This study investigates the stress wave force balance technique for the measurement of forces on a fuelled hypersonic flight vehicle in an impulse-type test facility. A three component force balance for the measurement of lift, thrust and pitching moment on a supersonic combustion ramjet engine was designed, built, calibrated and tested. The force balance was designed using finite element analysis and consisted of four stress bars instrumented for the measurement of strain. Relative errors of less than 2% were obtained for the recovered simulated calibration loads, while errors of less than 3% were obtained for lift and thrust components for simulated fuel-on and fuel-off force loading distributions. Tests in a calibration rig showed that the balance was capable of recovering the magnitude of point loads to within 3% and their lines of action to within 1% of the chord of the model. Additional errors result when testing in a wind tunnel. The uncertainties for the experiments with fuel injection are estimated at 9%, 7% and 9% for the coefficients of lift, thrust and pitching moment. The scramjet vehicle was 0.566m long and weighed approximately 6kg. It consisted of an inlet, combustion chamber and thrust surface. Fuel could be injected through a series of injectors located on the scramjet inlet. The scramjet model was set at zero angle of attack. Experiments were performed in the T4 Free Piston Shock Tunnel at a total enthalpy of 3.3MJ/kg, a nozzle supply pressure of 32MPa and a Mach number of 6.6, with equivalence ratios up to 1.4. Fuel-off force coefficients were measured to within 2% of theoretical values based on predictions using CFD and hypersonic theory. The fuel-off centre-of-pressure was measured to within 4% of the predicted value. The force coefficients varied linearly with equivalence ratio. Good comparison of the measured lift and thrust forces with theoretical values was obtained with increasing flow rates of fuel. The lift-to-drag ratio increased from 3.0 at the fuel-off condition to 17.2 at an equivalence ratio of 1.0. Poor agreement between the measured pitching moment and theoretical values was obtained due to difficulties in predicting the pressure distribution with heat addition on the latter parts of the thrust surface. A shift in the centre-of-pressure of approximately 10% of model chord was measured as the equivalence ratio varied from 0.0 to 1.0. For the design tested, the thrust produced was not enough to overcome drag on the vehicle, even at the highest equivalence ratio tested. Tests at higher stagnation enthalpies (up to 4.9MJ/kg) showed the lift and pitching moment coefficients remained constant with an equivalence ratio of 0.8 but the thrust coefficient decreased exponentially with increasing stagnation enthalpies. Good agreement of experimental values of lift and thrust force with predicted values was obtained for equivalence ratios of 0.0 and 0.8. Choking occurred at stagnation enthalpies of less than 3.0MJ/kg and a nozzle supply pressure of 32MPa with fuel injection at an equivalence ratio of approximately 0.8, resulting in a drag force of approximately 2.5 times the fuel-off drag force. Tests at a nozzle supply enthalpy of 3.3MJ/kg and nozzle supply pressures of 32, 26 and 16MPa were performed at equivalence ratios of 0.0 and 0.8. The fuel-off lift coefficient remained constant but the thrust coefficient increased. This is attributed to a reduction in skin friction associated with longer lengths of laminar boundary layers as the Reynolds number was decreased. The measured fuel-off lift and thrust coefficients agreed with the predicted values to within the known test flow and force prediction uncertainties. Combustion did not occur at a nozzle supply pressure of 16MPa. This work has demonstrated that overall scramjet vehicle performance measurements (such as lift-to-drag ratio and shifts in centre-of-pressure) can be made in a free piston shock tunnel.

