We have performed linear scanning (B-scan)by keeping a transmitter fixed and receiver scanned away over the defective region in composite. After wavelet denoising, we have to find location and size of defect. One way is to compare the normalized amplitudes of all A-scans along the distance. What are the other ways if amplitude detection cannot be applied ( amplitudes in defect-free regions are not similar)? Do we need to compare spectral components after applying EMD/EEMD or STFT?
My advice would be to use matched filtering with a linear frequency modulation (sweep/chirp) excitation signal. This will increase your spatial resolution and your SNR.
Here is an interesting report: http://k-space.org/ymk/sprab12.pdf
And a very interesting article: Article Use of modulated excitation signals in medical ultrasound. P...
Hello Costin,
Thank you for your advice. The experiment is already performed. After interaction of guided waves with the defective region, the change in amplitude is normal due to reflection, refraction, mode conversion or other wave phenomenon. The defective and defect-free regions are visible in B-scan but we have to find the location and size of defect. We already performed by the amplitude detection and cross-correlation in previous articles.The question is : if amplitude correlation/comparison is not possible, what are the alternatives? The solution is related to the post-processing of experimental B-scan signal.
Hi Kumar,
would probably need to analyze frequency and beam width (if possible). Also consider the whole signal, without denoising, process by either STFT or sliding DFT.
Regards,
- Badges
- Science topic
- Similar topics
- Physical Sciences
- Materials Science
- Materials
- Composites