Official S65 Bearing Specification/Clearance Wiki

[regular guy]

Originally Posted by Yellow Snow

It's a shame that eccentricity wasn't mentioned earlier in the thread. It would have saved sooo much time.

Great job RG.

Eccentricity is sooo hot right now.

[catpat8000]

Originally Posted by Yellow Snow

It's a shame that eccentricity wasn't mentioned earlier in the thread. It would have saved sooo much time.

Great job RG.

It was discussed months ago. For one example, see: http://www.m3post.com/forums/showthr...3#post15004073

[whats77inaname]

RG, you rock. The amount of time and dedication you have devoted to this is outstanding.

While I can appreciate a good "debate" and differing opinions, this thread is FULL of insults, name calling, and general maliciousness. Now that there is conclusive proof that the bearings are the problem, I personally find it *hilarious* that none of the antagonists can man up and simply say, "You were right, I was wrong, I apologize."

[FogCityM3]

Right, at the beginning I was the one who proposed the "deliberate ambiguity" thesis with respect to oil changes.

However, after thinking about this more, the best way to actually reduce litigation risk is to prevent it. BMW already came out with the LL-1 recommendations, which already flagged that there is a potential problem (in conjunction with the awesome work done on this thread). Now, by not actively making it clear to anyone that LL-1 is okay to use, no one will try to change the oil and not only have you increased the chances for failure (an economic impact), the chances for litigation (economic and safety issue), but also increased the chances for a gross negligence argument (knew was a safety issue, but was not clear in giving the right recommendations to customers that could have prevented damage to property and/or individuals). So if BMW continues to stand by the "deliberate ambiguity" approach with respect to oils, I think that is a grave error and creates more potential problems than it solves (predicated under the assumption that there is a higher than avg risk of bearing failure).

Originally Posted by regular guy

Others have pointed out that the ambiguous oil change wording is intentional to AVOID litigation. They say if BMW were to make the announcement unambiguous and say it's to mitigate a design "flaw" or "limitation" then it would INCREASE the risk of litigation and class action lawsuit. Hench the ambiguous wording.

I would subscribe to that same belief.

[FogCityM3]

verrrry interesting tidbit!!

Originally Posted by kawasaki00

Also highly coincidental that the president of BMWNA was tooling around a "shop" asking questions.

[kawasaki00]

Originally Posted by FogCityM3

Right, at the beginning I was the one who proposed the "deliberate ambiguity" thesis with respect to oil changes.

However, after thinking about this more, the best way to actually reduce litigation risk is to prevent it. BMW already came out with the LL-1 recommendations, which already flagged that there is a potential problem (in conjunction with the awesome work done on this thread). Now, by not actively making it clear to anyone that LL-1 is okay to use, no one will try to change the oil and not only have you increased the chances for failure (an economic impact), the chances for litigation (economic and safety issue), but also increased the chances for a gross negligence argument (knew was a safety issue, but was not clear in giving the right recommendations to customers that could have prevented damage to property and/or individuals). So if BMW continues to stand by the "deliberate ambiguity" approach with respect to oils, I think that is a grave error and creates more potential problems than it solves (predicated under the assumption that there is a higher than avg risk of bearing failure).

I know a car showed up at our local dealer with a rod out, it was out of warranty and had a tune and the guy got a new engine. I think it has been a hush hush but if it goes out they will replace it as long as no S/C has ever been installed.

[swamp2]

Originally Posted by whats77inaname

RG, you rock. The amount of time and dedication you have devoted to this is outstanding.

While I can appreciate a good "debate" and differing opinions, this thread is FULL of insults, name calling, and general maliciousness. Now that there is conclusive proof that the bearings are the problem, I personally find it *hilarious* that none of the antagonists can man up simply say, "You were right, I was wrong, I apologize."

Let's please not forget what a theory is. I've certainly beat the horse a bit on this in this thread. It combines, hypothesis, measurement, and testing. We are still lacking quite a bit in the latter most point.

Anyway, you may or may not call me one of those antagonists but as to whether the new bearings show a different clearance than the old ones, I "manned up" IMMEDIATELY (just a couple of posts back FYI).

