|
Robert Fares and Xiaomin Chen
Washington University in St. Louis Department of Mechanical, Aerospace, and Structural
Engineering
Advisor: Dr. Ramesh K. Agarwa
Abstract
In order to reduce fluid energy losses through a diffuser, it is important to maximize the
recovery of fluid static pressure through the diffuser. The shape optimization of an axially
symmetric diffuser to achieve maximum pressure recovery is considered in this paper by
using a Genetic Algorithm (GA). A genetic algorithm developed in Java is employed
with the CFD software Fluent to perform the optimization of the diffuser shape. The
fitness function for the optimization is the pressure recovery factor (Cp) of the diffuser,
for prescribed inlet velocity and outlet pressure. Results from GA optimization are
compared to the optimization results for the same diffuser given by Ghosh et al.
employing an evolutionary optimization algorithm. After the validation of the GA
optimization approach for the axisymmetric diffuser, GA is applied to optimize the shape
of a 3D hydro-turbine draft tube of a hydroelectric power plant in Keokuk, Iowa. The
fitness function for the optimization of the draft tube was the pressure at the inlet, which
was minimized to increase the performance and efficiency of the hydro-turbine power
plant. Results from the optimization show a marked increase in performance. These
results demonstrate that the GA based shape optimization method is effective and can be
successfully employed to optimize complicat ed real world configurations used in
industrial applications for increasing performance, reducing cost, and saving energy.
Robert L. Fares
Robert is in his last year of undergraduate study in the Department of Mechanical, Aerospace,
and Structural Engineering at Washington University in St. Louis. He will attend the University
of Texas at Austin in pursuit of a PhD in Mechanical Engineering as a Cockrell Fellow
beginning fall of 2010. He aspires to contribute to the thermal fluids field at the University of
Texas and find a career at the research and development level in the energy industry.
Xiaomin Chen
Xiaomin is a doctoral student in the department of Mechanical, Aerospace and Structural
Engineering in Washington University in St. Louis. He is working on his doctoral dissertation in
the area of shape optimization.
|
