Americas Cup Sail Design

INTRODUCTION

Upon embarking on a sail programme for an America’s Cup campaign, The mind of a sail designer may be filled with ideas of developing methods to calculate the flow field over sails to the nth degree, or to invent some method to solve the aero-elastic behaviour of the sail iteratively.  Alternatively, one may wish to tackle the as-yet unexplored areas of sail/rig interaction, or solve the dynamic flow around the sails when sailing downwind.

In reality, there is not enough time to do the things you may wish to.  There are production schedules and materials that need to be ordered well in advance.  The designer must also not forget that the views of the sail trimmers on the boat (the “clients”) should never be underestimated.

Like all aspects of the Team New Zealand campaign for the 1995 America’s Cup, the sail programme was based on a Team concept.  North Sails New Zealand was contracted to build all the sails for this campaign.

North Sails is a commercial entity,  and so much of what was learnt during the America’s Cup campaign is proprietary in nature and cannot be discussed at length in this paper.
 

DESIGNING SAILS FOR A PARTICULAR BOAT

Boat design has evolved over the years along many development paths.  Unlike the aircraft industry where development has been essentially driven by commercial considerations,  yacht design has, in most circumstances, had to comply to rules set down by the administrators of the sport.  Throughout the past 100 years,  there have been certain classes set aside for “development” purposes.  These range from the various skiff classes (12 foot,  18 foot) to the America’s Cup classes.  J-Boats were used prior to World War II,  and the 12 metre class was used until 1987.  Unfortunately,  despite being a development class, the 12 metre rule proved quite restrictive,  and by the 1980s the boats were quite outdated in terms of their hull form.  This lead to the development in 1989 of the International America’s Cup Class (IACC).  This rule was first adopted for the 1992 America’s Cup regatta.  In this regatta most importance was given to hull design and evaluating the correct relationship between the three main parameters in the rule:  Waterline Length, Displacement and Sail Area.  The 1995 America’s Cup regatta was more focussed on refining the use of each individual parameter.

Sail designers usually are given the task of designing sails for a given hull and rig design.  In the forum of the America’s Cup however,  the sail designer will probably have some input into the basic parameters of the boat.

Whatever the hull form and rig size of the boat is,  all yachts experience the following three “modes” when sailing upwind.  These can be loosely termed under-powered,  powered up and overpowered.  Each mode is defined by the actual Velocity Made Good to windward (VMGboat) relative to the maximum VMG of the yacht.  VMG is the component of the boatspeed in the direction that the true wind is coming from.

The hull speed can be determined by the following empirical relationship.

 VMG_max = (Waterline length in feet)^0.5     (1)

Bear in mind that in its simplest form,  the Lift force generated by the sails is a function of sail area and  Lift coefficient (itself a function of sail camber).  Wind speed and the angle of attack of the sails also affect the Lift force.

 Lift = k₁(t) * A * AWS² * a      (2)

where  t is some function of the effective depth of the sail combination
 A is the sail area of the rig
 AWS is the Apparent Wind Speed
 a is the effective angle of attack of the sails
 k₁ is a function of t with Lift Coefficient

Similarly,  Drag is also a function of sail area and sail camber (due to skin friction),  but also the effective aspect ratio of the rig.

 Drag = k₂(t) * A * AWS² * a + k₃(AR-1)     (3)

where AR is the effective aspect ratio
 k₂ is a function of t with skin frictional Drag Coefficient
 k₃ is a function of 1/AR with induced Drag Coefficient

The first term of (3) refers to the skin friction drag,  while the second term refers to the induced drag of the rig.

If the VMG_boat < VMG_max then the boat is said to be underpowered.  In this situation,  the primary consideration is to develop the maximum possible force from the sails.  This is achieved by increasing the camber of the sails,  at the expense of skin frictional drag.

If the VMG_boat = VMG_max at the minimum possible windspeed then the boat is powered up. Here the boat is sailing in its most efficient mode.  The camber of the sail is prescribed to generate enough force to achieve this boat speed.  Any more sail depth would lead to excessive heel angle and loss of both sail and keel efficiency.  If the sails are not deep enough,  then there will not be sufficient force to attain VMG_max.

