Jet noise is an extensively studied phenomenon since the deployment of the first civil jet aircraft more than 50 years ago. Jet noise makes up a considerable portion of the total noise of jet aircraft, and the expansion of the numbers of airplanes and airports has only been possible by keeping the noise of airplanes under control. However, while large efforts have been made to measure jet noise and to predict the features of the noise theoretically, the mechanisms by which turbulent jet flows produce noise are still not fully understood. The continued development of low noise jet engines will certainly be aided by improved understanding of these mechanisms. In the past, directly testing acoustic theory has been difficult, because experiments only provided limited information about the jet mechanisms. Recent developments in computational aeroacoustics have shown a promising avenue to provide the required information. The main aim of this thesis is to develop methods to simulate high speed jet flows and to study the results of completed simulations for special cases. These methods should provide an accurate solution of the equations of motion with only limited introduction of simplifying assumptions about the nature of the flow. Then simulation databases could be obtained for interesting cases to investigate the flow and acoustic features of these jet flows.

For the simulation of the jet flows, a staggered high order finite difference scheme was used. The implementation of robust boundary conditions that simulated free jet conditions proved to be particularly challenging. The simulations presented use a collection of different techniques which are designed to resemble physical jet conditions in the interior portion of the domain, with a surrounding region used to resemble the effect of an infinite extension of the finite interior.

This work took place as part of a larger project that also involved the creation of an experimental facility for investigating high speed, low Reynolds number jets. Results from this part of the project, compared with data obtained by the simulation data demonstrated that the simulations were able to accurately predict the flow and acoustic features of these jet flows.

Investigation of the acoustic features of the flow took place by means of solutions of Lighthill's acoustic analogy and the porous Ffowcs Williams and Hawkings method. The implementations of both of these methods required the development of special techniques to optimize the order and number of I/O operations, which was found to be essential for efficiency. Solution of Lighthill's acoustic analogy were found to become more spectrally limited as the spatial integration order was increased. No such effect was seen with solutions of the porous Ffowcs Williams and Hawkings method. The high frequency corruption of the signal was linked to Runge's phenomenon, and sound pressure levels were only accurate when the acoustic signals were appropriately filtered. In the region of resolved frequencies, it was found that the effects of non-linear propagation are contained in both the ``Entropy'' source term and the ``Reynolds Stress'' source term of Lighthill's acoustic analogy for emissions at low angles to the downstream direction of the jet axis, even for essentially isothermal jets.

The effect of inflow temperature was investigated independently of nozzle Reynolds number, by varying the temperature of the inflow for three jet simulations, with the nozzle Reynolds number fixed. The effect of increasing temperature was found to increase the sound pressure level of the acoustic emissions at all angles and to broaden the spectrum of the acoustic emissions. Investigation of the terms of Lighthill's acoustic analogy at 90 degrees reveal that the Entropy term makes a significant contribution to the noise emission at these angles for the heated jets, but very little for the isothermal jet. This gives some evidence for the view that this term is related to an independent emission mechanism in these jets.

Finally, a detailed simulation is performed at a larger Reynolds number in the final chapter. This simulation gave emissions that were around 5 dB higher than the lower Reynolds number jet. It was found that due to an interplay of the various boundary conditions, a recirculation developed in the flow. It is believed that this is responsible for the increase in noise emissions in this case. In other words, the acoustics of jet flows are greatly impacted, not only by nozzle conditions, but also on the surrounding environment, that can alter the way in which the jet develops.