This study investigates the stress wave force balance technique for the measurement of forces on a fuelled hypersonic flight vehicle in an impulse-type test facility. A three component force balance for the measurement of lift, thrust and pitching moment on a supersonic combustion ramjet engine was designed, built, calibrated and tested. The force balance was designed using finite element analysis and consisted of four stress bars instrumented for the measurement of strain. Relative errors of less than 2% were obtained for the recovered simulated calibration loads, while errors of less than 3% were obtained for lift and thrust components for simulated fuel-on and fuel-off force loading distributions. Tests in a calibration rig showed that the balance was capable of recovering the magnitude of point loads to within 3% and their lines of action to within 1% of the chord of the model. Additional errors result when testing in a wind tunnel. The uncertainties for the experiments with fuel injection are estimated at 9%, 7% and 9% for the coefficients of lift, thrust and pitching moment. The scramjet vehicle was 0.566m long and weighed approximately 6kg. It consisted of an inlet, combustion chamber and thrust surface. Fuel could be injected through a series of injectors located on the scramjet inlet. The scramjet model was set at zero angle of attack. Experiments were performed in the T4 Free Piston Shock Tunnel at a total enthalpy of 3.3MJ/kg, a nozzle supply pressure of 32MPa and a Mach number of 6.6, with equivalence ratios up to 1.4. Fuel-off force coefficients were measured to within 2% of theoretical values based on predictions using CFD and hypersonic theory. The fuel-off centre-of-pressure was measured to within 4% of the predicted value. The force coefficients varied linearly with equivalence ratio. Good comparison of the measured lift and thrust forces with theoretical values was obtained with increasing flow rates of fuel. The lift-to-drag ratio increased from 3.0 at the fuel-off condition to 17.2 at an equivalence ratio of 1.0. Poor agreement between the measured pitching moment and theoretical values was obtained due to difficulties in predicting the pressure distribution with heat addition on the latter parts of the thrust surface. A shift in the centre-of-pressure of approximately 10% of model chord was measured as the equivalence ratio varied from 0.0 to 1.0. For the design tested, the thrust produced was not enough to overcome drag on the vehicle, even at the highest equivalence ratio tested. Tests at higher stagnation enthalpies (up to 4.9MJ/kg) showed the lift and pitching moment coefficients remained constant with an equivalence ratio of 0.8 but the thrust coefficient decreased exponentially with increasing stagnation enthalpies. Good agreement of experimental values of lift and thrust force with predicted values was obtained for equivalence ratios of 0.0 and 0.8. Choking occurred at stagnation enthalpies of less than 3.0MJ/kg and a nozzle supply pressure of 32MPa with fuel injection at an equivalence ratio of approximately 0.8, resulting in a drag force of approximately 2.5 times the fuel-off drag force. Tests at a nozzle supply enthalpy of 3.3MJ/kg and nozzle supply pressures of 32, 26 and 16MPa were performed at equivalence ratios of 0.0 and 0.8. The fuel-off lift coefficient remained constant but the thrust coefficient increased. This is attributed to a reduction in skin friction associated with longer lengths of laminar boundary layers as the Reynolds number was decreased. The measured fuel-off lift and thrust coefficients agreed with the predicted values to within the known test flow and force prediction uncertainties. Combustion did not occur at a nozzle supply pressure of 16MPa. This work has demonstrated that overall scramjet vehicle performance measurements (such as lift-to-drag ratio and shifts in centre-of-pressure) can be made in a free piston shock tunnel.

