Fragility definition of reinforced concrete buildings by a method based on nonlinear pushover analysis

A tremendous amount of economical and life losses, due to damages after experienced earthquakes, is related to seismic performance of the existing buildings. In earthquake prone countries as Turkey with high seismic risk, it is inevitable to experience major earthquakes and associated losses in the future, as well as in the past. In addition to the design of new buildings in accordance with seismic codes, evaluation of seismic performance and vulnerability of existing buildings is necessary to estimate losses in possible future earthquakes and to determine precautions to reduce the losses.Fragility curves are widely used in estimating seismic risk posed by the existing buildings. In this dissertation, analytical fragility curves, that may have an important role for seismic risk assessment for the metropolitan cities, are derived. Fragility curves are sketched by using architectural and structural details of low- and mid-rise reinforced concrete buildings, which constitute the majority of the existing buildings and are generally used for residential and commercial purposes. Pushover analyses are performed by generating detailed 3D models of the buildings. Spectral displacement demand of the buildings is determined for three different earthquake levels and two local soil sites by using the successive approach given in the Turkish Earthquake Code-2007.Four limit states in terms of spectral displacement are defined for considered building classes by using idealized capacity curves. The generated fragility curves are expressed in the form of two-parameter lognormal distribution functions. Probability density functions for different damage states of building classes are formed by using the median of spectral displacements of predefined limit states and the standard deviation of lognormal distribution values. Consequently, fragility curves are constructed by calculating cumulative probabilities of reaching or exceeding the corresponding damage state for different values of spectral displacement, which is an earthquake parameter.Furthermore, population parameter is estimated by using the statistics obtained from building data, and 90% confidence intervals are calculated by using sampling distribution for median values of fragility curves, which are expressed in terms of spectral displacement.