DYNAMIC SITE-CITY INTERACTION ANALYSIS OF STRUCTURES IN URBAN ENVIRONMENT

Abstract

The Site-City-Interaction (SCI) is a complex phenomenon in which closely spaced structures interact with each other as well as with the underlying basin and finally alter the seismic response of both the free field motion and structures of the city (Clouteau and Aubry, 2001; Kham et al., 2006; Semblat et al., 2008; Lu et al., 2018). The anomalous pattern of structural damages in the Mexico City during 1985 Michoacán earthquake indicated that the complex structure-soil-structure interaction and the resonance of structures with the underlying soft sediments of lake-bed are responsible for the complete collapses of structures and long duration of vibrations (Wirgin and Bard, 1996; Chavez-Garcia and Cardenas-Soto, 2002). In this thesis, the SCI effects on the response of structures under double resonance condition considering city situated in the 2D and 3D basins with shape-ratio greater than 0.2 (response of such basins are dominated by 2D/3D resonance phenomenon) have been studied. Thesis addresses a problem which is common in most of the metro cities to construct infrastructure on filled dried ponds and land depressions, which behave as 2D/3D small basins. The role of parameters of structures of the city in insulation of the Rayleigh wave, which is caused by the metastructure behavior of the structures of the city falling in the path of Rayleigh waves, is also presented in this thesis. The fourth-order accurate staggered-grid viscoelastic 2D and 3D finite-difference (FD) algorithms developed by Narayan and Kumar (2013, 2014) and Narayan and Sahar (2014) are used for simulating the seismic responses of the various considered 2D and 3D site-city models. In order to evaluate the effects of SCI under double resonance condition, one need to comprehend the fundamental frequencies of both basin and structure beforehand. Therefore, empirical relationships are developed to predict the fundamental frequency of both the basins and structures based on their shape (elliptical, rectangular and trapezoidal), shape-ratio and polarization of the incident S-wave. The time domain responses of the 2D and 3D basins revealed a drastic increase in amplitude and duration due to the multiples of S-wave and back and forth horizontal propagation of the basin generated surface/diffracted waves (Bard and Bouchon, 1985; Kumar and Narayan, 2018). The simulated results revealed an increase of fundamental frequency of basins as well as spectral amplification factor (SAF) at fundamental frequency with an increase of shape-ratio of the 2D and 3D basins. It is inferred that the lateral and vertical resonances are the dominant phenomenon in the case of basins of large-shape ratio, which have resulted in an increase of SAF in the cases of SH- , SV- and 3D S-wave responses ii of the rectangular basin of shape-ratio 0.67 up to 2.0, 2.89 and 5.0 times of that in the 1D basin, respectively. A correlation between the spatial variations of spectral amplifications at different modes of vibrations of 3D basin along EW- and NS-arrays with the SH- and SV-wave responses, respectively of the equivalent 2D basin is observed. However, the obtained maximum SAF at fundamental mode in the 3D basin is around 2 and 3 times larger than that obtained in the SH- and SV-wave responses at the center basin, respectively. The obtained fundamental frequency of the elliptical basin was largest and least in the rectangular basin. Further, S-wave responses of standalone 2D and 3D homogenous structural block on rock are simulated to obtain the dynamic response of structures considering different size and shape ratios since such homogenous structural blocks are going to be used in place of real structures in the SCI studies. There is reduction of fundamental frequency of standalone structure on rock as compared to that predicted using the well know relation F0=Vs/4H for the SV-wave only. In order to quantify the SCI effects on the response of structures under double resonance condition, the seismic responses of the two site-city models with 9 and 25 structures placed on the 3D closed basin are simulated with a NS-polarization of the S-wave and the results obtained along the NS- and EW-arrays are compared with the results of SH- and SV-waves responses of the considered cities along the EW- and NS-arrays, respectively. Analysis of the results depicts that there is an increase in amplitude and duration of the structural response when placed on the 1D, 2D and 3D basins under double resonance condition, as compared to the standalone structure on rock. The spectral amplification factors (SAF) at the top of 2D and 3D standalone structures situated on 1D basin is 390% and 580% larger than that of standalone structure on rock and it increased to 594% and 950% when placed on closed 2D and 3D basins, respectively. A drastic decrease of amplitude and duration of motion at the top of structures of the city due to the SCI effects as compared to the standalone structure situated at the centre of basin is inferred. The SCI effects have resulted in the splitting of bandwidth of fundamental mode of vibration of structures as well as very large reduction of SAF at that frequency. A reduction of about 7% in the fundamental frequency of structure of a city of 25-structures placed on soft soil is also observed due to the presence of nearby structures. The achieved reduction in SAFs at fundamental frequency of the structure of the 2D city (5-structures) and 3D city (25-structures) due to the SCI effects is of the order of 40% and 63%, respectively. This finding presents the importance of 3D SCI studies in an urban environment and need for further research in this field for the cost effective earthquake engineering. An endeavor is also made to infer whether the structures falling in the path of Rayleigh waves may behave as a seismic metastructure. The insulating behavior of structures on the Rayleigh iii wave characteristics is quantified considering the factors like parameters and number of structures in the city, location of city between epicenter and the recording station and city-heterogeneity. The obtained drops in the spectral-ratios for both the components of Rayleigh wave are matching with the flexural and longitudinal modes of vibrations of the structures of the city. The insulation caused by a city with 30 structures has resulted in the reduction of the spectral-ratio up to 28% and 31% in the horizontal and vertical components of the Rayleigh wave, respectively. It can be inferred that the high-rise buildings falling ahead to low and medium height buildings may reduce the impact of Rayleigh wave on them. It is recommended to the town planners and earthquake engineers to consider the insulating effects of structures (falling in the path) on the Rayleigh wave characteristics during estimating the seismic hazard for the development of an urban environment in a tectonic setup causing shallow earthquakes