|
Abstract
|
This paper addresses the shortcomings of studies due to the effects of temperature changes and random noise contamination of modal data in the structural damage detection process. In this study, three structures have been modeled using the finite element method in MATLAB. The first two structures are an experimental steel beam and an experimental 3D Five-story steel frame. Because the effects of temperature variations on the modal data of concrete structures are more significant than those of steel structures, the third structure is a real Z24 concrete bridge. To formulate an optimization problem, a sensitivity analysis was done on the effects of damage, temperature variations, and noise on frequencies and the modal assurance criterion. Sensitivity analysis results reveal that in concrete structures, the frequencies are more affected by temperature variations, and the modal assurance criterion is a more reliable variable to formulate an objective function. So, an appropriate objective function has been formulated according to sensitivity results, which has a weight adjustment parameter to adjust the influence of the modal parameters used in the structural damage detection process. To investigate the performance of the proposed method, ten different scenarios consisting of only damages in structures, damages under temperature variations, and damages under temperature variations and noisy modal data were applied to the studied structures. Finally, the structural damage detection process was performed using the Marine Predator Algorithm, and the effectiveness of the present study was proven.
|