A study of the vortex flows of downwind sails

Abstract

A leading-edge vortex (LEV) can be a robust lift generation mechanism on both the wings of natural fliers and delta wings. A spinnaker-type of sail is a thin wing that promotes the formation of LEVs due to a sharp leading edge. Recent numerical simulations (Viola et al., 2014) have demonstrated that this type of sail can prevent LEV shedding and instead, keeps it trapped near the leading edge. In such cases, the LEV could enhance lift generation (Saffman and Sheffield, 1977; Huang and Chow, 1982), and so there is a need to investigate the existence of the LEV and its role for sails. To study the LEV in the context of sails, a rigid model scale spinnaker was tested in water at low Reynolds numbers and uniform flow. It was found that the flow separates at the leading edge, followed by turbulent reattachment, forming an LEV. For finite periods the LEV breaks down into weaker LEVs that are shed downstream; otherwise, the LEV remains coherent at the leading edge. On the lower half of the sail, the LEV has negligible diameter, and trailing edge separation occurs after the first quarter of the chord. To understand whether there is a benefit from having the LEV trapped near the leading edge, as opposed to being shed downstream into smaller LEVs, the local circulation was measured and its value utilised in a complex potential model. The model maps a circular arc into a rotating cylinder and assumes the Kutta condition, to provide a bound circulation value that is a function of the position and circulation of each LEV (Pitt Ford and Babinsky, 2013; Nabawy and Crowther, 2017). It is found that when the LEV is trapped near the leading edge, the LEV provides a marginally higher lift than when it breaks down and sheds. Surprisingly, with the conservative assumption of the Kutta condition, the LEV contributes between 10% to 20% to the sail’s sectional lift. In actual sailing conditions, the spinnaker experiences a twisted onset flow, that could not be replicated in the water flume, such that the angle of attack varies along the span of the sail. To explore this effect three spinnaker models were made, where the original sail was twisted from top to bottom by different angles. PIV and force measurements were compared. It was observed that a low twist sail allows the LEVs to remain close to the body of the sail, whereas a high twist sail causes them to drift away and generates counter vorticity on the surface of the sail. This viscous effect results in a marginal reduction in lift, but significant reduction of induced drag. The results presented in this PhD thesis aim to provide an improved understanding of the aerodynamics of downwind sails, where vortex flow is a dominant feature. The existence of trapped and shedding LEVs is demonstrated and an attempt is made to model LEVs through a complex potential model in order to assess their contribution to the sectional lift of the sail. Finally, the effect of twist is evaluated with regard to the aerodynamics of sails.