Statistical modeling of a seismic isolation object under random seismic exposure

Abstract

Using probabilistic methods, the reliability and seismic resistance of a large-panel building on kinematic foundations (KF) at the seismic isolation site in Almaty is analyzed. The building is modeled by a nonlinear single-mass system with an experimental deformation diagram. Seismic impact is modeled by a random process with a given correlation function. To calculate probabilistic characteristics, the method of statistical tests is used (Monte-Carlo method). The seismic impact parameters are set in accordance with the “Map of the seismic zoning of the Republic of Kazakhstan” by the median values of peak accelerations. 2 accelerograms of the 1990 Baysorun earthquake were chosen as the impact model. Therefore, these accelerograms were taken as the basis for modeling the seismic effects of a random process. Generation of realizations of a stationary random process is performed by recursive filtering. The influence of the number of statistical tests (artificial accelerograms) on the accuracy of calculating the moments of the distribution function of displacements was investigated. According to the results of a Monte Carlo calculation using 5000 realizations, the reliability of the building is determined (probability of failure-free operation). Conclusions about system effectiveness of seismoisolation of kinematic type for the influence specified like buildings at regional parameters are drawn.