Creep-rupture prediction for steels and alloys by parametric methods and the base diagram method. Part 1. Interpolation analysis of experimental data

Abstract

Results of interpolation analysis of over 1000 creep-rupture diagrams by the Larson-Miller, Manson-Haferd, and Orr-Sherby-Dorn parametric methods and the base diagram method are presented. Prediction accuracy is shown to be essentially improved by applying special systems analysis of known experimental data.Introduction. All the world knows over 100 creep-rupture prediction methods. However, only several of them have received wide acceptance, in particular the Larson-Miller (LM) [1], Manson-Haferd (MH) [2], and Orr-Sherby-Dorn (OSD) [3] parametric methods, despite numerous critical remarks, which have never been considered to date, and repeated attempts of their improvement [4,5]. It is determined by the three reasons: the majority of practical problems is comparatively simple, long-lived items from structural steels and alloys have already operated tens of years and in this case the object of prediction is simplified, prediction is a rather complicated process.Many investigators [6,7] believe that the problem can be settled by interrelated and comprehensive accounting for the breakthrough in mechanics and physics, more specifically in mechanics and materials science. However, in practice this important scientific concept has never been realized. In this connection A. J. Kennedy's statement [8] deserves special attention: "The growth of more concise and detailed theories of the solid state has provided, for the first time, a physical appreciation of many features of crystal behavior: this will be evident from the previous discussion of dislocations. It has been less successful, however, in dealing with aggregates. While the laws which govern the behavior of small groups of dislocations may be appreciated in physical terms, and are supported by the results of specially-designed experiments on single crystals, an aggregate of dislocations (if the term may be allowed) provides much more profound difficulties. Taken further, we find that the laws which appear to express the behavior of aggregates cannot, as yet, be successfully derived from the laws relating to the behavior of their individual units. This is a difficulty of great significance to the future of materials engineering and perhaps too lightly regarded at the present time. The fact is that the understanding of the nature of the deformation process which physical science has provided is not, in itself, sufficient. For many practical purposes, the earlier empiricism is far more valuable". This at least suggests how varied and complicated the role of different imperfections of the crystal lattice is in the behavior of heat-resistant steels and alloys under long high-temperature loading. Despite impressive progress in the dislocation theory, sufficient scientific grounds for solving the problem are still lacking.The object of the present investigation is the substantiation of possibilities for improving calculation results by solving both direct and inverse prediction problems. Among the first ones are conventional simulation methods of metallic material sta...

Methodological aspects of creep prediction for heat-resistant steels and alloys. Part 2. Prediction over wide ranges of temperatures and loading times

“…The latter are responsible for high β 3 values ranging up to three or more, while in analysis of large bodies of experimental data, β e > 2 3 . were not practically found [11].…”

“…Moreover, the difference can be so essential that normal data presentation for one or two random states of the material in many reference editions may be considered inadequate, thus, it is necessary to refine the above characteristics, including refinements by simultaneous creep-rupture prediction. The development of such prediction is associated with many difficulties, which can be judged from the following results obtained by the equation [6,10,11] 373 …”

“…The use of BDM and β i values is caused by the fact that under the most characteristic loading conditions, β i values for different materials are well-established. For example, β i values are 0.9-1.3 for solution-hardened alloys and 1.3-1.7 for dispersion-hardened alloys over corresponding temperature ranges [11]. Data (Tables 1 and 5) correspond to the first and second ranges, respectively.…”

Studies on the chemical composition–creep-rupture strength relation for heat-resistant nickel alloys

“…However, most investigations were carried out by BDM-2 [16,17], which is used for two-three log-time prediction cycle, i.e., at q = 1-3 or more. At the same β β − e difference, prediction accuracy can show a decrease due to an increase in the number of log-time prediction cycles, as well as due to stress-dependent β values.…”

“…1 [18]. The results of correlation (interpolation) analysis of more than 1000 creep-rupture diagrams, obtained by parametric methods and BDM at close S 1 values in the National Research Institute for Metals (NRIM, Tokyo, Japan) were presented earlier [16]. For analysis by parametric methods, 5-8 numerical constants were used, as compared to BDM with the β 1 , β 2 , and β 3 constants.…”