Identification of system parameters of multi-storey buildings with limited sensors

No Thumbnail Available
Journal Title
Journal ISSN
Volume Title
Evaluations of system parameters of civil engineering structures using system identification techniques are attracting considerable attention in recent years. Recorded dynamic responses induced in the structures due to ambient or earthquake excitations are utilized for identifications of system parameters of the structures. The potential of system identification procedures have been explored by number of researchers for identification of system properties of buildings. In this thesis, system identifications strategies have been adopted for identification of system parameters of two existing multi-storey shear buildings. The system identification strategies have been initially verified by identifying system parameters of laboratory test models. N4SID subspace identification technique has been employed for identification of modal parameters of a symmetric-plan nine storey shear building using on recorded acceleration histories. Similarly, identification of modal parameters of a four storey torsionally coupled shear building has been carried out using ITD method based on recorded acceleration histories. An iterative approach has been used for the determination of complete modal matrix incase of availability of only limited sensor data. Both the modal parameters as well as structural parameters identification techniques have been elaborated in the thesis. Further, detailed instrumentations adopted for acquiring acceleration histories from laboratory test models and existing buildings have also been described in this thesis. The verification through scaled laboratory models is done by exciting them on a uni-axial shake table. The acceleration responses are recorded from different floor levels and identification techniques are applied. The N4SID technique applied to symmetric plan shear building has been initially verified through tests on three numbers of two storey bare frame reinforced concrete 1/5th scale models. Variations of column dimensions in ground. storey of the test models have been deliberately introduced to verify the effectiveness of the identification scheme. Additional loads have been attached to each floor of the test models for gravity load simulation. Studies on laboratory test models with and without infill walls have been carried out separately. Identification of modal parameters of all the three test models have been carried out using N4SID algorithm based on acquired acceleration histories during shake table test. Three dimensional finite element models of all the scaled laboratory buildings have been developed using SAP 2000 Nonlinear (version 12). Modal analyses of all the numerical models have been carried out for the evaluation of natural frequencies. Computed modal frequencies of all the numerical models have been compared with the corresponding identified frequencies of the laboratory test models. The numerical models have been updated by applying appropriate adjustment in material properties so that the computed frequencies match with those of the identified natural frequencies of the test models. Thereafter, time history analyses have also been carried out using updated numerical model subjected to table acceleration histories. Floor acceleration histories from the numerical models have been compared with the corresponding recorded floor acceleration histories from the test models. The studies on all the three...
Supervisor: Anjan Dutta AND S. K. Deb