Uncertainty Propagation And QMU Application For Flutter Analysis Of Wing Structures

Flutter analysis is a particularly important component of airframe aeroelastic design due to its destructive effect.The generation of flutter is strictly prohibited in flight envelope.However,uncertainties are inevitable during the flight.Therefore,in the design process,uncertainty quantification based on modeling and simulation(M&S)in flutter analysis has received substantial attention.In the practical system,neither the aleatory nor the epistemic uncertainty can be ignored.Traditional uncertainty quantification methods generally focus on single uncertainty and bring high computational cost when handling with complex simulation model.In this paper,Probability Box is introduced to represent the comprehensive propagation.Then,the point collocation non-instructive polynomial chaos expansion(NIPC),an effective and efficient methodology is presented for mixed uncertainty quantification in flutter analysis.Based on the prior work,a practical quantification of margins and uncertainties(QMU)framework methodology is introduced for safety assessment of flutter boundary.Built upon the M&S and uncertainty quantification,the QMU is utilized as a direct indication for the safety of a flutter boundary and a guidance for parameter design.Different wing structures(a two-dimension airfoil,the AGARD445.6 wing model and a low-aspect-ratio hypersonic wing)are employed here for the following research.1.Dynamics modeling and flutter analysis of wing structures:Flutter analysis is conducted by numerical method.In allusion to the AGARD445.6 model,dynamic characteristics of different models are compared and the best model is chosen for the subsequent analysis.2.Uncertainty quantification:The theory and analysis process of Monte Carlo method and point collocation non-instructive polynomial chaos expansion are illustrated first.An aleatory uncertainty quantification example is presented to prove the applicability and superiority of the second method.As for mixed uncertainty quantification,the probability box(p-box),a special type of cumulative distribution function,which contains information of both aleatory and epistemic uncertainties,is introduced to express the contribution of two types of uncertainty simultaneously,but not confound them.Then,a two-phase MCS/NIPC framework serves as an efficient means of uncertainty quantification for problems with mixed uncertain inputs.For different wing structures,the uncertainties exist in the material properties,flight altitude and temperature have been quantified for a safer flutter boundary.3.QMU application for flutter assessment and parameter design:a QMU framework is implemented for flutter analysis with aleatory and mixed uncertainty.Based on the concept of the margin design,a confidence ratio is employed to decide the safety of the aeroelastic system.Then,a parameter design method based on the QMU framework is illustrated.