Probabilistic Design to Resist Earthquakes

The status of probabilistic design to resist earthquakes is surveyed. Studies of local and regional seismicity are covered, with applications of Bayes' theorem, as well as studies of structural responses to random motions such as white noise, Gaussian processes, and computer-simulated earthquakes of more general nature. The probabilistic approach to earthquake resistant design is developing in the following steps: Probability distributions of (a) parameters that define (incompletely) the earthquake characteristics, (b) parameters such as stiffness, damping, etc., and (c) parameters entering criteria of failure; combination of these distributions though rather general dynamic analyses to arrive at the probability of survival of structures; and decisions on the adequacy of the proposed design alternatives. The Monte Carlo approach is shown to be particularly promising. Several areas in which much work remains to be done are suggested.