Sailboat Performance Testing Techniques

Sailboat Performance Testing Techniques By Arvel Gentry Boeing Commercial Airplane Company Proceedings of the Eleventh AIAA Symposium on the of Sailing September 12 1981 Seattle Washington Abstract Testing methods data reduction techniques and data analysis programs used in the performance December 1999 testing of racing sailboats are reviewed and recording equipment are also discussed 1 Introduction 2 Personal Experiences in Sailboat Testing Sanding the bottom checking and rechecking the sails Sailboat performance testing has evolved rapidly over with the sailmaker and studying the local wind and tide the last ten years as a result of both intensive activity on the conditions are standard practice for the serious racer As 12meter boats plus the advancement of electronic and the boats get bigger the stakes get higher or the desire to computer technology My own personal experiences win increases so does the search for other factors that illustrate this evolution might improve performance However a still untapped resource for many sailors is the knowledge of how his boat My interest in boat performance started with attempts should perform in given sailing conditions to measure the effect of heel and bowdown trim on a 14 foot dinghy under very light wind conditions Small pieces Knowledge of boat performance can sometimes have of paper were dropped in the water and timed with a stop an important influence on your chances of winning One watch to measure changes in boatspeed with trim changes simple example is the knowledge of your boats tacking Later when I moved to a Cal 20 and had a knotmeter angle in different wind and sea conditions In light winds available I collected data to determine the optimum the tacking angle will be quite large on some boats greater downwind tacking angles Reference 1 than 100 degrees In smooth water and higher winds the tacking angle may get down into the low 70s Critical In 1970 I found out that the 12meter Intrepid was using tactical decisions such as tacking to the layline close boat an onboard strip recorder to record boat performance I crossing situations and hitting the proper point on the then set out to develop a similar device for my own boat If finishing line will be more accurate if you know your boats the 12meter boats made use of new technologies in order tacking angle to increase their chances of winning then why couldnt I My objectives were simple I did not have the experience of Selection of sails for the next leg of a course requires the people that I was racing against I certainly didnt have either good guesswork or a good understanding of boat the resources like the 12meter boats did but maybe my performance and the apparent to true wind relationships technical skills could help make up for my lack of sailing Many boats carry the wrong sails or fail to reef at the experience My approach was certainly smalltime by proper time because of a lack of knowledge of their boats comparison to the 12s but it is more typical of what most performance sailors might be able to achieve All of these problems can be solved with experience When I purchased Kittiwake a Ranger 23 in 1972 I However proper boat testing can shorten this learning equipped it with full sailing instruments The first strip period The history of the Americas Cup races shows a recorder that I developed for use on this boat is shown in continuous concern for achieving the best possible boat Figure 1 and was described in detail in Reference 2 all performance Performance testing with a pace boat and References Figures are at the end of the text This device the use of onboard electronic equipment and computers recorded boatspeed as the primary signal and periodically have become standard practice on Americas Cup boats for switched to record a few seconds of wind speed Wind todays maxiocean racers and to a lesser extent for boats angle and heel angle data were recorded by hand or on the grand prix racing circuit written directly on the strip of paper as it came out of the This paper presents a basic summary of the methods recorder A sample strip of output is shown in Figure 2 used in modern sailboat testing The information pre In 1974 Jim Kilroy asked me to build a recorder for his sented is based on the authors personal experience with new maxiboat Kialoa III and to participate in the trials of his own boats plus what he has learned in sailing and the boat off St Petersburg Florida The requirement was to testing on other boats develop a device that could be used on board during sea C 1981 Arvel Gentry trials that23 Ranger would furnish an Newsletter onboard record of perfor sets of onboard Even local racing boats mance and that would help to determine both average are sporting these expensive arrays of dials and digital sailing data as well as dynamic performance during readouts The basic instruments required for normal boat tacking The device that I developed was more compli testing are cated than that described in Reference 2 but still used a Boatspeed strip recorder as the recording media The Kialoa III Apparent wind speed recorder is shown in Figure 3 Apparent wind angle 0 to 360 degrees Compass After the experience on board Kialoa III I built a similar Heel angle recorder for my