That being said, I think it is still just a bit premature to conclude that the bearing dimensions and or materials are a definitive problem. If the changes to the new bearings "fixed" a "problem" we should not see any (or radically fewer) failed engines with the new bearings. Although I'm sure there have been reported failures in the new 702/703 bearings, I'd also virtually guarantee we can't conclude that the 702/703 failure rate (for only lightly modified engines) is conclusively lower than that for the older 088/089 bearings. Hence we really have no testing. We can certainly jump to some conclusions based on this observation but we still can't damn BMW as completely and fully as many wish to.

[Yellow Snow]

Originally Posted by catpat8000

It was discussed months ago. For one example, see: http://www.m3post.com/forums/showthr...3#post15004073

A few more posts here that were met with scorn at every opportunity, starting from page 1 right through the whole 80+ pages. Not one of the self professed experts could comment with any knowledge.

Perhaps with the help of Google some of the experts will now emerge with the knowledge to comment with authority

Originally Posted by Yellow Snow

How does the size between the parting lines compare to a regular lower revving BMW motor such as an M54?

The S65 shells only give a .0016" clearance across the smallest point ie 90 degrees to part line. The clearance across the part line is at least .002" greater, thus giving an oval bore with .0035" - .004"clearance at the widest point.

From this I'm assuming that BMW did some stress analysis at 8,400rpm and found the rod bore stretched by .001". This would mean the bearing bore would be perfectly round at 8,400rpm resulting in a clearance of approx .0025" all around the bearing?

Originally Posted by Yellow Snow

The ovality of the standard bearing is .0048". (Taken from an actual bearing) That's a lot more than most. Add on the planned .003" clearance and it's pushing .008"

Originally Posted by Yellow Snow

Not sure if I was clear enough

If a stock bearing has .0015" clearance at the tight spot, adding another .0015" clearance results in the planned .003" clearance at 90 degrees to the part line.

Adding on the .00048" extra at the parting line equals .0078" clearance across the split. That seems like a lot of clearance.

Surely an ovality of .0015" is more the norm, not .0048".

Originally Posted by Yellow Snow

What components were you getting a .004" variation on? That seems a ridiculous amount.

My sizing was taken from an actual shell, not an assembled rod.

The eccentricity does seem greater than on any regular engine. I imagine this eccentricity has been designed in to give tightish clearance at low revs and normal clearance at high revs as the rod stretches and the bore distorts into an almost perfect circle.

Let's face it, giving a bearing enough clearance for a 3 hour race with a fully warmed up engine is an absolute piece of cake.

Compare that to working out a rod/bearing design that covers daily cold starts and shopping duties etc then to be caned the life out of on track at weekends is a different kettle of fish.

Originally Posted by Yellow Snow

In the interest of clarity, I have re measured the eccentricity of the bearing against the same 2.000mm slip gauge at 10mm up from the parting line which is the officially recognised measuring point.

The shell is .0012" (.03mm) thinner here than at the 90 degree point. This equates to a .0024" eccentricity which is bigger than the norm.

Do you have an 088/089 bearing that you can measure at 10mm up from the part line for comparison?

Originally Posted by Yellow Snow

That's the problem with statistics. If Dinan admitted to just one failure they would have at least double the failure rate of stock motors at 1%. Just like BMW, would they release this info?

Anyway back on topic.

Did you get the measurements of the 088/089 bearings at 10mm above the parting line?

Originally Posted by Yellow Snow

We still don't know for sure that there is anything wrong with the clearance. The M designers obviously built in the clearances and eccentricity for a reason. I think it's fair to say they have more technical knowledge and R&D support than anyone on the forum.

Is the extra eccentricity of the bearings part of their deliberate design criteria? Can you do a comparison to a regular BMW shell

Grinding a thou off the journals really does seem like a crazy idea when the facts aren't proven.

Originally Posted by Yellow Snow

Is the extra eccentricity of the S65 relevant? The resulting .0065" clearance at the part line is plenty big enough to spew out the 10w60. I wouldn't want to make it any bigger

I would be very surprised If Clevite are saying that BMW don't know how to size a journal to suit their bearings. Surely that's professional suicide?

Originally Posted by Yellow Snow

Measure the shell at 10mm from the parting line. The S65 is .0012" thinner there than at the 90 degrees point which is significantly more than the normal .0004" of regular shells.

