Numerical modeling of 3D site-city effects including partially embedded buildings using spectral element methods. Application to the case of Viña del Mar city, Chile

Abstract

In recent years, seismic wave propagation analyses have become a powerful tool to evaluate the site effects in a given region. Among several approaches, Spectral Element Method (SEM) has been widely used with that purpose because its flexibility and computational efficiency. In addition to other effects than basin shape, material nonlinearity and heterogeneity, the multiple interactions between the soil and structures, denominated site-city effects (SCI), can play a crucial role in densely populated areas. There are many options to model this kind of interaction, especially if the buildings are partially embedded on the soil. This paper evaluates the importance of the proper SCI modeling against more standard uncoupled approaches, focusing on the local interaction between the soil and a group of buildings including inelastic soil behavior. We focus our work on the case of downtown Viña del Mar, a touristic coastal city of central Chile, where the observation of a reiterated distribution of damage in reinforced concrete buildings during two major earthquakes has motivated numerous studies. For that purpose, a realistic 3D numerical model of the area is created, considering the existing buildings and using micro-vibration as a main calibration tool. The open-source code SPEED was used to perform the wave propagation simulation, which combines the spectral element method with a discontinuous Galerkin approach. A geophysical study was conducted to estimate the model parameters, shear modulus degradation and damping curves are extracted from laboratory tests to account for the non-linearity of the soil. In general, the results indicate that the inclusion of the SCI is beneficial to the structure’s response in most cases, and that SCI modeling needs to considerate the level of embedment to obtain more precise results. Indeed, in buildings of 12 or more stories, the response would not be affected by the level of embedding of the base and the inclusion of site-city effects is beneficial, while for buildings lower than 5 stories, the total embedment of the base generates a significant decrease of the response.