Originally Posted by
Mike Mills
All things fatigue. Steel structures fatigue. Aluminum structures fatigue. Even graphite composite structures fatigue.
Fatigue life is usually dominated by flaws and flaw growth/propogation. Flaws are inherent in the materials and can also be induced during manufacturing and use.
Fatigue life must be dealt with statistically (Single samples are not statisically significant. It's like drawing a straight line through a single data point).
Good structural design anticipates flaws and is flaw-tolerant.
Fatigue life curves (cycles to failure as a function of strain level) are available for most common structural materials.
It is possible to estimate fatigue life using "cumulative damage" theory.
I thought fatigue life was based on the size of stress/strain cycles relative to the elastic range of the material, and the number of stress/strain cycles. If this isn't the case, how can a cumulative damage theory be used?
Failures due to fatigue begin at places where there is a microcrack, inclusion, or other flaw where stress is concentrated during normal use and stress cycling.