Inflated contours for extreme response prediction in complex structural systems

Abstract

This study investigates the technique of environmental contour inflation to account for statistical uncertainties in either the loading or the response. Complex structural systems are treated as simple systems and the additional loadings are treated as omitted random variables in what is essentially a forward First Order Reliability Method (FORM) problem. An empirical environmental model is used to map the dominant, or known, random variables from physical space to standard normal space. The inverse-FORM technique is then applied to these standard normal variables to form environmental contours. A Monte Carlo simulation is then performed using a structural finite element model to calculate the probability of failure given occurrence of a particular sea state. From this analysis, design concern regions can be determined for the tendon's contour and several illustrative examples presented. This type of methodology is quite general and is therefore applicable to many types of systems.