S65 Rod Bearing Failure - New Theory? - Worn Spark Plugs + No Knock Sensors

[Theodore]

This is a very interesting theory from Nick Saran of VF Engineering @ 1:08:46 minute mark talking about S65 rod bearing failures. I've not heard this theory before:

Nick Saran Interview on The Smoking Tire Podcast

[Dave07997S]

Originally Posted by Theodore

This is a very interesting theory from Nick Saran of VF Engineering @ 1:08:46 minute mark talking about S65 rod bearing failures. I've not heard this from anyone else:

Nick Saran Interview on The Smoking Tire Podcast

This is something I thought of as well. As I have seen a lot of uneven wear on RBs. This can be caused by other things as well.

Dave

[///M Power-Belgium]

This was discussed before . Bad detonation , bad gas , worn spark plugs , ect...
Interesting *THEORY* . But actually *proven facts* are required for/in our situation !
And ...Cylinders running unevenly ? NO way !

[M.Hagen]

I mentioned this in the RB thread a few years ago.

If you see upper RB shell wear and little to no wear on the lower shell, that can be a symptom of pre-ignition (knock).

The ion sensing system is not robust, and is model based. Meaning its purpose is to predict knock based on measurable variables (airflow, rpm, EGT, etc). Since it actually does not measure knock, we have no way to determine using our sensor suite on the S65 whether or not it works.

Minor to moderate pre-ignition can beat the sh!t out of bearing shells, and you wouldn't even feel it or hear it. Only severe pre-ignition is audible and typically feels like the combustion is imperfect/breaking up.

The physics is very basic here. When pre-ignition occurs, the torque applied to the piston/rod is not timed correctly and also uneven. Typically a minor or moderate pre-ignition event will have multiple flame fronts/pressure waves, applying force on the piston multiple times, which can cause cavitation of the oil film between the upper shell of the RB and the crank journal.

[jcolley]

Originally Posted by M.Hagen

I mentioned this in the RB thread a few years ago.

If you see upper RB shell wear and little to no wear on the lower shell, that can be a symptom of pre-ignition (knock).

The ion sensing system is not robust, and is model based. Meaning its purpose is to predict knock based on measurable variables (airflow, rpm, EGT, etc). Since it actually does not measure knock, we have no way to determine using our sensor suite on the S65 whether or not it works.

Minor to moderate pre-ignition can beat the sh!t out of bearing shells, and you wouldn't even feel it or hear it. Only severe pre-ignition is audible and typically feels like the combustion is imperfect/breaking up.

The physics is very basic here. When pre-ignition occurs, the torque applied to the piston/rod is not timed correctly and also uneven. Typically a minor or moderate pre-ignition event will have multiple flame fronts/pressure waves, applying force on the piston multiple times, which can cause cavitation of the oil film between the upper shell of the RB and the crank journal.

Curious on your statement it does not measure knock. The mss6x DMEs use MPC563s which have Time processing Units which correlate events to crank angle. In the code for mss6x, there are variables for knock base frequency and bandwidth which are used for the model, but for actual knock detection (both "light" and "heavy" knock conditions), the time of the peak ion current as well as the integral of current actual measured is referenced to the TPU offset for the cylinder. In this way, the mss6x actually knows exactly where the pressure peak occurs in relation to degrees KW (kurbenwelle) and references that to an offset angle for the cylinder being measured.

To me, that seems to be absolutely an actual measurement of knock. Having mounted a set of knock sensors in an S85 connected to a MoTeC M150 for measurement (still running under mss65 control), I've seen intentional knock induced measured on the M150 and correlated by the mss65 pulling timing from the targets. This would also be easily reproduced by a simple set of knock ears...the most accurate and easy to use means of all.


Now, that's all fine and good for detonation, but not pre-ignition. Since pre-ignition is defined as a combustion event which occurs prior to spark plug discharge, the ionic current system is not aligned (switched to measurement mode) at this point, so unless the even overlaps into intentional ignition and the programmed delay for ionic switching, the ionic system can't measure it.

