With the high costs of flight testing, especially at hypersonic speeds, ground based facility testing of scramjets becomes an attractive option. The expansion tube is the only facility currently that can offer full flight property duplication at the total pressures and total enthalpies required, while maintaining correct chemical composition. Due to difficulties with short test times and unsteady flow phenomena, scramjet testing in these facilities has not been thoroughly investigated. This study examines these issues, in order to explore the practicality of testing a full ’tip to tail’ scramjet engine at a true flight replication condition in an expansion tube facility. An investigation was initially undertaken on the large X3 expansion tunnel facility to maximise test time and core flow, aimed at producing a 30 km altitude, Mach 10 flow condition. This was identified as the limitation point of the T4 reflected shock tunnel, which has generally been accepted to produce reliable scramjet data for propulsion tests. Using a condition that is also able to be produced in the T4 facility, will permit direct comparison of data between the two facilities in the future, providing confidence in results from expansion tube facilities. Both experimental measurements and numerical calculations showed that the limitation of the test time was due to large boundary layer growth after transition, which engulfed the entire core flow 200 μs into the test time. This phenomenon is likely to affect all scramjet duplication conditions in expansion tubes, as the flow properties are conducive to boundary layer transition occurring. Two solutions where proposed and investigated in order to overcome the flow disruption caused by boundary layer transition; the use of a steady expansion nozzle at the acceleration tube exit; the use of hydrogen as an accelerator gas. Since the smaller X2 facility had a Mach 10 steady expansion nozzle and X3 was decommissioned for the free piston driver to be upgraded, the investigation was shifted to X2. Due to a restricted test time of 550 μs, the static pressure of the flow condition was increased to allow a reduction in the length of the scramjet (pressure-length scaling). A combination of experimental and numerical calculations of the facility was used to define the flow properties. With the confidence of overcoming the phenomenon associated with boundary transition in the X2 facility, numerical modelling of the X3 facility with a steady expansion nozzle was then undertaken to show a 1 ms condition could be produced. Although initially promising, the hydrogen accelerator gas solution requires further investigation. A two dimensional scramjet was designed with upstream injection for testing in X2. This was a three shock inlet with a constant area combustor and a planar thrust surface. Since the flow condition involved changes in flow properties during the test time, aninvestigation of the appropriateness of a quasi steady analysis was undertaken. Using a fuel off simulation of the scramjet duct with the transient inflow properties from the X2 facility nozzle exit, the convective terms for pressure were shown to be two orders of magnitude larger than local terms indicating the dominance of the convective terms change in flow properties at any location allowing quasi-steady flow to be assumed. A normalisation procedure was developed to deal with the transient nature of the data and to accurately represent the axial progression of the gas through the duct. The numerical simulations were also used to show that both flow establishment was achieved and that impulsive starting of the intake would occur. Experimentation with the scramjet using static pressure measurements throughout the body side of the engine provided verification of supersonic combustion. This was verified by the doubling of the static pressure from the start to the end of the combustor for an air test gas, whereas with a nitrogen test gas no significant change in pressure occurred. Effects of fuel equivalence ratio, injector size and cowl position were also investigated. A net inviscid thrust was predicted, using the quasi-steady flow analysis, indicating a specific impulse of 183 s. This work provides evidence to validate the use of expansion tube facilities for experimental testing of scramjets at flight duplication conditions. Limitations due to boundary layer transition flow effects has been shown to be avoidable. Numerical simulations of the facilities showed good agreement with experimental measurements, allowing definition of freestream properties and can now be applied to further scramjet conditions with confidence. Stable, supersonic combustion was shown to be produced for these expansion tube conditions. Coupling the transient simulation of the flow condition with a numerical calculation of the fuel off experimental scramjet has been useful in both verification of the design and performance predictions. Appropriate techniques have been presented to analyse scramjet pressure and thrust measurements where transient effects are present in the freestream.

L’exploration des quatre planètes géantes gazeuses, Jupiter, Saturne, Neptune et Uranus, est importante pour comprendre l’évolution de notre système solaire et plus généralement de l’univers. Les sondes entrant dans l’atmosphère des géantes gazeuses ont des vitesses de 20 à 50 km/s, largement supérieures aux vitesses d’entrée atmosphérique sur les autres planètes du système solaire. Il s’agit d’un problème complexe car les conditions d’entrées sont brutales et les vitesses associées dépassent largement les capacités des installations d’essai au sol actuelles. Cette thèse examine la possibilité de simuler expérimentalement les conditions d’entrées proposées pour Uranus et Saturne à 22.3 et 26.9 km/s avec un tube d’expansion à piston libre. D’abord, la possibilité de simuler les conditions directement en recréant la vitesse d’entrée réelle a été étudiée. Il a été trouvé qu’il était possible de simuler l’entrée d’Uranus mais seulement avec de grandes incertitudes. Pour cette raison, il a été proposé d’utiliser une substitution du gaz d’essai établie, dans lequel soit le pourcentage d’hélium dans l’atmosphère H2/He est augmenté, soit l’hélium est remplacé par du néon, un gaz noble plus lourd. Cela permet de simuler uniquement les conditions postchoc des entrées. Théoriquement, il a été constaté que ces substitutions permettaient de simuler l’entrée Uranus ou Saturne, ce qui a été confirmé expérimentalement à l’aide d’hélium. Notant l’intérêt actuel d’envoyer des sondes d’entrée atmosphérique vers ces deux planètes, cette étude a démontré que les capacités expérimentales requises sont disponibles pour la réalisation d’expériences simulées avec les modèles d’essais. / Exploration of the four gas giant planets, Jupiter, Saturn, Uranus, and Neptune, is important for understanding the evolution of both our solar system and the greater universe. Due to their size, flight into the gas giants involves atmospheric entry velocities between 20 and 50 km/s. This is a complex issue because the entry conditions are harsh but the related velocities are mostly beyond the capabilities of current ground testing facilities. As such, this thesis examines the possibility of experimentally simulating proposed Uranus and Saturn entries at 22.3 and 26.9 km/s in a free piston driven expansion tube, the most powerful type of impulse wind tunnel. Initially, the possibility of simulating the conditions directly by re-creating the true flight velocity was investigated. It was found to be possible to simulate the 22.3 km/s Uranus entry, but not without large uncertainties in the test condition. For this reason, it was proposed to use an established test gas substitition where the percentage of helium in the H2/He atmosphere is increased, or the helium is substituted for the heavier noble gas neon. This allows just the post-shock conditions of the entries to be simulated. Theoretically it was found that these substitutions allowed both Uranus or Saturn entry to be simulated, which was confirmed experimentally using helium. Noting the current interest in sending atmospheric entry probes to both of these planets, this study has demonstrated that the required experimental capabilities are available for performing simulated experiments using test models.