ABSTRACT
A discussion is presented of various alternatives for the estimation of the seismic reliability with respect to collapse of complex nonlinear systems. They do not require the use of a previously determined deformation capacity. The state of collapse failure is defined as in terms of a damage index, measured by the secant-stiffness reduction, that reaches the value of unity at failure. A maximum likelihood approach is presented that uses this index in the estimation of the reliability functions of the systems considered. An illustrative example is presented, based on the assumption that the damage index has a lognormal probability distribution. Some alternatives to account for different forms of probability distributions are discussed.
