DEVELOPING STRUCTURAL DESIGN CRITERIA WITH SPECIFIED RESPONSE RELIABILITY

The objective of developing response-based risk-consistent design crit eria is to ensure that design combinations of environmental or load va riables are chosen in such a way that one or more structural response variables meet a target level of reliability. Four general approaches are identified: the method of response statistics, methods based on it erative forward reliability analysis, methods based on inverse reliabi lity analysis, and contour methods. The last approach is the only one that provides uncoupling between the probabilistic processing of the e nvironmental and loading variables and the structural response analysi s. Two contour methods are described, one based on the inversion of FO RM spheres and the other on linearization of an arbitrary response fun ction. Contour methods can easily provide compensation for the effect of uncertain statistical loading, analysis, and resistance models as w ell. This is the technique of ''inflated contours.'' Examples are prov ided describing the use and the quality of the iso-reliability contour techniques. A valuation of criteria such as computational ease, uncou pling between environmental statistics and structural response, repeat ed use, and incorporation of secondary uncertainties leads to the conc lusion that the method of iso-reliability contours is the most effecti ve tool for developing response-based risk-consistent design criteria.