The final mode occurs when the windspeed is in excess of that required for the the Powered up situation.  Now there is ample sail force to sail at VMG_max so the key to optimal sail performance is to maintain a total sail force close to that experienced when the boat is powered up,  despite the fact that the onset windspeed has increased.  This is achieved by initially decreasing the sail camber.  However,  as the wind strength increases,  it becomes necessary to decrease sail force further.  From (3) above this can be achieved by decreasing sail area.  Also,  if possible,  this should be done by minimising drag as well.  This has lead to the development of high aspect ratio or blade sails for high wind applications in order to minimise induced drag.

When designing sails for a particular yacht,  it is important to know at what windspeed a boat will sail at VMG_boat = VMG_max.  Often a Velocity Prediction Program (VPP) can help predict at what windspeed a yacht will be powered up.  Once this is known then it is possible to design sails with the correct depth for a given wind range.

PLANNING

Contrary to popular opinion, America’s Cup syndicates do not operate on unlimited budgets.  Therefore an important aspect of the Team New Zealand sail programme was to establish what sails would be needed and when they could be built.  This had to tie in with the proposed sailing programme of the yachts and of course the racing schedule.

For the initial sails, North Sails aimed to take the best sails from the 1992 campaigns and apply ideas that it had learnt in the intervening two years.  The most significant development for the mainsail and genoas (upwind sails) was the ability to use North Sails’ 3DL (Three Dimensional Lamination) manufacturing process.

The 3DL process was invented by two Swiss engineers;  JP Baudet and Luc Dubois who brought their idea to North Sails who subsequently developed it.  Rather than using the traditional method of developing flat panels into a surface with absolute curvature,  the 3DL process builds the sail over a full size physical mould with the shape of the mould being determined by the sail designer.  A layer of polyester film is placed over the mould,  then a six axis plotter lays high modulus yarns (Kevlar,  Carbon Fibre etc) yarns over the surface of the mould to a layout prescribed by the sail designer.  A second layer of film completes the sandwich and infra-red radiation activates glue that is on the films and yarns and the shell of the sail is cured.

The yarn layout used and the types of yarns selected affects the aero-elastic behaviour of the sail membrane so these areas are very important to the sail designer.

Team New Zealand was the first group to make a commitment to 3DL, and it was also the only group to have all its mainsails and genoas built using this technique.

3DL sails are inherently smoother than panelled sails and are more efficient structurally.  This in itself meant that the sails for the 1995 America’s Cup would be significantly better than those from 1992.  However, the 3DL process was available to all America’s Cup syndicates, so the design that the sail is moulded to would still be a critical factor and an area were one group could gain an advantage. Also the yarn layout used and the types of yarns selected would affect the aero-elastic behaviour of the sail membrane.  The understanding of the difference between the unstressed moulded shape and the flying shape of the sail has been the main challenge of sail designers over the years (Figure 1).

UPWIND SAIL DESIGN

As mentioned earlier the IACC rule has three main inputs:  Waterline Length, Displacement and Sail Area.  It is the job of the boat designers to come up with these parameters.  The sail designers must work to the sail area prescribed by the designers.  For the Genoas, the dimensions are fixed as the foot length must be no greater than the fore-triangle base plus 3 metres.  The Mainsail area (MSA) is calculated to be a sail area which is determined by using Simpson’s Rule at five girth stations up the sail.  This means that the sail designer can opt for a long foot and moderate roach, or a short foot and high roach yet still end up with the same rated MSA.

MSA = (P-0.5) * (E1+4*E2+2*E3+4*E4+E5)/12 - C0*E5/2      (4)

where  P is the distance between the black bands on the mast
 E1...E5 are girths across the sail
 C0*E5/2 is a term that allows for the fact that the tack angle may not be 90  degrees

The full set of IACC rules are available from the class measurers.

