Uncertainty analysis techniques using approximation concepts in multidisciplinary structural design environment

One of the primary goals of structural engineers is to assure proper levels of safety for the structures they design. This seemingly simple task is complicated by uncertainties that result from material data, physical modeling, finite element mesh, linear/nonlinear theories, solution algorithms, etc. Information is available either as sparse data points, intervals, expert opinions, or as probability distributions. In multidisciplinary integration, depending on the uncertain information available, uncertainty propagates from one-step or discipline to another. Structural reliability and uncertainty quantification (UQ) are tools that can be employed to quantify these uncertainties and inaccuracies to produce designs that meet the safety requirements. These issues are the focus of this research work. It was important to investigate some methods that would not necessarily require the rigorous and intense testing procedures that prototypes must typically endure. Of added advantage would be the ability to apply these methods during the preliminary design stages. These methods would not replace the prototyping and actual testing but keep the number of actual tests to a minimum. In today's competitive world, major emphases is on reducing the actual testing of prototypes and certifying the systems analytically. The techniques developed in this research provide new tools to aid the complex task of analytical certification. These techniques help build safer systems more economically and in a relatively shorter period.