But, if pre-ignition is truly the cause, then why are the wear patterns of increased clearance bearings absent the characteristic of cavitation shown on OEM clearance bearings? Since oil clearance has no bearing on the conditions leading to pre-ignition (carbon build up and sharp piston edges being more causatory), it follows that it would still occur regardless of bearing clearance and we would see the same pattern.

[Theodore]

Originally Posted by jcolley

But, if pre-ignition is truly the cause, then why are the wear patterns of increased clearance bearings absent the characteristic of cavitation shown on OEM clearance bearings? Since oil clearance has no bearing on the conditions leading to pre-ignition (carbon build up and sharp piston edges being more causatory), it follows that it would still occur regardless of bearing clearance and we would see the same pattern.

Not saying I know this theory is true, but to keep the conversation going I wanted to point out owners who have replaced their rod bearings preemptively are a biased sample and perhaps the most likely group of owners to preemptively change their spark plugs well before the recommended interval. Also, is it possible the additional oil clearance masks or minimizes the issue Nick presents?

[Bartledoo]

Originally Posted by Theodore

Not saying I know this theory is true, but to keep the conversation going I wanted to point out owners who have replaced their rod bearings preemptively are a biased sample and prrhaps the most likely group of owners to preemptively change their spark plugs well before tge recommended interval. Also, is it possible the addition oil clearance masks or minimizes the issue Nick presents?

More cushion for the pushin?

[Malek@MRF]

Originally Posted by jcolley

Curious on your statement it does not measure knock. The mss6x DMEs use MPC563s which have Time processing Units which correlate events to crank angle. In the code for mss6x, there are variables for knock base frequency and bandwidth which are used for the model, but for actual knock detection (both "light" and "heavy" knock conditions), the time of the peak ion current as well as the integral of current actual measured is referenced to the TPU offset for the cylinder. In this way, the mss6x actually knows exactly where the pressure peak occurs in relation to degrees KW (kurbenwelle) and references that to an offset angle for the cylinder being measured.

To me, that seems to be absolutely an actual measurement of knock. Having mounted a set of knock sensors in an S85 connected to a MoTeC M150 for measurement (still running under mss65 control), I've seen intentional knock induced measured on the M150 and correlated by the mss65 pulling timing from the targets. This would also be easily reproduced by a simple set of knock ears...the most accurate and easy to use means of all.


Now, that's all fine and good for detonation, but not pre-ignition. Since pre-ignition is defined as a combustion event which occurs prior to spark plug discharge, the ionic current system is not aligned (switched to measurement mode) at this point, so unless the even overlaps into intentional ignition and the programmed delay for ionic switching, the ionic system can't measure it.

But, if pre-ignition is truly the cause, then why are the wear patterns of increased clearance bearings absent the characteristic of cavitation shown on OEM clearance bearings? Since oil clearance has no bearing on the conditions leading to pre-ignition (carbon build up and sharp piston edges being more causatory), it follows that it would still occur regardless of bearing clearance and we would see the same pattern.

Good post.

The theory on pre-ignition, is a plausible theory in the upper shell wear of the connecting rod bearings (A topic another forum member speculated on a 2-3 years back), it doesn't explain how this abnormal upper shell wear goes away once the rotating assembly clearances are corrected.

In many engines I have yet to see the typical upper shell wear be visible once the clearances are corrected.

[last1left]

Maybe the 36k plug replacement requirement is to long for this engine. Need more data on this.

[BRNDNSLVRA]

Originally Posted by last1left

Maybe the 36k plug replacement requirement is to long for this engine. Need more data on this.

I feel like all the factory advised intervals for service are too long and detrimental to our cars

[pbonsalb]

They are short enough to reduce failures during the warranty period.

