AI Creep life-prediction assisted by creep damage evaluation
e-Journal of Nondestructive Testing · 2026-07 · 31 (7)
DOI: 10.58286/33536 ↗Kurz gefasst
Fachartikel in der Zeitschrift e-Journal of Nondestructive Testing, 2026-07. Verfasst von Arsalan Nazim und 2 weiteren Autor:innen.
Abstract (Originalsprache)
Ferritic Cr-Mo steels are used to a large extent in high-temperature applications, such as pressure vessels. However, prolonged exposure to extreme stress and temperature conditions results in material degradation, mainly through creep. Traditional methods of assessing remaining life use the Larson-Miller Parameter and Non-Destructive Testing, such as replica metallography. Nonetheless, each method suffers from intrinsic limitations that reduce the reliability and accuracy of the predictive process. The present work utilises several AI models, including Support Vector Regressors, Gaussian Process Regressors, and Neural Networks, to predict the remaining safe life (using the component's service time). The novelty of the present work lies in the integration of microstructural damage evolution information from metallographic replicas into the AI models as an added feature. The work also compares the performance of two different models (with or without replica information) strategies in assessing the remaining safe life.
Zitieren
APA
Nazim, A., Tonti, A., & Gariboldi, E. (2026). AI Creep life-prediction assisted by creep damage evaluation. e-Journal of Nondestructive Testing, 31(7). https://doi.org/10.58286/33536
Vancouver
Nazim A, Tonti A, Gariboldi E. AI Creep life-prediction assisted by creep damage evaluation. e-Journal of Nondestructive Testing. 2026;31(7). doi:10.58286/33536
BibTeX
@article{Nazim2026,
author = {Nazim, Arsalan and Tonti, Andrea and Gariboldi, Elisabetta},
title = {AI Creep life-prediction assisted by creep damage evaluation},
journal = {e-Journal of Nondestructive Testing},
year = {2026},
volume = {31},
number = {7},
doi = {10.58286/33536}
}
EUVIMED