Abstract

Development of effective discrete film cooling is recognized as an essential task in gas turbine design since it has been shown necessary to ensure an acceptable turbine component life-time. Traditional design of film cooling schemes is driven by a single objective, maximizing the cooling performance. The aerodynamic penalty resulting from the coolant injection is generally neglected. Achieving an aero-thermal design of film cooling is a challenging task as it is based on two competing objectives; attaining a high thermal protection of the airfoil from the hot mainstream gas usually results in deteriorating the aerodynamic efficiency.
The present research addresses this challenge and investigates the complex flow underlying the aero-thermal interaction on a film cooled airfoil. The simultaneous effects of film coolant flow parameters and film hole geometric variables on the aerodynamic loss and the cooling effectiveness are examined. Trends in design variations to minimize the aerodynamic penalty, while maintaining a high cooling performance are established.
The research objective is achieved by implementing an automated optimization procedure that consists of a non-dominated sorting genetic algorithm coupled with an artificial neural network. The latter is used to provide prediction of the objective function at every optimization iteration and reduces computation time. It is constructed based on numerical flow simulations where the three dimensional Reynolds-Averaged Navier-Stokes equations are solved.
The optimization methodology is applied on a typical high-pressure turbine with two staggered rows of discrete film cooling on the suction side. The Pareto front of optimal solutions is generated. The thermal, aero-thermal, and aerodynamic optimums are identified and investigated numerically. A subsonic wind-tunnel facility available in Concordia University is used to re-create experimentally the optimum design points. Measured experimental data allowed verification of the CFD model, and substantiated the optimization methodology as a reliable design tool for film cooling in turbomachinery applications.