[last1left]

Originally Posted by 954MM

I feel like all the factory advised intervals for service are too long and detrimental to our cars

Like the oil interval's

[M.Hagen]

Originally Posted by jcolley

Curious on your statement it does not measure knock. The mss6x DMEs use MPC563s which have Time processing Units which correlate events to crank angle. In the code for mss6x, there are variables for knock base frequency and bandwidth which are used for the model, but for actual knock detection (both "light" and "heavy" knock conditions), the time of the peak ion current as well as the integral of current actual measured is referenced to the TPU offset for the cylinder. In this way, the mss6x actually knows exactly where the pressure peak occurs in relation to degrees KW (kurbenwelle) and references that to an offset angle for the cylinder being measured.

To me, that seems to be absolutely an actual measurement of knock. Having mounted a set of knock sensors in an S85 connected to a MoTeC M150 for measurement (still running under mss65 control), I've seen intentional knock induced measured on the M150 and correlated by the mss65 pulling timing from the targets. This would also be easily reproduced by a simple set of knock ears...the most accurate and easy to use means of all.


Now, that's all fine and good for detonation, but not pre-ignition. Since pre-ignition is defined as a combustion event which occurs prior to spark plug discharge, the ionic current system is not aligned (switched to measurement mode) at this point, so unless the even overlaps into intentional ignition and the programmed delay for ionic switching, the ionic system can't measure it.

But, if pre-ignition is truly the cause, then why are the wear patterns of increased clearance bearings absent the characteristic of cavitation shown on OEM clearance bearings? Since oil clearance has no bearing on the conditions leading to pre-ignition (carbon build up and sharp piston edges being more causatory), it follows that it would still occur regardless of bearing clearance and we would see the same pattern.

Excellent post - thank you.

I say that the ionic measurement method is not directly measuring knock because it is using combustion quality and model-based parametrics to count knock events and frequency of knock. That isn't to say that the models aren't doing their job in most instances. For that knock control strategy to pass through development, and through multiple engine builds (S85 then S65), it must work in most conditions.

Both of my points are very much anecdotal since I do not have access to the funktionsrahmen or have any direct experience as you do measuring knock events both on a Motec and the DME at the same time... however those points are (1) while models are both predictive and reactive, they are blind to unforeseen variables. And (2), if the ionic sensing system was more robust than conventional wideband knock sensing, then most high-performance ICE's and worldwide automakers would transition to this strategy. Since ionic knock sensing systems are not widely seen, my assumption is there are gaps in the modeling strategy.

As for the data points where we see that larger clearance RB's are holding up better... I think we can optimistically say that the fix is known. Barring any longevity issues with the new RB's, I would recommend that fix to anyone looking to have preventative service done. With that said, the root cause for why there is excessive bearing wear is likely multi-factored. Could the primary cause be too thin of oil film/tight clearance? Sure. But what are the factors that cause that too thin film to degrade during performance? Maybe pre-ignition is a factor. Maybe it is cold start RPM's. Maybe it is oil viscosity and operating temperature variances outside of intended use range. Maybe it is all of those factored together that create the conditions responsible for bearing wear.

My "hypothesis" is: BMW's ionic knock sensing system allows for some degree of pre-ignition before intervention. While not audible, the pre-ignition/knock is severe enough to cause the very thin oil film between the RB shell and the crank journal to cavitate, as evidence of the visibly higher degree of wear on upper bearing shells compared to lower bearing shells. The factor of thin oil film may have such a high p-value in this large array problem, that simply creating a stronger oil film (larger bearing clearance) can in most cases eliminate the longevity issue with S65 and rob bearing wear.

[Patrón]

Originally Posted by ///M Power-Belgium

This was discussed before . Bad detonation , bad gas , worn spark plugs , ect...
Interesting *THEORY* . But actually *proven facts* are required for/in our situation !
And ...Cylinders running unevenly ? NO way !

proven facts, do you mean like the less than 2% failure rate?

