![Sample our Physical Sciences journals, sign in here to start your FREE access for 14 days]({"type":"image" , "src" : "/sda/110819/TOC-Subject-Sample-2021-Physical-Sciences.jpg"})
ABSTRACT
Quantitative evaluation of stress corrosion cracking (SCC) using eddy current testing (ECT) signals often underestimates the depth of crack due to the complex geometry and conductivity characteristics of crack. It can be approximately regarded as a multi-variable nonlinear optimisation problem considering the cross-sectional feature and conductivity distribution of crack. For the multi-variable optimisation problem, how to quickly and efficiently find a solution close to the true value in a huge search space is a challenge. In this study, a model of crack conductivity distribution and an intelligent optimisation algorithm were proposed for the quantitative evaluation of the SCC profile using the ECT technique. An SCC numerical model with a linear distribution of crack conductivity along the crack depth was adopted and validated for the numerical calculation of ECT signals due to SCC. Subsequently, a scheme based on the support vector regression method and particle swarm optimisation algorithm was proposed and implemented to improve the precision and efficiency of SCC sizing. The profiles of several SCC samples were quantitatively evaluated from both the simulated and measured ECT signals. The reconstruction results of SCC length and depth show good precision and stability, demonstrating the validity and efficiency of the proposed scheme.
Disclosure statement
No potential conflict of interest was reported by the authors.