This will effectively give a lot more eccentricity than a regular shell when installed in a rod no matter how you want to look at it.

If you are actually doing the measurements yourself, why can't you check it and report back

Originally Posted by Yellow Snow

If you assemble and torque a rod together the bore should be perfectly round.

Fitting a bearing shell which tapers slightly to the part line gives the oval shape.

Measuring the shell at 10mm from the split ignores any crush distortion when fitted torqued into the rod.

If one type of bearing has a .0012" radial difference (.0024" total)at 10mm and the other has .0004" radial difference (.0008" total) it equates to more oil clearance.

It's a significant difference

Originally Posted by Yellow Snow

Does anyone have a theory yet on why the standard bearings have so much eccentricity when compared to other cars?

Originally Posted by Yellow Snow

No one has come up with a reason on why the S65 bearings have a much higher eccentricity than most.

Originally Posted by Yellow Snow

Here is an explanation of effective diameter when eccentricity is taken account of. Taken from the Bearing geometry link. http://www.substech.com/dokuwiki/dok...ngine_bearings

Eccentricity

The inside bearing surface is not round. It has a lemon shape due to the varying thickness of the bearing wall having maximal value at the centerline (T) and gradually decreasing towards the parting line. It is accepted to measure the minimal value of the bearing wall thickness (Te) at a certain specified height h in order to exclude the zone of the crush relief.
Bearing eccentricity.png
The difference between the maximal and minimal wall thickness is called eccentricity:

eccentricity = T - Te

Eccentricity produced by the varying wall thickness is added to the eccentricity “e” caused by the displacement of the journal from the concentric position. Such increased total eccentricity allows to establish a more stable regime of hydrodynamic lubrication.

With regards to the hydrodynamic conditions the bearing with an eccentricity is equivalent to the bearing with an increased diameter (or increased oil clearance). The oil wedge of the bearing with eccentricity is the same as the wedge formed by the bearing with an increased diameter (”effective bearing diameter”). On the other hand the actual bearing diameter is not changed, thus the adverse effect of the bearing clearance on the vibration and noise is prevented.

Bearing eccentricity is designed to compensate distortions of the bearing housing bores caused by the forced applied to the connecting rod and to the crankcase. Under the forces the housing bore is stretched in the vertical direction. As a result the bearing diameter measured along the parting line decreases (close-in) changing the shape of the oil wedge. The bearing eccentricity allows to retain the wedge shape of the oil gap, which is required for hydrodynamic regime of lubrication. High performance engine bearings operating at high rotation speed and high loads have an increased amount of eccentricity.

Advanced quality control of machining operations allows to produce bearings with tight tolerance. Such tight tolerances result in superior consistency of the oil clearance and eccentricity.

Recommended values of eccentricity:

For passenger cars: 0.0002 - 0.0008” For high performance cars: 0.0006 - 0.0012”

Originally Posted by Yellow Snow

Could someone with an open mind comment on the following paste from this link regarding extra oil clearance when eccentricity is factored in.
Eccentricity. The S65 appears to be right at the top of the eccentricity scale.

Taken from here.

http://www.substech.com/dokuwiki/dok...ngine_bearings

The inside bearing surface is not round. It has a lemon shape due to the varying thickness of the bearing wall having maximal value at the centerline (T) and gradually decreasing towards the parting line. It is accepted to measure the minimal value of the bearing wall thickness (Te) at a certain specified height h in order to exclude the zone of the crush relief.
Bearing eccentricity.png
The difference between the maximal and minimal wall thickness is called eccentricity:

eccentricity = T - Te

Eccentricity produced by the varying wall thickness is added to the eccentricity “e” caused by the displacement of the journal from the concentric position. Such increased total eccentricity allows to establish a more stable regime of hydrodynamic lubrication.

With regards to the hydrodynamic conditions the bearing with an eccentricity is equivalent to the bearing with an increased diameter (or increased oil clearance). The oil wedge of the bearing with eccentricity is the same as the wedge formed by the bearing with an increased diameter (”effective bearing diameter”). On the other hand the actual bearing diameter is not changed, thus the adverse effect of the bearing clearance on the vibration and noise is prevented.

