Creep crack growth experiments and their numerical simulation.

Creep crack growth in 21 CrMoNiV 5 7 steel at 550 degree C can be described after first transients within a broad scatterband a power law using the C * intergral. Within the scatterband, some indications of dependencies on specimen size (triaxiality), crack length and crack growth increment can be explained. The initial crack growth rate values of all experiments can be described by the inital values of the strees intensity factor K. Under these circumstances, the elastic part of the deformation in the specimen prevails. A difference between creep crack growth rates under vacuum and laboratory air conditions was not observed. For a numerical simulation of creep fracture mechancis experiments, reliable material laws are essential. Then the experiments are fitted best by finite-element calculations under assumption of plane stress. Calculations on the basis of a Norton law for the minimum-creep rate of uniaxial tension specimens are already in rather good agreement with experimental resu lts. But the consideration of primary creep and secondary creep according to a Garofalo-type creep law is necessary for a realistic simulation and a better understanding of the whole specimen behavior. A stationary-crack model is adequate to simulate the first part of an experiment where crack growth no role. If crack growth is significant, calculations including crack growth are necessary for a reliable numerical description of the experimental results.