little Ranger 23 Kittiwake New electronic circuits were used but all of the basic functions of the Successful sailboat testing requires a thorough under Kialoa III recorder were retained The electronic circuits standing of each of these gadgets were designed by a friend Alan Sewell This recorder is 21 Boatspeed shown in use in Figures 4 and 5 It is described in more detail in a later section Boatspeed is usually measured by a sensor extending through the hull Present sensors fit into three classes In late 1980 Jim Kilroy asked me to help out during the 1 Paddle Wheel sea trials of his newest Kialoa maxiboat IV The trials were 2 Propeller again held out of St Petersburg Florida but this time the 3 Direct Force Measurement strain gauge professionals provided the computer recording equip ment The testing plan organization and engineering Figure 9 shows samples of these basic types of sensors support were provided by David Pedrick The recording The paddle wheel on the left is by Signet The center and computing equipment were furnished by 12meter propeller sensor is the part of the BG sensor that pro performance expert Richard McCurdy Most of the trudes outside of the hull On the right is the direct force equipment was the same as Pedrick and McCurdy had measurement sensor by Telcor used for the 12meter Clipper in 1980 where it was However the basic problem with all this equipment is identified as the Starship Nova system My role on the the sensor location on the hull The speed of the water past new Kialoa IV was as a performance testing engineer the sensor location is not necessarily the true speed of the The navigation station on Kialoa IV is shown in Figure boat The water changes both speed and direction as it 6 The CRT terminal at the left was used to control the flows past the hull Sensor position error can be quite onboard MicroNova computer and to enter sail trim sea significant depending on the type and location of the conditions etc The computer itself and the floppy disk sensor the size of the boat and the sailing conditions system is shown in Figure 7 McCurdy is obviously a If the shape of the boat and internal structure permits hardware man Figure 8 shows the equipment used to the best position for the sensor is usually on the centerline reduce and analyze the test data This equipment was ahead of the keel For the paddle wheel and strain gauge located in a shoreside trailer and consisted of a Data type of sensors this will give readings that do not change General Nova MiniComputer several terminals and a from tack to tack heeled readings may still be different plotter This shoreside computer equipment had also been from upright values It seems obvious that boat manufac used for the 12meter Clipper The floppy disk was not turers should provide an appropriate centerline speedo used on Clipper since the data was sent to a shore receiver thruhull location in with the design and construction of by radio The capability of this equipment is described in the boat but this is seldom done more detail in Reference 3 Sensors that use a small propeller or spinner however The MicroNova computer and terminal were still may read different between tacks even when the removed from Kialoa IV after the sea trials However centerline location is used This is caused by the fact that McCurdy has since developed the necessary interfaces the propeller rotates the same direction on both tacks The between the Brookes Gatehouse Hercules microcom effect of the prop support and the weed guard may cause puter system and an onboard Apple computer The Apple the prop to spin faster on one tack than on the other computer reads data from the Brookes Gatehouse data lines and the satellite navigation computer and then Large boats frequently use two sensors positioned on performs a variety of additional performance tactical and either side of the hull in front of the keel A gravity switch is navigational computations 4 The computer functions of used to automatically select the leeside sensor Much time the B G Hercules system are turned off and the Apple and effort is required to obtain consistent readings computer removed from the boat for races that do not between tacks for this type of installation Alignment of the permit this type of equipment two sensors may have to be different in order to give consistent readings on the two tacks This may introduce 2 Electronic Boat errors for the upright downwind sailing conditions Almost all grand prix racing boats have rather complete Manual switching may be required to select the most 2 reliable sensor signal depending upon the sailing condi The biggest problem with wind speed sensors is the tion location The masthead is subjected to flow distortions and speed errors due to the flow created by the sails The height The boatspeed sensors should be carefully calibrated of the sensor above the water must also be considered by sailing measured miles or by sailing close to a boat with when comparing data taken on different size boats well calibrated instruments Calibrations should be because of the wind speed gradient with height performed at various heel angles on both tacks and in the upright condition Uncorrectable errors should be 23 Wind Direction recorded and the proper corrections applied to all mea Several major