This thesis presents the study and development of optimization methods that can perform concurrent aerodynamic-aeroelastic blade optimization in a multi-bladerow environment, and for realistic turbomachinery blade geometries. The Nonlinear Harmonic Phase Solution Method has been chosen as the flow solution method of this work because of its capability to calculate the aeroelasticity features of interest (blade flutter) and the main flow aerodynamic performance in steady flow timescales. The first optimization method that is shown is a more generic non-gradient method with an improved version of a quadratic Response Surface Model. The new Re-Scaled Response Surface Model has shown marked convergence and performance improvements against traditional surrogate models. However, the computational cost of this method for cases with a large number of design variables limits its real applications. A gradient-based adjoint method is presented next as a cost-independent alternative that can accomplish efficient multi-bladerow optimization for a large number of variables within the current levels of computational power. The continuous adjoint system has been developed based on the same methodology as the flow solution method and it shows a more consistent relation between the flow field and its corresponding adjoint field, in agreement with the "anti-physics" information path. An adjoint interface treatment has been developed as an extension of the flow harmonic interface treatment. This unique treatment allows capture of the damping sensitivities of the vibrating blade to shape changes in adjacent rows. The application of this method to the design optimization of compressor and turbine stages has shown its capability to perform efficient multicomponent and multi-disciplinary design optimization of turbomachinery blades.
| Identifer | oai:union.ndltd.org:bl.uk/oai:ethos.bl.uk:719875 |
| Date | January 2015 |
| Creators | Valero Ricart, Omar Ruben |
| Contributors | He, Li |
| Publisher | University of Oxford |
| Source Sets | Ethos UK |
| Detected Language | English |
| Type | Electronic Thesis or Dissertation |
| Source | https://ora.ox.ac.uk/objects/uuid:b89d8f80-9134-4856-8223-5f55967c0bde |
Page generated in 0.0082 seconds