Because of the danger and difficulty of studying sailing yacht behaviour in breaking seas, model tests have been used. To conduct such tests, it is necessary to model an entire system; boat, drogue, line, wind, and sea conditions. This was accomplished according to the scaling principles discussed in the previous section.
The model tests conducted by Jordan, reference 1, examined the effect of boat length and design capsize propensity For these tests he modelled three boats which span over 50 years of yacht design, including; Tally Ho, winner of the 1927 Fastnet Race and a heavy traditional cutter design, a New York 30 design from the 1930's, and a modern yacht, the well-known Stand fast design. The three models used are shown in Figure 1. The models were 12 inches long representing a scale of 1/43. In addition, the New York 32 and Stand fast were built in 16" lengths, representing a scale of 1/32. All models were constructed of balsa and weighted to give the correct dynamic characteristics.
The models were tested under simulated breaking wave conditions. To do this, a horizontal jet of water was discharged into a static pool of water. The horizontal jet was generated by permitting a quantity of water to fall vertically and then deflecting the water from a vertical to a horizontal direction by a curved ramp. A schematic of the set-up is shown in Figure 2. Several hoppers of different heights were used during the test to provide different jet velocities.
Results of these early model tests indicated that size played an important role in the ability of a sailboat to resist capsize. This is not surprising if you consider that the kinetic energy necessary to capsize any boat design will vary as the fourth power of the boat length. Thus a 60-foot sailboat requires sixteen times as much kinetic energy from a moving wave crest as a 30-footer in order to capsize.
Another aspect of the early model work was to investigate the effect of design variations on capsize; variations in mast weight, displacement, freeboard, keel design, and the relationship of beam and centre of gravity were examined on the previously described models. Using the hopper arrangement, the models were tested to examine the differences in behaviour with respect to hull design. The Tally Ho, New York 32, and Stand fast models were used. The three designs had about the same capsize performance, despite varying design features. it should be recognized that this type testing is not sufficiently sensitive to pick up small differences in capsize vulnerability.
Similar studies have been carried out by the joint SNAME/USYRU Project for Safety from Capsizing (Refs. 3 and 4) and by the University of Southampton in England under the sponsorship of the Royal Ocean Racing Club (Ref. 2). These tests were conducted in a towing tank using computer-generated breaking waves. Models with various hull forms and ballast configurations were tested. Although certain design parameters such as beam-to- length ratio appeared to have a measurable effect on capsize vulnerability, no major improvements were found. A slightly larger wave would capsize all the designs.
Research done by the British has led to the conclusion that "although discernible trends in resistance to capsize have been determined, no form or ballasting combination consistently resisted capsize in the 0.42 m. high wave. [This corresponds to a 18-foot wave full scale.] This suggests that alterations in form which improve capsize resistance may be rendered ineffective by a relatively small increase in breaking wave height..." (Ref. 2).
This conclusion is the same one reached in Jordan's early work, i.e., that moderate design changes could not produce significant resistance to capsize. Therefore, it was decided .to investigate the use of sea anchors and drogues, devices which would hold the vessel in a safer orientation to the wind and waves and thus prevent capsize.
For the initial drogue testing, the horizontal water jet was used to simulate the breaking wave. The boat was positioned so that the wave front struck at 45 degree. from astern and the drogue was deployed 15 degree. from the wave direction. The model drogue consisted of a simple plastic disk with a wire shaft. It was found that a drogue with a diameter less than 10 to 15% of the length of the boat, i.e., 3 to 4-1/2 ft. for a 30ft. boat, would not exert enough force to pull the stern into the wave face. As a result the boat would broach and capsize However, a drogue with a diameter equal to or greater than 10 to15% of the length of the boat would pull the stern into the wave and prevent capsize.
These tests were run with a relatively stiff towline. When the model towline was provided with elasticity simulating a full-scale nylon line, the load would not build up as quickly and the model would often capsize. In the actual case the towline would be somewhat pre stretched at the time of wave strike. it was apparent that a better method of testing was needed to study this effect.
For the second series of tests, which are described in detail in Ref. 1, the models were struck by a breaking wave formed by the wake of a towed dinghy Without a drogue all the models would be capsized. When struck abeam they would of ten roll through 360 degree. When struck on the quarter they would sometimes pitch pole end over end. Various types of drogues were tested . For the conditions under which these tests were conducted, it appeared that a disk drogue 3 to 4-1½ ft .in diameter or an equivalent cone, parachute or series drogue would prevent capsize of a 30 ft . boat in almost cases whereas a 2 ft ; diameter drogue would permit the boat to broach and capsize in approximately half of the wave strikes. however the boat could be capsized with any drogue if there was too much slack in the towline at the time of the wave strike. As discussed later in this report, it is felt that in the real case is likely that there would be too much slack in the towline; particularly if a series type drogue is used.
The overall conclusion of this early testing was that a properly designed drogue could prevent capsize. Continuing effort was devoted to object of these tests was to investigate the dynamic behaviour of a sailing yacht model when it was struck by a confirming this result with a larger model and developing and testing a full-scale drogue with optimum characteristics.