Signal detection and noise suppression using a wavelet transform signal processor: application to ultrasonic flaw...

Area of Science:

• Nondestructive Testing
• Ultrasonic Signal Processing
• Applied Mathematics

Background:

• Signal processing is crucial for analyzing ultrasonic nondestructive testing (NDT) data.
• Techniques like signal averaging and frequency spectrum analysis are commonly used.
• The Wavelet Transform (WT) offers advanced capabilities for signals with time-varying spectra.

Purpose of the Study:

• To utilize the Wavelet Transform (WT) for improved ultrasonic flaw detection in noisy signals.
• To present WT as an alternative to the Split-Spectrum Processing (SSP) technique.
• To demonstrate the effectiveness of WT for analyzing ultrasonic pulses.

Main Methods:

• Applied the Wavelet Transform (WT) to analyze ultrasonic signals.
• Compared WT's performance against the Split-Spectrum Processing (SSP) technique.
• Utilized a filter bank with a self-adjusting window structure inherent to WT.

Main Results:

• The Wavelet Transform (WT) demonstrated effective flaw echo detection in noisy ultrasonic signals.
• Successful detection was achieved even at low signal-to-noise ratios (SNR) of -15 dB.
• Experimental verification using steel samples with simulated flaws confirmed the improved detection capabilities.

Conclusions:

• The Wavelet Transform (WT) is a powerful tool for enhancing ultrasonic flaw detection.
• WT provides superior performance in noisy environments compared to traditional methods.
• The self-adjusting resolution of WT is highly beneficial for identifying flaws in NDT.