Papers by Keyword: Frequency Sensitivities

Abstract: Damage detection using changes in global dynamic characteristics has been a hot research topic in recent years. In the present paper, based on natural frequencies and mode shapes, a numerical tudy locating and assessing damage in a circular cylindrical shell is presented. The axial position of the damage can be easily obtained by comparing the theoretical and measured frequency changes due to damage. Additional information of mode shapes is introduced to locate the exact damage position. And the damage extent can be estimated by the first order approximation method. The feasibility and practicality of the damage detection scheme are evaluated for several damage scenarios by locating and sizing damage in free-clamped shells. The results illustrate that the proposed detection scheme can confidently locate the single or multiple positions of damage. It is also observed that damage extent can be estimated with a relatively small error.

Abstract: In the present paper, the Damage Location Assurance Criterion (DLAC) is extended to locate and assess damage in a circular cylindrical shell based on natural frequencies and mode shapes. Frequency sensitivities computed from a defect-free finite element model are applied to calculate the theoretical frequency changes. The axial position of the damage can be easily obtained by comparing the theoretical and measured frequency changes due to damage. For the shell is axisymmetric, additional information of mode shapes is introduced to locate the exact damage position. The damage extent can be estimated by the first order approximation method. The feasibility and practicality of the damage detection scheme are evaluated for several damage scenarios by locating and sizing damage in the free–free, simply-supported and free-clamped shells, respectively. Results from simulation examples show that the proposed detection scheme can confidently locate the single or multiple positions of damage. It is also observed that damage extent can be estimated with a relatively small error.