Degree Name

Doctor of Philosophy


University of Wollongong. School of Mechanical, Materials and Mechatronic Engineering


The work described in this thesis was focused on computational fluid dynamic (CFD) modelling of a range of axial flow air turbines designed to service Oscillating Water Column (OWC) wave energy converters. In particular, the research concentrated on: a) analysis of the performance of turbines such as the Denniss-Auld turbine with variable pitch blades; b) the influence of the design of components such as the nozzle-diffuser, the nacelle and the OWC chamber on turbines coupled to OWC. This work was primarily motivated by the fact that very little research and data had been previously reported on these issues and that much work has yet to be done to develop design methods to optimise such systems by increasing their efficiency and extending their operational range and technical reliability.

Preliminary investigations into simplified turbine analysis via the Blade Element Momentum (BEM) method revealed a number of “cascade” issues having particular importance for analysis of axial flow turbines, which had not been previously examined with the aid of CFD. Thus, one of the significant outcomes of the present study was that CFD analysis confirmed the applicability of Weinig’s analytical and inviscid predictionof the cascade lift interference factor, k0, for linear cascades of practical aero foils (e.g.NACA0012 and NACA0021) providing angles of attack are less than about 10°. Weinig’s theory holds that the cascade lift interference factor, k0, for an infinite cascade is independent of the angle of attack, a, and is only a function of the stagger angle, g, and ratio of s/c (inverse of solidity). However, through the present CFD study it was found that the cascade interference factor does indeed depend on the angle of incidence due to practical issues such as finite aero foil thickness, stall, drag, etc, which could not be accounted for in Weinig’s inviscid formulation. Thus, Weinig’s inviscid model must be used with caution to predict the lift of blades in a cascade at angles of incidence, am >10º.

A second important outcome of the present study was the development of a new non-dimensional formulation of blade element analysis which can be used to predict the performance of OWC axial flow turbines using only generic/non-dimensional input parameters such as hub-to-tip ratio and non-dimensional axial velocity or flow factor, f. Other results from the CFD study were that the efficiency of the Denniss-Auld turbine is strongly influenced by variation of design parameters such as tip clearance, tc, hub-to-tip ratio, h, and number of blades, N. Adjustment of these design parameters so that tc = 2.3mm, h = 0.62, and N = 13 resulted in a predicted increase in turbine efficiency by up 5.5% compared to the actual/baseline rotor configuration. CFD modelling to ascertain the influence of the non-dimensional thickness of the blades in a Denniss-Auld turbine demonstrated that thinner blades produce higher turbine efficiencies for large blade stagger angles. Moreover, in general thinner blades will result in higher maximum turbine efficiencies for a given stagger angle. This was in contrast to previous research on Well’sturbines.

Three dimensional CFD modelling was carried out on a full scale OWC wave energy converter case study, which was based on a demonstration plant built and commissioned by Oceanlinx at Port Kembla, Australia, in 2006. This case study demonstrated that the CFD technique can be successfully applied to the analysis and optimisation of the major components of a turbine/OWC system, such as the nozzle-diffuser, nacelle and rotor. It was found that the pneumatic efficiency, hpneu, of the case study system depended on the volume flow rate through the system. Other issues investigated include the influence ofthe 90° turn that the air flow must make between the OWC chamber and the inlet to the turbine for OWC systems employing a turbine with a horizontal axis of rotation. CFD analysis demonstrated that the horizontal air ducts with converging nozzle section appeared to have no detrimental effect on the uniformity of the axial velocity as a function of circumferential position around the turbine.

Analysis of the experimental data obtained during sea-trial tests of the full scale Denniss-Auld turbine in 2006 revealed that the most efficient turbine operation was achieved with blades staggered at an angle of approximately 60° relative to the plane ofthe rotor. It was also found that the 3D CFD analysis improved the correlation between numerical results and experimental efficiency data for the full scale Dennis-Auld turbine by approximately 11% as compared to the BEM method.