Numerical Modeling of RC Bridges for Seismic Risk Analysis

Introduction

It is during the occurrence of earthquakes that deficiencies, which cause a bad behavior of structures, has been highlighted and that can be drawn lessons with regard to design aspects that allow to provide it with a good seismic behavior.

Reports of seismic effects on reinforced concrete bridges, especially those built according to the codes of practice of the seventies, show that a lot of them behave poorly and have an inadequate safety assessment. Some efforts, therefore, must be made to accurately define bridge assessment and hence evaluate the need for reliable solutions to improve safety levels and behavior during seismic activity in the future.

In recent earthquakes it has been demonstrated that bridges and viaducts are structures that suffer extensive damage. Even in “moderate magnitude” earthquakes, the consequences in these structures have been very important, on many occasions causing their partial destruction or even total collapse, with correspondingly heavy costs. When compared, these consequences have been greater than those observed in current structures. Some examples of these earthquakes in the United States are the Loma Prieta (Soberón et al., 1999; Priestley et al., 1996) and the Northridge (Oliveira et al., 1995), in Japan the Kobe earthquake (Legeron, 2000) and in China the Sichuan earthquake (ISSN, 2008), with a magnitude of 7.9 on the Richter scale and more than seventy thousand deaths.

The evaluation of seismic vulnerability of bridges, considering their nonlinear behavior (a necessary consideration when a realistic analysis of the seismic response is intended), involves enormous computing resources and corresponding calculation efforts. On the other hand, the development of research studies in this field has enabled a better understanding of their behavior, and the availability of numerical and experimental results allows the possibility of the calibration of simplified procedures.

However very high refined models could be adopted, involving a 3D space modeling and the spread of the plasticity along the member length as well as across the section area, in this work a substantially more simplified model (without losing accuracy) is also explored. This simplicity is evidenced by a much lower number of parameters involved and an enormous lesser computing time – about 50 times less when compared with a more refined model. Therefore, the adoption of such simplified model shows the great advantage, from the practical usage point of view, when a large number of analysis is needed, as is the case of vulnerability analysis.