Following the procedure outlined by Pandya and Williams , a circumferential notch was introduced into molded CDNT specimens with a single-point cutting tool of tip radius [less than or equal to]20 [micro]m.
Figure 4 shows how the time to failure in the CDNT test varies with static applied stress for the different PE materials assessed.
5 show scanning electron images of the failure ligament from CDNT tests on PE 3 (AUS) and PE 1 (AUS) samples.
While craze strength data was readily obtained using the CDNT test, the problem is to relate the uniform tensile stress state that exists in CDNT specimens to the actual stress state in the craze zone that is formed in other component geometries.
Following the reasoning outlined earlier, the time dependence of the reference stress is assumed to be the same as the time dependence of craze stress determined using the CDNT test.
4 and 5 it can then be equated to the measured craze stress from the CDNT test, and the time to crack initiation for a pipe in service should then be given by