Combined optimization of the blade shape and operating point of a cross-flow tidal turbine for increased performance and reduced fatigue loadings

Hydro-kinetic cross-flow tidal turbines (CFTT) are attractive for tidal energy exploitation due to their omni-directional nature and high area-based power density, when compared with horizontal-axis tidal turbines (HAT). However, the continuous variation of the angle of attack, caused by the rotating motion around the vertical axis, induces alternating loads. This may cause fatigue failure and structural damage over time. By overcoming this challenge and increasing the overall efficiency of CFTT, these turbines can be a strong contender for tidal energy exploitation. The OPTIDE project aims at optimizing CFTTs with regards to the aforementioned problems numerically and experimentally. Therefore a turbine flume model has been developed, which features a height and diameter of 400 mm. The studied turbine is a Darrieus-type turbine with three straight blades, where the blade tips are connected through an upper and lower hub to the central shaft. This design has been chosen to allow for easily interchangeable blades in the process of the project. The blade shape of CFTTs has a big influence on the efficiency and loadings of the turbine. For the same reason, the optimal operating point (Tip Speed Ratio, TSR) is different for every blade shape and size combination. Therefore, in the framework of the OPTIDE project a combined optimization is conducted in order to find an optimal blade shape, while simultaneously finding the optimal operating point. This multi-objective optimization does not only maximize the power coefficient (CP), but also minimizes the structural loads experienced by the blades. The blade shape optimization process couples a Genetic Algorithm (GA) with 2D Computational Fluid Dynamics (CFD) simulations. Ten design variables, which define the hydrofoil shape, chord length and operating point, are set by the optimizer. After creating a smooth hydrofoil and generating an automatized mesh, Unsteady Reynolds Averaged Navier Stokes (URANS) simulations deliver the turbine CP and blade loads, converted into a stress coefficient via a model. This allows to assess the blades overall performance. After 5 generations of individuals, 2000 different blade shapes have been created and examined as part of the optimization process, providing several optimal shapes for the present application. A spot analysis delivers insight on the effect of specific blade characteristics on the turbine flow field, as well as on its performance and fatigue loadings. We will present the current state of the optimization and analyze the results. Based on the huge dataset of available case files, the impact of the design variables will be correlated to the optimization objectives.