Abstract In this study, a numerical approach has been implemented to find the optimal shape of a two dimensional turbulator above an isothermal flat plate and parallel to the flow. The turbulent boundary layer over the plate was disrupted at various distances by inserting a quadrilateral bar where the boundary layer thickness was grown to the three times greater than the insert\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\'s height. As a result, the overall wall heat transfer was altered. Heat transfer coefficient was defines as a cost function. The single objective Genetic Algorithm (GA) with the modified Teach-t code were employed to optimize variables such as the shape and distance of the obstacle from the plate. Applying the optimal geometry demonstrated that the mean heat transfer coefficient is enhanced by more than 51% over the affected area comparing with the case of undisturbed boundary layer. Downstream flow comprises of the stagnation point which formed on the frontal face of the body and accounts for the slight decrease in the heat transfer coefficient. The jet underneath the obstacle followed by the developing boundary layer is responsible for the sudden peak of the heat transfer coefficient. However, the downstream wake zone makes the main contribution.

Keywords