Ultrasonic arrays have been widely used and developed for defect detection and characterization over the last 10 years. In this paper the question of how to inspect and characterize near-surface defects that are small with respect to the wavelength is addressed. The challenge is to overcome the effect of the proximity of these scattered signals to the high amplitude signals reflected from structural features, such as planar surfaces. Here, a method is proposed to extract the scattered signals from a near-surface defect which overlap with signals from a structural feature in both the time and frequency domains. The extracted signals are then used to generate a scattering coefficient matrix, from which it is possible to characterize the defect. In the proposed method, the location of the defect is first approximately identified from an image. The arrival time difference between the signals from the defect and the nearby planar surfaces for each combination of transmitter and receiver array elements is then calculated. In some cases the scattered signals can be directly separated in time from the structural features, and in other cases they are extracted by subtracting the data with reference signals, obtained in the absence of a defect. Finally, the proposed method is used to experimentally detect and characterize three different near-surface defects by extracting their corresponding scattering matrices.
- ultrasonic arrays
- scattering coefficient matrix
- defect classification
- defect characterisation