Relative Importance of Uncertain Structural Parameters

A novel procedure for the estimating the response sensitivity to input parameters of a complex FE model is presented. The method is specifically directed toward problems involving high-dimensional input parameter spaces, as they are encountered during uncertainty analysis with complex, refined FEmodels. In these cases one is commonly faced with thousands of uncertain parameters and traditional techniques, e.g. finite difference, are unfeasible. In contrast, the presented method quickly filters out the most influential variables. Hence, the topic is not how to compute the sensitivity but how to identify all those parameters for which probable variations have the biggest influence on the response.. This is achieved by generating a set of samples with direct MCS, which are closely scattered around the point at which the sensitivities are sought. From these samples, estimators of the sensitivities are synthesized and the most important ones are refined with a finite difference calculation. In this paper, the underlying theory as well as the resulting algorithm is presented. Practical software issues will be also addressed.