Experimental, analytical and numerical studies of the hypervelocity flow over flat plates and rearward-facing steps have been performed. Experiments were conducted in flows of air in the X2 superorbital expansion tube and the T4 shock tunnel. The total enthalpies of the flows ranged from 5 to 54 MJ/kg. Heat flux and base pressure measurements obtained downstream of the rearward-facing step were made for flows at higher total enthalpies than have been reported previously. Also, the spatial resolution in the heat flux distributions downstream of the step is higher than in previously reported experiments in hypersonic flows.
The main focus of the experimental work is on the results from the X2 expansion tube. A suborbital flow condition at a total enthalpy of 26 MJ/kg and a superorbital flow condition at a total enthalpy of 54 MJ/kg were targeted. The uncertainties in the freestream flow conditions in expansion tubes at such flow enthalpies have been high in the past. Working in conjunction with Stewart , numerical modelling of the flow in the shock tube and acceleration tube of X2 was performed for equilibrium and finite-rate chemistry. It is shown that the techniques developed give good agreement with measured static and Pitot pressures at the exit of the acceleration tube. There are shot-to-shot variations in conditions for nominally repeat shots in X2 and it is time consuming to do the numerical simulations for each test. Therefore, a technique was used to perturb the flow parameters calculated using the numerical simulations at nominal conditions to account for the shot-to- shot variations.
The short duration of the test flow in an expansion tube (approximately 50 to 85 μs at the present conditions in X2) means that it is important to confirm that both the boundary layer flow and the separated flow behind the step can be established. The establishment of the flow is investigated using results from the thin-film heat flux gauges installed in the test models. The features observed in the experimentally measured heat fluxes during die flow establishment are confirmed with results obtained from transient Navier-Stokes simulations. The results show that flat-plate boundary layers can be established in X2 at up to 130 mm from the leading edge for both the suborbital and superorbital flow conditions. A new empirically based flow establishment criterion is proposed for the flat plate flow in an expansion tube. The experimental heat flux distributions and simple separated flow models indicate that steady separated flows downstream of a step can be established in high enthalpy expansion tube flows.
In preparation for the rearward-facing step flow experiments, heat flux was measured over a 140 mm long flat plate at the X2 suborbital and superorbital conditions. The heat flux distribution over a flat plate was measured with a higher spatial resolution, at higher total enthalpies, than in previous investigations. The influence of real gas effects, such as recombination and surface catalycity, were found to be minimal at the X2 flow conditions, with the experimental results agreeing well with perfect gas models of the heat flux.
Dimensional analysis was used to identify the important parameters influencing the flow over a rearward-facing step in high enthalpy flows of a given gas. The non-dimensional parameters required for similarity are the freestream Mach number, M∞, the length upstream of the step L to step height h ratio, L/h, the Reynolds number at the step, ReL the wall to stagnation temperature ratio, Tw/T0, and the gas-phase Damkohler number, Ωg. In high Mach number and low Reynolds number flows, such as those encountered during re-entry, an analysis shows that the important parameters to match are the hypersonic viscous interaction parameter,*, the hypersonic small disturbance parameter, M∞T and the DamkÖhler number, Ωg. By matching these parameters, it is shown that the heat flux distributions downstream of the step can be correlated between different experimental facilities. This is the first known instance that such a correlation has been achieved between different experimenters' data. The influence of these parameters on the heat fluxes, base pressure and reattachment distance downstream of the step are examined both theoretically and experimentally.
*, M∞T and Ωg were matched between a 26 MJ/kg condition in X2 and a 5.6 MJ/kg condition in T4. The heat flux distributions behind the steps at the two flow conditions agreed to within experimental error, indicating the efficacy of the scaling parameters and that real-gas effects on separated flows at these conditions are minimal.
The effect on heat flux distributions of varying the Reynolds number at the step is examined using results obtained in the T4 shock tunnel. The results show that varying the Reynolds number did not significantly change the form of the heat flux distribution downstream of the step for the flow conditions analysed.
*NB: Equation can not be replicated from original thesis.