Anyone with an S54, S62, S85, or S65 will at some point encounter the question of rod bearing wear. Hopefully that encounter will be in the form of text on the Internet rather than a loud bang and a massive repair bill. This question has generated a lot of debate that, in most circles, has degenerated into tribal squabbling that produces far more heat than light. It's hard for a non-expert to understand what's really worth worrying about. In the absence of real data and fully honest communication from BMW, it comes down to value judgments based on incomplete data. This is one of my attempts to explain why informed and reasonable people often come to different conclusions.

Rod bearings are supposed to operate entirely in the hydrodynamic regime, which means a continuously-maintained oil film keeps the bearing from touching the crank journal while the engine is running. That obviously works most of the time, or engines would seize far more often than they do. But it equally obviously doesn't work all the time, or there'd be no need for replaceable bearings and all the related headaches. Why spend all the time designing, testing, sourcing, and assembling rod bearings, with all the possibility of bad tolerance stacking and assembly error they introduce, when you don't expect any wear? You could just bolt the rods straight to the crank and be on your way. Replaceable rod bearings exist because there's always some probability of freak stresses that break through that oil film. That probability is different for different engines and in different circumstances. Obviously it's higher on average in these engine than in most others, at least with their stock bearings and certain aftermarket alternatives.

To the extent that we see wear, the next question is what that wear means, and this is where things can get tricky.

Sometimes things are pretty cut-and-dry. If a rod bearing wears to the point of rod knock or seizure while everything else in the engine is perfectly serviceable, that's obviously a problem. If a rod bearing shows no significant wear until something else in the engine reaches a condemnation point, that’s obviously good (for the bearing at least).

What if the bearing wears noticeably but doesn’t get to the point of allowing knock or seizure until after the engine needs a rebuild for some other reason? That seems to be the case for many if not most of these engines in service. Is that amount of wear acceptable? That's not a hard case to make. If you're tearing down the engine anyway, you'll probably replace the rod bearings whether or not they're worn, so it won't matter whether they're new-looking or close to death. If we accept that, what does it mean on the scale of tens of thousands of engines? Statistically, even if the vast majority of engines’ rod bearings last until something else prompts a rebuild, a few won’t. How many premature failures is acceptable before we call it a design flaw? Ten percent? One percent? One-tenth of one percent?

These are all questions on which reasonable people can disagree. That, plus the lack of reliable stats and all the usual Internet-conversation shenanigans, means it's no surprise that there’s so much disagreement over what’s normal for these bearings.

Bearing manufacturers themselves seem to think some amount of wear is perfectly normal and permissible. As far as I can tell, all major bearing manufacturers have guidelines that include what constitutes “normal wear”: a consistent wear pattern with no hard boundaries over the majority of the bearing surface, without evidence of some other problem (scoring, corrosion, etc.). Of course, that doesn't settle anything because:

1. It doesn't say anything about wear rate, and
2. Not everyone trusts their motives.

Then there’s the following quote from Glyco, the manufacturer of the updated BMW "702/703" bearings for the S85 and S65 (source attached below):
If the clearance is minimal there is good conformability between the bearing and crankshaft journal. This conformability is a result of material that is worn in some parts of the bearing in the order of magnitude of μms. This process leads to less local stress on the sliding layer, a better absorption of shock loads and less wear.
That seems to suggest that with little clearance, as we all know some of our rod bearings have, there exists some form and amount of wear that is not only normal but beneficial. This seems to line up well with the fact that some rod bearings, particularly the 702/703, have been known to show wear patterns after just a few thousand miles.

Unfortunately, Glyco hasn’t posted pics of what this normal micro-wear might look like. We also don't know the thickness or metallurgy of the bearings' layers, and no one's measuring 702/703 bearings after they're removed from service. In fact, no one's really measuring any bearings after they're removed. Essentially all of the evidence of rod bearing wear comes from assessing them visually. That works just fine for lead-copper trimetal bearings, like the 088/089 bearings for the S85 and S65, because we know what they are and how they work. We know things can go south pretty quickly once that top layer is gone, and it’s easy to tell visually when that has happened. We don’t have the same knowledge for the 702/703 bearings. There are some cases in which they're obviously wrecked, and a few cases in which they're obviously fine, but beyond that we don't know where to draw the line between tolerable and problematic.

Those of us in the “rod bearings should never wear” camp can just dispense with all of this hand-wringing. If it looks worn, it doesn’t matter how long it’d last past that point; it’s sub-optimal by definition, and thus too risky, and that’s that. I’m not so sure, but maybe that’s a different topic.
Attached Files