Accurate prediction of performance is an important aspect of modern sailing yacht design and provides a competitive advantage on the racecourse and in the marketplace. Although wind tunnel testing of yacht sails is a common tool for obtaining input data for Velocity Prediction Programs, its results have not been validated against aerodynamic full-scale measurements as quality full-scale data is rare. Wind tunnel measurements are conducted at the Twisted Flow Wind Tunnel of The University of Auckland and are compared to the full-scale aerodynamic force measurements from the Berlin Sail-Force-Dynamometer. To realise this comparison wind tunnel techniques and aerodynamic force models for yacht sails are enhanced; this in turn also improves the accuracy of Velocity Prediction Programs.

Force and surface pressure measurements were conducted demonstrating that the interaction of the hull/deck with the sails has a significant effect on the side force and the force perpendicular to the deck plane, and that this should be considered in aerodynamic analysis of sails and the performance prediction of yachts.

The first Real-Time Velocity Prediction Program for wind tunnel testing has been developed and implemented as an additional module of FRIENDSHIP-Equilibrium. Model sails can now be trimmed based on the full-scale performance of the yacht, and at the correct heel angle, which makes the trimming process in the wind tunnel much more similar to the real life situation.

Improved aerodynamic force models have been developed from realistically depowered sail trims obtained with the Real-Time Velocity Prediction Program. An empirical model that describes the force and moment changes due to depowering in detail has been developed and implemented. The standard semi-empirical trim parameter model, which expresses depowering in a more generic way, has been enhanced based on aerodynamic principles and validated against the wind tunnel results.

Utilising the enhanced wind tunnel techniques and aerodynamics force models, a generally good qualitative and quantitative agreement with the full-scale data is achieved. Remaining challenges associated with full-scale and wind tunnel tests are however also highlighted and, based on this work alone, a conclusive judgement that scaling effects are negligible cannot be made.