Research On Wing Flutter Boundary Prediction Based On Atmospheric Turbulence Excitation Response

Considering the risk of flutter test,if there is an effective flutter boundary prediction technique which could judge the flutter boundary at a subcritical velocity before flutter point or even at a very low velocity,it will be helpful to protect the safety in test,reduce test cost and improve the accuracy ot flutter boundary prediction.Therefore it will have great significant and prospect on engineering application to develop practical flutter boundary prediction method.Previous studies have been carried out on flutter test data processing technique and flutter boundary prediction method.However,further researches are needed as well on how to identify the modal parameters of turbulence response for flutter boundary prediction.This dissertation investigates on the problem of modal hybrid and low signal-to-noise ratio for turbulence excitation response.The spectrum analysis and empirical mode decomposition method are used on noise reduction and data reconstruction,then signal processing technique and modal parameter identification method that suits for atmospheric turbulence excitation response are developed.Secondly,two effective stability analysis methods for flutter boundary prediction on vibration response are investigated,and the relationship between these two criterions is given.Finally,this dissertation develops the flutter margin method and proposes a new stabilitycriterion based the dynamic equation of the three degree of freedom wing and the Routh's stability theory.The main contents of this dissertation are shown as follows:(1)This dissertation proposes a series of signal processing technique and modal parameter identification method for atmospheric excitation response.Random decrement technique,natural excitation technique and Hilbert-Huang transform are used and results are compared with numerical simulations.Then theoretical derivations are adopted on complex exponential method,ARMA model time series analysis method,stochastic subspace identification and matrix pencil method.The results of the flat plat wing model show that the matrix pencil method has better modal parameter identification accuracy and thus will identify good results when combined it with Hilbert-Huang transform and natural excitation method.(2)This dissertation derives flutter margin method and ARMA system stability analysis method theoretically which suit for vibration responses in flutter boundary prediction,then gives the difference and connection between these two methods upon binary wing model pre-twist coupling.The critations of these two methods have approximate curves but differ four power of sampling frequency.The results in example of flat plat wing model and beam frame wing model show that the critations of these two methods have good downtrend with the decrease airspeed so that will help to predict flutter boundary early.And the flutter margin is less effected by damping identification error than ARMA system stability analysis method that will have better robustness.(3)A new stability criterion is proposed based on the dynamic equation of the three-dof wing and Routh's stability theory and its variation with airspeed to describe numerical experimental data.Further more,the influence of modal parameter errors is discussed through different combinations of maximum allowable errors.The results show that,the criterion can help to predict an approximate flutter boundary with linear fit at low airspeed and thus will improve the security of flutter test.There is a “first prediction” in the extrapolation process,and the significance of the “first prediction” is that,if airspeed is less than it,the prediction is toward to the “first prediction”,and when airspeed is larger than it,the prediction is reaching true flutter boundary gradually.(4)Four wind tunnel flutter tests,especially flutter test of full-aircraft,are analyzed to prove the effectiveness of the techniques and methods developed in this dissertation.The results show that through the EMD technique the stationary random response of every mode can be decomposed respectively.And then the free decaying response can be extracted by the natural excitation technique.Thus it will be helpful to identify modal parameters at any airspeed through the matrix pencil method.Inverse amplitude tracking method can be used to determine dangerous mode.And finally good flutter boundary can be predicted at a low airspeed by the ARMA stability analysis method and at the subcritical airspeed through the velocity-damping method.