Abstract
The aim of this thesis is to develop a combined analytical, computational and experimental methodology for characterisation and optimisation of the hydrodynamic performance of a novel vertical-axis tidal turbine concept. A primary objective is to develop experimentally validated, efficient, numerical predictive tools with the capability to allow the design and assessment of different turbine configurations for optimal power performance. The two numerical modelling approaches adopted and developed are (1) blade element momentum (BEM) theory, developed within MATLAB®, and (2) computational fluid dynamics (CFD) methods, using the Reynolds-averaged Navier Stokes (RANS) equations within ANSYS Fluent®. The model predictions are supported and validated by physical test data from two facilities: (1) a new small-scale tow tank facility setup at NUI Galway and (2) a state-of-the-art medium-scale recirculation flume at IFREMER, Boulogne Sur Mer. The methodologies and tools developed are more generally applicable for assessment of other existing and novel turbine designs. The primary focus of the work is the patented Brí Toinne Teoranta (BTT) vertical-axis tidal turbine. This concept’s novelty arises from the blades’ spiral geometry, which is intended to overcome highlighted limitations of existing vertical-axis turbines, potentially giving increased power efficiency and improved self-starting capabilities over current vertical-axis designs. The design concepts encompass various blade shapes, including an ellipsoidal design, varying height-to-diameter ratios, and a cylindrical design. A set of geometric equations is established here for applicability to different blade design concepts. A superior BEM method, including significant improvements in relation to finite aspect ratio effects, dynamic stall, and flow expansion, is presented for hydrodynamic power prediction of vertical-axis turbines. The results are validated against existing experimental data for SANDIA Darrieus and straight-bladed wind turbines. A coupled design optimisation study for the BTT concept is presented to identify peak power performance, with crucial turbine parameters investigated. The optimised turbine gives a 24% higher maximum power coefficient than the base case design; 18% higher than an equivalent straight-bladed design. A RANS approach, based on transitional flow turbulence modelling, to determine lift and drag coefficients for the NACA0015 profile is developed. Results are validated against published experimental data for a broad range of angles of attack and Reynolds number. Flow visualisations highlight the distinct strengths of this modelling approach in the critical stall region for hydrofoils. Significantly improved BEM predictions are obtained using the RANS lift and drag coefficients, compared to the traditional panel-method dataset, for straight-bladed turbines. A 2D RANS model is also developed and validated for straight-bladed turbines. The new BEM approach achieves a similar accuracy level to the latter with significantly faster run-times, thus providing a viable design tool for vertical-axis tidal turbines. A procedure for assessment of power performance of a scaled vertical-axis turbine is presented based on testing at the recirculation flume at IFREMER. The power curves for two BTT design concepts are compared using dimensionless parameters, mean power coefficient and tip-speed ratio (TSR) at equivalent operational Reynolds numbers. A new cylindrical BTT design (with horizontal support arms) gives more than 25% higher maximum power coefficient than the spherical BTT profile. An accurate 3D RANS model of the optimised BTT turbine is presented based on a rigorous model development approach. The converged model uses sliding-mesh RANS modelling approach with SST Transition turbulence. The model predicts the peak power coefficient within 6%. Excellent correlation of downstream CFD-predicted and measured flow velocities is demonstrated, providing further confidence in the 3D CFD model. A comparative assessment of the different modelling approaches developed with the experimental data for the novel BTT turbine design showed that the BEM model, with CFD dataset for NACA 0015 profile, (i) reduced the root means square error by 30%, over the traditional panel method dataset and (ii) gave comparable levels of accuracy to the full 3D CFD simulations for peak power and optimal TSR, at a considerably lower computational cost. A towing tank was designed, developed and fabricated for in-house testing of small-scale prototypes. The effect of turbine orientation on measured torque under constant flow conditions was studied. BEM and CFD models of rotationally constrained turbines gave general agreement with experimental measurements in terms of the effect of towing velocity on torque.