Signal processing for ultrasonic imaging of lossy composites

Abstract

Although ultrasonic imaging is popular for visualization of defects within composite materials, its application to certain types of composites produces unsatisfactory results if conventional methods are used. High resolution imaging, which would require using high frequency signals and sharply focused transducers is not possible because of high signal loss. If lower frequencies are used, there will be problems with both poor axial and lateral resolutions. There is also the problem of high structural noise due to scattering from the fibres, even at the lower frequencies. A new signal processing scheme has been proposed for ultrasonic imaging of high loss composite structures. The new scheme incorporates a 'frequency-selective loss compensation' and Wiener filtering within a split-spectrum processing algorithm. It was shown that new approach greatly enhances the signal-to- noise ratio and improves the depth resolution for images obtained in a specimen of high acoustic noise level. Wiener filtering, if applied in the conventional sense would be unsatisfactory in the case of high loss material, because of the difficulty in obtaining a suitable reference 'wavelet'. In comparison with other signal processing methods new approach shoved the best performance. The good performance of the algorithm can be explained by time-frequency dependent signal components proper correction by FSLC, which then contributes to good SSP algorithm performance and excellent interaction of Wiener and SSP properties. The X-ray radiography confirmed the results obtained by ultrasound.