Reliability Evaluation for Creep Life Prediction and Creep Crack Growth Rate of Modified 9Cr-1Mo steel

Abstract

Modified 9Cr-1Mo steel (ASME Grade 91, hereafter referred to as Gr. 91 steel) is a promising candidate for structural materials of Gen-IV reactor systems, such as steam generators, intermediate heat exchangers, and hot pipes in sodium-cooled fast reactors, and reactor pressure vessel in a very high temperature reactor, owing to its high creep and low cycle fatigue resistance than those exhibited by its counterparts such as 9Cr-1Mo and 2.25 -1Mo steels. Since these structures are designed for up to 60 years at elevated temperatures, they may cause creep failure, and a significant portion of the component’s life will be spent in crack propagation during the long service life. Therefore, it is necessary to reasonably establish the reliability on the long-term creep life prediction and creep crack growth rate (CCGR) behavior in the components that are subjected to creep loading for a safe design life of 60 years. To do so, this study deals with above two subjects on their reliability evaluation, as follows. First is to evaluate a reliability of the long-term creep life prediction of Gr. 91 steel. In the creep life prediction, a new master function of a hyperbolic sine (“sinh”) form, instead of a polynomial equation which has been conventionally used well in the Larson-Miller parameter, was newly proposed. A number of creep rupture data were collected through literature surveys, and using these data, the long-term creep life was predicted by the L-M parameter. The results showed that a master curve of the “sinh” function was found to have a wider acceptance with good flexibility in the low stress ranges beyond the experimental data. In a reliability evaluation of the predicted creep rupture life, a Service Condition-creep Rupture property Interference (SCRI) model based on the Z parameter was demonstrated. The Z parameter followed a normal distribution well. From the normal distribution, a number of random variables for Zs and Zcr were generated using Monte-Carlo simulation technique. As examples for application of Gr. 91 steel, the prediction for the creep rupture life under specific service conditions was reasonably drawn from the viewpoints of reliability. The reliability of the creep rupture life decreased with increase in service time, and a higher temperature and stress conditions caused a faster deterioration. Also it decreased rapidly at the higher temperature/stress fluctuations and the large scattering of the creep rupture data. Second is to investigate a probability in the CCGR evaluation of Gr. 91 steel. A series of CCG data was obtained from the CCG tests under various applied loads at 600℃. In the characterization of CCGR, the C* fracture parameter was used, and material constants, B and q in the CCGR equation, da/dt=B(C*)q were analyzed by means of three methods; a least square fitting method (LSFM), a mean value method (MVM) and a probabilistic distribution method (PDM). The results showed that the PDM and MVM were more useful than the LSFM, because it could assess the CCGR lines from the probabilistic viewpoints. It was supposed that the B and q followed two-parameter Weibull distribution. Using the Weibull distribution, a number of random variables were successfully generated by Monte Carlo simulation. The CCGR lines for 10 % to 90 % ranges were drawn from the viewpoint of reliability. Consequently, it is expected that the results of this study can be utilized to various heat resistance steels to evaluate a reliability in the long-term creep life prediction. Moreover, it is suggested that they are useful to achieve the correct and economic designs because a probability method in the CCGR evaluation has a good rationale for reasonable conservatism.