|About this Abstract
||Materials Science & Technology 2019
||Data Science for Material Property Interpretation
||Application of Artificial Neural Networks to Low Cycle Fatigue and Creep Data Processing for Power Plant Materials
||Kuk-Cheol Kim, Jhin-Ik Suk, Byeong-Ook Kong
|On-Site Speaker (Planned)
The objective of this study is to obtain more reliable method for material data processing using artificial neural network. For low cycle fatigue, the design curve is necessary to ensure integrity against cyclic loading. To secure Coffin-Manson design curve in low cycle fatigue testing, tests are carried out under specific conditions such as temperature and stress ratio. However, to simulate various situations, fatigue design curve at arbitrary temperature and stress ratio conditions are also required, which have not been tested. To solve this problem, an artificial neural network technique has been proposed which can derive general fatigue design curve equation. For creep properties, survival analysis is best method which can utilize the ongoing test data. In this study, we applied artificial neural network technique to existing survival analysis method and to improve the reliability of analysis so that creep rupture curve can be used even when it is nonlinear.
||Definite: At-meeting proceedings