Application of a vectorized particle simulation to the study of plates and wedges in high-speed rarefied flow

The scientific community has placed a great deal of emphasis on understanding the unique phenomena associated with hypersonic, rarefied flowfields in order to successfully design future high-speed, high-altitude flight vehicles. Scientists are interested in using particle simulation methods to recreate the combinations of speeds and rarefaction associated with flight through the upper atmosphere. Particle methods build up a macroscopic flowfield based on the interactions of thousands or millions of particles that behave as molecules on the microscopic level.; Unfortunately, most particle methods employ basic algorithms that are incompatible with the architecture and resources of most high-speed computers. The Stanford Particle Simulation Method, developed by Baganoff and McDonald, overcomes these disadvantages by paying strict attention to details incorporating the advantages of these machines while retaining correct macroscopic physical behavior.; Since this method is still quite new, it must be adequately tested in order to gain acceptance in the scientific and engineering communities. The purpose of the current investigation is to use the advantages afforded by the Stanford Particle Simulation Method to help establish the method's legitimacy by thoroughly studying a relatively simple two-dimensional flowfield: high-speed, low-density, viscous flow over a flat plate at zero incidence. The freestream Knudsen number is varied such that the simulation's solutions range from continuum flow all the way to the free-molecule limit.; Issues concerning proper implementation of boundary conditions for simulation wind tunnels and geometry models are discussed, and novel approaches are used to develop a diffuse adiabatic boundary condition and make necessary alterations for highly-rarefied and free-molecule conditions. A review of experimental data is presented, and a new parameter is developed using aspects of kinetic theory and demonstrated to correlate hypersonic, rarefied drag and heat transfer data better than commonly-used alternatives. Surface-flux quantities, such as drag and heat transfer, obtained by the simulation are compared to experimental data whenever possible. Aspects of the simulation's surface-flux and flowfield results are compared to theories, where available. In addition, a more complicated hypersonic, rarefied, two-dimensional flow over a wedge was studied and compared to the experimental results of Batt.