This study investigates the stress wave force balance technique for the measurement of forces on a fuelled hypersonic flight vehicle in an impulse-type test facility. A three component force balance for the measurement of lift, thrust and pitching moment on a supersonic combustion ramjet engine was designed, built, calibrated and tested. The force balance was designed using finite element analysis and consisted of four stress bars instrumented for the measurement of strain. Relative errors of less than 2% were obtained for the recovered simulated calibration loads, while errors of less than 3% were obtained for lift and thrust components for simulated fuel-on and fuel-off force loading distributions. Tests in a calibration rig showed that the balance was capable of recovering the magnitude of point loads to within 3% and their lines of action to within 1% of the chord of the model. Additional errors result when testing in a wind tunnel. The uncertainties for the experiments with fuel injection are estimated at 9%, 7% and 9% for the coefficients of lift, thrust and pitching moment. The scramjet vehicle was 0.566m long and weighed approximately 6kg. It consisted of an inlet, combustion chamber and thrust surface. Fuel could be injected through a series of injectors located on the scramjet inlet. The scramjet model was set at zero angle of attack. Experiments were performed in the T4 Free Piston Shock Tunnel at a total enthalpy of 3.3MJ/kg, a nozzle supply pressure of 32MPa and a Mach number of 6.6, with equivalence ratios up to 1.4. Fuel-off force coefficients were measured to within 2% of theoretical values based on predictions using CFD and hypersonic theory. The fuel-off centre-of-pressure was measured to within 4% of the predicted value. The force coefficients varied linearly with equivalence ratio. Good comparison of the measured lift and thrust forces with theoretical values was obtained with increasing flow rates of fuel. The lift-to-drag ratio increased from 3.0 at the fuel-off condition to 17.2 at an equivalence ratio of 1.0. Poor agreement between the measured pitching moment and theoretical values was obtained due to difficulties in predicting the pressure distribution with heat addition on the latter parts of the thrust surface. A shift in the centre-of-pressure of approximately 10% of model chord was measured as the equivalence ratio varied from 0.0 to 1.0. For the design tested, the thrust produced was not enough to overcome drag on the vehicle, even at the highest equivalence ratio tested. Tests at higher stagnation enthalpies (up to 4.9MJ/kg) showed the lift and pitching moment coefficients remained constant with an equivalence ratio of 0.8 but the thrust coefficient decreased exponentially with increasing stagnation enthalpies. Good agreement of experimental values of lift and thrust force with predicted values was obtained for equivalence ratios of 0.0 and 0.8. Choking occurred at stagnation enthalpies of less than 3.0MJ/kg and a nozzle supply pressure of 32MPa with fuel injection at an equivalence ratio of approximately 0.8, resulting in a drag force of approximately 2.5 times the fuel-off drag force. Tests at a nozzle supply enthalpy of 3.3MJ/kg and nozzle supply pressures of 32, 26 and 16MPa were performed at equivalence ratios of 0.0 and 0.8. The fuel-off lift coefficient remained constant but the thrust coefficient increased. This is attributed to a reduction in skin friction associated with longer lengths of laminar boundary layers as the Reynolds number was decreased. The measured fuel-off lift and thrust coefficients agreed with the predicted values to within the known test flow and force prediction uncertainties. Combustion did not occur at a nozzle supply pressure of 16MPa. This work has demonstrated that overall scramjet vehicle performance measurements (such as lift-to-drag ratio and shifts in centre-of-pressure) can be made in a free piston shock tunnel.

This study investigates the stress wave force balance technique for the measurement of forces on a fuelled hypersonic flight vehicle in an impulse-type test facility. A three component force balance for the measurement of lift, thrust and pitching moment on a supersonic combustion ramjet engine was designed, built, calibrated and tested. The force balance was designed using finite element analysis and consisted of four stress bars instrumented for the measurement of strain. Relative errors of less than 2% were obtained for the recovered simulated calibration loads, while errors of less than 3% were obtained for lift and thrust components for simulated fuel-on and fuel-off force loading distributions. Tests in a calibration rig showed that the balance was capable of recovering the magnitude of point loads to within 3% and their lines of action to within 1% of the chord of the model. Additional errors result when testing in a wind tunnel. The uncertainties for the experiments with fuel injection are estimated at 9%, 7% and 9% for the coefficients of lift, thrust and pitching moment. The scramjet vehicle was 0.566m long and weighed approximately 6kg. It consisted of an inlet, combustion chamber and thrust surface. Fuel could be injected through a series of injectors located on the scramjet inlet. The scramjet model was set at zero angle of attack. Experiments were performed in the T4 Free Piston Shock Tunnel at a total enthalpy of 3.3MJ/kg, a nozzle supply pressure of 32MPa and a Mach number of 6.6, with equivalence ratios up to 1.4. Fuel-off force coefficients were measured to within 2% of theoretical values based on predictions using CFD and hypersonic theory. The fuel-off centre-of-pressure was measured to within 4% of the predicted value. The force coefficients varied linearly with equivalence ratio. Good comparison of the measured lift and thrust forces with theoretical