When turbines are started up in power plants, the procedures recommended by the manufacturer should be carefully followed. However, many utilities often start turbines so fast that they accelerate the crack growth of existing cracks. Hence, in order to accurately predict the crack growth life of a turbine rotor, a crack growth prediction model is needed in which the effects of load increase time on the crack growth rate during the succeeding load hold period are considered.
In this study, creep-fatigue crack growth tests were performed at 538°C using C(T) specimens of 1Cr-1Mo-0.25V rotor steel. Triangular and trapezoidal fatigue waveshapes with 0.5, 1, 10, and 100-s load rise times and 100-s load hold times were employed. The time-dependent crack growth rate during the load hold period, (da/dt)avg, was correlated with estimated (Ct)avg. Large data scatter was observed due to the effect of various load increase rates reaching the hold load. A new Ct, estimation equation was proposed in which the effects of load increase rate are considered. The effectiveness of the proposed equation is argued by showing that the scatter of the measured (da/dt)avg data was reduced when the new equation was adopted. Also, characteristics of the initial transient crack growth behavior are studied and show that the cause is the oxidation-dominated crack growth mechanism during the transient period.