Seismic response analysis of Ganga sand deposits using shake table tests

Abstract

An accurate estimation of seismic behavior of a soil deposit is important from various aspects, namely, predicting the influence of the deposit in altering the earthquake waves propagating through that medium, quantifying the amplification or attenuation of the seismic shaking intensities, development of site-specific design spectra for seismic design of infrastructure to be built on that deposit, evaluating the possibility of ground failure and so on. The present study focuses on dynamic response analysis of local Ganga sand through performing a number of shake table experiments using a flexible laminar container. Uniform, dry Ganga sand of 30 % relative density exhibited acceleration amplification of 20, 40 and 68 % for input motions with 1, 2 and 5 Hz frequencies with amplitudes 0.2, 0.36 and 0.56 g, respectively. Further, analytical investigation on seismic response of the same soil deposit has been carried by assuming the behavior of the soil deposit as: (1) linear and (2) equivalent linear. The acceleration amplifications are estimated as 1, 4 and 28 % using the linear analysis and 3, 30 and 72 % using the equivalent linear analysis for the 1, 2 and 5 Hz motion, respectively. Compared to the linear analysis, the equivalent linear analysis predicts the surface amplification of the experiment with better accuracy (a maximum deviation being less than 6.7 %). This may due to the strain dependent behavior of the soil bed during strong shakings, which is compatible with the assumptions of the equivalent linear analysis procedure.