Evaluation of Nusselt number and friction factor in zigzag channel heat exchangers by neural networks
Volume Title: 1
Department of Chemical Engineering, Faculty of Engineering, University of Kurdistan, Sanandaj, Iran
In the work, the ability to estimate Nusselt number (Nu) and friction factor (f) in zigzag channels using artificial neural networks (ANNs) was investigated. The computational fluid dynamics (CFD) technique was used to obtained data points related to the zigzag channels with different geometries. The wall temperature was determined constant 353 K and water was employed as the working fluid. The channels with various bend angles between 5° to 45° were evaluated for turbulent flow and Reynolds number (Re) between 4000 to 40,000. Two ANNs were developed for predicting Nu and f as functions of Re, bend angle (Ɵ), and Prandtl number (Pr). The results show the mean relative error of 0.25% and 0.42% for the prediction of Nu and f, respectively, which indicates the high accuracy of the model.
Heat transfer; Zigzag channels; Artificial neural network (ANN); computational fluid dynamics (CFD); Geometrical parameters