Physico–statistical models based on the residual defectiveness of structural materials - Probabilistic Safety Assessment for Optimum Nuclear Power Plant Life Management (PLiM) - page 143

by the PISC members;
• preparation of technical reports on issues of rules and standards and
transfer them to the authorities responsible for the development of
normative documents.
The main results of these studies are summarised concisely below. The
results of studies of the reliability of the linear dimensions of discontinuities
in the direction of the wall thickness in inspection on the full-scale model
of the reactor vessel are described in Fig. 5.23. As can be seen, the error
in determining the depth of the discontinuity may reach significant values,
both in the direction of understating the size and in the direction of
overstatement. Figure 5.24 and 5.25 show defects No. 8 and No. 9 (see
Fig. 5.23) for which the errors in detection were the highest.
The results of the study of the probability of detecting defects in
austenitic steel are shown in Fig. 5.26. 5.27, and the test samples on which

a)

￿ ￿ ￿ ￿ ￿

b) Actual size, mm

5.23 Variation of the size of defects (except for defects Nos. 3
and 4) according to the results of inspection by all groups of NDT
inspectors (a) and dependence of the error in determining the
size of defects in the direction of the wall thickness of the reactor
vessel on the actual size of the defect (b).