Hypervelocity scramjet engines promise to be an efficient air-breathing propulsion technology capable of performing as part of an access-to-space system. However, difficulties arise when operating a scramjet engine at high altitudes and flight Mach numbers, notably above Mach 10.0. These difficulties arise due to a reduction in captured air flow by the engine, residence time of flow within the engine, and turbulence intensity which facilitates effective fuel-air mixing. Oxygen enrichment is a technique proposed to augment scramjet thrust under these adverse conditions, whereby a small amount of oxygen is mixed with fuel before injection into the engine combustor. Previous experimental studies conducted by Razzaqi and Smart (2011) have shown that Oxygen Enrichment (OxEn) enhances the combustion efficiency of a scramjet by an amount equal to the enrichment percentage. These findings suggest that oxygen enrichment should be considered in the design of scramjet engines. However, the previous studies do not offer sufficient information about the effects of OxEn on the turbulence characteristics of the hypervelocity flow and consequently the physical mechanisms that drive the mixing of fuel and oxygen. Effective design of a scramjet engine which utilises oxygen enrichment requires a thorough understanding of these physical processes. A simplified scramjet engine configuration ground tested at Mach 12.3 and a dynamic pressure of 40.5 kPa, based upon the aforementioned experiments, has been selected to perform as the basis for several numerical simulations to assess the benefits of oxygen enrichment upon fuel-air mixing driven by compressible shear. The Reynolds number of the free shear flow, based upon the core flow conditions ingested by the isolator and the height of the centre-body is 56,650. This engine configuration comprises a planar duct with an intrusive centre-body wherein combustible propellant is injected through a rear-facing slot. The fluid flow which forms down-stream of injection is referred to as a mixing wake
. Reynolds Averaged Navier-Stokes (RANS) is the current standard for the computational analysis of the characteristic turbulent, supersonic fluid flows within a scramjet engine during flight. RANS provides information about the mean flow field as well as key modelled turbulence parameters. Large Eddy Simulation (LES) is a high fidelity numerical methodology which resolves the large scale unsteady behaviour of turbulence flow, enabling a statistical description of the turbulence to be established, but at increased computational expense. Both Reynolds-Averaged Navier-Stokes and Large Eddy simulations of this mixing wake flow are performed in this work. Results from RANS simulations of the simplified scramjet engine have shown that combustion between fuel and ingested oxygen is promoted by oxygen enrichment. This enhancement has been attributed to greater shearing velocity difference across the mixing layers that form between the ingested and injected fluid streams, as oxygen enrichment results in a slower moving injectant. Consequently, a greater production of turbulence kinetic energy enhances mixing of fuel and ingested air within the mixing layers. Analysis of the turbulence model parameters indicates compressibility effects influence the turbulent flow behaviour, and that turbulence within the mixing layers may not be fully developed.
To test the validity of the previous findings and gain more insight into the effects of oxygen enrichment on the turbulent flow structure, LES of the hypervelocity mixing wake were performed. The LES results showed that the initial mixing layers which develop between the injected fuel and ingested air streams is strongly influenced by compressibility. This is evident from the near-stream-wise oriented vortices which form and slow transverse growth rate of the layers and visible shocklets in the near field. However, reducing compressible effects coincide with a transition in the behaviour of the wake, characterised by dominant large scale span-wise vortices similar to those expected within an incompressible wake. This turbulent regime significantly boosts the spreading rate of the wake and entrainment of oxygen. LES of oxygen enriched fuel injection found that the mixing of ingested oxygen and injection hydrogen was slightly reduced in comparison to the pure fuel injection case. Despite increasing the turbulence production within the initially developing mixing layers, oxygen enrichment caused a delay in the wake roll-up. The dynamic effect of injecting a greater flow rate is to further displace the mixing layers away from the fuel rich core flow. Additionally, greater compressible effects suppress the generation of cross-stream turbulent disturbances that activate the wake roll-up. Mixing of ingested oxygen and fuel remains limited whilst wake roll-up is delayed. Direct comparison of RANS and LES results has shown that the former method reproduces the mean turbulent behaviour of the initial, highly compressible mixing layers. However, the RANS method fails to predict the effects of wake roll-up and subsequently under predicts the mixing of ingested oxygen with injected fuel. This under prediction is exacerbated in the pure fuel injection case as the wake roll-up occurs more rapidly than in the oxygen enriched fuel case. Consequently, RANS predicts that oxygen enrichment augments fuel-air mixing which conflicts with the findings of LES.