A novel approach to surface defect detection

Defects or flaws in highly loaded structures have a significant impact on the structural integrity. Early inspection of faults can reduce the likelihood of occurrence of potential disasters and limit the damaging effects of destructions. According to our previous work, a novel approach called as Quantitative Detection of Fourier Transform (QDFT) using guided ultrasonic waves is developed in this paper for efficiently detecting defects in pipeline structures. Details of this fast method consist of three steps: First, an in-house finite element code has been developed to calculate reflection coefficients of guided waves travelling in the pipe. Then, based on boundary integral equations and Fourier transform of space-wavenumber domain, theoretical formulations of the quantitative detection are derived as a function of wavenumber using Born approximation. This lays a solid foundation for QDFT method, in which a reference model in a problem with a known defect is utilized to effectively evaluate the unknown defects. Finally, the location and shape of the unknown defect are reconstructed using signal processing for noise removal. Several examples are presented to demonstrate the correctness and efficiency of the proposed methodology. It is concluded that the general two-dimensional surface defects can be detected with high level of accuracy by this fast approach.