The response surface methodology for BLISS achieves the desired improvements while retaining key attributes of previous versions of BLISS: the autonomy of the black box optimizations and the clear separation of the system variables from the potentially numerous local variables.  The response surface formulation of BLISS was successfully demonstrated on a simplified conceptual design of a supersonic business jet (1995 AIAA graduate team aircraft design topic).

In addition to changes in the overall optimization methods of BLISS, subsystem fidelity was enhanced, accompanied by the necessary modifications to the data flow between subsystem analyses.  Documentation of these modifications that have not as yet been tested is included as a reference for future research.

This research was conducted in partial satisfaction of the requirements for the degree of Master of Science with The George Washington University.  Professor Robert R. Sandusky (ext. 41982) is coordinating this presentation which is expected to last approximately 1 hour.

Implementation of the Lighthill theory in Jet3D centers on the modeling of two-point space-time correlations.  Mean flow correlations for velocity and density are modeled using a Taylor series expansion, written in terms of local mean flow gradients.  Turbulent velocity correlations are separated into space and time factors, and modeled using a combination of Gaussian-type exponential functions and quadratic functions.  The proposed model satisfies basic correlation requirements and scales with characteristic turbulence length and time scales. 

Two cases were used to validate and test Jet3D, based on the experimental jet noise and flow field data Yamamoto et al. obtained for a supersonic axisymmetric nozzle.  In each case, Jet3D noise predictions were in very good agreement with experimental data for observer angles from 50° to 140°, but diverged from experimental data at shallow angles to the jet axis, overpredicting the OASPL by 8­12 dB at an observer angle of 20°.  1/3 octave band frequency spectrum SPL results were in good agreement with experimental data at forward and sideline observer angles, but diverged away from experimental data for observer angles less than about 70°.  As the observer point approached the jet axis, Jet3D overpredicted SPL levels and the SPL spectra were biased (or shifted) to higher frequencies than the experimental data. 

The disparity between Jet3D predictions and experimental data at rearward observer angles is likely due to sound-flow interaction effects (not modeled in Jet3D).  Simple ray acoustics analysis suggests that mean flow convection and refraction affects the sound radiated to observer angles less than about 73-75°, which falls in line with the observed region in which experimental data and the Jet3D prediction differ.

This research was conducted in partial satisfaction of the requirements for the degree of Doctor of Science with The George Washington University.  Dr. John L. Whitesides (ext. 41982) is coordinating this presentation which is expected to last approximately 1 hour.

This research was conducted in partial satisfaction of the requirements for the degree of Master of Science with The George Washington University.  Dr. John L. Whitesides (ext. 41982) is coordinating this presentation which is expected to last approximately 1 hour.

This thesis presents a method to size a solar concentrator-Stirling engine system for a conceptual High Altitude Long Endurance (HALE) aircraft and to determine the power available.  Concentrated solar energy is input to the hot side of a Stirling Engine and a radiator rejects heat to the environment.   The development of an optimization approach incorporating the existing AirCraft SYNThesis program, ACSYNT, is explained with emphasis on concentrator-Stirling engine integration. The new Propulsion/Energy Storage module and modifications to the existing Weights and Aerodynamics modules are all incorporated within the ACSYNT program to permit analysis of all input parameters for the lowest weight aircraft.

To test the method, a design study is performed for an example HALE Unmanned Aerial Vehicle (UAV) having a payload of 9.0 kilograms and an endurance of 30 days at 32 degrees N latitude and 17,000 - 21,000 meters altitude.   Both, a Stirling-powered baseline vehicle and a PV cell-powered baseline vehicle are developed.  The PV cell-powered baseline is used to validate sections of the methodology and serve as a reference for comparison.  Sensitivity studies are presented for key independent design parameters. 

This research was conducted in partial satisfaction of the requirements for the degree of Master of Science with The George Washington University.  Professor Robert Sandusky (ext. 41982) is coordinating this presentation which is expected to last approximately 1 hour.

This research was conducted in partial satisfaction of the requirements for the degree of Master of Science with The George Washington University.  Dr. Michael K. Myers (ext. 41982) is coordinating this presentation which is expected to last approximately 1 hour.

The simulation consists of a MATLAB program of the mathematical model that describes the dynamics of the FASER aircraft using the rigid body equations of motion in standard form.  The main body of work performed for this thesis was obtaining the FASER characteristics necessary to develop the non-linear simulation, which included measuring the moments of inertia accurately and calculating the aerodynamic characteristics through theoretical methods.  To calculate the aerodynamic coefficients, a MATLAB program that uses theoretical methods to estimate the stability and control derivatives and the propulsion effects based on the geometry of the aircraft was created. 

The simulation performs as expected and it received a positive response from the people that tested it at the Langley Research Center.

This research was conducted in partial satisfaction of the requirements for the degree of Master of Science with The George Washington University.  Dr. V. Klein (ext. 41982) is coordinating this presentation which is expected to last approximately 1 hour.