An engineering method using a 2D unsteady potential formulation (called the free vortex model or FVM) has been developed to predict the normal force, centre-of-pressure and vortex position for cruciform wing-body combinations in the “plus” orientation, at supersonic speeds and cross flow Mach numbers less than or equal to 0.55 up to angles of attack 20◦. The wings are of very low aspect ratio ( ≤ 0.1), have taper ratios greater than 0.85 (or significant side edges) and have low span to body diameter ratios ( ≤ 1.5). The method predicts the position and subsequent loads imposed by the vortex along the length on the wing-body combination by determining the shed vorticity using Jorgensen’s modified Newtonian impact method. The vortex position is well predicted for angles of attack from 4◦ until symmetric vortex shedding occurs, whilst the normal force is well predicted from 0◦. The centre-of-pressure is predicted further aft at the low angles and further forward at the high angles of attack. If this method is used in combination with the single concentrated vortex of Bryson applied to cruciform wing-body combinations the vortex positions prediction limitations at angles of attack less than 4◦ can be overcome. An investigation of the lee side flow field of cruciform wing-body configurations was also performed, and revealed that the vortex position is dependent upon the lee side secondary vortex separation characteristics. Other features revealed that symmetric vortex shedding occurs when both the region of flow outside the shed vortex sheet and reverse flow region are supersonic and a termination shock exists. The thesis also investigated the applica- tion of the discrete vortex model (DVM) method to cruciform wing-body combinations and found that the potential only formulation overpredicts the normal force, whilst the inclusion of boundary layer separation (and therefore modeling the secondary separation vortex) predicted the normal force very well. The application of the concentrated vortex method of Bryson was also investigated and found to be only applicable at low angles of attack (< 4◦).

The reduction of drag on commercial aircraft is an active field of study especially with environmental pressures to reduce the carbon emissions associated with climate change. To this end, the AEROMEMS-II project was commissioned by the EU with a view to investigate methods for reducing drag by using MEMS devices for controlling separation. One method for investigating flow control devices is to use the field of Computational Fluid Dynamics (CFD) to simulate the flow interactions produced in flow control applications and assess their effect. Simulating such flows can be computationally expensive so a number of methods have been investigated here to assess their use in flow control simulation applications. The first of these is the Immersed Boundary Method (IBM) which allows complex geometries to be simulated using simple cartesian grid CFD codes. IBMs are found to reduce requirements whilst maintaining flow resolution and accuracy. Next is the use of turbulence modelling with wall functions to reduce the need for fine grids near any solid surfaces. This method is found to work well and can allow the grid spacing near the wall to be 100 times coarser than with no wall functions applied. Finally, Detached Eddy Simulation (DES) has been considered as a method for allowing unsteady flow control structures to be simulated without being damped by conventional turbulence modelling. Each of these methods is presented, implemented and validated against known flow cases to assess their abilities fully. All three methods have then been applied together to a known experimental turbulent flow-control set-up at the University of Lille (fellow partners in the AEROMEMS-II project) in order to assess the feasibility of using all of these methods together to simulate flow control. All three of these methods are seen to work well together although not always with the same effect.

The jet acting perpendicular to a cross-flow boundary layer is a commonly studied complex turbulent flow. Our research was motivated by their potential application in separation delay devices, where jets can be used to produce streamwise vortices in a similar manner to conventional solid vortex generating vanes. This thesis addresses two problems; firstly the generation of inflow conditions for the simulation of a spatially developing turbulent boundary layer, and secondly the simulation of low velocity ratio jets interacting with the boundary layer. Our approach involved refining a popular turbulent inflow generation technique, validating the accuracy of our improved method against well established direct numerical simulation data. This turbulent boundary layer was used to simulate a low velocity ratio perpendicular jet test-case, which was validated against experimental data. Finally, a pitched and skewed jet model was investigated. Our modifications to the turbulent boundary layer inflow generation method were successful, addressing problems described by various authors regarding the stability and accuracy of the technique. Secondly we have found excellent agreement in our perpendicular jet in cross flow test-case, and have produced what we believe to be the first documented unsteady numerical simulation of the flow field behind a low velocity ratio pitched and skewed jet.