How do you define cracks in metallic structure? At what instance do you define crack initiation? Isn't it always a crack if you just magnify it enough?
I'm not a materials engineer but here's my take on this:
-I define a crack as any location with a stress riser that has grown into the thickness of the part due to applied stresses at that point.
-I'd define crack initiation as the point where stress causes the as-fabricated defect/stress riser to significantly expand up (crack) beyond the initial geometry in the original part.
-Yes and no. Obviously there will always be microscopic defects and even a perfect re-entrant corner can open up into a crack due to stress risers. As the stress theoretically hit an asymptote at the tip of these stress risers you could argue that yes, there is always a crack. However, I would say that locations that have not significantly expanded from their initial conditions are not cracks; just defects or geometric stress risers.
In practicality, if it hasn't caused a significant loss of section I wouldn't really consider the part "cracked".
Professional and Structural Engineer (ME, NH, MA)
American Concrete Industries
Thsnk you for the explanation. I think there must be some definition somewhere. In fracture mechanics, you are comcerned about crack propagation. If one can measure or calculate crack propagtion, where is the starting point? Is it just an abitrarily picked (small) crack length? I might just be satisfied that parhaps there is no good answer to that.
Typically, a minimum detectable crack/flaw is defined (based on the testing procedures in place or reasonable expectations based on manufacturing method), and this minimum flaw is analyzed for acceptability. In aerospace applications, this may be very small. In low-risk applications, a larger flaw (and less expensive test to detect such flaw) may be acceptable.
Flaws larger than the minimum may be rejected out of hand or analyzed on a case by case basis, depending on the application.
A crack is detectable (if it can't be detected, then is it there ? ... philosophy majors anyone ?)
If your installation is critical, then you need to figure out what your critical crack size is (the longest crack that will sustain the largest load expected in service, which should be less than your maximum design load).
Now you can have a discussion about inspection methods ... visual ? NDI ? different methods detect different minimum size cracks; how critical is it ?
another day in paradise, or is paradise one day closer ?
Crack propagation is usually only critical in dynamic structures. Does the crack grow in width? Length? What happens if the crack grows?
In general, a crack is defined as a visible anomaly. Yes, there are microcracks and sometimes they make a difference; however, in most cases we are only concerned about propagation and growth of cracks that eminate from a dynamic source. If it's visible, it's a crack!
