|
|
|
|
|
What is meant by "failure" must be defined early since avoiding failure is the objective both of design and prediction. "Failure," even for the same material, means different things to different industries and even for the same industry depending on the component:
For the architecture industry, rusting of exterior stainless steel may be a failure.
For post-tension applications, failure may not occur until up to 30% of the tendons are broken, depending on the distribution of breaks and the design basis for numbers of failed tendons that are acceptable.
For a pipeline, extensive small SCC may not be a failure if they do not propagate.
For some aircraft designs a crack may not be a failure until it approaches KIc .
For the chemical industry extensive corrosion is not a failure if it is judged not to perforate before the next outage.
For some pressure vessels any propagating SCC is a failure.
Once failure is defined, it is possible, then, to define what information is necessary to develop a predictive model and to define what elements of feedback are necessary during operation in order to prevent failure. Defining failure at the outset also provides a basis for the margin necessary to avoid failures in the future.
Lifetime Prediction, Roger W. Staehle, Adjunct Professor, Department of Chemical Engineering and Materials Science, University of Minnesota, Staehle Consulting Co.