Multidisciplinary Optimization of Radial Outflow Turbines With Highly Deformable Blades

Abstract

Abstract Miniature radial outflow turbines are used to generate onboard power for various aerial devices using the ram effect. The incoming air flow axially entering the inlet conduit of the turbine housing spins the turbine rotor, which may have excessive tip gap and a sizable hub cavity, and radially discharges from the turbine flow passages. The current class of turbine rotors under investigation are typically made of polymeric materials for serial production at reduced cost. The turbine airfoils are designed with significant overhanging sections over the large hub cavity. Such configuration allows regulating the rotational speed of the rotor, at high inlet velocity conditions, by enabling the rear part of the blades to bend under the effect of centrifugal and fluidic forces. However, such deflections may lead to material non-linearities and strongly affect the flow development through the blade passages, which are not yet documented in the open literature. This paper investigates the complex flow field inside such turbines for a wide range of operating conditions. A multidisciplinary design optimization routine was utilized to assess the conflicting influence of the aerodynamical and structural performance for this class of turbines. The strategy utilizes an in-house metamodel-assisted optimizer (CADO) based on artificial neural networks and genetic algorithms to maximize the power output while evaluating the impact of the rotational speed on blade displacement and flow behaviour. A decoupled fluid-structure interaction approach was established in the computational fluid dynamics solver integrated with the optimizer. The methodology validation was achieved through the experimental results on high-speed blade deformation of the initial turbine mockup. The results show striking changes to turbine geometry mainly dictated by the centrifugal force as compared to the fluidic loads play when a hub cavity exists. A change in the blade loading is identified due to the three-dimensional geometric modifications of the rotor under load solicitations and connected with a modification of the incidence angle and profile curvatures. Such a process decreases the turbine’s efficiency, especially at very high rotational speeds. The results of this work provide a comprehensive design space and guidelines on the development of radial outflow turbines through the extensive optimization study for preliminary design choices for turbines with similar geometries.