Global sensitivity methods in engineering analysis

Sensitivity Analysis (SA) can be considered as the study of the effects of the input variables over the output of a model. It can be used to identify the critical variables in a process and make decisions based on this information. There are several methods available in the literature for sensitivity analysis. These methods can be classified in three major groups: local methods, global methods and screening methods.

Global sensitivity analysis (GSA) quantifies the effect of a variable in a model based on its contribution to the total variance of the model. The methodology and properties of GSA were studied using some academic examples and the results obtained were compared with the results provided by other sensitivity methods.

The sensitivity methods studied were also applied to two realistic applications: sensitivity analysis of risk assessment in a general aviation problem and sensitivity analysis of a laser peening process to improve the fatigue life in a turbine disk. The results of the sensitivity methods in all the cases were consistent.