Browsing by Author "Ferguson, Alex"

Now showing 1 - 2 of 2

Accurate True Direction Solutions to the Euler Equations Using a Uniform Distribution Equilibrium Method
(2010-07-01)
Ferguson, Alex
;
Smith, Matthew R.
;
Wu, J. -S.
;
機械工程學系
;
Department of Mechanical Engineering
A novel approach for the use of multiple continuous uniform distributions for reconstruction of the Maxwell-Boltzmann equilibrium probability distribution function is used for the solution of one and two dimensional Euler equations. The Uniform distribution Equilibrium Flux Method (UEFM) is a kinetic-theory based flux solver which calculates true directional, volume to volume fluxes based on integration (over velocity space and physical space) of a sum of uniform probability distribution functions working to approximate the equilibrium distribution function. The resulting flux expressions contain only the Heaviside unit step function and do not require the evaluation of the Exponential or Error Functions. The proposed method is verified using a series of one and two dimensional benchmarks and is shown to provide a higher level of accuracy (for a given computational expense) when compared to the similar Quiet Direct Simulation (QDS) method.
Rapid optimization of blast wave mitigation strategies using Quiet Direct Simulation and Genetic Algorithm
(2010-06-01)
Smith, Matthew R.
;
Kuo, Fang-An
;
Hsieh, Chih-Wei
;
Yu, Jen-Perng
;
Wu, Jong-Shinn
Presented is a rapid calculation tool for the optimization of blast wave related mitigation strategies The motion of gas resulting from a blast wave (specified by the user) is solved by the Quiet Direct Simulation (QDS) method - a rapid kinetic theory-based finite volume method The optimization routine employed is a newly developed Genetic Algorithm (GA) which is demonstrated to be similar to a Differential Evolution (DE) scheme with several modifications lit any Genetic Algorithm, individuals contain genetic information which is passed on to newly created individuals in successive generations The results from unsteady QDS simulations are used to determine the individual's "genetic fitness" which is employed by the proposed Genetic Algorithm during the reproduction process The combined QDS/GA algorithm is applied to various test cases and finally the optimization of a non-trivial blast wave mitigation strategy Both QDS and the proposed GA are demonstrated to perform with minimal computational expense while accurately solving the optimization problems presented (C) 2010 Elsevier BV All rights reserved