PROBABILISTIC STRUCTURAL ANALYSIS METHODS AND APPLICATIONS

Abstract

An advanced algorithm for simulating the probabilistic distribution of structural responses due to statistical uncertainties in loads, geometry, material properties and boundary conditions is reported. The method effectively combines an advanced algorithm for calculating probability levels for multivariate problems–-Fast Probability Integration (FPI)–-together with a general purpose finite element code for stress, vibration and buckling analysis. The combination provides the analyst with the ability to calculate probability levels for output variables with speed and accuracy over the full range of probability levels. Application is made to a space propulsion system turbine blade for which the geometry and material properties are treated as random variables. The results also demonstrate the ability of the code to define the relative ranking of the random variables in terms of their influence on the response variable uncertainty. Extension of the code to nonlinear problems and to other analysis modules is being made.