The proportion of the overall drag force on a flight vehicle that can be attributed to skin friction drag is much higher on hypersonic vehicles than for vehicles operating at lower speeds. Consequently, controlling skin friction drag is a major obstacle to the successful development of an operational scramjet-powered flight vehicle. To date, many techniques have been proposed for the reduction of skin friction. Of the many methods available, the combustion of hydrogen in supersonic turbulent boundary layers has been experimentally, numerically and theoretically shown to be a promising method of reducing skin friction drag in hypervelocity applications. However, one key issue that can affect the implementation of the boundary layer combustion technique in a realistic scramjet is the presence of highly non-uniform flow entering the combustion chamber from the inlet. The aim of this project is investigate if and how flow non-uniformities entering a scramjet combustor can affect the potential for boundary layer combustion to reduce skin friction. To achieve this aim, the project is approached experimentally and numerically.
Experiments were conducted on a circular constant-area combustor that was attached downstream of a Rectangular-to-Elliptical Shape Transition (REST) inlet and an injector designed to deliver hydrogen into the boundary layer. Using a stress wave force balance, the integrated skin friction drag on the internal surface of the combustor was measured for three scenarios - one, where fuel is not injected, two, where fuel is injected but combustion is suppressed, and three, where fuel is injected and allowed to burn. The REST inlet was used to produce flow disturbances that are typical of those to be expected in operational scramjet inlets. When the experimental model was tested at on- and off-design conditions, the experimentally measured drag coefficients for the fuel-on tests were 28% - 30% lower than those for the fuel-off tests. The levels of skin friction reduction that are measured in the current experiments are similar to those measured in experiments conducted without a realistic scramjet inlet upstream of the combustor, thus demonstrating that the drag reduction brought about by boundary layer combustion is not significantly affected by the flow disturbances generated from the REST inlet.
In addition, the experimental model was also tested with vortex generators attached in the inlet. These vortex generators were used to generate flow disturbances similar to those brought about by conventional cross-stream fuel injection techniques. The experimental results show that the level of skin friction reduction is similar to that measured in the tests without the vortex generators, thus demonstrating that flow disturbances similar to those from cross-stream fuel injection in the inlet do not affect the drag reduction potential of boundary layer combustion. These experiments also demonstrate that the levels of drag reduction brought about by film-cooling effects are more easily affected by flow disturbances than those brought about by boundary layer combustion.
To support the analysis of the experiments, non-reacting RANS CFD simulations of the internal flowfield in the experimental model were conducted using an in-house code called Eilmer3. To facilitate the simulations of the types of flows relevant to this study, the latest version of Wilcox's k-ω turbulence model was implemented in Eilmer3. However, because this version of Wilcox's k-ω model has yet to be thoroughly validated for hypersonic applications, a series of simulations had to be conducted to validate it against test cases that have flowfields representative of those to be expected in this project. A generally good agreement between the numerical and experimental results is obtained for all the six cases that were tested.
Simulations of the internal flowfield in the experimental model were conducted with varying levels of inflow disturbances to assess the levels of flow disturbances that may be required to sweep the fuel layer out from the boundary layer. The numerical results demonstrate that, even in the presence of strong flow disturbances, most of the injected fuel remains in the region of the flowfield that will induce a reduction in skin friction if the fuel burns. Additional simulations also demonstrate that for the drag reduction potential of the boundary layer combustion technique to be realised, heat addition must be occur in regions of the boundary layer where there are large near-wall values of Reynolds shear stress.