Browsing by Author "Estay Espinoza, Monserratt Alejandra"

Now showing 1 - 1 of 1
  • Thumbnail Image
    Item
    Calibración de un modelo numérico para un edificio no convencional de hormigon armado.
    (Universidad de Concepción, 2024) Estay Espinoza, Monserratt Alejandra; Silva Muñoz, Rodrigo Antonio; Opazo Vega, Alexander; Aguilar Vidal, Víctor Hugo
    The development of numerical methods and monitoring tools has encouraged the study of the behavior of existing structures, leading to the generation of updated digital models that reduces differences between observed and analytical behavior. This, together with the interest in their application in medium-rise buildings of architectural value, is the main motivation for studying the Edificio Innovación of the Universidad del Bio-Bio. For these reasons, the main objective of this project was to calibrate its numerical model based on experimental measurements. To achieve this objective, the building was modeled using the ETABS software. The model error was calculated based on the differences between numerical and experimental frequencies, resulting in 12% error for the preliminary baseline model. From the model, 18 potentially significant variables were identified for its dynamic behavior. Among these, twelve were modules of elasticity (stiffness) and six were loads distributed on each floor (masses). Then, a sensitivity analysis was applied, using the quoFEM software, employing the LHS sampling method. A total of 100 iterations per variable were performed, and from this analysis, three variables were identified as significant using Sobol’s Indices. Two of the variables pertained to the modules of elasticity of walls and columns on the first and last floors, while the last one corresponded to the distributed load applied on the fifth floor. Regarding these results, the influence of the elastic modulus of the first floor is expected because these represent a fundamental part in the transmission of loads from upper floors to the base. As for the load applied on its last floor, its influence can be explained due to the displacements relative to the building’s base increasing in height , and the mass at this level may amplify torsional moments, affecting the behavior of the building. With these three variables, the building calibration was conducted, also using the quoFEM program, along with the NL2SOL optimization method. The calibration produced different optimal solutions (local optima), which could be due to a limited amount of experimental data with which to discriminate solutions. To overcome this problem, it was decided to calibrate the model with Boxplot Diagrams, resulting in an 5% error with respect to experimental measurements.
Síguenos en...