Structural Parameters Identification Using FRF of Incomplete Strain Data

Abstract : Structural health monitoring by estimation of structural parameters changes such as stiffness and mass parameters by monitoring dynamic characteristics has attracted much attention in recent decades. Generally used dynamic characteristics to locate and quantify structural damages are natural frequencies, mode shapes, mode shape curvature, modal strain energy, frequency response function (FRFs) and so forth. Using FRF data has the advantage of avoiding modal analysis errors included due to indirectly extraction from the measured FRF data. Moreover, past studies showed that strains are more sensitive to localized damage compare to displacement. So, in this study FRF of strain data are utilized to identify unknown structural parameters using a sensitivity-based model updating approach. In this paper a quasi-linear sensitivity equation which diminishes the adverse effects of incompleteness of FRFs data is proposed for model updating. The efficiency of the proposed method is validated through a numerical 2D-frame example using FRF of strain data considering the side effects of noise and incompleteness of measurements. The results indicate that this method can locate and quantify the severity of damage precisely.