|Title:||Identification of an Unsteady Aerodynamic Model up to High Angle of Attack Regime|
|Degree:||Doctor of Philosophy|
|Department:||Aerospace and Ocean Engineering|
|Committee Chair:||Dr. Frederick H. Lutze|
|Committee Members:||Dr. Mark R. Anderson, Associate Professor|
|Dr. Eugene M. Cliff, Reynolds Metals Professor|
|Dr. Wayne C. Durham, Associate Professor|
|Dr. Frederick H. Lutze, Professor|
|Dr. Roger L. Simpson, Jack E. Cowling Professor|
|Keywords:||Signal Processing, Unsteady Aerodynamic Characteristics, Aerodynamic Model, Parameter Identification|
|Date of defense:||November 20, 1997|
|Availability:||Release the entire work immediately worldwide.|
The harmonic oscillatory tests for a fighter aircraft configuration using the Dynamic Plunge-Pitch-Roll (DyPPiR) model mount at Virginia Tech Stability Wind Tunnel are described and analyzed. The corresponding data reduction methods are developed on the basis of multirate digital signal processing techniques. Since the model is sting-mounted to the support system of DyPPiR, the Discrete Fourier Transform (DFT) is first used to identify the frequencies of the elastic modes of sting. Then the sampling rate conversion systems are built up in digital domain to resample the data at a lower rate without introducing distortions to the signals of interest. Finally linear-phase Finite Impulse Response (FIR) filters are designed by Remez exchange algorithm to extract the aerodynamic characteristics responses to the programmed motions from the resampled measurements. These data reduction procedures are also illustrated through examples.
The results obtained from the harmonic oscillatory tests are then illustrated and the associated flow mechanisms are discussed. Since no significant hysteresis loops are observed for the lift and the drag coefficients for the current angle of attack range and the tested reduced frequencies, the dynamic lags of separated and vortex flow effects are small in the current oscillatory tests. However, large hysteresis loops are observed for pitch moment coefficient in the current tests. This observation suggests that at current flow conditions, pitch moment has large pitch rate and alpha-dot dependencies. Then the nondimensional maximum pitch rate q_max is introduced to characterize these harmonic oscillatory motions. It is found that at current flow conditions, all the hysteresis loops of pitch moment coefficient with same nondimensional maximum pitch rate are tangential to one another at both top and bottom of the loops, implying approximately same maximum offset of these loops from static values. Several cases are also illustrated.
Based on the results obtained and those from references, a state-space model is developed to describe the unsteady aerodynamic characteristics up to the high angle of attack regime. A nondimensional coordinate is introduced as the state variable describing the flow separation or vortex burst. First-order differential equation is used to govern the dynamics of flow separation or vortex bursting through this state variable. To be valid for general configurations, Taylor series expansions in terms of the input variables are used in the determination of aerodynamic characteristics, resembling the current approach of the stability derivatives. However, these derivatives are longer constant. They are dependent on the state variable of flow separation or vortex burst. In this way, the changes in stability derivatives with the angle of attack are included dynamically. The performance of the model is then validated by the wind-tunnel measurements of an NACA 0015 airfoil, a 70 degree delta wing and, finally two F-18 aircraft configurations. The results obtained show that within the framework of the proposed model, it is possible to obtain good agreement with different unsteady wind tunnel data in high angle of attack regime.
List of Attached Files