Design Solutions for Reducing Losses in Aeronautical Low-Pressure Turbines

The present thesis deals with the generation of innovative design solutions for reducing aerodynamic losses in Low-Pressure Turbines through the combined use of Artificial Neural Networks and CFD (Computational Fluid Dynamics). In particular, a lean model for optimizing blade airfoils in a multi-row environment and the development of a passive control device for reducing secondary losses are presented. The first procedure allows building a 1.5 stage by starting from a single blade profile, which is one of the main focuses of low TRL (Technology Readiness Level) design activities. In the obtained setup, two airfoil aerodynamic optimizations are performed by considering steady and unsteady conditions separately. The evaluation of the unsteady phenomena proves to be critical already at the considered blade design stage. Indeed, such practice allows achieving better aerodynamic performance and reducing the blade count per row. The second part of the thesis explains the blade fence development, which has been performed in collaboration with Avio Aero and Università degli Studi di Genova (UniGE). Such shelf-like devices are applied onto the blade surface to hinder the generation of secondary flows. The blade fence shape is the result of numerical aerodynamic optimizations so that the devices determine the reduction not only of the secondary loss but also of the flow turning deviation at the vane outlet. Such beneficial effects are also confirmed by the results of the experimental campaign conducted by UniGE on linear cascades.