Inviscid approach for upwind sails aerodynamics. How far can we go?

Abstract

a b s t r a c tThis work presents a full-scale experimental study of a yacht rig and sails in real upwind sailing conditions and a comparison with Fluid Structure Interaction (FSI) simulations with the ARAVANTI model (Finite Element Method for the structure and Vortex Lattice Method for the fluid). A specific on-board instrumentation system simultaneously measures loads in the rig and sails, sailing data (wind, boat attitude and speed) and the shape of sails in real navigation conditions (flying shape). Flying shape parameters are extracted using the camera-based VSPARS system to characterize the effects of sail trims and to be compared with the results of the simulation. The potential flow solver gives fast and accurate predictions of both the flying shape and the loads in the rig in most conditions. The inviscid approach, commonly used in the early stage of design, must be checked, as in particular cases where the sails are heavily loaded, flow separation is significant and results from a potential flow solver are inaccurate. A new version of the model including the heel angle as an additional degree of freedom in the structural solver enables to detect when the inviscid flow approach overestimates the aerodynamic load. This upgrade improves the utility and reliability of the inviscid flow approach which remains relevant at the early stages of design as it is much more cost-effective than RANS models.

Dynamic Fluid Structure Interaction of NACRA 17 Foil

“…Some studies [2,7,9] found bend-twist coupling of the foils to be important. Experience with large flow induced deformation of thin structures can also be found on the aerodynamic of sailing boats, where recent work [15][16][17][18][19] investigated FSI simulation of yacht sails.…”

Efficient optimization procedure in non-linear fluid-structure interaction problem: Application to mainsail trimming in upwind conditions

“…Otherwise, the numerical optimization procedure can produce an unphysical solution, even if the model is actually fairly accurate in the neighborhood of the physical optimum. In the present case, the problem could be remedied by means of an ad hoc penalization of the performance function; for instance one could add a righting moment penalty term or artificially decrease the performance for trimmings likely to produce a detached flow (Aubin et al, 2016).…”