Detection and Characterization of Hidden Crack Inside a Stainless Steel Material Using Shear Vertical Waves and PLSR Technique
Résumé
Industrial equipment can be exposed to various types of damage during their long exploitation in harsh environments which might lead to corrosion cracking. Typically, a hidden crack is very dangerous since it is difficult to be detected by Non-Destructive Testing (NDT) techniques. Moreover, the integrity requirements of metallic structures cannot be satisfied by simply detecting the existence of cracks. However, the quantitative description of the cracks is still an issue that should be solved. This paper introduces a new approach to estimate the dimensions of hidden cracks inside a stainless steel plate based on Shear Vertical (SV) waves generated by an Electromagnetic Acoustic Transducer (EMAT). The Finite Elements Method (FEM) has been used to model the acquired signal by EMAT. Then, the characterization of hidden cracks is carried out by combining two methods; Time-of-Flight (ToF) technique that has been used to evaluate the crack height, and the Partial Least Square Regression (PLSR) method that has been used to estimate the crack width. The obtained results demonstrate the efficiency and the reliability of the adopted approach concerning the quantitative evaluation of the crack dimensions.