Recognition and evaluation of corrosion profile via pulse-modulation eddy current inspection in conjunction with improved Canny algorithm
2019
Abstract For the anti-corrosion measure, the nonconductive protection coating is normally deployed over the surface of the conductive structure which is widely utilized in the engineering fields including aerospace, nuclear energy, etc. Nevertheless, due to hostile and corrosive environments, hidden corrosion buried right beneath the non-conductive coating, namely Corrosion Under Coating (CUC) has been found occurring in the conductor surface, and severely undermines the structural integrity. Therefore, it is imperative to non-invasively interrogate CUC via Non-destructive Evaluation (NDE) techniques without removing the protection coating. Pulse-modulation Eddy Current technique (PMEC) has been identified to be advantageous over Pulsed Eddy Current testing (PEC) in higher inspection sensitivity to hidden corrosion and accuracy of corrosion imaging. It could be a promising and preferable NDE technique for evaluation of CUC. In this paper, PMEC for recognition and evaluation of the opening profile of CUC is intensively investigated through simulations and experiments. Based on the Extended Truncated Region Eigenfunction Expansion (ETREE) modelling, closed-form expressions of PMEC responses to CUC are formulated. A series of simulations are conducted to compare the evaluation sensitivity of PMEC to CUC with that of PEC. Following this, experiments for profile recognition and evaluation of CUC via PMEC are conducted. The improved Canny algorithm is proposed to enhance the accuracy in identification and assessment of the CUC profile. The results from simulations and experiments reveal that: (1) PMEC is superior to PEC in terms of evaluation sensitivity to CUC; and (2) in conjunction with the improved Canny algorithm, PMEC is capable of identifying and evaluating the CUC profile without much loss in accuracy.
Keywords:
- Correction
- Source
- Cite
- Save
- Machine Reading By IdeaReader
18
References
14
Citations
NaN
KQI