values was obtained with increasing flow rates of fuel. The lift-to-drag ratio increased from 3.0 at the fuel-off condition to 17.2 at an equivalence ratio of 1.0. Poor agreement between the measured pitching moment and theoretical values was obtained due to difficulties in predicting the pressure distribution with heat addition on the latter parts of the thrust surface. A shift in the centre-of-pressure of approximately 10% of model chord was measured as the equivalence ratio varied from 0.0 to 1.0. For the design tested, the thrust produced was not enough to overcome drag on the vehicle, even at the highest equivalence ratio tested. Tests at higher stagnation enthalpies (up to 4.9MJ/kg) showed the lift and pitching moment coefficients remained constant with an equivalence ratio of 0.8 but the thrust coefficient decreased exponentially with increasing stagnation enthalpies. Good agreement of experimental values of lift and thrust force with predicted values was obtained for equivalence ratios of 0.0 and 0.8. Choking occurred at stagnation enthalpies of less than 3.0MJ/kg and a nozzle supply pressure of 32MPa with fuel injection at an equivalence ratio of approximately 0.8, resulting in a drag force of approximately 2.5 times the fuel-off drag force. Tests at a nozzle supply enthalpy of 3.3MJ/kg and nozzle supply pressures of 32, 26 and 16MPa were performed at equivalence ratios of 0.0 and 0.8. The fuel-off lift coefficient remained constant but the thrust coefficient increased. This is attributed to a reduction in skin friction associated with longer lengths of laminar boundary layers as the Reynolds number was decreased. The measured fuel-off lift and thrust coefficients agreed with the predicted values to within the known test flow and force prediction uncertainties. Combustion did not occur at a nozzle supply pressure of 16MPa. This work has demonstrated that overall scramjet vehicle performance measurements (such as lift-to-drag ratio and shifts in centre-of-pressure) can be made in a free piston shock tunnel.

Inlet-injection was motivated by the possibility for skin-friction reduction in the combustion chamber of a flight style, three-dimensional, scramjet engine. High Mach number flight, where skin friction in the combustion chamber is a significant proportion of the overall drag, is the regime of interest for this type of reduction. This is a result of high Mach number supersonic flow within the combustion chamber, coupled with high densities due to the compression process. The flight condition of interest was chosen to be Mach 8.0 at an altitude of 30km. This choice was dictated by near-term flight-testing capabilities. The approach was to design an inlet with a reduced contraction ratio. This would produce a relatively low-density combustion-chamber flow, that would, in turn, lead to lower viscous drag. Due to low temperatures in the combustion chamber, as a result of the reduced compression, a novel method of ignition was required. This fluid-dynamic ignition technique made use of inlet injection together with flow non-uniformities generated by the inlet. The inlet chosen for this purpose was a rectangular-to-elliptical-shape-transition inlet or REST inlet. The focus of the investigation, was therefore, to determine the potential for performance improvement using inlet injection of fuel. The general approach to the investigation was experimental, using a scramjet model consisting of inlet, combustion chamber and a truncated nozzle. Flow-path thrust-potential was used as the primary performance parameter, where the term `thrust-potential' is used to indicate the lack of full expansion. A secondary performance metric was combustion efficiency, determined by matching one-dimensional analysis to experimental pressure distributions. In addition to inlet-injection, conventional injection into the combustion-chamber was tested as the performance baseline. Based on findings from these tests, two additional methods of injection were investigated both having a combination of inlet and combustion-chamber injection. The general findings showed that inlet injection, in comparison to combustion-chamber injection, produced an increase in performance in terms of thrust-potential and combustion efficiency for supersonic combustion. This occurred over a range of equivalence ratios up to 1.0. However, the maximum thrust developed by inlet injection was limited by engine unstart. In terms of the maximum thrust-potential, combustion-chamber injection exceeded that of inlet injection but significantly higher fuelling was required and poor combustion efficiency persisted. In order to offset the limit in thrust production due to unstart, an alternative fuelling method was implemented. This took the form of partial injection of the fuel in the combustion chamber in combination with inlet injection. An increase in thrust-potential and combustion efficiency as a result of increased fuel coverage in areas of the combustion chamber, which were fuel lean under inlet-injection. A thrust potential level similar to that of combustion-chamber injection was achieved with significantly higher combustion efficiency and consequently a lower fuelling level. This type of combined-injection is an attractive option for fuel delivery at the nominal flight condition. An additional finding for combustion-chamber and combined injection was that very high equivalence ratios led to separated flow in the combustion chamber and isolator. This was a result of excessive heat release producing an adverse pressure gradient in the engine. This mode of operation showed high levels of thrust-potential at equivalence ratios in excess of 1.0. Although interesting, these findings were outside the scope of the investigation since the flow within the combustion chamber is no longer purely supersonic.