Bearing eccentricity is designed to compensate distortions of the bearing housing bores caused by the forced applied to the connecting rod and to the crankcase. Under the forces the housing bore is stretched in the vertical direction. As a result the bearing diameter measured along the parting line decreases (close-in) changing the shape of the oil wedge. The bearing eccentricity allows to retain the wedge shape of the oil gap, which is required for hydrodynamic regime of lubrication. High performance engine bearings operating at high rotation speed and high loads have an increased amount of eccentricity.

Advanced quality control of machining operations allows to produce bearings with tight tolerance. Such tight tolerances result in superior consistency of the oil clearance and eccentricity.

Recommended values of eccentricity:

For passenger cars: 0.0002 - 0.0008” For high performance cars: 0.0006 - 0.0012”

[swamp2]

Originally Posted by SenorFunkyPants

Indeed...the recent engine failure poll gave:
60% of the total votes were for the years 2008/9/10
against
40% of the votes for the years 2011/12/13.
Factor in the difference in mileage implies no sudden downturn in the failure rate.

Thank you for adding that. I've browsed that thread and should have extracted this simple stat myself. If all of the changes don't "fix the problem" empirically then RGs conclusions just prior in the long eccentricity post (below for clarity) seem still yet to be a bit premature, especially the part I placed in bold.

Now that being said I have not attempted to analyze which of these engine failures have potentially anything to do with bearings, clearances or oil.

Originally Posted by regular guy

Conclusions and my $0.02

So to me, the dimensional changes of the new bearing weren't based on a simple materials change, but were much more deliberate.

The S65 started production with 088/089 bearings. But something was wrong: relatively new engines were puking connecting rods. The clearances were too tight, the side clearance was too tight, the eccentricity was too tight, and the oil was too thick. As a result, it seems like tolerance stack up with a bad luck of the draw, and your engine might end up looking like the photos shown above. So BMW decided to do something about it.

When BMW designed the newer 702/703 bearings to comply with lead-free regulations, they made changes. I believe they took steps to mitigate these problems. BMW increased the rod bearing clearance, and they increased the eccentricity by 250%. Those changes mean two of the four possible "trouble spots" I mentioned above have now been addressed.

...

Some will say it's all a coincidence. They are just as entitled to their opinion as I am mine. I don't think it's a coincidence. For whatever reason, BMW changed (coincidence or not), BMW had three years with old bearing; then three years with new bearings; then changed to allow thinner oils. Sure it may all be a coincidence, but to me, it seems like they were chasing something and were making incremental changes to mitigate what they saw as a problem.

Last but not least we still don't yet have much of an idea if the total overall bottom end catastrophic engine failure rate for lightly to unmodified cars is closer to 0.1%, 0.5% or 1%... At what number is this a real problem?

And truly lastly (at least in this post...) RG - don't your measurements overall support the contention that BMW achieves pretty remarkably tight tolerances? In some cases your equipment can barely capture part size deviations (e.g. rod journals). Even with purposefully chosen worst case limits the bearing clearances have a really small range. Certainly some tolerance stack up exists but it begs the question if the failures are explained by rare cases of extreme tolerance stack up. It's a very reasonable point of view. It is unlikely for measurement of such a small number of parts to exhibit any outlier types of results. However, tolerance stack up is also something that most likely would not be fully remedied by adjustments to mean (design) values. Failures would decrease but almost for sure wouldn't be eliminated. This is sort of nature of the beast of manufacturing (variation will exist) and probably explains a huge number of failures across many parts of many different car models.

[aussiem3]

It has been one epic thread on all counts but no conclusion yet. A special thank you to RG and also to all others who made this discussion informative, argumentative, heated and character-assignation But it was all in good fun.

Also not to forget all those who financially contributed and made it possible to make the thread eccentric

... ...

[JRV]

Originally Posted by regular guy

Eccentricity is sooo hot right now.

Lol Omg it's everywhere. BTW thx for doing the work

[FogCityM3]

Agree, the content of this thread is excellent.

Would it be premature to at least draw the conclusion that the risk of rod bearing wear can be mitigated if you have 702/703 bearings (higher clearance/greater eccentricity) and you switch to thinner LL-1 approved oils?

