The present project investigates a new technique for using an expansion tube to produce rarefied hypervelocity gas flows. The XI expansion tube at The University of Queensland that has an exit diameter of 38 mm has been used to perform the experiments. The technique requires low initial filling pressures in the shock tube and the acceleration tube/dump tank sections. The flow at the exit plane of the acceleration tube is expanded into the dump tank as a free jet. As the flow expands into the dump tank, its density decreases rapidly but its flow speed increases only slightly. This means that rarefied high speed gas flows can be produced using this arrangement.
Two new test conditions (I and II) were established to investigate the generation of hypervelocity rarefied flows. Both conditions used nitrogen as the test gas and flows at 8.9 and 9.6 km/s with test flow durations of 55 and 50 µs respectively were generated.
Modified bar gauges were developed for the measurement of Pitot pressures in low density expansion tube flows. The bar gauges developed here are modified from the conventional designs for bar gauges. A PCB impact hammer was used to calibrate the bar gauges. Test flows with the tunnel modified to operate as a straight-through shock tunnel, generating speeds around 1.3 km/s and Pitot pressures close to 650 kPa, were used to check the calibrations of the bar gauges. The bar gauges have a flat measurement disc of 9 mm diameter and their measurement is the average pressure acting on the disc. Calibration results indicate that the true Pilot pressure could be obtained by multiplying the measured average disc pressure by a factor of 1.08. The overall uncertainty of the Pi tot pressure measured is estimated to be ± 7% for Pitot pressures of order 600 kPa.
Computational Fluid Dynamics (CFD) packages were used to model the XI expansion tube in order to calculate the free-stream conditions in the test section. L1d is used to simulate the flow along the compression tube and shock tube. The acceleration tube is modelled using an axisymmetric MB_CNS code. The flow in the test section and dump tank is also simulated using MB_CNS. Calculations are made for chemical equilibrium and non-equilibrium with and without viscous effects. Computed results indicate that the established conditions can produce rarefied hypervelocity gas flows starting at 200 mm and 150 mm downstream from the tunnel exit for conditions I and II, respectively.
The axial Pilot pressure distributions in the test section were obtained by placing the modified bar gauges at different locations along the centre line of the tunnel inside the dump tank. Reasonable agreement was obtained with the Pilot pressures calculated using the CFD simulations. This was taken as a validation of the simulations and the free-stream conditions from the simulations were used for analysis of all further results.
The direct measurements from the bar gauges were also interpreted as the drag on a flat-faced cylinder. By combining the drag data and the estimated free stream conditions using computational calculations, drag coefficients were determined. The results are compared with theoretical continuum and free-molecular drag coefficients for flat-faced cylinders. Results indicate that the drag coefficient deviates more from the continuum level for the more rarefied Condition II than for Condition I.
Experimental measurements of stagnation point heat transfer rate were made on two geometric models- hemispheres and bluff-faced cylinders. Thin-film gauges were installed in the test models for measurements. The measured data were combined with the CFD free-stream conditions to determine the Stanton numbers at different degrees of rarefaction. The results were compared with other, low-enthalpy rarefied flow results to see if any high temperature rarefied gas effects were present. The present results agree with the equivalent low enthalpy rarefied flow data to within experimental uncertainty, indicating that real-gas effects are not significant for the present rarefied flow conditions.
A scaled test model based on the re-entry flight vehicle, Fire II, was manufactured. A test condition was established to simulate the flight conditions and test model was placed in the test section for drag measurements. The free-stream conditions were estimated using the techniques developed for conditions I and II. The measured drag coefficients agreed with the flight data to within ± 10% in the simulated flow regime.