In this article, the recently proposed approach known as 'risk targeting' for the development of national seismic design maps is investigated for mainland France. Risk targeting leads to ground-motion maps that, if used for design purposes, would lead to a uniform level of risk nationally. The Eurocode 8 design loads currently in force for France are used as the basis of this study. Because risk targeting requires various choices on, for example, the level of acceptable risk to be made a priori and these choices are not solely engineering decisions but involve input from decision makers we undertake sensitivity tests to study their influence. It is found that, in contrast to applications of this methodology for US cities, risk targeting does not lead to large modifications with respect to the national seismic hazard map nor to changes in the relative ranking of cities with respect to their design ground motions. This is because the hazard curves for French cities are almost parallel. In addition, we find that using a target annual collapse probability of about 10-5 for seismically designed buildings and a probability of collapse when subjected to the design PGA of 10-5 leads to reasonable results. This is again in contrast to US studies that have adopted much higher values for both these probabilities.
展开▼