The remaining questions for me personally are 1) what is the bearing clearance under load at 8,400rpm (probably unanswerable); 2) as a partial corollary to #1, does raising rpm limit increase risk of bearing wear? 3) did the crankshaft measurements change?; 3) the nature of the 2011+ engine failures- notwithstanding the small sample size, does seem to be too many given the bearing changes.

Originally Posted by aussiem3

It has been one epic thread on all counts but no conclusion yet. A special thank you to RG and also to all others who made this discussion informative, argumentative, heated and character-assignation But it was all in good fun.

Also not to forget all those who financially contributed and made it possible to make the thread eccentric

... ...

[SenorFunkyPants]

Originally Posted by FogCityM3

3) the nature of the 2011+ engine failures- notwithstanding the small sample size, does seem to be too many given the bearing changes.

I'm in the process of messaging and re-messaging the voters in the latest engine fail poll to try and get a clearer idea of whats going on.
The poll question must have been quite ambiguous going by the fairly large number of people who voted by mistake (14 with still quite a few replies to arrive).

[BPMSport]

Raising the rev limit will definitely cause more bottom end stress.

To build onto what FogCityM3 said, it seems that since the newer bearings allow for more clearance, that switching to the thinner oils would mainly benefit those with the older bearings. Is the switch to 0-40 any less recommended than it would be to those with the old bearings? Seems like 10w60 is fine for those with the newer ones given there is more room there for lubrication than there was previously?

Any thoughts on this are much appreciated.

[swamp2]

Originally Posted by FogCityM3

Agree, the content of this thread is excellent.

Would it be premature to at least draw the conclusion that the risk of rod bearing wear can be mitigated if you have 702/703 bearings (higher clearance/greater eccentricity) and you switch to thinner LL-1 approved oils?

My two cents are that this basically follows from the engineering. However, there are a variety of very complex phenomena. Tribology is not always really hard science with black and white conclusions. Thus I'd say this is advisable as "insurance" but it certainly is not firmly established by the work in this thread. Without more empirical data or controlled test data we can't achieve certainty. Hopefully we can firmly conclude that BMW would not make a change in the bearings for the worse in any way whatsoever.

Originally Posted by FogCityM3

The remaining questions for me personally are 1) what is the bearing clearance under load at 8,400rpm (probably unanswerable); 2) as a partial corollary to #1, does raising rpm limit increase risk of bearing wear? 3) did the crankshaft measurements change?; 3) the nature of the 2011+ engine failures- notwithstanding the small sample size, does seem to be too many given the bearing changes.

The clearance at redline is only known by BMW and can be determined with careful finite element simulation by a 3rd party.

Increasing redline will likely increase loads but to be sure a rigid body kinematic simulation model is required to answer the question firmly. There can be cases where loads can be smaller at higher rpms due to cancellation of combustion pressure and inertial loads (I posted graphs showing this prior)at this rpm range (redline to redline plus some small amount) the load increase would be almost entirely from inertial components.

[catpat8000]

Originally Posted by swamp2

Increasing redline will likely increase loads but to be sure a rigid body kinematic simulation model is required to answer the question firmly. There can be cases where loads can be smaller at higher rpms due to cancellation of combustion pressure and inertial loads (I posted graphs showing this prior)at this rpm range (redline to redline plus some small amount) the load increase would be almost entirely from inertial components.

The peak combustion pressure occurs at peak torque. This occurs at 4000 rpm. The torque curve is pretty flat on the S65 and the engine develops roughly the same amount of torque up to 7500 rpm, whereupon torque starts to fall.

So from 7500 rpm to 8400 rpm, torque (combustion pressure) is falling and inertial loads are climbing with the square of the rpm.

So why couldn't we say conclusively that loads will increase substantially as the rpm limit is raised upwards of 8400 rpm?

Pat

[regular guy]

Originally Posted by aussiem3

It has been one epic thread on all counts but no conclusion yet. A special thank you to RG and also to all others who made this discussion informative, argumentative, heated and character-assignation But it was all in good fun.

Also not to forget all those who financially contributed and made it possible to make the thread eccentric

... ...

