Non-iterative eigenstructure assignment based finite element model updating of a Mindlin–Reissner plate in Duncan form of state space using ambient vibration respnose

International Conference on Theoretical, Applied, Computational and Experimental Mechanics

Subhamoy Sen, & Baidurya Bhattacharya

2014-12-31

Eigenstructure assignment (ESA) based model updating is a control based technique for systematic calibration of finite element models using measured response from real structure. Application of this technique in physical space restricts simultaneous updating of stiffness and damping matrices of any mechanical system. On the other hand ESA when used in state space domain demands state space eigenstructure to be identified which is a challenging job. It is not certain that the identified state space eigenstructure will be in the same order and orientation as desired by the ESA algorithm. In this paper we used Duncan form of state space model of the mechanical system so that assignable eigenstructure in this form can be easily constructed using modal properties of the system in its physical space and thus problems regarding orientation is avoided. To achieve compatibility between assignable state space eigenstructure and state space model the later has been reduced using structural equivalent reduction expansion program (SEREP). Assignable eigenstructure is then used along with ESA algorithm given by B.C. Moore to update the reduced primary model of the mechanical system to simultaneously update the stiffness and damping matrices. Proposed algorithm is tested on a Plate modeled using Mindlin-Reissner plate element and updated model demonstrated a good agreement with the desired result.