At Mystic Seaport Museum a project was begun in 1992 to determine sail coefficients for schooners for sailing technology research and historical vessel research. Since computational fluid dynamics techniques and wind tunnel techniques have not yet been developed to the point where sail coefficients can be developed accurately, the experimental approach was adopted. Full-scale sailing tests of schooner Brilliant were performed at Mystic, Connecticut. Tow-tank tests of a 1/9 model were completed at Davidson Laboratory, Hoboken, New Jersey. Three previous experimental programs for sloops were reviewed, notably Gimcrack (Davidson, 1936), Bay Bea (Kerwin, 1974), and Standfast (Gerritsma, 1975).
One goal of the Brilliant program was to reduce uncertainty in sail coefficient measurement. The uncertainty in previous programs illustrates the inherent difficulties. Despite the uncertainty the results were extremely useful and provided a benchmark for further improvements in sail coefficient programs.
Procedures used are described. Preliminary results indicate that the uncertainty is reduced to about 10%.
The topics covered are: (1) a description by Olin Stephens of previously unpublished details of the equipment and methods used to take data on the sloop Gimcrack, as applied by Davidson reported in his landmark 1936 paper, and some thoughts on the way the information gained from early studies was used; (2) extension of the Gimcrack sail coefficients to the heel plane; (3) a review by Howard Grant of the measurements in the schooner Brilliant program, including correction for ship/wind interference, use of cross checks including chase boat true wind solutions and internal consistency techniques (Ockam, 1992), and unique time-correlations to handle non-steadiness; (3) a hindsight view of program shortcomings, which included the omission of measurements at several heights above the deck of wind and sail shape that would have been helpful to CFD studies of sail aerodynamics.