APPLYING ASME AND BRITISH STANDARD WELDED FATIGUE METHODOLOGIES TO THERMAL FATIGUE CASE STUDIES AND INVESTIGATING SENSITIVITY TO FATIGUE FLAW GROWTH MODELS

摘要

Predicting fatigue cracking at welded locations in fixed process equipment due to cyclic loading is of great interest to the chemical, petrochemical, and nuclear industries. Welded fatigue methodologies have been incorporated into both ASME and international pressure vessel Codes and Standards. These fatigue methodologies are based on welded fatigue tests. Some of the challenges associated with evaluating in-service equipment for susceptibility to fatigue failure are considering the overall scatter associated with the test data that forms the basis of the welded fatigue methodologies and using Code-based fatigue predictions to make practical recommendations relating to establishing inspection intervals and facilitating overall life cycle management. In this study, the Battelle Structural Stress welded fatigue methodology outlined in ASME Section Ⅷ Division 2 is utilized. Additionally, the welded fatigue assessment techniques of British Standard PD 5500 are included for comparison. In this paper, several case studies are investigated where transient thermal-mechanical finite element analysis (FEA) is used to develop fatigue predictions for several pieces of in-service equipment subjected to cyclic operation that experienced significant cracking in certain regions of the pressure boundary. These case studies focus on the thermal fatigue aspect of the observed cracking. In addition to Code-based fatigue methodologies, fracture mechanics calculations are performed to determine critical flaw sizes, and sub-critical fatigue flaw growth calculations are performed to investigate the propensity for a given initial flaw to propagate to critical dimensions. Furthermore, flaw propagation rate sensitivity to varying fatigue flaw growth models is quantified and discussed. This paper highlights several real- life examples of thermal fatigue cracking and how computational and analytical methods can be used to perform fatigue and flaw growth predictions to understand the root-cause of the observed damage and to determine the likelihood of fatigue crack propagation going forward.