|M.Sc Student||Safe Kamal|
|Subject||Structural Damage Identification for Plates|
|Department||Department of Civil and Environmental Engineering||Supervisor||Professor Yiska Goldfeld|
|Full Thesis text - in Hebrew|
The current study deals with damage identification and quantification procedures in a rectangular plate, via the non destructive global dynamic testing approach. The study opens with an introduction that presents the motivation for damage identification of rectangular plates. Then a literature survey of the exiting knowledge in this field is presented. Unlike the extensive literature available on beams, very few research reports can be cited for plates. Generally, they are divided into the following two main distinct approaches: the “model- based” and the “response based”.
The identification procedure is divided into two main steps: at the first step the damaged elements are identified, and then their severity evaluated. The inverse problem algorithm is based on an FE model and a subset of measured vibration frequencies and vibration modes from the damaged plate.
The first step uses the direct stiffness evaluation method; it calculates directly the dynamic stiffness distribution of the plate, based on the fact that the curvature mode shape is more sensitive to local damage than that of the displacement. Theoretically, the stiffness distribution could be directly evaluated only by the first step, however since some inaccuracies occurred, the first step succeed in locating only the damaged elements with a high level of reliability and not their severity.
The second step evaluates the severity of the damaged elements using an optimization procedure which minimizes the error of the obtainable pseudo static equation (used in the first step). In order to assure global minimum, a genetic algorithm is chosen using the Differential Evolution algorithm.
The current procedure is verified using both analytical and distorted mode shapes. Parametric study is presented for various cases of different: boundary conditions; number of damaged element, location and damage severity by using analytical and distorted modes.
For the analytical mode shapes, the parametric study shows that the proposed method can identify, with a high level of reliability, the damaged locations for all tested examples. The severity of the damage was also identified accurately.
For the distorted mode shapes, the first step identifies most of the damaged elements. Estimation of the damage severity by the second step is higher than the real value. In all cases the damaged zone is clearly identified.