problems plague boat wind direction sured boatspeed data devices The first is that most systems on boats that I have Sensors that use paddle wheel or prop rotation counter been on are not aligned properly so that they read the circuits are usually quite stable once the instrument is same on both tacks Careful alignment at the masthead properly calibrated With time however wear or damage together with small electrical adjustments at the naviga to the bearings can affect the readings The electronic tion table should give consistent readings circuits themselves can frequently be checked at the dock The next problem is that the wind direction sensor by placing a 60 Hz signal near the sensor such as a solder measures what it sees the wind direction at its location ing iron This may not be the correct apparent wind angle because A direct force measurement sensor such as that of flow distortion due to the sails the upwash effect and manufactured by Telcor Instruments is very sensitive at because of heel angle Means of correcting for these effects low speeds and is not affected by local flow angles will be covered later Calibration is accomplished just as with other sensors Most wind direction sensors are integrated with the However subsequent checks of the calibration can be wind speed device so that the rotating cups are located accomplished by simply hanging a small weight on the under the wind vane This means that the most practical retracted sensor tip and checking the reading This can position for the unit is on a rod extending at an angle out in even be accomplished underway front of the masthead In this position it is subjected to The type of boatspeed cockpit display depends upon strong sail upwash effects These effects may be corrected personal preference An analog display can usually be for windward conditions but are more difficult to account averaged by eye better than the digital display The digital for in the running and reaching conditions On large boats display can give a more accurate instantaneous reading in broad reaching conditions the removal of a staysail may but since the readings are almost always changing significantly affect the wind direction reading average values are harder to read Recording data manu The last problem inherent in wind direction measure ally requires some care Either record an average reading ments is the fact that the reading may be fluctuating quite a or record many readings and determine the average bit Average readings may be difficult to obtain Most mathematically systems have an adjustable electronic dampening control Some boats make use of automatic speed recording to slow down the system response so that the readings are devices The type of sensor may influence the selection not always jumping all over the place This will cause and design of the recorder equipment A digital circuit may problems if you are studying dynamic maneuvers such as require a DA converter if the data is to be recorded on an tacks It also means that attempts to sail by a VMG meter analog device such as a strip recorder may lead to bad results since the VMG computations use the apparent wind angle 22 Wind Speed 24 Compass There are several different types of wind speed sensors Figure 10 shows three examples The rotating cup sensor is Little new can be said about compasses except that most frequently used although it does have its problems they should be carefully adjusted before serious testing They are nonlinear at the low speeds the basic calibration starts Any errors should be noted and corrections applied may be affected by heel angle and they have bearings that to the readings before they are used in the data reduction wear out process Keep magnetic objects such as pliers screw drivers and portable radios away from the compasses The wind speed sensor made by Telcor Instruments is a during testing just as you would during a race solid state device with no moving parts and avoids most of these problems The wind blowing past a thermistor tends If you plan on using one of the more sophisticated to cool the unit The amount of current necessary to heat instrument and data recording systems you will need a the sensor back up to the balanced condition can be compass with an electronic readout measured and converted to wind speed The thermistor 25 Heel Angle tip must be cleaned occasionally to remove spider webs that degrade the sensitivity Heel angle is an important parameter that is frequently 3 left off sailing data sheets However it is required if the is the multifunction display units that can be positioned proper corrections are to be applied in the data reduction about the boat where they are needed one boat is reported process Heel angle will usually have to be recorded by to have twenty of these units Figure 12 shows the cockpit hand from readings taken off of small bubble indicators of Kialoa IV with its two sets of five readouts on either side The more sophisticated systems use an electronic heel of the center hydraulic control panel On board Kialoa IV angle device However none of the presently available the Hercules 190 system produces data that is read by the microcomputer based boat systems include a heel angle Apple computer and the Apple computer in turn puts input output data back on the BG