Optimal active control of laminar flow over a circular cylinder using Taguchi and ANN

Abstract

Increasing the lifetime of engineering structures is dependent on the unsteady forces exerted by vortex shedding. Flow controllers can be used for suppression of vortex shedding behind the cylindrical structures. The main objective of this paper is to minimize the wake generated behind a circular cylinder in laminar flow, Re=150. Two counter rotating controllers inject the momentum in the wake region or actively control the vortex shedding. The important parameters for defining the effectiveness of flow control are the lift and drag fluctuations besides the drag coefficient. The computational methodology is based on the finite volume method using the SIMPLE algorithm. The parameters that affect on the wake control are the rotation rate, diameter of controllers, radial and the angular distances. Determining the optimum governing parameters is important. A neural network is used for finding these optimum parameters. The input data for training this neural network were provided by the Taguchi method. Moreover the Taguchi method clarified the influence of each parameter on the wake generation. At the optimal obtained conditions the drag coefficient reduced up to 99.99% and the vortex shedding suppressed appropriately.