Development of a process and toolset to study UCAV flight mechanics using computational fluid dynamics

Vallespin, David Development of a process and toolset to study UCAV flight mechanics using computational fluid dynamics. Doctor of Philosophy thesis, University of Liverpool.

PDF VallespinDav_Sep2011_6593.pdf - Author Accepted Manuscript Available under License Creative Commons Attribution No Derivatives. Download (13MB)

Abstract

The work carried out during this project used a computational Fluid Dynamics code to generate aerodynamic tabular models and aircraft manoeuvre simulations. As an outcome of this work, a validation of the aerodynamic prediction tools and an assessment of tabular models for aircraft flight dynamics applications was made. The Stability and Control Unmanned Combat Air Vehicle has been used as a demonstration case. Validation of computational fluid dynamics methods was carried out for highly nonlinear flow topologies using wind tunnel measurements. Integral data, pressure tap measurements and particle image velocimetry information was compared against the predictions over two configurations. Each one had a different leading edge shape distributed along the span of the model. One was sharp throughout with varying leading edge thickness and the other one was mainly rounded. Results showed a good agreement in longitudinal force and moment predictions for low angles of attack. High angles were dominated by a double vortex structure which was very sensitive to incidence angle and leading edge shape. Some wind tunnel effects were noticed in the measurements when predictions were made with and without sting. Overall the numerical predictive capabilities for low and high angles of attack were deemed good for the purpose of flight dynamics model generation. Two methods for predicting manoeuvering flight aircraft loads are presented in this thesis. A tabular aerodynamic model based on numerical predictions was generated for the sharp configuration. Kriging interpolation was used to populate a model consisting of tables of lateral and longitudinal aerodynamic characteristics. Further to this, longitudinal dynamic derivatives were predicted for the test case in hand using forced oscillation numerical predictions. Aircraft geometric characteristics were approximated based on real aircraft data. A set of controls were designed and implemented for the purpose of manoeuvering flight predictions. A code was implemented to predict realistic aircraft manoeuvres based on an existing program. At the core of this method was a commercial optimisation Matlab code called DIDO. Using this and the nonlinear, six degree of freedom equations of motion, purposedly designed aircraft manoeuvres were predicted. The motions were then replayed using time-accurate simulations and the predicted loads were compared against the tabular predictions. In this manner, the validity of the tables of aerodynamic data were benchmarked against a more reliable and expensive numerical method. The static based predictions showed good agreement with the replays for slow manoeuvres at low angles of attack. As manoeuvres became more aggressive, noticeable disagreement was present in the aircraft loads, particularly in the lateral characteristics during periods of large rates of change in attitudes. Hysteresis effects during manoeuvering flight were seen to produce large spreads in data in the angle of attack domain which the predicted dynamic derivatives were unable to capture.

Item Type:	Thesis (Doctor of Philosophy)
Additional Information:	Date: 2011 (completed)
Uncontrolled Keywords:	computational fluid dynamics, cfd, aircraft manoeuvres, manoeuvering trajectory, Kriging, DIDO, vortical flow, nonlinear flow, delta wing
Divisions:	Faculty of Science and Engineering > School of Engineering
Depositing User:	Symplectic Admin
Date Deposited:	03 Aug 2012 09:20
Last Modified:	16 Dec 2022 04:36
DOI:	10.17638/00006593
Supervisors:	Badcock, Kenneth J
URI:	https://livrepository.liverpool.ac.uk/id/eprint/6593