Health monitoring of stiffened plates structures based on modified Lamb wave energy
摘要
Structural Health Monitoring (SHM) of stiffened plates is challenged by wave scattering and mode conversion caused by stiffeners, which often produce false damage indications in reference-free algorithms. This study presents a semi-reference-based method that integrates Continuous Wavelet Transform (CWT), probabilistic imaging, and a hybrid optimization scheme to extract correction coefficients that compensates stiffener-induced energy distortion. An array of eight piezoelectric transducers is modeled on a defect-free stiffened aluminum plate in ABAQUS, and guided-wave responses are processed in MATLAB to derive optimal correction coefficients using a combined genetic algorithm and fmincon refinement. These coefficients are stored and reused during in-service monitoring to distinguish geometric effects from real cracks. Six crack cases including five single-crack and one dual-crack case were simulated to evaluate the method. The proposed framework eliminated false detections along stiffener paths and improved localization accuracy compared with the conventional instantaneous baseline method. Across all cases, the mean localization error was 3.65 mm, with a maximum of 12.5 mm in worst case and successful separation of simultaneous cracks. The results demonstrate that the proposed semi-reference-based method provides reliable and accurate crack localization in stiffened plates, offering a practical path toward SHM of geometrically non-uniform structures.