| Abstract: | The applicability of accepted engineering procedures to predict the amplitude of singular crack tip fields during the transitional regime prior to steady state has been investigated. Estimates of both transient and steady state values of the C(t)-integral for simplified geometries are compared with results obtained from detailed finite element analyses. Different creep exponents and both primary (mechanical) and secondary (displacement-controlled) loadings are considered.Estimates of the steady state value C* were obtained from Nuclear Electric's high termperature assessment procedure, R5, which is based on reference stress concepts, and from the EPRI's fully plastic solutions for J-integrals. The transient behaviour of C(t) was estimated using approximations given by Ehlers and Riedel (Advances in Fracture Research, Proc. Fifth Int. Conf. on Fracture, Vol. 2, Pergamon, New York, 1981, pp. 691–698), Saxena (Fracture Mechanics: Seventeenth Volume, ASTM STP 905, ASTM, Philadelphia, 1986, pp. 185–201), and the R5 procedure (Ainsworth et al., Fatigue Fract. Eng. Struct. 10 (1987) 115–127). In most cases, the transient estimates given by the latter two were found to conservatively predict the finite element results, although some underpredictions were encountered in the planar geometries shortly after loading. The recommended use of plane stress reference stresses with the R5 procedure, however, provides overall conservatism in the values of C(t). Ehlers and Riedel estimates, though, are generally non-conservative, except for the lowest crack-tip constraint configuration analysed (i.e. centre cracked plate). |