data line for display on the multifunction units The Hercules System 190 itself My own electronic heel angle system consists of an contains 32 channels of data amplifier circuit with a weighted potenti ometer furnishing the heel angle signal This unit is on the Rochester Instruments makes a microcomputer based right side of the photo in Figure 5 boat system that was used on Freedom in the 1980 Americas Cup A photograph of the system is 26 Leeway Angle shown in Figure 13 The system computes speed made There is presently no completely satisfactory leeway good upwind or downwind true wind speed and true angle measuring device Various leeway angle measure wind direction off the bow One nice feature is the output ment techniques have been tried with varying success the port for a cassette tape recorder so that the basic sailing local flow angles measured on the boat are not the same as parameters can be recorded automatically Rochester the true leeway angle Sometimes a line is towed behind provides a service of converting the cassette tape to the boat and a large protractor used to record the angle printed output form that the line makes with the boat centerline However this Signet also produces a microcomputer based boat system is difficult to use because of the normal small angle system This system computes speed made good VMG changes in the boat heading as the wind and sea change true wind speed and direction and has a start timer Careful navigation from fixed sea markers can also be used but again accurate results are difficult to obtain In The present boat microcomputer systems have only the data reduction procedures used in this paper we will limited capacity for the more complex computations In use an empirical equation to account for leeway effects my opinion several practical implementation problems have not been solved As stated previously none of the 27 Navigation Instruments systems have a heel angle input and none provide a Sophisticated modern navigation instruments may be means for correcting for upwash This makes their VMG of some help in sailboat testing and their use should be and true wind results suspect investigated if you have them on your boat Satellite A display of relative boat performance would be navigation systems coupled with Omega systems have helpful as compared to stored polars The Hercules 190 been used to help detect water current variations that system uses a builtin set of data that represents general would affect testing boat performance using your input IOR 28 Microcomputer Based Systems rating The boat performance is compared with informa tion stored in the computer and a performance percentage The microcomputer chip is presently causing a revolu number displayed Data is provided for either windward tion in the sailboat instrument business Several manufac or reaching conditions turers have systems that use microcomputer circuits The boatspeed wind speed and direction and compass Ideally the user should be able to determine his own sensors send information to a central computer processor boats performance and to load it into an EPROM for use The information is then sent out to the cockpit display by the boat microcomputer system Another possibility instruments would be to have the data prepared by a home computer or by a service provided by the instrument manufacturer The new microcomputer based systems have tremen and then loaded into the boat microcomputer through a dous potential However I find the available systems still cassette tape lacking in some important areas Most systems compute what is called speed made good to windward VMG An onboard computer that is separate from the boat Accurate computation of VMG requires that corrections be instruments such as the Apple computer on Kialoa IV applied for both the upwash at the masthead sensor and provides a powerful system to assist and supplement the for heel angle None of the presently available microcom normal boat microcomputer instrument puter based systems have a heel angle input sensor They However the use of something like the Apple com also do not provide a means for correcting for upwash on puter requires a number of difficult interfaces with the different boats boats plus some sophisticated software The Brookes Gatehouse Hercules 190 system is 4 And last it would require a boat owner plus probably a shown in Figure 11 One of the best features of this system navigator who could understand and make maximum 4 use of such a system and that the output is not immediately available for use within modern computers 29 Automatic Data Recorders 292 Electronic Recording Performance data can always be recorded by hand onto data forms However this means of gathering data The electronic recording and processing of sailboat depends upon the judgment and diligence of the person performance was used extensively in the 1980 Americas writing down the numbers Automatic recorders are a Cup season David Pedrick and Richard McCurdy devel more reliable means of recording the number data but oped a rather sophisticated system for Clipper in an have their own problems which must be solved The attempt to shorten the learning and boat tuning time 3 primary one is the recording of data that is not available by Much of the equipment from Clipper was used