Ground-motion amplification factors (GMAFs) are used to characterize amplification of a ground motion propagating from the bedrock to the ground surface. They are usually determined by ground response analysis, in which the soil parameter variabilities and input motion uncertainties contribute to their uncertainty. The construction of design response spectra requires mean GMAFs or GMAFs with different probability levels. Thus, it is significant to study the sensitivity of soil parameter variabilities and the number of random soil profiles for the estimation of GMAFs. This study investigates the minimum number of random soil profiles required to represent the extent of the epistemic uncertainty in the GMAFs obtained from ground response analysis. It shows that at least 20 and 60 random soil profiles are respectively required to estimate the mean and standard deviations of GMAFs with the maximum relative difference below 10%. In addition, potential reasons for a reduction in the mean GMAFs resulting from randomization of the soil column properties are discussed.

You do not have access to this content, please speak to your institutional administrator if you feel you should have access.