Imaging damage in plate waveguides using frequency-domain multiple signal classification (F-MUSIC)

Area of Science:

• Structural Health Monitoring
• Non-Destructive Evaluation
• Wavefield Modeling

Background:

• Conventional MUSIC algorithms are limited to linear arrays and monochromatic excitation, hindering in situ health monitoring with sparse sensor networks.
• The ameliorated MUSIC (Am-MUSIC) algorithm improved upon conventional methods but incurred high computational costs and was restricted to single frequencies.
• Existing methods often ignore scattered signal features outside the excitation frequency band, limiting comprehensive damage assessment.

Purpose of the Study:

• To develop a computationally efficient and broadband-capable algorithm for structural health monitoring.
• To overcome the limitations of pixel-based calculations and monochromatic excitation in previous MUSIC algorithms.
• To enhance the precision of damage imaging in structures using sparse sensor networks.

Main Methods:

• A multiple-damage-scattered wavefield model was developed for signal representation in the frequency domain, creating the frequency-domain MUSIC (F-MUSIC) algorithm.
• F-MUSIC quantifies orthogonal attributes between signal and noise subspaces to generate a spatial spectrum for damage visualization.
• The algorithm was validated using simulations and experiments on an isotropic plate waveguide with a sparse sensor network.

Main Results:

• F-MUSIC successfully imaged single and multiple damage sites with high precision, even with a highly sparse sensor network.
• The frequency-domain approach enabled the fusion of broadband scattered information, enhancing imaging accuracy.
• Computational costs were significantly reduced compared to pixel-based methods.

Conclusions:

• F-MUSIC offers a robust and efficient solution for structural health monitoring, overcoming key limitations of previous algorithms.
• The algorithm's effectiveness is not compromised by the number of damage sites or the sparsity of the sensor network.
• F-MUSIC provides enhanced imaging precision and broad applicability for non-destructive evaluation and in situ monitoring.