Section 7.5.2. Initial and Continuing Damage
A 0.05 inch initial crack is assumed in slow-crack growth
structure and in fail safe structures.
If the specimens for design development testing are not provided
with artificial defects, the cracks, once initiated, will grow through the sizes
mentioned above. Crack-growth records
would automatically cover the span of the requirements, provided the cracks can
be detected. Otherwise the recorded
crack-growth curve would have to be extrapolated backwards. If initial flaws are provided, it is
recommended to make them the size assumed in the requirements or close to the
size for analysis substantiation.
Continuing damage, from a testing standpoint, is more difficult
to make as a result of the small initial sizes and the different growth requirements
for different cases. Consider the
example configuration shown in Figure 7.5.1 where A
is the primary damage site, and B, C, D, & E are continuing damage
sites. The four parts of the figure
show (a) the initial damage assumed in the panel per JSSG-2006 paragraph 3.12.1
and Tables XXX and XXXI, (b) the initial damage and growth until the primary
damage terminates at the edge, (c) the continuing damage that starts at B, the
opposite side of the primary damage site which terminates in hole 2, and (d)
the growth of continuing damage at C until termination at hole three. While the
analysis can follow the assumptions required by JSSG-2006, it would be
difficult (if not impossible) to manufacture the necessary continuing sizes
either prior to test or after the primary damage (segment A) terminated at the
edge. Therefore common practice is to
put in the primary damage and continuing damage starter flaws as shown in Figure 7.5.1.a and let the specimen crack growth
proceed without additional perturbations.
Post test analysis of the crack growth data and fracture surface
striation morphology will document this logic.
Figure 7.5.1. Primary and Secondary Damage Sites and Continuing Damage
In a AFRL/VA contracted study, Brussat et al.  were able
to show that the experimental fatigue lives of built-up structure with and
without continuing damage flaws were about the same and that the primary crack
damage chose the most effective path through the structure. Secondary cracks
developed in a natural way during the test. Dormant periods when a crack ran
into a hole could be estimated and
subtracted if the results were used for a check of the analysis.