Seminarium AMAS Estimating crack closure characteristics - an example of optimal experimental design dr T. Svensson środa, 18 lutego 2004

Topper&Lam have developed a method for estimating crack closure characteristics without observations of the very phenomenon. This is done by postulating that

1. the mean value influence on fatigue life depends solely on crack closure and
2. that the crack closure level can be described by the Lam/Topper formula, including two parameters.

Fatigue tests on different mean values give then information about the closure characteristics and the parameters can be estimated. Topper&Lam propose that the tests should be performed at the R-ratios R=0 and R=-1. I will present the use of optimal statistical experimental design on the present problem, which leads to a test design that reduces the number of tests needed to about 60% of the original design.