Free Publication

Multi-Objective Design Optimization of a Mixed-Flow Pump

Written by Yumiko Takayama and Hiroyoshi Watanabe (Ebara Corporation).

What's inside?

In most cases of high specific speed mixed-flow pump applications, it is necessary to satisfy more than one performance characteristic such as deign point efficiency, shutoff power/head and non-stall characteristic (no positive slope in flow-head curve). This paper presents the results of the multi-objective optimization strategy of mixed-flow pump design by means of three-dimensional inverse design approach, Computational Fluid Dynamics (CFD), Design of Experiments (DoE), response surface model (RSM) and Multi Objective Genetic Algorism (MOGA).

Pump efficiency, maximum slope in flow-head curve and shut-off power/head are selected as objective functions. Objective functions of pumps, designed by design variables specified in DoE, are evaluated by using CFD.

Multi-Objective Design Optimization of a Mixed-Flow Pump Cover

In this publication, you will:

  1. Discover how to optimize these performance characteristics with a 3D Inverse Design method.
  2. See an efficient way to overcome the performance curve instability characterized by positive slope in flow-head curve (stall characteristic) at low flow rate.
  3. Understand the design strategy of coupling the parameterization of blade surface by using a three-dimensional inverse design method, DoE to reduce the number of design parameters, CFD calculations to predict pump performance characteristics, RSM to correlate design parameters with objective functions, and MOGA to derive Pareto front for enabling the trade-off designs.