This thesis investigates the application of CFD techniques to the aerodynamic analysis of a U-shaped JETC. Investigations were carried out to determine the flow patterns present at a number of locations within the structure of a full U-shaped JETC. The CFD solutions produced in these investigations used recommendations from the literature in the set-up of the CFD solver, and provided the computational component towards problem-specific validation of the CFD techniques used. A structured series of CFD-aided investigation and design processes were then performed. These processes were based around a series of analyses that evaluated the influence of a number of cell parameters in terms of cell airflow efficiency and velocity distortion. Four cell components; the inlet and exhaust stack baffle arrangements, the turning-vanes, the rear of the working section and augmenter entrance, and the lower exhaust stack, including the BB, were investigated in individual analyses. Throughout the investigations the value of CFD as a design tool was constantly assessed. Overall, the findings suggest that aerodynamic optimisation of the baffle arrangements would provide the greatest gains to cell airflow efficiency. As some cells contain as many as three baffle arrangements, the potential increases made to cell airflow capacity are sizable. Through implementing the findings of the baffle arrangement investigations, static pressure loss across the five-row baseline arrangement was reduced by 79%. For low levels of velocity distortion in the upstream region of the working section, the need to design the inlet stack baffles in the turning-vane arrangement was highlighted. Mid-baffle vane alignment, consistent flow channels, and sufficiently low chord to gap ratios should be incorporated into a turning-vane design to maximise flow uniformity. The need for the baffle and vane components to combine with the geometry of the cell to limit adverse pressure gradients was found as a requirement to minimise inner corner separation, and the downstream threat it creates to a safe testing environment. CFD proved to be a valuable analysis tool throughout the investigations performed in this thesis. The number of design iterations analysed, and the detail of data that could be extracted, significantly exceeded what could have been achieved through an isolated experimental testing programme.

This thesis details an in investigation of two problems arising during the testing of a jet engine in a test cell, namely the formation and ingestion of vortices and the generation and propagation of infrasound. Investigation involved the use of computational fluid dynamic as well as analytical tools. The author extended the work of previous researchers by investigating the effect when a suction inlet in surrounded by four walls, (as it is in a test cell). A previously suspected but not documented small region of unsteady vortex was discovered to lie between the steady vortex and no vortex regions. The preferential attachment of the vortex, when formed, to a particular surface was investigated and a low velocity region near that surface has been proven as a possible cause. A cell bypass ratio > 90% was found to be necessary to avoid the formation of vortices in typical situations. Parametric studies (conducted cetaris paribus) on four different geometries and flow parameters were also conducted to determine how they affected the vortex formation threshold. Boundary layer thickness on the vortex attachment surface, upstream vorticity, size of suction inlet was found to have a direct relationship with probability of vortex formation whereas Reynolds number of flow was found to have an inverse relationship. Three hypotheses regarding the generation and propagation of infrasound in test cells were analysed. The first hypothesis states that the fluctuating of flow within the test cell led to a periodic fluctuation of pressure. The second hypothesis predicts a change in flow conditions can leads to a change in the acoustic reflection characteristics of the blast basket perforates. The final hypothesis proposes that changing engine location and size of augmenter, can lead to a reduction in the slip velocity between the engine exhaust jet and the cell bypass flow thus reducing the engine jet noise. The first hypothesis has been disproved using CFD techniques, although the results are as yet inconclusive. The second and third hypotheses have been proven to be potentially feasible techniques to be employed in the future. The changes proposed in the final hypothesis are shown to reduce the engine jet noise by up to 5 dB.