Mr Tom Schnackenberg,  one of the most respected sail designers in the world, experimented with a high roach design in the 1992 America’s Cup preliminary rounds aboard Spirit of Australia with limited success, with the problem being the inability to control the sail twist up high.  For the first mainsail on NZL-32, North Sails opted for what was at the time a high roach design.  The curvature in the sail was redistributed, and improved sail battens meant that this sail was a relative success.  As time went on, this concept was pushed further.

For ideal aerofoils the effective span can be defined as

 Span_eff = CL*(A/(p*CD)^0.5     (5)

where CL = Effective Lift Coefficient
 CD = Total Drag coefficient
 A = Total Sail Area

A high roach design increases the effective span of the sail which is a good measure of the efficiency of any foil.  This is achieved by reducing the tip vortex shed near the top of the span by making the sail less “triangular” and more “elliptical” meaning less induced drag when sailing upwind.  Induced drag considerations have been neglected in the past.  It is not generally realised that in moderate wind strengths the induced drag due to the sails may contribute as much as 15% of the total drag of the yacht.  Also, as the wind increases, a high roach design tends to twist more in the head which decreases the angle of attack in the top quarter of the sail and de-powers it.

While the upwind performance of the mainsail is paramount, one nice benefit with a high roach mainsail is that the downwind performance is also enhanced as there is more sail area up high.

The high roach of the Team New Zealand mainsail designs meant that the sail was quite complicated structurally, so the yarn orientations used had a large bearing on what the flying shape of the sail would be.  This was further complicated by the use of Carbon Fibre yarns with around twice the Young’s Modulus of Kevlar 49 which was the standard yarn used in these sails.  The low stretch materials certainly show up any inconsistencies in the mainsail structure.  It was North Sails aim to produce a sail with constant strain across the entire surface of the sail.  This enables the sail to respond uniformly in varying wind conditions.  However, as the stress map of the sail is constantly changing due to differing trim settings, this is difficult to achieve in practise.  The yarn layout used in the mainsails is shown in Figure 2.  In order to predict the stresses within the sail under certain conditions  a finite element stress analysis program written by Michael Richelsen of North Sails called MemBrain was used.  The main advantage of this program over other commercially available stress analysis packages is that it links directly with the North Sails design software which was used to determine the 3DL mould shapes.

While the MemBrain program gave the designers a pretty good idea of what was going on within the body of the sail,  full scale observations were still relied on to make the final modifications to our next generation of sail designs.  Improved structural models the sails will be sought for the sail programme leading up the America’s Cup defence in the year 2000.

The structural design for the genoas was a much simpler process as the triangular geometry of the sail is more stable under load.  The MemBrain program yielded results that could be relied on, and the results of initial studies proved adequate throughout the campaign.  The yarn layout used in the genoas is shown in Figure 3.

Aerodynamically, the two factors that most affect the performance of genoas are overall depth and the distribution of twist up the sail.  It is generally accepted that a sail that will set up as close as possible to the spreaders on the mast will enable the yacht to point higher.  A small amount of work was done with panel codes which suggested that the optimum performance is achieved if the sail is sheeted inside the spreaders,  but this cannot be done in practise.  At present the structural requirements of the rig are the limiting factor in achieving an optimum upwind sail performance.  The method which is used to get the twist distribution of the genoa as close as possible to the rig is proprietary and cannot be discussed here.

It was pointed out earlier that the overall depth of the sails for a particular windspeed is unique to each boat design, and while a VPP (Velocity Prediction Program) may give some insight to the required lift coefficients, fine tuning can only be done on board the yacht at full scale.  The point to realise here is that an IACC yacht reaches hull speed at a very low true wind speed compared to (say) an IMS boat.  Therefore a genoa with 15% camber at the middle girth of the sail will perform best at different wind speeds on different classes of boats.  An important spin-off of the America’s Cup is that the designs developed during this regatta can be applied to other racing yachts.  The key is finding just where these designs will fit into other inventories and what adjustments need to be made to account for differing mast configurations.

North Sails have undertaken this project in the months following the Team New Zealand victory in May 1995 and these designs have performed well on the European IMS racing circuit.

