Comparative Study of 1D Codes for Site Response Analyses

During the last decades, significant number of study has been done on the effects of local soil conditions on ground motion. These effects are commonly evaluated by experimental and/or numerical methods. The numerical analyses are generally performed either in frequency or time domain. Time domain analyses usually incorporate in the equations of motion viscous damping which is generally assumed to be of the Rayleigh type, i.e. stiffness- and mass-proportional, or in a simplified form, only stiffness-proportional. Both formulations result in a frequency-dependent damping, in contrast to the soils experimental behavior. In this study, several linear analyses were conducted for an idealised soil deposit with two different time domain computer codes in order to explore the influence of Rayleigh damping formulation on site response analyses. Moreover, nonlinear analyses were performed for the selected Orvieto site in Central Italy. The results of the analyses show that the use of the simplified Rayleigh damping formulation can significantly underestimate the seismic response, especially for deep deposits and high frequency input motion. In addition, it is shown that with the use of full Rayleigh damping formulation, the results of analyses are more consistent with those obtained with a frequency-independent analysis.