Water tunnel testing of highly-cambered plates and wings for downwind yacht sail applications

Abstract

The spinnaker is the most powerful yacht sail. However, its complex aerodynamics, governed by substantial flow separation, are yet to be fully understood and are modelled with difficulty with numerical simulation. Moreover, full-scale flow visualisation data is not available. Only recently have quantitative flow field measurements been achieved through particle image velocimetry (PIV) in water tunnels. Novel insights facilitated by PIV could lead to new breakthroughs. Consequently, this thesis investigates to what extent sail-like plates and wings can be tested in water tunnels for downwind yacht sail applications in order to provide guidelines for low Reynolds number testing of spinnakers.

First, the aerodynamics of an idealised, two-dimensional section through a downwind yacht sail, namely a circular arc with a camber-to-chord ratio of 0.22, is characterised at chord-based Reynolds numbers from 53 530 to 218 000. For incidences above the ideal angle of attack (where the stagnation point lies at the leading edge) but below stall, a leading-edge separation bubble (LESB) forms on the suction side of the arc, and the reattached boundary layer separates further downstream. The LESB is always turbulent at the tested conditions. Notably, blockage effects are shown to vary linearly with the blockage ratio, and a blockage-independent force crisis is evidenced and associated with the suppressed relaminarization of the boundary layer downstream of the LESB. Relaminarization is shown to occur up to a combination of critical angles of attack and critical Reynolds numbers, varying linearly with each other.

Then, six rigid models of a three-dimensional symmetric spinnaker are tested at average-chord-based Reynolds numbers from 5 870 to 61 870. Blockage effects are shown to vary linearly with the blockage ratio for the spinnaker as well. The results also demonstrate the existence of a blockage-independent critical Reynolds number. This is significant as it is found at a Reynolds number of the same order of magnitude as that of all previous water tunnel tests for spinnakers. Evidence of the suppressed relaminarization of the boundary layer downstream of the LESB is provided for the spinnaker. Additionally, applying a rotation angle to an isolated spinnaker is shown not to replicate the effect of the mainsail’s upwash despite yielding the same lift coefficient. Ultimately, water-tunnel testing of downwind yacht sails is demonstrated to be a valid experimental methodology, and guidelines are provided to ensure model-scale experiments are representative of full-size.

These results provide novel insights into the experimental testing of highly-cambered plates and wings. It is anticipated these results may contribute to the design of cambered wings, yacht sails and blades at transitional Reynolds numbers, such as the wings of micro-aerial vehicles, axial compressor blades and the sails of autonomous sailing vessels.