during electronic means the sea trials of Jim Kilroys new Kialoa IV in 1981 For Kialoa Data is useless if you do not know what was happening IV McCurdy had to develop interfaces incorporate the on the boat when the data was taken This should include onboard floppy disk system and develop new shoreside such data as the sail configuration all the sail trim parame computer software As a result this sophisticated equip ters genoa car location outhaul halyard tension etc ment was not ready for the first part of the sea trials backstay pressure running backstay pressures babystay During this period it was necessary to record data by hand pressures helmsman sea conditions etc and to do all of the data reduction on an HP41C program mable calculator that was my job 291 The Strip Recorder During the Kialoa IV sea trials performance polars were The strip recorder provides one means of recording updated daily as new data was gathered The performance both the electronically generated data and the other testing on Kialoa IV was probably the most complete and information mentioned above Notes can be made right on sophisticated yet applied on a racing yacht including the the strip of paper as it comes out of the machine The 12meters recorder developed by the author for use on Kialoa III and on his own boat Kittiwake was a rather sophisticated 3 Testing Techniques instrument for its time The most important time for serious testing is right This recorder as it was used on Kittiwake is shown in after the boat is completed and before the first race Most Figures 4 and 5 The recorder was kept below during races owners want to get the most out of their boat as soon as but used in the cockpit during other testing periods possible and careful testing can aid significantly in During short races an audio cassette recorder was started accomplishing this However the performance testing with the strip recorder and recorded all of sail trim must not interfere with other equally important aspects tacking and tactical information A typical record from such as crew training general boat sail this recorder is shown in Figure 14 inventory checks and rig tuning On the Kialoa III and Kittiwake recorders a total of six Initial sea trial testing provides the first indications as to data signals could be input to the control unit Only two how the boat will perform under various conditions Data signals could be recorded at a time but a combination of gathered during this period should be considered as being automatic and manual switching permitted the effective preliminary since significant improvements in perfor recording of six parameters on a single strip of recorder mance will usually be obtained during actual racing paper Boatspeed VS was the primary signal on the lower conditions These initial tests however will usually channel The two secondary parameters on the lower provide a chance to obtain general trends that will be channel were the apparent wind speed and a spare useful in correlating the data obtained during racing channel used for the Brookes Gatehouse Horatio conditions computer output on Kialoa III Accurate performance polars require hundreds of data Any signal could be recorded fulltime or the primary points If possible the data gathering process should signal and one of the selected secondary signals could be continue throughout the racing life of the boat This will automatically alternated The upper channel had two provide an excellent baseline for comparison if modifica primary signals that were selected by a switch on the tions are subsequently made to the boat control unit These were the apparent wind angle scaled from 0 to 180 degrees and the closehauled wind angle The It is important that the maximum amount of data be signal from an electronic heel angle indicator was the gathered for the boat sailing in smooth water conditions single secondary signal for the upper channel This This usually gives the maximum performance characteris recorder was used on board Kittiwake for all of its races and tics for the boat If you know what the boat should be able practice sessions for over two years to do under ideal smooth water conditions you are better able to judge how the boat should be sailed as the sea The disadvantages of the strip recorder are that it needs conditions deteriorate Rough sea conditions degrade the someone to write all of the notes on the recorder paper performance of the boat Eventually you will want to 5 prepare two or more boat performance polars for different The area at the right is used for general notes as to sea sea conditions conditions sail trim rigging pressures comments on the estimated value of the data point etc 31 Data Gathering It is usually a good idea to record the data numbers Since most boats will not have the sophisticated close to the bottom line on each row This will leave the top equipment of a Clipper 12meter or a Kialoa IV maxiboat part of the row for subsequent data corrections ie for the rest of this discussion will assume that the data are to be calibration errors computations using different upwash gathered by the hand recording method constants etc A typical set of data is shown in Figure 16 Data to be used for constructing speed polars should 312 Windward Testing normally be recorded when the boat