Modeling and optimization of impinging jet pressure using artificial intelligence

Abstract

Artificial Intelligence (AI) can be used to model efficient processes. In this paper, AI and CFD are employed to maximize the wall pressure of the impinging jet, which has a wide range of industrial and technological applications. Firstly, the CFD model is validated with the experimental results for various geometrical and flow rate configurations (H/D= {1, 2, 3, 4, 5, 10, 20}, where H is nozzle to flat plate distance and D is nozzle diameter, and Reynolds Number (Re) ranges {Re= 15,000, 25,000, 30,000}. In explanatory data analysis, Pressure and D are slightly negatively correlated. Re and D show a negative relation of −0.2 whereas a slight negative relation appears for D vs H/D. Re vs H/D have a positive correlation of 0.2. Various activation functions were explored to find that tangent hyperbolic performed best model fit under AI. The Artificial Neural Networks (ANNs) are applied to train the model with minimized least squared error. Finally, the trained model is optimized for the maximum wall pressure over the distance. The maximum pressure obtained from the trained model is 383.83 KPa.

Keywords

• Impinging jet
• Pressure distribution
• Nozzle to Plate Distance
• Artificial Intelligence