I think the character assassination was amusing. Seriously some have thrown up the same two 3 year-old articles 5 or 6 times now and ignored the response every single time. I keep asking myself what they hope to gain by bringing it up a seventh time if the first six times didn't work for them? And what does any of that have to do with the data presented in this thread anyways? Truly mind boggling.

[M3Captain]

All this nonsense about bearings and no one is talking about detonation, timing, and octane.

The fact that the top bearing is taking all the abuse and not the bottom looks like detonation. How many S65 engines are out there? And how many have failed due to bearing failure?

Discuss.....

[regular guy]

Originally Posted by M3Captain

All this nonsense about bearings and no one is talking about detonation, timing, and octane.

The fact that the top bearing is taking all the abuse and not the bottom looks like detonation. How many S65 engines are out there? And how many have failed due to bearing failure?

Discuss.....

We hashed out that topic quite a bit in here.

[catpat8000]

Originally Posted by regular guy

We hashed out that topic quite a bit in here. Maybe more than any other topic.

Indeed we did. And the general result (also true for most of our discussions) has been that we lack the appropriate data to make concrete assertions. This is why RG decided to create data by doing all the measurements he performed over the holidays. And which is the major positive outcome of this thread - RG's data.

Pat

[aussiem3]

Originally Posted by Mike Benvo

Raising the rev limit will definitely cause more bottom end stress.

To build onto what FogCityM3 said, it seems that since the newer bearings allow for more clearance, that switching to the thinner oils would mainly benefit those with the older bearings. Is the switch to 0-40 any less recommended than it would be to those with the old bearings? Seems like 10w60 is fine for those with the newer ones given there is more room there for lubrication than there was previously?

Any thoughts on this are much appreciated.

Mike, I just cannot make sense of this oil issue. BMW NA is the only market an alternate oil has been suggested but that too is vague. Non of the other markets are even talking about it. Something is not right.

As you know, my car is a Sept '07 and probably one of the earliest build cars on this board and probably with the least mileage - 41,500 km. It has had BPM Stage I for a year now, and as of 24th Dec 2013, it's had BPM Stage II with TPs. I am running TWS 10w-60 and the engine hasn't skipped a heartbeat. I religiously warm up the car and then I do push to its limit in every single drive. According to BMW you start the car and drive carefully until all oils in the car gradually reach maximum operating temperature. There is a practice that, you start the car and leave it idling until the engine oil warms up and then push. I think this practice may (please note the use of may here) put unnecessary load on the engine. Changing to a lower viscosity oil is only a Band-Aid solution in my opinion. I have TWS stocked up to last me at least five oil changes.

I think like SFP pointed out, there is more to this bearing failure like poor fuel quality. I was pumping fuel yesterday and inside the filler it recommends RON 95-98 and AKI 91. So if people are put in anything less you have to factor that it in. It may not be the case but combine that with other production factors and bad luck = explosion.

So, it's just not eccentricity, bearing clearance, poor gas, oil change intervals, money shift ... who knows what else. I don't think even BMW can figure this out if not there would have been a fix already like the E46.

So lets document it and hope when we blow that BMW are still in the market and there are enough S65s on the shelf in Munich.

I just can't see why people here think that they know better than BMW ... that's to drive the car at high rpm from the time you turn the key. There is a reason why you are instructed to follow the oil temp as you gradually increase the load.

We do everything BMW tell you not to, and then when things go wrong we blame the manufacturer. These points have been raised and discussed in the 80-page record breaking thread but I am just reminding including myself.

And, how many of these blown engines have the rev limiter increased? If they have they will not acknowledge it on this thread. BMW might not be able to detect a tune but increasing the rev-limiter and the top speed are absolute giveaways that there has been some kind of tinkering going on. I am sure in those cases engine swap under warranty have been denied.

[GIdriver]

RG, a big Thank You Sir!!!

The effort you have put into this is impressive!

Thanks also to all who contributed financially and anyone who contributed to the discussion.

Does anyone here thinks we have enough information and/or enough bearing related failures to print this thread, take it to a lawyer and start a big ass class action lawsuit?

PS: I believe that the title of the last table on post #1786 should read:
Complete 702/703 Bearing Specifications with Eccentricity