Stochastic 1d site response analysis at the archaeological site of kancheepuram

Student: Heidy Sanchez Lizarraga
Supervisors: Dr Carlo Giovanni Lai


Historical buildings, like Ekambaareswara at Kancheepuram, constitute and essential component of the cultural heritage of a country. And identification of vulnerability has to be done in order to develop seismic risk mitigation techniques for these buildings. To do so, a proper assessment of the seismic input and possible amplifications of the ground motion due to localized lithographic or geo morphological characteristic of the soil at the site of the building is needed. In current engineering practice, site response analysis is performed in a deterministic way. However, the execution of a deterministic analysis of the seismic amplification does not represent properly the response of a site in which the uncertainties of the parameters used to develop the subsoil model are high. Moreover, the quality and completeness of the data used to construct the subsoil model strongly affects the reliability of the results. On the other hand, the uncertainty and reliability of the results can be determined through stochastic procedures that allow the validation of the sensitivity of the analysis to the variability of the input parameter and seismic input. Therefore, for this study, a stochastic approach was taken in order to consider the uncertainties when performing site response analysis.

First of all, a research investigation has been undertaken to determine the seismic hazard at the site, this was done in a different study [Ornthammarath et al. 2008], performed as part of the Indo-Italian research project currently underway between the University of Pavia, Italy and the Indian Institute of Technology Madras, funded by the Italian Ministry of foreign Affairs and the Indian Department of Science and Technology, to carry out Probabilistic Seismic Hazard Assessment (PSHA) of the urban nucleus and monumental structures at Kancheepuram in the Tamil Nadu State of South India. The result of the PSHA will be used in this study to determine the seismic input at the site. Additionally, a procedure to select spectra compatible input records has been used.

After integrating the results of the geotechnical investigation performed at the site, the soil deposit was characterized by a 1D model. The uncertainties associated to the subsoil model were taken into account using Monte Carlo technique. The input parameters as well as the input motion were defined by appropriate probabilistic distributions, from which each simulation was generated. Site response was assessed using the software SHAKE91, which was ran for every simulation. The moderate sesimicity of the site does not justify the adoption of more advance constitutive models. The output of these analyses was then evaluated statistically and an acceleration mean response spectrum was computed as final result.

Furthermore, a set of 7 records to be used for future structural/geotechnical dynamic analysis was selected by imposing the compatibility with the mean spectra obtained from the stochastic analysis and by using the results of the simulations as data base. Finally, the dynamic impedances that characterize the soil are computed and given as a final output of the study, to be used in soil structure interaction analysis.

You may download a digital version of this MSc dissertation here.