DOWNWIND SAIL DESIGN

In the 1995 America’s Cup, the course on which the races was held was modified from the one used in the 1992 America’s Cup to include three upwind legs and three downwind legs of approximately equal length.  This increased the importance of the downwind legs as they made up a larger portion of the race.  Tactically it is important to hold a speed advantage downwind as there are more opportunities to pass a boat on the running legs because it is more difficult for a leading yacht to cover the opposing boat.

Another factor that was evident from the 1992 campaigns was that there was still a lot to be learnt about the downwind sails.  There are two main types of downwind sail.  Spinnakers are the more traditional sail and by definition they are symmetrical in cross section and size.  A Gennaker is an assymmetrical version of a Spinnaker.  In general, Gennakers are better suited to lighter wind conditions as the gybe angles are larger and the apparent wind angles are smaller. In extremely light air conditions the apparent wind angle may be as small as 35 degrees when sailing downwind.  However under normal circumstances, it is more typical for the apparent wind angle to be in the range of 70 degrees to 100 degrees.  As such,  the best downwind sails simply have to generate the maximum amount of lift,  while flying in a stable manner.  Remember that the downwind sails are only held at the corners and,  unlike mainsails or genoas, are not supported along any of their edges.

As the wind strength is increased, the apparent wind moves aft and the Spinnaker becomes relatively more efficient.  It is still a matter of debate whether or not a Gennaker is faster than a Spinnaker in these conditions in a straight line, but a Spinnaker is easier and quicker to gybe which is of advantage in a tactical sense.

Figure 4 shows the relationship between apparent wind angle and true wind angle in differing wind strengths.

As it was not practical in terms of both time and budget to embark on a full scale testing schedule,  North Sails chose to base its design development on both full scale testing as well as model testing in a specially commissioned wind tunnel.  The tunnel was a collaboration between Team New Zealand, North Sails New Zealand and the University of Auckland and 1:15 scale Spinnakers and Gennakers were tested in it.  The details of the design and construction of the tunnel are covered elsewhere in this issue in the paper by R G J Flay.

The wind tunnel was built to simulate the actual sailing conditions on board the yacht, complete with a boundary layer and windshear experienced on the race course.  Once the wind tunnel results were calibrated against the full scale testing performed in San Diego, it was found that the model results could be relied upon with a high degree of accuracy despite the fact that we did not attempt to match the Reynold’s Number of the model with that of the full scale sail.  In all over 300 model sails were built and over 2000 tests were performed in the near ideal conditions of the wind tunnel.

At the beginning of the sail programme the designers were able to use their favoured Gennaker designs of the time, then improve on them by trying different ideas in the tunnel. The force in the direction that the boat is travelling (thrust) was chosen as the performance criteria when sailing downwind.  The side force generated is small so heeling considerations are not as important as when sailing upwind.  In some cases the driving force increased by 15% from first generation sails to those that were used in the America’s Cup races.

In all there were five different codes of downwind sail which covered the expected wind range from 0 to 20 knots (0-10 m/s) of true wind.  Over the nine month period of development,  several design breakthroughs were made on each code or style of downwind sail. The result was big improvements in downwind boat speed at critical times in the campaign.

SUMMARY

Much was learnt over the nine months leading up to the America’s Cup.  Already North Sails is applying some of the lessons learnt to IMS boats and to other classes of yacht. Although sail design is still controlled primarily by empirical observations,  numerical results are becoming more important although most of these solutions lack the resolution required to base sail designs solely on their results.

There were big improvements made in the structural aspects of sail design between 1992 and 1995 primarily because of the 3DL manufacturing process.  Improved structural models for the sails and the aerodynamic properties of the sails will be scrutinised more in the next five years.  All marine related designers are looking forward to the defence early in the year 2000 in Auckland - Home of the America’s Cup.
 

 List of Captions
 
 
 

Figure 1.   Sail design development process showing the relationship between moulded sail shape and flying sail shape.
 
 

Figure 2.   Team New Zealand 3DL yarn layouts for Mainsails.
 
 

Figure 3.   Team New Zealand 3DL yarn layouts for Genoas.
 
 

Figure 4.   Relationship between Apparent Wind Angle (AWA) and True Wind Speed when sailing downwind.