Quantitative mapping of service process-microstructural degradation-property deterioration for a Ni-based superalloy based on chord length distribution imaging process

Directionally solidified (DS) and single crystal (SC) Ni-based superalloys inevitably underwent microstructural degradation induced by the harsh operating environment. For the safety service and economic overhaul, con-structing a quantitative mapping chain from service process to microstructural degradation and to property de-terioration is critically essential. The present work started with stress-free and stress-assisted pre-service treatments of a DS Ni-based superalloy to obtain microstructures with different degraded states. An imaging pro -cess based on two-phase rotary chord length distributions was established to extract the high dimensional sta-tistical information for identifying the morphology and size features of microstructures. To reduce the dimension of the statistical information and quantitatively characterize the microstructural states in fewer pa-rameters, principal component analysis was employed to capture the core microstructural indicators, which was utilized to establish the response surface between the deterioration of fatigue resistance and the microstruc-tural degradation. Finally, a multi-output support vector regression (SVR) model was constructed to map be-tween service process and microstructural degradation. The results showed acceptable accuracy to estimate the microstructural degradation of pre-serviced alloys. Meanwhile, the framework provides a technical chain for the waste determination and microstructural degradation estimation of the hot section components made by DS and SC Ni-based superalloys. ? 2021 Published by Elsevier Ltd. This is an open access article under the CC BY-NC-ND license (http://creativecommons.org/licenses/by-nc-nd/4.0/).

Automated summary

The study focused on obtaining different degraded states of DS Ni-based superalloy microstructures through stress treatments, extracting core microstructural indicators using principal component analysis, establishing a response surface, and constructing a multi-output support vector regression model for mapping service process to microstructural degradation. The results showed acceptable accuracy in estimating the microstructural degradation of pre-serviced alloys and provided a technical chain for determining waste and estimating microstructural degradation of hot section components made by DS and SC Ni-based superalloys.