The overall performance of a scramjet-powered vehicle not only depends on the performance of individual components but also on how the components interact with one another. Because scramjet engines must be integrated into the vehicle design, the optimization of the design of one component may detrimentally offset the performance of other components. This thesis addresses the optimization of the thrust nozzle of a scramjet-powered vehicle and shows how the optimization must include the integration of the nozzle into the overall vehicle design. The basic scramjet vehicle configuration chosen for the study is the somewhat conventional design of a vehicle with an axisymmetric centrebody, where fuel and payload may be stored, and a quasi-axisymmetric cowl. Six internal intake-combustor-nozzle modules are arranged around the centrebody. Paull et al. (1995) tested a configuration of this type (referred to as model 0) and demonstrated that, at some conditions, a net positive thrust could be produced. They suggest that the thrust nozzle of the vehicle has potential for design change that could lead to significantly improved performance. The main aim of the present study is to test this hypothesis. An important decision that was made early in the present project was that the research would focus on unfuelled operation of the vehicle design. This still allowed the influence of integration of the optimized nozzle to be studied but removed the complications introduced by combustion processes. In order to integrate optimization of design of the thrust nozzle, it was necessary to analyze the performance of the complete scramjet vehicle design. A simple analysis methodology that captures the important physical processes occurring in the flow through and around the vehicle is necessary so that rapid calculations can be made in an iterative optimization program. Therefore, the force prediction methodology that was developed used a combination of simple hypersonics theories and inviscid 3-D CFD modelling instead of using a full 3-D Navier-Stokes equation solver to minimize the calculation time. The pressure forces and viscous forces on the model were calculated separately for each component of the scramjet vehicle designs. The theory of van Driest (1956) for skin friction drag in turbulent boundary layers was found to produce best agreement with measurements. Therefore, it was employed to estimate the turbulent skin friction drag in the present research. To validate the current force prediction methodology, the net axial drag force on an unfuelled quasi-axisymmetric scramjet model derived from the design of Paull et al. (1995) and designed for operation at Mach 6 (model 1) was measured in the T4 Stalker tube at The University of Queensland using a single component Stress Wave Force Balance. (The design of Paull et al. (1995) is referred to here as model 0.) Tests were performed with Mach 6, Mach 8, and Mach 10 nozzles attached to the end of the shock tube. In order to get a wide range of flow conditions the nozzle-supply enthalpy was varied from 3 to 10 MJ/kg and the nozzle-supply pressure from 35 to 45 MPa. A reduction of the drag coefficient of model 1 was observed with decreasing nozzle-supply enthalpy for each of the tunnel nozzles tested. The performance of model 1 was analyzed using the force prediction methodology. Generally, the force prediction results were in good agreement with experimental results. The results indicate that the internal intakes provide 50% of the total drag. The skin friction drag in the combustion chambers and the nozzles account for 30% of the total drag. In order to investigate the influence on the overall performance of the vehicle obtained by improving the nozzle performance, optimization and parametric studies of quasi-axisymmetric scramjet nozzle designs were conducted. The vehicle which was optimized in this study is of a similar configuration to the model used in Paull et al. (1995). The vehicles are optimized for minimum fuel-off net axial drag for a design flight Mach number of 8 using the force prediction methodology and the Nelder and Mead (1965) optimization algorithm. The optimization studies focused on the combustion chamber and the nozzle. Therefore, the shape of the conical forebody and the intake were not changed. The external flow over the cowl was taken into account during the optimization studies. The results showed that a long nozzle with a large external cowl deflection angle, which allowed the nozzle area ratio to be increased, did not give better performance than a short nozzle with a smaller area ratio. This was due to the competing effects of increased external drag on the cowl and increased nozzle thrust as the nozzle area ratio increased. The optimum shape gave limited improvement compared with that of Paull et al. (1995). While fuelled performance of the vehicle was not the focus of the present investigation, a preliminary theoretical study of fuelled operation was performed. A parametric study to vary the nozzle length and external cowl deflection angle was performed for different flight Mach numbers. The results indicate a larger nozzle and higher external cowl deflection angle are appropriate for fuelled cases compared with unfuelled cases. The net axial force on a model with a geometry close to the optimum design (model 2) was measured in the T4 shock tunnel in order to check that the optimization procedure was valid. Model 2 showed generally better performance than other models experimentally. For the Mach 6 nozzle tests, although model 2 has some performance losses due to the spillage of flow around the intakes, model 2 shows approximately a 20% lower drag coefficient than model 1 and shows slightly better performance than model 0. For all test conditions, a break-down of the components of the drag coefficient indicates that the nozzle of model 2 produces approximately three times more thrust than the nozzle of model 0 and approximately twice more than that of model 1. For the Mach 8 nozzle tests, model 2 has approximately a 20% lower drag coefficient than model 1. However, for the Mach 10 nozzle tests, no significant differences between the models were observed in the measurements. Finally, the measurements and optimization study indicate that when model 2 is fuelled, it could be expected to be capable of cruise up to Mach 8 because of its very effective nozzle.
| Identifer | oai:union.ndltd.org:ADTP/254016 |
| Creators | Katsuyoshi Tanimizu |
| Source Sets | Australiasian Digital Theses Program |
| Detected Language | English |
Page generated in 0.0025 seconds