Laser Ultrasound Inspection Based on Wavelet Transform and Data Clustering for Defect Estimation in Metallic Samples

Area of Science:

• Materials Science
• Non-Destructive Testing
• Signal Processing

Background:

• Classical ultrasonic methods are standard for non-destructive testing (NDT).
• Laser-generated ultrasound (LGU) offers a modern alternative for industrial maintenance and quality control.
• Accurate detection and characterization of internal defects are crucial for material integrity.

Purpose of the Study:

• To detect and reconstruct internal defects in metallic samples using LGU.
• To evaluate the efficacy of wavelet transform and clustering for defect analysis.
• To establish a reference-free NDT method for industrial applications.

Main Methods:

• Laser-induced thermal mechanism generated ultrasonic waves.
• Wavelet transform was employed for time-frequency analysis of ultrasonic signals.
• Density-based spatial clustering was applied to time-frequency maps to identify defect echoes.
• Clustering algorithms processed geometric solutions for defect outline reconstruction.

Main Results:

• Successful detection and sizing of internal defects in a metallic sample.
• Wavelet transform provided multi-resolution time-frequency characteristics for signal analysis.
• Clustering effectively isolated defect echoes and determined defect positions.
• A two-dimensional projection of the defect outline was obtained.

Conclusions:

• The proposed wavelet transform ultrasonic imaging method is effective for defect detection and sizing.
• The technique operates without requiring reference information, enhancing its industrial applicability.
• LGU combined with advanced signal processing offers a valuable NDT solution for quality control.