Dispersion curve estimation via a spatial covariance method with ultrasonic wavefield imaging

Area of Science:

• Materials Science
• Acoustics
• Non-Destructive Evaluation

Background:

• Lamb wave dispersion curve estimation is vital for structural health monitoring (SHM) and non-destructive evaluation (NDE).
• Existing methods require robust feature extraction from ultrasonic wavefield imaging (UWI) data.
• Accurate velocity measurements of different Lamb wave modes are essential for SHM applications.

Purpose of the Study:

• To develop and validate a new method for estimating Lamb wave dispersion curves using covariance matrices.
• To extract phase and group velocities of S0 and A0 modes from UWI data.
• To assess the efficacy of the proposed method for NDE/SHM.

Main Methods:

• Utilized a laser ultrasonic interrogation system to capture full-field ultrasonic signals in a 2-mm aluminum plate across five frequencies.
• Processed signals in the three-dimensional space-time domain to compute time-dependent covariance matrices.
• Constructed spatial covariance maps and variance maps to extract S0 and A0 mode properties and estimate velocities.

Main Results:

• The covariance matrix method successfully extracted features from UWI scans.
• Variance maps effectively identified S0 and A0 mode properties.
• Estimated phase and group velocities showed good agreement with theoretical dispersion curves for both modes.

Conclusions:

• The covariance matrix approach provides an effective means for Lamb wave dispersion curve estimation.
• This method enhances feature extraction from UWI, supporting advanced NDE/SHM strategies.
• The validated technique offers a reliable tool for assessing material integrity and detecting damage.