ANN and RSM approach for modeling and optimization of designing parameters for a V down perforated baffle roughened rectangular channel

Abstract

The turbulence promoters are widely used to enhance the performance of rectangular channel which were used for turbine blade passage cooling. In the present study, the influence of design parameters of the V down perforated baffle roughened rectangular channel on the heat transfer and friction factor was investigated using RSM and ANN. The quadratic model generated by RSM is used to predict the performance parameters, i.e. Nusselt number and friction factor with reasonably good accuracy. The optimum values of the design parameters of the V down perforated baffle roughened rectangular channel are relative roughness pitch of 2.6, relative roughness height of 0.33, open area ratio of 18% and Reynolds number of 18,500, in the desirable range of the order of 0.95. The training of the experimental data is carried out using 4-10-2 neural network and the predicted values are compared with the experimental values and found deviation in the range of ±10% among predicted and experimental values. The comparison of predicted values by RSM and ANN with the experimental values was carried out for each run of experiment and it was observed that the RSM predicted values are in accord with the experimental values in the uncertainty range of ±5%.