Prediction of Round Jet Flow with Artificial Neural Networks

Abstract

In this study, an artificial neural networks model was established by using experimental measurement values at low speed subsonic wind tunnel of which length was 75 cm and experiment test section was 32 cm × 32 cm, and model results were compared with experimental values and then, the prediction was made for the unmeasured tunnel stream values. In the wind tunnel, when the jet velocity was 25 m/s, four tunnel velocities, 0, 5, 10, and 20 m/s were used. At the four measurement stations; x/D = 0.3, x/D = 12.5, x/D = 31.2, and x/D = 50, experimental measurements were made by using hot wire anemometer. Plain circular jet flows at x/D = 0.3 and x/D = 50 stations with average tunnel flow velocities of 7.5 m/s and 15 m/s were studied by using artificial neural networks. The data was obtained and evaluated by graphics.