AbstractThe lifespan and structural integrity of various high temperature low stress applications are limited by creep damage. For most metals and alloys, cavitation damage is understood to be the dominant damage mechanism. Currently, there is no reliable and precise method for industries to model creep deformation and predict rupture time. The situation is even more challenging due to the strong stress level dependency of the minimum creep strain rate and creep lifetime.
This research focuses on developing a cavitation-based method for predicting creep rupture time using early-stage creep data. For accuracy, cavitation data obtained using x-ray synchrotron tomography are chosen for the study. Two methods of calibrating damage criteria based on the cavitated area and volume fractions have been reported. From the calibrated damage criteria, relationships between creep exposure time and cavitation damage are developed to enable a time-based extrapolation method for predicting rupture time. This approach offers traceability as quantifiable physical changes in the material (cavity nucleation and growth rates) are calibrated.
The cavitated volume fraction method offered better accuracy (87%) in assessing the level of damage in the material. The cavitated area fraction method produced 82% accuracy and was developed for comparison with current practices. The impact of cavity coalescence on the total number of cavities was illustrated, a result that suggests that evaluating damage via number of cavities per unit area can be erroneous. Furthermore, the research clearly identified the need of excluding measurements taken at the point of final rupture in creep lifetime modelling. This thesis contributes to the specific knowledge of modelling creep cavitation and rupture. It also offers a theoretical foundation and support for a time-based extrapolation approach of predicting creep lifetime. In addition, the cavitation modelling approach used in this study is expected to find application in other failure modes like fatigue.
|Date of Award
|18 Nov 2023
|Qiang Xu (Main Supervisor), Vladimir Vishnyakov (Co-Supervisor) & Joan Lu (Co-Supervisor)