Análise do comportamento de arcos e abóbadas através de modelos físicos e computacionais

José Maria Salema Garção Ferreira Martins

Key information

Authors:

José Maria Salema Garção Ferreira Martins (José Maria Salema Garção Ferreira Martins)

Supervisors:

António Manuel Candeias de Sousa Gago (António Manuel Candeias de Sousa Gago)

Published in

July 3, 2025

Abstract

The present dissertation is dedicated to the study of the structural behaviour of masonry arches, with special focus on the vaults of the Igreja Matriz de Bucelas. The main objective is to understand the mechanisms that ensure the stability of these historical structures, using an approach that combines theory, computational modelling, and physical experimentation. The work began with a historical and theoretical analysis of arches, employing the concept of the line of thrust, which is fundamental to understanding the behaviour of arched structures. Subsequently, the application of different analysis methods was explored: the graphical method, using the funicular polygon; the experimental method with physical models, printed in 3D; and the numerical method through the Discrete Element Method (DEM), using the UDEC software. The results of these methods were compared to assess their respective capacities to predict the structural behaviour of arches, particularly collapse mechanisms and the formation of hinges. The case study of the vaults of the Igreja Matriz de Bucelas allowed the validation of the methods and the identification of critical structural vulnerabilities. Some reinforcement solutions to address existing damage were also explored. The tests carried out showed that physical models provide immediate visual perception but have quantitative limitations. On the other hand, DEM offers precise quantitative predictions, although it requires rigorous data and good calibration. The comparison between the methods revealed that combining both is a robust approach, as practical tests validate and enrich the interpretation of simulations, while numerical analyses deepen understanding and can provide quantitative data.