Estimation des demandes sismiques en deplacements elastiques et inelastiques au Quebec et au Canada - application aux ponts

The northeastern part of the American continent is located in the middle of a tectonic plate. Nevertheless, ground motions of significant magnitude occur frequently in this region. Being densely populated, the area is associated with considerable seismic risk. There is very little historical seismic data corresponding to the Eastern North American seismic hazard and the underlying earthquake mechanisms remain unclear. Nevertheless, the infrastructures are constantly built and renovated considering no harm from the seismic risk. The main objective of the research presented in this thesis is to estimate elastic and inelastic seismic displacement demands in Quebec and Canada.;First, a literature review presented the objectives of research in earthquake engineering in Canada and the measures prescribed in the current Canadian Highway Bridge Code (CSA-S6- 06). The study of two different displacement-based design methods, permitted identification of the ingredients depending on the applied ground motion records, to be evaluated in the seismic context of eastern Canada. The literature review showed that several codes in the world, except the Canadian Code, set a limit for residual displacement.;A performed research led to characterization of elastic displacement demands of earthquakes in eastern and western Canada. The historical records in eastern Canada were filtered and corrected. The elastic displacement spectra of these ground motions were generated using SeismoSignal (version 5.0.0) and RSpec (Bouaanani 2005). It was found that the corrections to accelerograms had very little influence on the displacements when working with fundamental periods of vibration ranging between 0.02 and 10 s. Although this analysis was performed to highlight certain trends, given the low number of historical earthquakes, we naturally made a comprehensive study of simulated earthquakes. The database used is the set of accelerograms generated using the ground motion prediction equations developed by Atkinson (2009). The elastic displacement spectra from this database were generated for different levels of viscous damping ranging from 5 to 30%. To be analyzed and characterized, these spectra were grouped by families.;The development of elastic displacement spectra has many usages. One of them is to develop coefficients to directly obtain the spectrum of elastic displacement or acceleration of a certain damping level from the elastic displacement or acceleration spectrum generated for 5% damping. This ratio is commonly called damping adjustment factor. Some equations exist in the literature, which is why it seemed interesting to verify them against simulated earthquakes studied. The response spectra of elastic displacements were calculated for damping levels of 10, 15, 20, 25 and 30% using these equations and then compared to spectra developed using RSpec software (Bouaanani 2005). Given the large margin of error, more than 30% on average, a new formulation has been developed significantly reducing the error.;A study of displacement spectra concerning maximum inelastic displacements and residual displacements was performed. These spectra were generated using the database of simulated earthquakes. The study of inelastic displacements was first performed on the basis of the initial period and next considering the effective period. The effect of the reduction coefficient of seismic force was evaluated. The results showed that it plays a very important role in the maximum inelastic displacement. The demand for residual displacements is about three times smaller than that of maximum inelastic displacements. However, the residual displacements are not of lower significance and having them under control seems to be important in the seismic context of eastern Canada. A study of the sensitivity of the residual displacement to the reduction factor shows that this factor plays a much more important role for structures subjected to earthquakes in western Canada than those subject to earthquakes in the east.;A final chapter is devoted to applications to bridges. A first study focused on the damping reduction factor. A bridge was studied using the direct displacement-based design method (Priestley 1996) considering the existing damping reduction factors and the developed equation. As a reference, the method recommended by the Canadian bridge code (CSA-S6-06) was also addressed. Another displacement-based design method was also used for the analysis of the bridge: The Yield Point Spectrum method. This method requires an important ingredient, the R -- mu -- T coefficient. The coefficients introduced by Miranda and Bertero (1994) and Nassar and Krawlinkler (1991) have simply been used. A second study is devoted to the study of inelastic displacement and maximum residual displacement of a bridge modeled by a bilinear system with viscous damper. It was found that the viscous damper significantly reduces maximum inelastic displacements as well as residual displacements.