Publication

Hyperautomation Artificial Intelligence

2020 Operation-Adaptive Damage Assessment of Steam Turbines Using a Nonlinear Creep-Fatigue Interaction Model

본문

Journal
IEEE Access
Author
Woosung Choi, Heonjun Yoon, and Byeng D. Youn*
Date
2020-07
Citation Index
SCIE (IF: 3.4, Rank: 34.5%)
Vol./ Page
Vol. 8, pp. 126776-126783
Year
2020

Abstract


For sustainable operation, power plants require accurate damage assessment of the steam turbines used for power generation. In the past few decades, steam turbine designs have considered only steady-state operation, without accounting for thermal cyclic loads. However, steam turbines designed for steady-state operation are often subjected to operation under cyclic and transient operation as well; this situation can cause both creep and fatigue damage. Creep and fatigue damage can interact with each other, thereby significantly impacting performance and leading to premature failure. Thus, this study proposes operation-adaptive damage assessment for steam turbines using a nonlinear creep-fatigue interaction model. The three-fold novel aspects of this study include methods to: 1) adaptively determine the hyper-parameters embedded in the nonlinear creep-fatigue interaction model depending on the operation mode (i.e., base-load and peak-load); 2) incorporate actual operation data acquired from field-deployed steam turbines; and 3) interpret the creep-fatigue damage interaction diagram with respect to the operation modes. It can be concluded from the results that the nonlinear creep and fatigue damage interaction is significantly affected by the operation mode as well as by the type of dominant damage mechanism.