a comparison for number of cycles at prescribed crack size, will not
reflect the actual error.
The error in predictions by scheme-B and scheme-C is given in
Table 7. Scheme-C i.e. KRMS based approach works better for FCG life
estimation except for test SSPB 6-9.
Crack shape (front) of the growing crack has also been predicted
using scheme-C. This has been plotted with respect to 2c/a and a/t
and shown in Fig. 16.
During the experiments on pipes, crack depths have been
measured all along the crack front at different number of loading
cycles. A typical comparison between the predicted and tested
crack profiles with loading cycles has been shown in Fig. 15.
Plot of 2c/a and a/t (Fig.16) shows that analytical predictions and
experimental results compare well. SSPW 6-1 with initial notch
aspect ratio (2ci/ai) as two, exhibits relatively faster growth in
circumferential direction and thereafter 2c/a is maintained as 3 till
through thickness crack (Fig. 13). The pipe with initial aspect ratio
as 4 (SSPW 6-3) shows comparable growth rates in depth (a) and
length (2c) directions after a/t reaches 0.4. This is clear from the
plateau region as shown in Fig. 16. This has occurred due to the