Performance Analysis and Optimization of a Darrieus VAWT

Increasing research efforts are being devoted to understand where small-size wind turbines can effectively represent an alternative for delocalized power production and emissions reduction, particularly when applied in new installation contexts (e.g. the built environment). In current eolic research, Darrieus VAWTs are particularly appreciated, thanks to the possibility of producing small machines with high reliability and almost no noise. In addition, a wider range of design solutions can be exploited in order to obtain an appealing exterior design and consequently to reduce the visual impact of the installation. As a result, many efforts are being made to develop industrially-designed turbines with a reduced environmental impact and a competitive cost; moreover, in order to ensure an effective diffusion of small Darrieus turbines, the overall performance, and especially the start-up and transient behavior, still need to be improved. Within this context, pushed by several industrial Italian partners, interested in developing innovative Darrieus turbines, a multipurpose research activity was carried out. In particular, the research project was aimed at analyzing in depth the physics and the functioning behavior of small-size Darrieus turbines, whose design process is affected by several additional criticalities in comparison to that of greater machines or of more conventional horizontal-axis rotors. Inter alia, small rotors are in fact deemed to be affected by a reduced power capacity and by lacks of self-starting capabilities, mainly due to the their reduced dimensions (i.e. small swept areas, chord lengths and diameters) and to frequently unfavorable conditions (i.e. low towers heights and complex orography of the installation contexts). To this purpose, was preliminary undertaken a theoretical analysis of the physical fundamentals which regulate the functioning of Darrieus VAWTs, dedicated to highlight the critical characteristics of these machines and the relative impact of the main design parameters. Efforts were therefore directed to create a reliable numerical tool which could be able to effectively model the behavior of H-Darrieus turbines during both the normal functioning and the transient phases. Within this activity, a numerical code, denoted VARDAR, was developed on the basis of the Momentum Theory; more in details, a Double Multiple Streamtubes approach was exploited for the turbine modeling, together with the possibility to account for the dynamic stall of the airfoils during the revolution and the width expansion of the streamtubes. In addition, due to the relevant impact of the secondary and parasitic effects on the performance estimation of small VATWs, several original sub-models to account for these effects were purposefully developed; inter alia, specific models for the shadowing effect of the central tower and for the parasitic torque of the rotating struts were realized. Furthermore, an innovative numerical routine was created in order to to simulate the time-dependent functioning of the rotor under a generic oncoming wind, with particular reference to the startup transitory. All the theoretical conjectures and the developed numerical models have been experimentally validated by means of several extensive test campaigns in two large wind tunnels in Italy on properly deigned 1:1 scale models of H-Darrieus rotors with either two or three blades. An impressive agreement between experimental evidence and simulated data was constantly found and all the developed models were therefore satisfactorily validated. Moving from this theoretical background, the attention was focused on developing new aerodynamic solutions which could partially solve the main critical aspects in designing a new turbine. Within this context, a new aerodynamic airfoil was designed by means of a CFD optimization, aimed at improving the airfoil performance for the working conditions (i.e. Reynolds numbers and incidence angle) on board of a Darrieus rotor. The static polar curves of the new profile over the whole range of the angle of attack (0°-360°) were measured in the wind tunnel, achieving interesting results in terms of lift increase and stall angle delay. Specific analyses were also directed to evaluate the parasitic torque of non-aerodynamic structures during the revolution; by means of both numerical analyses and experimental tests, an extended experimental study was carried out to investigate the most convenient shapes of the turbine struts. In particular, different hulls geometries were compared in terms of torque detriment and of impact on the self-starting behavior of the machine. Finally, the experience collected during the research activity, was synthetized in two original theoretical studies. In the first analysis, the startup behavior of a three-bladed H-Darrieus turbine was investigated and the self-starting capabilities of the rotor were related to its main design parameters. In the second contribution, a new model to account for the virtual camber effect on the performance estimation of H-Darrieus rotors using the momentum theory was developed. In particular, a transition point, based on the flow field characteristics around the airfoil, has been highlighted, below which a different estimation of the virtual camber effect must be adopted to ensure a more realistic description of the first part of the characteristic curve of the machine: on the basis of experimental evidence, the proposed model seems to represent a necessary condition for a correct description of the transient behavior of the turbine, especially in the startup phase.