Structural damage identification in plates using spectral strain energy analysis

In this paper, a vibration response parameter known as spectral strain energy (SSE) is proposed for structural damage identification in plate-like structures in the context of a non-model-based damage identification approach. The SSE presented in this study is derived from moment–curvature response in which all the modal parameters (namely natural frequency, mode shapes and modal damping) are taken into account. First, displacement response power spectral density obtained from a plate element is used to determine moment power spectral density (MSD) and curvature power spectral density (CSD). A statistical parameter known as mean-square value (MSV) of the response spectral density is then obtained and used to derive SSE based on the assumption that the excitation force is a stationary, ergodic random noise. Consequently, various damage indices are computed and employed for damage detection and localization in simply supported plate and beam elements. The level of sensitivity to damage and performance of the SSE method is illustrated using extensive numerical simulation studies and by comparing the results with those obtained from the modal strain energy (MSE) method. Moreover, the performance and robustness of the SSE method is verified using the experimental modal data obtained from a full-scale bridge structure.