ASME STP-PT-080:2016 pdf free download – DEVELOPMENT OF AVERAGE ISOCHRONOUS STRESS-STRAIN CURVES AND EQUATIONS AND EXTERNAL PRESSURE CHARTS
The two parameters used for the tertiary creep are the initial creep rate and a . As can be seen in Figure 1.7, the initial creep rate is fairly well-behaved within the data range and well represented by a Larson-Miller type expression. Extrapolation, especially to very low stress values, could prove problematic. The a -parameter displays significantly more variability as can be seen in Figure 1.8. In addition, the values generated from the NIMS and those reported in the Ellis tables seem to form two distinct populations. The ASME FFS-1 tertiary creep equation is of the same form as the Ellis equation and the parameters, initial creep rate and omega, have nearly consistent meaning as the initial creep rate and alpha in this study. Both parameters in ASME FFS-1 are related to time and stress using Larson-Miller type expressions, though the Larson-Miller expressions are presented in logarithmic form. On this basis, we can compare the initial creep rate and alpha parameter determined from the regressions of the digitized NIMS data for “virgin” material to the published curves in ASME FFS-1 recommended for service aged material. In order to make the comparison, we must use the same Larson-Miller Constants for the data set as the ASME FFS-1 tables, which differ slightly from the constants we found, optimized the current data set. The resultant parametric fit determined for the initial creep rate in this study is similar to the curve developed for ASME FFS-1 curve (see Figure 1.9). This was not expected since the ASME FFS-1 curves were based on service aged materials. The Larson-Miller constants in the parametric fits are remarkably close. The sensitivity of the R 2 value to the constant is very low within the range of 30-40 and so either a value of 35.5 (current study) or 34 produce nearly the same results. The ASME FFS-1 expression appears to be better suited for extrapolation, but is less accurate a description of this data set.
The tertiary creep fit of the form of equation (1.9), “tertiary (with F),” provides the best fit of the final part of the creep curve (strains above 2-3%). The “tertiary only” fit ignores the integration constant, F, but uses the same parameters for initial creep rate and a in equation (1.2). This fit over-predicts the time to rupture and under-predicts strain. The “primary only” fit is the power-law model and it performs well at low strains (less than 2%) but, of course, does not include rupture. Simple attempts to fit the primary creep into the tertiary creep curve tend to cause the rupture time to either become consistently over-conservative or consistently under-conservative. A linear combination model of the two models fits well up to 3% strain but afterwards deviates from the true behavior. A simple first attempt of the non-linear combination fit which uses the primary creep to predict the integration constant under-predicts the time to rupture and over-predicts the strain and begins to deviate at around 2% strain. Only one example is shown here, but the same results hold true for all the cases investigated. The linear combination under-predicts strains and the non-linear combination over-predicts strains.