STS304 스테인리스강의 크리프 파단 데이터 통계적 성질에 관한 연구

Abstract

Recently, it is being used the elevated temperature heat-resisting materials and becoming severed their used environment. It is necessary to study of creep behavior for the elevated temperature heat-resisting materials. During the past three decades, numerous studies of individual uni-axial tension creep behavior of elevated temperature heat-resisting materials have been reported. However there are very few about the study on statistical properties of creep rupture data, and the scatter. There is a probabilistic nature to strength, load, geometry, and the mathematical models of underlying phenomena that has to be identified and described in order to design to a reliability goal. Therefore, the aim of this paper is the investigation of the statistical properties of creep rupture characteristic values (for example, creep rupture time, steady state creep rate, total creep rate, initial strain, etc.) in STS304 stainless steels. The creep rupture tests of STS304 stainless steels were performed at three different elevated temperatures of 600, 650 and 700℃ under constant stresses, respectively. It seems clear that the rupture time has a correlation with various characteristics of elevated temperature creep. The effect of temperature on the statistical scatter for creep characteristics was the smallest at 700℃. And, the effect of stress on the statistical scatter for creep characteristics was smaller with increasing stresses. The probability distributions of creep rupture data were well followed 2-parameter Weibull distribution. The mean Weibull shape parameter values for rupture time obtained 7.2, 9.5 and 17.1 at 600℃, 650℃ and 700℃, respectively. The shape parameter of rupture time, initial strain and total creep rate increased with increasing temperature. The probability distribution function of the creep rupture test data seems to show a dependence on the temperature and stress level. For creep rupture test data, confidence bounds of Weibull parameters was estimated by Fisher Matrix(FM) confidence bounds. The presented approach provides background necessary for the statistical modeling of creep rupture data, and for making long-term creep failure predictions concerning the reliability.