Avaliação da representatividade de modelos numéricos para análise do comportamento de pontes de espaguete

Abstract

The model uncertainties affect several steps for a structural reliability analysis, from the description of the loads to the system and the methodology by which the effects of the load in the system are evaluated. Extensive literature has been written hypothesizing on model uncertainties and their effects on the assessment of structural reliability. A recurrent problem in these studies is the lack of resources for the development of real-scale prototypes that adequately represent structural behavior. Based on these studies, this current work presents the design and construction of an arch bridge prototype, built using spaghetti and epoxy resin. The dimensioning of the structure was crafted considering the mechanical properties of the material available in the literature. At a reduced scale but with the same purpose, the effort was consistently maintained to test the constructed prototype and compare its structural response with a theoretical structure modeled three-dimensional in finite element software. Numerical modeling is performed considering the mechanical properties of the real structure sections tested with the goal to provide data as accurately as possible to the finite element software used for numerical simulation. By comparing the theoretical rigidities and the models tested, we aimed to verify if the numerical simulation was adequate in representing the behavior of the actual structure. In the end, suggestions are made regarding the methodology of bridge construction and the search for new numerical routines that can represent the structure more faithfully.