Flapping-foils offer an alternative to rotary turbines in generating power from fluid flows. This thesis examined the flapping-foil concept using numerical simulations with the aim of improving performance, by considering streamlined foils and a flat plate of different leading and trailing edge configurations, control of the effective angle of attack, and optimum kinematic parameter values in laminar and turbulent flows. The flat plate, anticipated to be cheaper and easier to manufacture, has been found to generate higher power compared to streamlined foils using sinusoidal plunge and pitch motions. This is largely a result of (i) the ability of a flat plate to harvest the energy of LEVs, by the uniform thickness along its chord length. Investigation into the active control of the effective angle of attack is implemented with sinusoidal plunge motion, with the pitch motion adjusted to achieve the desired effective angle of attack profile. A square wave effective angle of attack profile is found to be superior for a flat plate with the maximum effective angle of attack 21⁰ or less. This is a result of the reduction in the pitch rate and the increase in moment acting in the same direction as the pitch motion during the stroke reversals. A cosine variation is best for maximum effective angle of attack of at least 46⁰ because the moment of the flat plate nearer the end of the pitch-reversal-stroke and into the beginning of a new pitch stroke opposes the moment for the pitch motion, as the cosine profile moves towards a square wave profile. Optimisation of five kinematic parameters of a flat plate flapping-foil turbine is performed with a “multi-fidelity” evolutionary algorithm for a single objective function. The multi-fidelity algorithm is found to be good in predicting near optimum solutions with reduced overall computational cost compared to single fidelity algorithms. However, in the search for optimum solutions of power generation and efficiency and the values of five kinematic parameters, the surrogates did not predict the optimum solutions and five kinematic parameter values accurately as result of a non-linear relationship between the five kinematic parameters and each objective function.
展开▼
