IFASD-2011-71 REDUCING UNCERTAINTY IN AEROELASTIC FLUTTER BOUNDARIES USING EXPERIMENTAL DATA Richard P. Dwight 1 , Hester Bijl 1 , Simao Marques 2 , and Ken Badcock 3 1 Faculty of Aerospace Engineering Technische Universiteit Delft, 2600GB, The Netherlands r.p.dwight@tudelft.nl 2 School of Aerospace Engineering Queen’s University Belfast, BT9 5AH, Northern Ireland 3 Department of Engineering University of Liverpool, L69 3GH, United Kingdom Keywords: aeroelasticity, flutter, uncertainty quantification, data assimilation, in- verse problems, model updating, Bayes’ theorem, probabilistic collocation, Markov-Chain Monte-Carlo Abstract: Flutter prediction as currently practiced is usually deterministic, with a sin- gle structural model used to represent an aircraft. By using interval analysis to take into account structural variability, recent work has demonstrated that small changes in the structure can lead to very large changes in the altitude at which flutter occurs (Marques, Badcock, et al., J. Aircraft, 2010). In this follow-up work we examine the same phe- nomenon using probabilistic collocation (PC), an uncertainty quantification technique which can efficiently propagate multivariate stochastic input through a simulation code, in this case an eigenvalue-based fluid-structure stability code. The resulting analysis pre- dicts the consequences of an uncertain structure on incidence of flutter in probabilistic terms – information that could be useful in planning flight-tests and assessing the risk of structural failure. The uncertainty in flutter altitude is confirmed to be substantial. Assuming that the structural uncertainty represents a epistemic uncertainty regarding the structure, it may be reduced with the availability of additional information – for example aeroelastic response data from a flight-test. Such data is used to update the structural uncertainty using Bayes’ theorem. The consequent flutter uncertainty is significantly reduced across the entire Mach number range. 1 INTRODUCTION Flight flutter tests can be dangerous and costly, whereas computational methods may not accurately predict the flutter boundary [1]. As such there is a long history in aeroelas- ticity of combining limited flight-test data with physical models to estimate the flutter envelope. A successful and widely-used approach was proposed by Zimmerman and Weis- senburger [2] based on an analytic model of the aeroelastic phenomena, and has been modified multiple times to account for uncertainty information in order to evaluate ro- bustness [1,3]. The present work takes one step towards the goal of performing a com- parable stochastic analysis of flight-test data on the basis of modern high-fidelity PDE simulations. The increase in modeling accuracy should lead to a reduction in the quantity of flight-test data necessary, but a major challenge is the high computational expense of the basic fluid-structure interaction (FSI) simulation. 1