(59n) Momentum Transfer in TUBE FLOW By Using Artificial Neural Networks

This paper presents the use of artificial neural networks (ANNs) to describe the momentum transfer enhancement in tube flow with an insert promoter. Entry region coil is an insert promoter and it is placed coaxially in the tube. An experimental work was carried out to examine the effects of pitch of the coil, length of the coil and diameter of the coil on mass and momentum transfer enhancement. The experimental data sets were extracted and were tested within a geometrical range pitch of the coil 0.015 < P_c < 0.035 m/turn; length of the coil 0.035 < L_c < 0.125 m, Coil diameter 0.02 < D_c < 0.04m and Reynolds numbers are varied from 1200 to 14,500. The experimental data sets have been used in training and validation of ANNs in order to predict the momentum transfer coefficient with entry region coil as insert promoter. The 78 experimental data sets have been used in training and 26 data sets for the validation of the Artificial Neural Networks by using MATLAB 7.7.0, particularly tool boxes. The training of neural network is used to minimize the error function with a learning rule. The generally used learning rule is gradient-based such as the popular back propagation algorithm. The results of this study reported that ANN configuration of 4_4_1 is recommended for the momentum transfer training for faster convergence and accuracy.