y rel y-component of the relative distance between two UAVs in formation z rel z-component of the relative distance between two UAVs in formation disadvantage: not only are they extremely computationally demanding to handle in real-time, but these databases can only be used for a specified air vehicle and a range of flight conditions (32) . Simple computational methods are rapid, but the results are not always realistic or accurate enough. However, with increasing accuracy comes increasing computation time. Hence choosing a computational method to model the wake vortex involves finding a compromise between accuracy and rapidity of execution. This paper describes the development of a computational method and simulation models that incorporate wake vortex effects associated with air vehicles flying in close proximity. The aim has been to develop a generic wake vortex model which could be used for any type of wing geometry, and still produce results which could be used in a real-time synthetic environment. Using the same VLM to model both vehicles has allowed them to exchange position during simulations. The non-uniform vortex-induced wind and wind gradients acting on the trail aircraft have been approximated as effective wind and wind gradients and directly used within dynamic simulations, following the method developed by Dogan and colleagues (32)(33)(34)(35)(36)(37)(38)(39) .In the following sections, first the WVM, then the model integration process are described, followed by some results and conclusions, and an outlook on further developments.
WAKE VORTEX MODELA one-lifting-line vortex lattice method for linear aerodynamic wing applications has been developed and implemented in MATLAB. The code (ELL) computes the steady-state velocity induced by the wake of one or more air-vehicles at a given location using Weissinger's extended lifting-line theory (40,41) , and supports 3D, subsonic multiwing designs with swept, tapered, twisted wings of any aspect ratio, with or without dihedral.Special care was used to maintain the symmetry between the leading and following air-vehicles in order to allow them to exchange positions during simulations. As a consequence, the modelling reference frame could not be attached to any particular air-vehicle. Hence, it was decided to work in the inertial NED (North East Down) frame: not only was this reference frame independent of UAV positions and orientations, but it presented advantages when interfacing the simulation with visualisation tools such as FlightGear (42,43) or AVDS (44,45) .