One of the major challenges of scramjet propulsion remains the generation of sufficient thrust to overcome the large drag of hypersonic vehicles. Since the viscous drag constitutes a large portion of the overall drag, its mitigation offers potential for performance improvement. Viscous drag is generated on all wetted surfaces of the vehicle but is largest in the scramjet combustion chamber, where the fluid not only has a high flow speed, but also a high density. Reduction of the skin friction drag in the combustor hence promises large improvements to the efficiency of the propulsive system. Stalker (2005) proposed a novel approach to skin friction reduction that is based on the combustion of hydrogen within the turbulent boundary layer of supersonic or hypersonic flow. An extension to the theory of van Driest II was developed that suggests that the effectiveness of this method is significantly superior to that of film cooling without combustion effects. In essence, the combustion heat release reduces the velocity gradient at the wall and the density in the boundary layer so that the momentum transfer to the wall is decreased. This work investigates the applicability of this skin friction reduction method to scramjet combustors that would operate at flight Mach numbers between 8 and 13 at altitudes between 34 and 39 km. The corresponding combustor Mach number is approximately 4.5 and the total enthalpies are between 3.6 and 7.8 MJ/kg. Experiments that directly measured the skin friction drag on the internal scramjet combustor surface were conducted in the T4 Stalker tube at The University of Queensland. A constant area, axisymmetric combustor was tested with a matching constant area, axisymmetric inlet that did not compress the oncoming flow. Therefore, the experiments were of a quasi-direct-connect nature where the inlet was used to condition the wall layer of the flow that enters the combustion chamber. The start of the combustor was formed by a step at the end of the inlet which contained an annular slot for the injection of the gaseous hydrogen fuel. Fuel was injected tangentially to the main stream flow into the circular combustor as a uniform layer underneath the established boundary layer from this annular slot. Combustion was monitored via the measurement of the axial pressure distribution in the combustor and viscous forces on the combustor were measured with a stress wave force balance. Two different inlet lengths were tested to assess the effect of the boundary layer state and thickness on the ignition and combustion of the injected hydrogen. The leading edge of the inlet was either sharp or blunt to investigate the effect of the hot gas that is contained in an entropy layer that is generated by a blunt leading edge. Finally, the diameter of the duct was varied to ensure that the experimental data was not subject to duct scaling effects. The effect on skin friction of the combustion of fuel in the boundary layer was assessed directly by measurement as well as analytically with several prediction methods. The experimental data show reductions of skin friction drag of up to 77% when stable combustion was established. A thick, turbulent boundary layer results in ignition for lower enthalpy conditions than a thin, laminar layer. The blunted leading edge configuration creates conditions that results in ignition of the injected fuel at all tested flow enthalpies and when a sharp leading edge configuration does not. Analytical predictions of the skin friction drag are in close agreement with the experimental data for fuel-off, film cooling and boundary layer combustion cases. It is demonstrated that the characteristics of boundary layer combustion do not change when the duct diameter is increased and the hydrogen mass flow rate per unit circumferential length is kept constant.

