Artículo de revista
Operational modal analysis and FE model updating of the Metropolitan Cathedral of Santiago, Chile
Fecha
2017Registro en:
Engineering Structures 143 (2017) 169–188
18737323
01410296
10.1016/j.engstruct.2017.04.008
Autor
Torres, Wilson
Almazán, José Luis
Sandoval, Cristián
Boroschek Krauskopf, Rubén
Institución
Resumen
Heritage buildings in Latin American countries possess high architectural value. Studying these constructions under extreme loads, particularly earthquakes, requires representative models for simulating expected response. At present, the non-invasive Operational Modal Analysis (OMA) tests offer interesting possibilities for obtaining modal parameters to update and validate a structural model for this type of structure. In this context, this article focuses on the calibration and adjustment process for a finite element model of the Metropolitan Cathedral of Santiago Chile, based on experimentally identified modal and mechanical material properties. Accordingly, an in situ experimental campaign, aimed at obtaining the response of the structure due to ambient vibrations is presented and discussed. Six high-sensitivity synchronous triaxial accelerometers were employed in this campaign. Enhanced Frequency Domain Decomposition (EFDD) and Stochastic Subspace Identification (SSI), system identification methods, were applied. Mechanical tests were performed on the Cathedral's stone blocks. The experimental data and derived modal properties were used to generate and update a finite element model. Several considerations were made in the model updating process: the most relevant was the homogeneous treatment of the stone masonry with their mortar interface, and the boundary elements restraining effect caused by adjacent structures. A preliminary model updating process was applied to define the boundary conditions and initial material properties. This optimization was based on minimizing an error function given by the difference between the experimental and analytical frequencies. A second step was then applied, in which models with different material properties were evaluated within a physically possible range. The final model selection was based on the distance between the experimental and analytical frequencies, and the mode shapes. The updated model allows an assessment to be made of the structure behavior in its current condition and models to be prepared for a wide range of possible future research scenarios.