Reliability analysis of creep rupture dataset extrapolation methods for 316H austenitic stainless steel

¹ College of Mechanical Engineering, Zhejiang University of Technology, Hangzhou 310023, China;
² Engineering Research Center of Process Equipment and Re-manufacturing of Ministry of Education, Zhejiang University of Technology, Hangzhou 310023, China;
³ College of Materials Science and Engineering, Zhejiang University of Technology, Huzhou, 313200, China;
⁴ Key Laboratory for Green Pharmaceutical Technologies and Related Equipment of Ministry of Education, Zhejiang University of Technology, Huzhou, 313200, China;
⁵ Key Laboratory of Pharmaceutical Engineering of Zhejiang Province, Zhejiang University of Technology, Huzhou, 313200, China;

^a) 2112102139@zjut.edu.cn
^b) 211122020177@zjut.edu.cn
^c) lsy330521@163.com
^d) hzhexh@zjut.edu.cn
^e) lijialuo@zjut.edu.cn
^f)  bsy@zjut.edu.cn

Abstract

The prediction of the service life of high temperature structural components is critical to the safety of structures. Due to the scarcity of long-term test data, it is necessary to extrapolate and analyse the creep rupture data set to predict the life of a component. In addition, the reliability and integrity of the extrapolation of the creep rupture data set has an impact on the accuracy of the life prediction of high-temperature structural components, which in turn has an impact on the safety of structures. Firstly, the existing time-temperature parametric life prediction methods, e.g. Larson-Miller, Manson-Haferd, Orr-Sherby-Dorn and Monkman-Grant, are investigated based on the creep rupture data set of 316H austenitic stainless steel published by NIMS (National Institute for Materials Science) in Japan. Secondly, the Z-parameter method is used to examine the regularity of the statistical distribution of the data set. Then, considering the overall fit effect and consistency with the creep rupture failure mechanism, the distribution is estimated by the great likelihood method. In addition, the selected hypothetical distribution is tested by means of the K-S test. A reasonable distribution of Z-parameter is determined. Finally, the effect of the number of creep rupture data sets on the determination of the Z-parameter distributions is analyzed. As a result, the consistency and disparity of Z-parameter distributions of low-stress, long-duration creep data sets and high-stress, short-duration creep data sets are presented detailed.