The process of shock-induced ignition has been investigated both computa- tionally and experimentally, with particular emphasis on the concept of radical farming. The first component of the investigation contained Computational Fluid Dynamic (CFD) calculations of an ignition delay study, a 2D pre-mixed flow over flat plate at a constant angle to the freestream, and through a generic 2D scramjet model. The focal point of the investigation however examined the complex 3D flow through a generic scramjet model. Five experimental test conditions were ex- amined over flow enthalpies from 3.4 MJ/kg to 6.4 MJ/kg. All test conditions simulated flight at 21000 metres ([symbol=almost equal to] 70000 ft), while the equivalent flight Mach number varied from approximately 8.5 at the lowest enthalpy, to approximately Mach 12 at the highest enthalpy condition. The presence of H2 fuel injected in the intake caused a separated region to form on the lower surface of the model at the entrance to the combustor. A fraction of the total mass of fuel was entrained in this separated region, providing long residence times, hence increased time for the chemical reactions that lead to ignition to occur. In addition, extremely high temperatures were found to exist between each fuel jet. Both fuel and air are present in these regions, therefore the chance of ignition in these regions is high. Streamlines passing through the recirculation zone ignited within this zone, while streamlines passing between the fuel jets ignited soon after entry into the combustor. The first instance of a pressure rise from combustion was observed on the centreline of the model where the reflected bow shock around the fuel jets crossed the centreline of the combus- tor. Upstream of this location the static pressure of the flow was too low for the chemical reactions that release heat to occur. The comparison between the experimental and computational results was lim- ited due to inaccuracies in modelling the thermal state of the gas in the CFD calculations. The gas was modelled as being in a state of thermal equilibrium at all times, which incorrectly models the freestream flow from the nozzle of the shock tunnel, and also the flow downstream of oblique shock wave within the scramjet model. As a result combustion occurs sooner in the CFD calculations than in the experimental result.

The process of shock-induced ignition has been investigated both computa- tionally and experimentally, with particular emphasis on the concept of radical farming. The first component of the investigation contained Computational Fluid Dynamic (CFD) calculations of an ignition delay study, a 2D pre-mixed flow over flat plate at a constant angle to the freestream, and through a generic 2D scramjet model. The focal point of the investigation however examined the complex 3D flow through a generic scramjet model. Five experimental test conditions were ex- amined over flow enthalpies from 3.4 MJ/kg to 6.4 MJ/kg. All test conditions simulated flight at 21000 metres ([symbol=almost equal to] 70000 ft), while the equivalent flight Mach number varied from approximately 8.5 at the lowest enthalpy, to approximately Mach 12 at the highest enthalpy condition. The presence of H2 fuel injected in the intake caused a separated region to form on the lower surface of the model at the entrance to the combustor. A fraction of the total mass of fuel was entrained in this separated region, providing long residence times, hence increased time for the chemical reactions that lead to ignition to occur. In addition, extremely high temperatures were found to exist between each fuel jet. Both fuel and air are present in these regions, therefore the chance of ignition in these regions is high. Streamlines passing through the recirculation zone ignited within this zone, while streamlines passing between the fuel jets ignited soon after entry into the combustor. The first instance of a pressure rise from combustion was observed on the centreline of the model where the reflected bow shock around the fuel jets crossed the centreline of the combus- tor. Upstream of this location the static pressure of the flow was too low for the chemical reactions that release heat to occur. The comparison between the experimental and computational results was lim- ited due to inaccuracies in modelling the thermal state of the gas in the CFD calculations. The gas was modelled as being in a state of thermal equilibrium at all times, which incorrectly models the freestream flow from the nozzle of the shock tunnel, and also the flow downstream of oblique shock wave within the scramjet model. As a result combustion occurs sooner in the CFD calculations than in the experimental result.

