The complex nature of uncertainty propagation in aeroservoelastic systems, which are inherently multidisciplinary, makes Monte Carlo simulation methods desirable tools for statistical analysis. With the growing efficiency of analysis methods, coupled with growing computing power, such Monte Carlo methods, which were computationally impractical in the past, can now be used to simulate real flight vehicles at industry level of complexity and accuracy and allow statistical analysis of aeroservoelastic behavior on a scope and with depth not possible before. This paper describes the development of a new, fully integrated computational capability for Monte Carlo simulations of uncertain aeroservoelastic systems and presents results that add important insights regarding their reliability and overall uncertainty. The new capability allows selection of uncertain variables at all levels of modeling in all disciplines involved. As an important example, the focus of the work described here is the sensitivity to uncertainty of the “notch-filter” approach, commonly used to decouple aeroelastic and structural dynamic behavior and stabilize unfavorable aeroservoelastic interactions in cases where such interactions lead to instability. It is shown that, if notch filters are used in the aeroservoelastic loop, attention has to be paid to the variability and uncertainties of the aircraft involved. Uncertain parameter prioritization and risk mitigation are proposed by using sensitivity analysis and Bayesian update.
Probabilistic Aeroservoelastic Reliability Assessment Considering Control System Component Uncertainty
AIAA Journal ; 54 , 8 ; 2507-2520
2016-06-13
14 pages
Article (Journal)
Electronic Resource
English
Aeroservoelastic Simulation Considering Control System Component Uncertainty (AIAA 2015-2056)
British Library Conference Proceedings | 2015
|