Blown engines - speculative explanation

[Parump]

Do you trust your tachometer and rev limiter to avoid over - reving the engine?

The following is only speculation.

Because the dashboard tachometer only indicates an average of the actual engine shaft revolutions per minute, is it sufficiently precise to provide warnings about impending, potential damage from a momentary over speed?

There have been a number of M3 engine failures attributed to over speed for both the manual and SMG transmissions. Whereas the manual transmission will allow excessive over speed when downshifting, the SMG transmission is programmed to prevent this type of occurrence. Yet, engine failures have occurred with both types of transmissions. Why?

I suspect that “momentary bursts” of excessive engine speed occur that fail to register on the analogue tachometer and fail to effectively engage the rev limiter. Because of the real time monitoring of the DME, the overspeed is recorded with no apparent warning to the driver. I believe that this may explain the reason why some engines have failed even though the driver may not be aware of exceeding the analogue tachometer’s redline or warning lights.

Furthermore, the changes recommended by BMWAG to address this problem by increasing the engine oil's viscosity and increasing the shell clearance - both implicate a condition where increasing friction due to lack of adequate lubrication, results in excessive temperatures leading to catastrophic engine failure.

Any comments?

[Kaz]

Now, I haven't been following these failures very closely. What is the typical mode of failure when this happens? Bearing spinout/burning? Valve crash? piston sieze? I think knowing how they fail would give a better idea of what the cause is. I assume BMW knows exactly what's going on by now.

[Parump]

Kaz,

See these links:

http://members.roadfly.com/jason/m3engines.htm

http://forum.e46fanatics.com/showthread.php?threadid=7264

[Kaz]

VERY interesting. Months before I bought my wagon (in fact before E46M3s were available in the US), a few of my car buddies and I had a few chats about early reports from Europe about blown E46M3 motors. 3 of my friends are M owners. E30M3, E36 '332is' and MCoupe. We had all heard from various sources about the engine failures, and speculation as to that being the reason why US intro was being delayed.

I'm no engine expert but I have a good idea of how they work, and how they can fail. If those posts are representative of all the failures out there, I think its relatively straightforward as to what the general failure is.

Crankshaft/rod bearing failure seems pretty obvious. At high loads and RPMs, a lack of proper lubrication in this area will no doubt cause a failure in the bearing. Frankly, if there are problems with lubrication getting into that area at all (from various possible reasons including sump starvation, a baffle limiting natural splash-circulation, lack of auxiliary oil spray, etc.) the engine wouldn't even have to be highly stressed for failure to occur.

BMW's apparent decision to increase bearing clearance and specify a higher viscoscity oil sounds to me like a 'quick fix', and that there are further inherent design issues.

FWIW, I've read that one of the reasons that Honda's DOHC VTEC engines can safely reach stratospheric RPMs is that they developed a special micro-finish on the bearing surface specifically to hold oil in at high stress/high RPM conditions. This finish is apparently so fine that if the bearing surface is even touched during assembly, it MUST be replaced.

[Kaz]

As for the DME overrev readings that BMWNA and dealers are using to deny repairs, some of them at least, sound suspect to me, especially the SMG ones.

Even IF the overrevs occured, I don't think a momentary (<1sec) overrev into the 8600rpm territory, especially in SMG cars where that's not even supposed to be possible to begin with) would take an engine from 'perfectly good' to 'blown'. Traditional spun bearing/thrown rod type failures come, I believe, from an instant of extreme overrevving (>20% of max?) or extended periods of mild overrevving.

In any case, this kind of close tolerance WRT the difference between 'good' and 'dead' isn't something that should be part of a mass production car. Vehicles that are to be driven by 'normal' people really should have greater safety margins against failures.

[scottn2retro]

I would think it has something to do with the RPMs exceeding the design of something in there. Almost all of the reports are from M3s. The Roadster and Coupe are limited about 500 RPM lower and there are virtually no reports of blown engines on those.