Optimization of louver fin geometries for miniature microchannel condenser by Taguchi and CFD method

In this work, optimization of louver fin geometries is performed to obtain better thermal hydraulic performance of multilouvered microchannel heat exchanger by CFD method. Five louver fin geometries are considered in this work namely fin pitch, fin height, louver pitch, louver angle and louver length. The air side performance is analyzed with the help of airside heat transfer coefficient and pressure drop. These parameters are determined using Colburn-j factor and f factor for Reynolds numbers range of 100–600. In this work, two factors are studied for the effect of individual and combined louver fin geometries. It is observed from the literature study that increase in Reynold number, increases the Colburn-j factor and hence increases the rate of heat transfer favorably. At the same time, increase in Reynold number, increase f factor in term increase the pressure drop which is not desirable. Hence, it is challenging to increase the heat transfer without increase the pressure drop characteristics for heat exchanger design. So, aim of this work is to maximize the heat transfer and minimize the pressure drop. To account for these two contradicting objectives, dimensionless number (JF factor) is considered to determine the thermal hydraulic performance for the heat exchanger and it accounts both Colburn-j factor and f factor simultaneously. Orthogonal array-based Taguchi analysis is performed to obtain optimized louver fin geometries. Taguchi-CFD analysis revealed that fin pitch is the most influencing parameter, that alone accounts for 94.33% of contribution ratio on JF factor. Taguchi-confirmation test showed that the enhancement of JF factor for optimal louver fin is 5.97% higher than that of the initial design parameter. Finally, CFD analysis is performed to compare the performance of optimal louver fin geometry with that of the default louver fin geometry. From this analysis, Colburn-j and JF factor of optimum fin geometry are found to be 24.42% and 18.23% higher than those of default fin geometry. Regression models are developed for optimum fin geometry to predict the Colburn-j, f and JF factor for the Reynolds numbers range of 100–850, whose adj. R² value is 99.05%.