[///M Power-Belgium]

Originally Posted by Patrón

proven facts, do you mean like the less than 2% failure rate?

Yes . And I never said less ..

[opihi5]

Originally Posted by last1left

Maybe the 36k plug replacement requirement is to long for this engine. Need more data on this.

If you drive the car hard cut all intervals in half or more. I replace sparks every 10-15k and oil every 5k.

[GORDON.M3]

Originally Posted by opihi5

If you drive the car hard cut all intervals in half or more. I replace sparks every 10-15k and oil every 5k.

Okay oil every 5k, fine. But spark plugs every 10-15k sounds overly excessive to me...

[Pig Farmer]

Is there a higher occurrence of RB failure in areas of the country that don’t have access to 93 octane fuel?

[Green-Eggs]

Originally Posted by Pig Farmer

Is there a higher occurrence of RB failure in areas of the country that don’t have access to 93 octane fuel?

No, there isn't, nor is there correlation with climate, temperature, etc.

There's nothing new about this theory. This theory has been discussed before...at length in the bearing wiki thread. It was shot down pretty hard in that thread. Anybody curious can find the discussion in the bearing wiki thread. This problem, if real, could have been fixed with a software update...but it wasn't.

We recommend to stick to getting advise from engine builders and the rod bearing manufacturers who can analyze failures and know exactly what caused them. To them, there's no real mystery about what's causing the problems.

[Patrón]

Originally Posted by GORDON.M3

Okay oil every 5k, fine. But spark plugs every 10-15k sounds overly excessive to me...

many tuners and 'shops' have recommended every 15k for all S motors going back to the s14 and s38. it is not that expensive considering the peace of mind, it's more important on the s65 because of the lack of knock sensors that the other s motors had.

[aswy6]

Originally Posted by M.Hagen

I mentioned this in the RB thread a few years ago.

If you see upper RB shell wear and little to no wear on the lower shell, that can be a symptom of pre-ignition (knock).

The ion sensing system is not robust, and is model based. Meaning its purpose is to predict knock based on measurable variables (airflow, rpm, EGT, etc). Since it actually does not measure knock, we have no way to determine using our sensor suite on the S65 whether or not it works.

Minor to moderate pre-ignition can beat the sh!t out of bearing shells, and you wouldn't even feel it or hear it. Only severe pre-ignition is audible and typically feels like the combustion is imperfect/breaking up.

The physics is very basic here. When pre-ignition occurs, the torque applied to the piston/rod is not timed correctly and also uneven. Typically a minor or moderate pre-ignition event will have multiple flame fronts/pressure waves, applying force on the piston multiple times, which can cause cavitation of the oil film between the upper shell of the RB and the crank journal.

My bearings and indeed most of bearings i have seen on this forum have this wear pattern. When i changed my bearings with deansbimmer he told me that its because explosion in the cylinder creates stress on the piston by moving it down suddenly. What do you think about that?

[deansbimmer]

Originally Posted by aswy6

My bearings and indeed most of bearings i have seen on this forum have this wear pattern. When i changed my bearings with deansbimmer he told me that its because explosion in the cylinder creates stress on the piston by moving it down suddenly. What do you think about that?

There's simply more force against the crankshaft during the power stroke a few degrees after TDC which correlates to the wear location we see in the upper shells. It stands to reason that there would be more wear on the upper shell. That's why upper shells are usually thicker than the lowers- They will wear longer.

When clearances are too small to begin with, it is easier for normal cylinder pressure stresses to overcome the oil film, causing the upper bearing wear. Pre-ignition as in the theory presented in this thread is a possible bearing wear culprit in all engines, however I personally believe if that were the real culprit here, we'd only see wear to the upper shells and no wear to the bottoms which is not the case. Could it be a combination of low clearances plus occasional pre-ignition? Possibly, but I'm not putting money on either theory until extra clearance bearings start coming out with similar poor wear after a bunch of miles.