When flaws are detected in pressure retaining components, assessments have to be done in order to demonstrate the fitness-for-service (FFS) of the component for continued operation. This FFS demonstration is performed in accordance with FFS Codes providing flaw assessment procedure and acceptance standards. The first step of the flaw assessment is the flaw characterization which aims at determining the flaw geometry to be used for the analyses. This key step is done according to flaw characterization rules provided in the FFS Codes and hence appears as essential for the rest of the assessment.
According to the flaw characterization rules of ASME B&PV Code Section XI, a nonplanar flaw (i.e., oriented in two or more intersecting inclined planes, curvilinear geometry, or combinations of nonplanar geometry) shall be resolved into two planar flaws by projection of the flaw area into planes normal to the maximum principal stresses. This approach allows to simplify the flaw assessment but should remain conservative. Therefore, the conservatisms due to the simplified projection approach for nonplanar flaws are investigated in this paper. Current computational tools have been clearly improved so that the modelling of nonplanar flaws does not present any significant difficulty. In this frame, this paper compares the stress intensity factors of projected nonplanar flaws and the mixed mode stress intensity factor of actual nonplanar flaws. This is carried out for multiple flaw sizes, flaw shapes, flaw orientations and different load cases. The final scope is to quantify how the flaw projection into planes normal to the maximum principal stresses is conservative and how this conservatism could be improved, if need be.