Thursday, February 17, 2011

Beech Marks, Fatigue Failure, and High Compression Pistons

"Designs will fail if subjected to overload...that's just the nature of efficient design--
they might not fail immediately; but some statistically determined time in the future--
they will not fail from overload but from a more insidious process called fatigue."

Lycoming Crankshaft Fillet - Fatigue Failure
Beech Marks are a sign that a crack progressed across the part and failure was due to fatigue. They are shown in the picture at the red arrow. The white arrow shows the crack initiation point.

 Fatigue occurs when the metal is subjected to repeated or alternating stresses not exceeding the material's static yield strength. A fatigue failure is a failure due to repeated stress BELOW the material or parts ultimate tensile stress. A part can operate normally and then suddenly fail in fatigue if cyclic stresses are above the fatigue strength of the metal. 

Even more interesting is that fatigue strength is a probability based on statistics and not one set value. The actual fatigue strength of a particular part might be less or might be more. There is no way of knowing unless you test it to failure.

So what has this to do with high compression pistons? Anytime you increase engine power above what the engine was designed for you assume that the original design is inefficient -- built stronger than need be and this extra strength caused by design ignorance is just waiting for some smart person to exploit.

But could it be that the original design is competent. That the engineer designed for endurance; the designer knew that statistically some of the parts would be slightly weaker in fatigue so he designed beyond 3 sigma as he had to be sure your crankshaft would not fail; that he designed for an infinite fatigue life by purposely limiting the stress. 

So how do you know if the high compression pistons place the fatigue life into the finite part of the fatigue curve without doing the stress analysis? You don't. Will it fail? You don't know. When will it fail? You don't know. Engine horsepower output overtime (endurance) can be no greater than crankshaft fillet strength.

Years ago when I asked a Lycoming engineer what he thought of a popular engine modification he said: 'ask me in 5 years; but of course it might fail in the 6th year." Who knows without the stress analysis.

19th century economist Frederich Bastiat:
"This explains the fatally grievous condition of mankind. Ignorance surrounds its cradle: then its actions are determined by their first consequences, the only ones which, in its first stage, it can see. It is only in the long run that it learns to take account of the others. It has to learn this lesson from two very different masters—experience and foresight. Experience teaches effectually, but brutally. It makes us acquainted with all the effects of an action, by causing us to feel them; and we cannot fail to finish by knowing that fire burns, if we have burned ourselves. For this rough teacher, I should like, if possible, to substitute a more gentle one. I mean Foresight."
Continental IO-520 thrown connecting rod

To summarize: it is better to learn from foresight than experience. 

No comments:

Post a Comment