Vincent Wheatley
Febuary 1999
The orbit of the Magellan probe around Venus was circularized using aerobraking in the upper atmosphere. This manoeuvre expended less than 5% of the propellant that would have been required if only thrusters were used. The flight speed was 9 km/s and the flow over the satellite was in the transition regime between continuum and free molecular flow [5]. Magellan was eventually allowed to fly deeper into the atmosphere of Venus. During this last phase of the flight when the flow had a higher density but was still rarefied, the thruster firing sequence required to maintain the stability of the satellite indicated that the aerodynamic forces differed from the predicted values. During the Mars Pathfinder mission similar aerobraking manoeuvres were used with flight speeds exceeding 10 km/s. In view of the anomalous Magellan results an improved understanding of the aerodynamics of high-speed, rarefied flow is required.
There is a lack of experimental data for the aerodynamics of rarefied gases in comparison with other areas of fluid dynamics. The Direct Simulation Monte-Carlo (DSMC) method [2], whereby the motions and collisions of gas molecules are simulated on a computer, is the standard numerical method for rarefied flows. DSMC has assumed the role of surrogate for experiments [4]. Molecule-molecule and molecule-surface collision models, which can involve exchange of energy amonst translational, rotational, vibrational and chemical energy modes have been developed for DSMC. Some assumptions of near equilibrium conditions are often incorperated into the derivation of these models. Aerobraking manoeuvres present extreme conditions where collision models can be expected to be severely tested, it is very important that the accuracy of DSMC be assessed in these conditions. Present experimental facilities for rarefied gas flows are limited to stagnation temperatures of about 2000 K and hence are limited to test speeds of 1.5 to 2.0 km/s. The development of a high-speed test facility which spans the range from continuum to rarefied flow will represent a significant increase in the range of experimental testing for DSMC [6].
A pilot study into the development of a rarefied hypervelocity test facility has recently been carried out [6]. The X1 expansion tube of the Centre for Hypersonics at The University of Queensland was proposed for producing the flow. A conical nozzle was attached to the exit of the tube to produce a flow of Argon in the transitional regime at 8.8 km/s with a test flow duration of 60 microseconds. A 50 mm diameter central core flow was generated with a Pitot pressure variation of 30%. Overall the flow was found to be not very useful. Also, there were found to be significant differences between the experimental data and the results from a CFD simulation.
The broad goal of my thesis is to continue the investigation into the development of a test facility that produces a uniform flow spanning the transition regime from continuum to rarefied. This will be acomplished primarily by developing adequate computational models of the flow through the X1 and small shock tunnel (SST) test facilities. The SST [1], with its simple operation and non-reacting flow, will be used as a test bed for experimental and computational techniques to be applied to the X1. As hypervelocity flows cannot be realised in the SST facility, experimental work in this facility is only a precursor to that in the X1.
Complete models of both the SST and X1 will initially be developed using the CFD code MB-CNS [3]. Although the tunnel geometries are already well defined, there are a couple of modelling issues that need to be explored.
On the experimental side, there are also a couple of issues to explore. Most of the instrumentation available has been designed for operation at much higher pressures and, at low pressures, the signals tend to drift and have large amounts of noise. Specifically, the Pitot probes used in the pilot study are not ideally suited to use in rarefied flow. An investigation will be carried out to establish which types of transducers will allow accurate measurements to be taken in the flows produced in the X1 and SST facilities. This will allow the results of the CFD simulations to be compared with accurate experimental data. So far, MB-CNS is unproven at simulating rarefied flows. If it proves inaccurate, one possibility for improving the computional model is to use MB-CNS where the flow is in the continuum regime and then continue the the simulation using DSMC.After initial simulations have been completed, a decision will be made on whether to proceede with the coupled simulation.
References