Multiobjective evolutionary optimization of a compressor stage using a grid-enabled environment

Abstract

Multiobjective and multidisciplinary optimization with high-fidelity analysis is becoming an essential factor in the design of turbomachinery blades. Grid-computing environments enable the solution of optimization problem requiring large computational resources. Here, the Geodise computing system is used as a Grid-enabled tool, which realizes the client functionalities for a Globus Grid service in the Matlab environment. It allows users to handle their computing jobs on Grid-enabled machines as Matlab functions. As Matlab includes various useful functions to analyze and visualize data, and to integrate several components via its scripting language, Matlab is used as the main framework of the work presented. In this research, single stage rotor/stator blades for a multistage compressor are optimized to improve aerodynamic performance in terms of efficiency, blockage and loss, while satisfying four aerodynamic constraints to maintain the flow similar to a baseline geometry. To identify the trade-off between three objectives with a reasonable number of function evaluations, the Adaptive Range Multi-Objective Algorithm is adopted as the optimizer. The benefits of constrained multi-objective optimization of single-stage blades by Evolutionary Algorithms using Grid-enabled environment are discussed.

More information

Identifiers

ORCID for A.J. Keane: orcid.org/0000-0001-7993-1569

Catalogue record

Export record