2004 Volume 44 Issue 8 Pages 1441-1450
In general, the ultimate creep failure mode of welds fabricated from low alloy ferritic steels can be considered to be Type IV cracking resultant from coalescence of grain boundary damage. Thus, observation of grain boundary feature using a replication technique has been one of the most common practices for the remnant life assessment of welds operated at high temperatures. The experimental results in the present work and those found in previous works, however, prove that the feature of grain boundary damage highly depends upon testing conditions. Therefore, the relationship between the extent of damage and remaining life would not necessarily be effective for the practical use if it were obtained under the extremely accelerated laboratory conditions.
In contrast, strain rate measurement predicted the rupture life with reasonable accuracy independently of the susceptibility to grain boundary cavitaion. The increase in the strain rate due to the presence of creep cavities was not observed in the experiments using heat treated specimens, which had the microstructure expected at the Intercritical HAZ in whole gauge length. This fact suggests the effectiveness of strain rate measurement for the life assessment and the necessity of reconsidering the physical meaning of grain boundary damage.
