The head bolts failed on my 93 750 after only 65,000 miles. There was no overheating, abuse, head crack, etc. One day I just found coolant in my oil. The seal between the head and the gasket had failed for no apparent reason. As will be seen below, I ultimately determined that the head bolts had failed in normal use.

A search of various web sites revealed that this problem, while not an epidemic like the V8 coating problem, infects a small but significant number of V12's.

For a variety of reasons, I did the fix myself (successfully, though with many a scraped knuckle). A big part of fixing a problem is knowing why the problem exists, so you cure the problem rather than just fix a symptom. So I did quite a bit of research into why the head bolts fail on V12's. I stopped short of having the bolts scanned by an electron microscope to determine the exact mode and sequence of failure, but I think I can give everyone a pretty solid theory on why these bolts fail (and, unfortunately, why you can't do anything to stop it). My apologies in advance to any metalugists who read this, as I'm using common english and not precise technical terms

1. The V12 heads are bolted to the block using a concept known as "torque to yield." A blue collar machinest or mechanic might call the bolt a "stretch bolt." The elastic properties of the bolt are a significant factor in clamping the head to the block.

2. Start with an analogy. As you stretch a rubber band, it will resist more and more and also will clamp tighter and tighter on whatever you are binding with the rubber band. This will be true up until the "yield point," where the rubber band starts to break down and can't return to its original position. Once you've reached the yield point, the band won't bind as tightly. Also, after the yield point, only a little more force is needed to reach the "modulus of rupture," (i.e., the point where the band breaks).

3. Just like a rubber band, a head bold will stretch and has a yield point. Up to the yield point, as the bold is streched it will clamp down harder and harder on the head. Unlike a typical head bolt, where the torque applied during tightening supplies the vast majority of the clamping force, the v12 uses the stretch in the bolt to apply a significant percentage of that force. Since the clamping force increases as the yield point is approached, the tightest fit theoretically is reached at the point just short of the yield point.

4. Thus, we have "torque to yield." The bolt is torqued and streched until it is just short of the yeild point, thus maximizing the available clamping force.

5. The head bold fails when, during service, the razor-thin difference between maximum clamping force and the yield point is exceeded, resulting in the bold "overstreching" and losing it ability to firmly hold down the head.

5. Here is some real-world data to show you what happens: on my V12 the head bolts were so overstretched that it never took more than 47 foot pounds of torque to remove the bolt. Some came off with as little as 35 foot pounds. When I put in new bolts, I used a torque-angle gauge per specification, but tracked foot pounds also. The final torque on each bold was about 65 foot pounds (i.e., not a lot). I accidently overtorqued one bolt by 30 degrees. The bold was so elastic, I torqued 30 degrees beyond maximum and never even hit 70 foot pounds! After about 65 foot pounds, the bolt just yields and is easily stretched further.

6. You can spot a "torque to yield" situation because the bold is threaded along its entire length, even though only a small section actually fits into the block. The full-length threading is required to make sure the stretch occurs evenly along the bolt's entire length. Another clue is the use of degree specifications rather than foot-pounds for torquing. The degree method is supposed to be more accurate and lets you get closer to that magic "just short of yield" torque value.

7. So, assuming BMW doesn't overtorque one or two bolts at the factory, why is there a failure? The most likely reason is that the margin of error is just too small. Once a bolt is torqued that close to yield, there is a statistically significant likelihood that something will occur to push one bold past the yield point. Once it fails, the load redistributes to the other bolts but, because they also were torqued close to the yield point, they do not have sufficient excess capacity to carry the load without exceeding their yield points and the result is a complete failure of the bolts. The "something" which causes one bolt to fail could be anything, from an isolated hot spot in the head to corrosion to simple fatigue.

8. I noticed in other web postings that some had attempted to get BMW to help out and it would not acknowledge the problem. It isn't hard to see why. Compare the v12 block to, say, a Jaguar six. The v12 has very little "meat" around the bolt holes, while you could just about double the size of the holes on the jag. My guess is that BMW got too fancy, tried to put too much into a lightweight casting, and left no margin for error. It is stuck with using "torque to yield" bolts and cannot, say, change to studs or any other less exotic system, because such systems would require way more torque than the 65-odd foot pounds the present system uses, the size and depth of the holes would have to be changed accordingly, and there just ain't room to do it. The problem is not easily fixable (have they changed the casting on the 5.4 liter?) and BMW is not about to admit such a major error on its top-of-the line motor. Also, lets face it, most of the people who buy these puppies new are leasing them and trading them in 4 years down the line. Longevity is not an issue for them and most people stuck with the problem are people like me, who bought used. Were not BMW's target market and thus not a likely recipient of corporate good will.

Anyway, I've said enough for now. Happy motoring.