Mod.9Cr-1Mo鋼のクリープ疲労損傷則の改良

抄録

A genetic algorithm (GA) is a mathematical search technique based on the principle of natural selection and genetic combination, and there are many problems in the research fields of the mechanical behavior of materials to which a GA can be applied. The present paper investigates the improvement of creep-fatigue damage rule for Mod.9Cr-1Mo steel by using the GA program, which was developed by the authors in order to estimate the values of material parameters required to describe the creep-fatigue damage growth models of materials. Three kinds of creep-fatigue damage growth models were studied, such as a model proposed for Mod.9Cr-1Mo steel, a model proposed for 316LC steel and a more general model proposed in the present study. Optimization of material parameters in each model by GA analysis was conducted based on the results of constant-strain amplitude creep-fatigue tests and two-step variable-strain waveform creep-fatigue tests. As the results, it is found that the 316LC type creep-fatigue damage growth model can predict more precisely creep-fatigue crack growth behaviors and remaining life of Mod.9Cr-1Mo steel than the other two models.