Optimum Parameter Design of Microbubble Drag Reduction in a Turbulent Flow by the Taguchi Method Combined With Artificial Neural Networks

Abstract

This study attempts to optimize parameters for the microbubble drag reduction in a turbulent flow based on experimental measurements. Five parameters were investigated: three are control factors (the area of air injection, bubble size, and the rate of air injection) and two are indicative factors (flow speed and the measured position of local shear stress). An integrated approach of combining the Taguchi method with artificial neural networks (ANN) is proposed, implementing the optimum parameter design in this study. Based on the experimental results, analysis of variance concluded that, among the control factors, the rate of air injection has the greatest influence on microbubble drag reduction, while bubble size has the least. The investigation of drag reduction characteristics revealed that the drag ratio decreases with an increasing rate of air injection. However, if the rate of air supplied exceeds a certain value, the efficiency of drag reduction can drop. In the case of optimum parameter design, a 21% drag reduction and an S/N ratio of 1.976 dB were obtained.