Modeling of Heat Transfer Characteristic of Serpentine Microchannels Using Artificial Neural Network and Genetic Algorithm Analysis

In the present research, an Artificial Neural network (ANN) is applied for estimating the heat transfer characteristic (Nusselt number) in serpentine microchannels. Experimental data were collected for developing the ANN. The experiments were carried out with cold fluid in serpentine micro tubes which placed in a hot bath. There are six serpentine micro tubes with diameter of 787.4 µm and various geometric parameters used in the experiments. The output (target) data of the ANN model is Nusselt number (Nu) and input data are Reynolds number (Re), ratio of straight length to diameter (Ls/d), and curvature ratios (RC/d ). The validity of the neural network modeling was evaluated through a testing data set, which were randomly extracted from the database and were not used in the training of the network. Furthermore, empirical correlation for prediction of Nu was developed in the form of classical power–law correlation and the equation constants were determined using genetic algorithm (GA) technique. The estimated results of the developed ANN model were compared with the presented correlation.