Bibliographic citations
Oroz, K., Zapaille, N. (2024). Detección de daños estructurales utilizando un modelo numérico de una viga de concreto armado mediante la metodología de algoritmos de Machine Learning (ML) [Trabajo de Suficiencia Profesional, Universidad Peruana de Ciencias Aplicadas (UPC)]. http://hdl.handle.net/10757/674705
Oroz, K., Zapaille, N. Detección de daños estructurales utilizando un modelo numérico de una viga de concreto armado mediante la metodología de algoritmos de Machine Learning (ML) [Trabajo de Suficiencia Profesional]. PE: Universidad Peruana de Ciencias Aplicadas (UPC); 2024. http://hdl.handle.net/10757/674705
@misc{renati/419447,
title = "Detección de daños estructurales utilizando un modelo numérico de una viga de concreto armado mediante la metodología de algoritmos de Machine Learning (ML)",
author = "Zapaille Neyra, Nevin Alfonso",
publisher = "Universidad Peruana de Ciencias Aplicadas (UPC)",
year = "2024"
}
Currently, the importance of knowing the structural health of a building is very relevant since with this knowledge it is possible to detect the problems that said building has and take corrective measures. Taking into account that there are external and internal factors that affect structural health, in the first aspect earthquakes clearly exemplify seismic events and the consequences that these lead to a building in poor condition, as an internal factor there are cracks caused by the different failures that affect a reinforced concrete structure. Therefore, in this research, a system is developed for the detection of structural damage (flexing and shear cracks) through vibration analysis and artificial intelligence algorithms in an embedded rectangular reinforced concrete beam. Using the Finite Element Method (FEM), the vibration frequencies of the beam are determined, and different damage states are simulated in SolidWorks. The frequencies obtained are analyzed with Machine Learning (ML) algorithms, specifically Multilayer Perceptron (MLP), to identify structural damage (cracks). The results indicate that the reduction in stiffness in the beam, simulated through cracks and removal of finite elements, causes a decrease in vibration frequencies. This confirms that the MLP is effective in detecting frequency variations and, therefore, structural damage in a numerical model of a doubly embedded rectangular beam. Furthermore, a direct relationship is established between the magnitude of the damage and the variation in vibration frequencies.
This item is licensed under a Creative Commons License