Computational analysis and design of fan blade configurations for optimal fan performance characteristics

This thesis describes the computational analysis of industrial centrifugal fans using incompressible Reynolds Averaged Navier Stokes equations in sliding meshes and the optimization of fan performance characteristics using genetic algorithm with a reduced order modeling technique called proper orthogonal decomposition (POD). POD modes created from flow simulation snapshots at the fan outlet are combined with interpolation method to estimate the large population sample needed for optimization.MASTER OF ENGINEERING (MPE