The process of shock-induced ignition has been investigated both computa- tionally and experimentally, with particular emphasis on the concept of radical farming. The first component of the investigation contained Computational Fluid Dynamic (CFD) calculations of an ignition delay study, a 2D pre-mixed flow over flat plate at a constant angle to the freestream, and through a generic 2D scramjet model. The focal point of the investigation however examined the complex 3D flow through a generic scramjet model. Five experimental test conditions were ex- amined over flow enthalpies from 3.4 MJ/kg to 6.4 MJ/kg. All test conditions simulated flight at 21000 metres ([symbol=almost equal to] 70000 ft), while the equivalent flight Mach number varied from approximately 8.5 at the lowest enthalpy, to approximately Mach 12 at the highest enthalpy condition. The presence of H2 fuel injected in the intake caused a separated region to form on the lower surface of the model at the entrance to the combustor. A fraction of the total mass of fuel was entrained in this separated region, providing long residence times, hence increased time for the chemical reactions that lead to ignition to occur. In addition, extremely high temperatures were found to exist between each fuel jet. Both fuel and air are present in these regions, therefore the chance of ignition in these regions is high. Streamlines passing through the recirculation zone ignited within this zone, while streamlines passing between the fuel jets ignited soon after entry into the combustor. The first instance of a pressure rise from combustion was observed on the centreline of the model where the reflected bow shock around the fuel jets crossed the centreline of the combus- tor. Upstream of this location the static pressure of the flow was too low for the chemical reactions that release heat to occur. The comparison between the experimental and computational results was lim- ited due to inaccuracies in modelling the thermal state of the gas in the CFD calculations. The gas was modelled as being in a state of thermal equilibrium at all times, which incorrectly models the freestream flow from the nozzle of the shock tunnel, and also the flow downstream of oblique shock wave within the scramjet model. As a result combustion occurs sooner in the CFD calculations than in the experimental result.

The process of shock-induced ignition has been investigated both computa- tionally and experimentally, with particular emphasis on the concept of radical farming. The first component of the investigation contained Computational Fluid Dynamic (CFD) calculations of an ignition delay study, a 2D pre-mixed flow over flat plate at a constant angle to the freestream, and through a generic 2D scramjet model. The focal point of the investigation however examined the complex 3D flow through a generic scramjet model. Five experimental test conditions were ex- amined over flow enthalpies from 3.4 MJ/kg to 6.4 MJ/kg. All test conditions simulated flight at 21000 metres ([symbol=almost equal to] 70000 ft), while the equivalent flight Mach number varied from approximately 8.5 at the lowest enthalpy, to approximately Mach 12 at the highest enthalpy condition. The presence of H2 fuel injected in the intake caused a separated region to form on the lower surface of the model at the entrance to the combustor. A fraction of the total mass of fuel was entrained in this separated region, providing long residence times, hence increased time for the chemical reactions that lead to ignition to occur. In addition, extremely high temperatures were found to exist between each fuel jet. Both fuel and air are present in these regions, therefore the chance of ignition in these regions is high. Streamlines passing through the recirculation zone ignited within this zone, while streamlines passing between the fuel jets ignited soon after entry into the combustor. The first instance of a pressure rise from combustion was observed on the centreline of the model where the reflected bow shock around the fuel jets crossed the centreline of the combus- tor. Upstream of this location the static pressure of the flow was too low for the chemical reactions that release heat to occur. The comparison between the experimental and computational results was lim- ited due to inaccuracies in modelling the thermal state of the gas in the CFD calculations. The gas was modelled as being in a state of thermal equilibrium at all times, which incorrectly models the freestream flow from the nozzle of the shock tunnel, and also the flow downstream of oblique shock wave within the scramjet model. As a result combustion occurs sooner in the CFD calculations than in the experimental result.