Combined experimental testing and numerical modelling of a novel vertical axis tidal turbine
MetadataShow full item record
This item's downloads: 288 (view details)
This research presents a study of a novel vertical axis tidal turbine (VATT) – the GKinetic turbine – using experimental testing of scale devices and numerical modelling. The primary aims of the thesis were: (1) to assess the mechanical power extraction performance of the turbine via experimental testing and (2) to develop a hydrodynamic model capable of predicting the power performance. During the process of numerical model development, two different modelling techniques were assessed for their suitability for modelling of VATTs: (1) the computationally-efficient blade element momentum theory (BEMT) approach and (2) the more computationally-intensive Reynolds-averaged Navier Stokes (RANS) approach. The former was developed by the author in MATLAB while the latter used the commercial computational fluid dynamics (CFD) code, ANSYS Fluent. The GKinetic turbine incorporates two novel design features; a mechanism for accelerating the inflows to its twin turbines and a blade pitch control mechanism to implement variable pitching. This research is the first detailed research study of the device and since the device is novel, the research is also novel and significantly advances the knowledge-base concerning the device. Experimental testing of the turbine was conducted at various scales. 1:40 and 1:20 scale devices were tested under controlled conditions in a recirculating flume and a 1:10 scale device was subsequently tested under controlled conditions in the field using tow-testing. The device was shown to be capable of accelerating the free-stream velocity by a factor of 2 and achieved a peak mechanical efficiency of 40 %. Following the 1:10 scale tests, a structural analysis of some of the device components was undertaken and recommendations for improved design were made; some of these recommendations have since been incorporated in subsequent iterations of the device. A BEMT design tool has been developed for prediction of the hydrodynamic performance of high solidity and highly loaded VATTs. The model utilises a graphical approach from the literature for determining induction factors, rather than the more common iterative approach. The research is significant as VATTs typically have a higher solidity than wind turbines, thus the iterative approach for determining induction factors, which is more suited to low solidity rotors will generally, not be valid. This is the first time that the graphical approach for determining induction factors has been implemented in a VATT model. The model also incorporates modifications to correct for processes such as dynamic stall, flow expansion and finite aspect ratio blades. The model reproduced measured peak power coefficient values to within 6.4 % for a low solidity case, and to within 27 % for a high solidity case. 2D and 3D CFD models of a 3-bladed VATT were developed to assess the accuracy of RANS models in prediction of power performance and near-wake properties. The turbine models were developed using the sliding mesh technique and model performance was assessed using the results of experimental tow tank testing from the literature. The primary aim of the research was to determine the suitability of the sliding mesh technique for modelling of standard VATTs before applying the technique to the more complex GKinetic turbine. A blockage correction approach was trialled for 2D models and shown to significantly improve model accuracy. The CFD methodology developed allowed the 3D model to accurately model the power performance curve and turbine near-wake velocities. An additional novelty of this research is the use of the Transitional SST turbulence model, which, to the author’s knowledge, has not yet been used in 3D modelling of vertical axis tidal turbines. A 2D CFD model of the GKinetic turbine was developed using a nested sliding mesh technique. The model includes the flow accelerating mechanism and the variable blade pitching facilitated by the nested sliding meshes. Model performance was assessed by comparison with measured data for mechanical power and near-wake velocities from the undertaken 1:20 scale experimental tests. The model was used to investigate various aspects of the current device setup including the number of turbine blades, the benefits of variable versus fixed pitch blades, shaft sizing, the location of the turbine relative to the bluff body and the effect of blade chord-length. The results of the design study will inform future development of the device.