Sensitivity analysis applied to a test rig model

Methods for sensitivity analysis applied to a test rig model are presented. On the one hand, the effect of parameter variations on system variables is analysed by means of parameter sensitivities of first and second order, while on the other hand - for the purpose of comparison - conventional Monte Carlo simulations is applied. The result evaluation leads to the conclusion that sensivity calculation using parameter sensivities - especially of second order and with respect to systems with a high number of parameters - is a serious alternative to Monte Carlo analysis.