Development of numerical procedures for turbomachinery optimizaion

This Doctoral Thesis deals with high speed turbomachinery optimization and all those tools employed in the optimization process, mainly the optimization algorithm, the parameterization framework and the automatic CFD-based optimization loop. Optimization itself is not just a mean to improve the performance of a generic system, but can be a powerful instigator that helps gaining insight on the physic phenomena behind the observed improvements. As for the optimization engine, a novel surrogate-assisted (SA) genetic algorithm for multi-objective optimization problems, namely GeDEA-II-K, was developed. GeDEA-II-K is grounded on the cooperation between a genetic algorithm, namely GeDEA-II, and the Kriging methodology, with the aim at speeding up the optimization process by taking advantage of the surrogate model. The comparison over two- and three-objective test functions revealed the effectiveness of GeDEA-II-K approach. In order to carry out high speed turbomachinery optimizations, an automatic CFD-based optimization loop built around GeDEA-II-K was constructed. The loop was realized for a UNIX/Linux cluster environment in order to exploit the computational resources of parallel computing. Among the tools, a dedicated parameterization framework for 2D airfoils and 3D blades has been designed based on the displacement filed approach. The effectiveness of both the CFD-based automatic loop and the parameterization was verified on two real-life multi-objective optimization problems: the 2D shape optimization of a supersonic compressor cascade and the 3D shape optimization of the NASA Rotor 67. To better understand the outcomes of the optimization process, a wide section has been dedicated to supersonic flows and their behavior when forced to work throughout compressor cascades. The results obtained surely have demonstrated the effectiveness of the optimization approach, and even more have given deep insight on the physic of supersonic flows in the high speed turbomachinery applications that were studied.