Vibration-Based Structural Damage Identification and Evaluation for Cylindrical Shells Using Modified Transfer Entropy Theory

Abstract

In this paper, modified transfer entropy theory is combined with a surrogate data algorithm to produce a new method in order to identify nonlinearity in the vibration data of a damaged cylindrical shell. The proposed identification method can eliminate the necessity of acquiring baseline statistics by comparing the transfer entropy of original vibration data and that of surrogate data. Moreover, a new index ξ is established to reflect the degree of nonlinearity by quantifying the discreteness of the entropy of each group of surrogate data. Vibration tests are conducted and experimental data are analyzed to confirm the effectiveness of this method. Then, a semi-analytical method based on a Galerkin method and the classic shell theory is used to precisely predict the linear and nonlinear vibration response of a cylindrical shell under different damage circumstances. The corresponding results show that the proposed method can not only identify the structural damage but also be further applied to the evaluation of such damage for cylindrical shells. In addition, the influence of different load pressures and degrees of damage on the effectiveness of the identification method is analyzed and discussed. As verified, the proposed methodology can be potentially used for structural damage identification and evaluation in areas such as civil engineering, mechanical engineering, and ocean engineering.