Optimization of film cooling hole configuration on leading edge plate of jet impingement/effusion cooling system

This study proposes an approach to enhance film cooling efficiency by optimizing the arrangement of film cooling holes using computational fluid dynamics (CFD), and artificial intelligence (AI)-based methods. Both adiabatic and conjugate heat transfer (CHT) numerical simulations were performed on the curved surface of the leading edge of a turbine blade vane. A feed-forward neural network (FFNN) was employed to evaluate and improve the cooling performance using a performance index that assigned equal weights to the area-averaged cooling effectiveness and its standard deviation. Experimental validation using non-contact phosphor thermometry was performed to assess heat transfer performance and to select an appropriate turbulence model by comparing temperature fields and Nusselt number distributions. The optimization process treated the hole positions as design variables and identified configurations that outperformed the original layout. The optimized adiabatic geometry demonstrated a 4.8 % improvement in performance, while an elliptical-hole configuration derived from the CHT-optimized design achieved a 6.9 % performance gain. These findings highlight the effectiveness of combining CFD, AI, and experimental methods to improve film cooling designs in turbine applications.