Probabilistic analysis of buildings with tuned mass damper

Abstract

Uncertainties in properties and input earthquake excitation can have significant influence on the response reduction of systems with Tuned Mass Damper (TMD). Hence, an effort has been made in this paper towards statistical characterization of tuning parameters and response of structural systems with and without TMD. Uncertainty in mass, stiffness and Peak Ground Acceleration (PGA) are modelled with mean, coefficient of variation and probability distribution functions adopted from literature. Sample size adopted is 5000 and time history analysis are carried out for a ten storey steel building frame(B1) and fifteen storey RC building(B2) with the random mass, stiffness properties and the samples of ground motions with random PGA. From the Cumulative Distribution Function (CDF) of peak responses and response ratios, it is observed that fitting lognormal distribution is found to pass the K-S goodness of fit test for 95% confidence for peak responses of ten storey steel frame. Whereas, fitting both normal and lognormal distribution do not found to pass the goodness of fit K-S test for fifteen storey RC building. Response ratios of peak displacement do not found to pass the goodness of fit K-S test for both normal and lognormal distribution for both the buildings.