Posted by frankc

Steve,
Thanks again for your wisdom.

To set the preload, it appears to me that the pinion can be disengaged from the ring gear by simply unbolting the diff housing from the center housing - is this correct? That is, there should be no need to tear into the internals of center housing. And then to check the backlash, bolt the two (diff & center housing) back together? Or am I being too optimistic here?

To replace just the bad bearing, you really don't have to do anything on the pinion depth. You can leave that alone. You are lucky enough that the bad bearing was on the flange cover.

Your ring gear backlash was already set by the shim on the left side bearing that is good.

You just have to make sure the new bearing will have the right preload using the same shims that came off the flange cover side to keep the differential locked axially and still be able to spin. Too much shim or preload, and the bearing will not spin and will overheat. Too loose, it will move axially and my bottom out and destroy the pinion.