is settled down and at maximum speed for the sailing conditions However you The highest priority should be placed on windward at times may find it useful to record data at other odd testing The general procedure at first will be to let the conditions as the situations arise For example odd relative helmsman sail the boat at what he thinks is the best bits of data may help in correlating VMG and performance windward point of sail Give the helmsman time to get numbers computed by the onboard instrument system used to the boat and to get what he thinks is the best with subsequent values computed by the data reduction performance the best seatofthepants condition program that will include upwash and heel angle Record a series of data points over several minutes I find that one complete set of data taken every minute to be a One of the major problems in the data correlation good procedure process is the determination of the true wind strength and direction As stated previously heel angle and upwash at Then tack the boat and repeat the data on the new tack the masthead complicate this process During the testing This is very important since the compass tacking angle it is usually helpful to periodically bring the boat to a data is necessary in the correlation process to determine complete stop and head to the wind Record the wind the proper empirical upwash constants for the particular speed and compass direction and use this data as a check boat It is difficult to obtain useful windward data correla against the true wind values calculated during the data tions without a sufficient amount of tacking data reduction process Next have the helmsman intentionally pinch the boat a 311 Data Recording Form degree or two closer to the wind than he normally thinks is best Again gather data on both tacks Have the helmsman Everyone has his own preference for the format and sail at conditions that are two degrees and four degrees arrangement of the data recording form One of the ones wider than he thinks is best Be sure that the sails are that I have used is shown in Figure 15 trimmed as good as possible for each data point The data point sequence number is recorded in the Later you will calculate the VMG for each of these Point No column This number may be useful in the data conditions and plot the results as a function of apparent correlation process The clock time is input in the next wind angle This will tell you what the proper apparent column The PS column is used to indicate the tack port or wind angle is for the given wind and sea conditions The starboard Next comes the basic boat performance data to be used in constructing the windward perfor parameters mance curves will be picked off of these plots at the VS Boatspeed maximum VMG speed The computed true wind direc VA Apparent wind speed tions should be approximately the same providing the bA Apparent wind angle wind is not shifting This can be checked by noting the f Heel Angle compass and apparent wind readings CH Compass heading 313 Reaching Conditions The remaining data columns will be filled out during Testing for reaching conditions has its own particular the data reduction process These parameters are problems Upwash effects may be smaller than the windward conditions but are more difficult to determine VT True wind speed You should frequently bring the boat to a stop heading into g Polar wind angle includes leeway the wind so that you will have some way of determining TackJibe Tacking angle for windward work or the upwash correlation constants jibing angle for running VMG Speed made good to windward Sail selection is very important for the reaching g2l True wind angle without leeway conditions The test apparent wind angles should cover WD Direction true wind is blowing from the range for each sail configuration Additional testing e Upwash correction beyond what you normally think is the limit angle for a sail l Leeway angle configuration will help identify the precise sail change point for best performance 6 314 Running Conditions vortex representation of the sail which is a quite conven tional approach in aerodynamic theory we can get some The primary purpose of testing in the running condi idea of how conditions at the measuring unit are affected tion is to determine the optimum jibing angles As always by the flow field caused by the sails sail trim is very important Be sure that you also have all of the proper sails up staysails or blooper First sail dead For the masthead unit position shown in Figure 17 the downwind according to the masthead wind direction bound vortex and the tip trailing vortex are additive in indicator Record data and then jibe over and repeat the producing an upwash flow field Without going into the dead downwind condition again You may find a differ details of the aerodynamic theory it is sufficient to just ence in compass readings This indicates that you have state that the upwash velocity at any point w is directly upwash at the wind direction indicator just as with the proportional to the sail lift coefficient other sailing conditions w 2 2 A CL V Next sail the boat 10 20 30 and 40 degrees off of the where w is the upwash velocity dead downwind condition If possible repeat the data on each jibe before going to the next apparent wind angle a V is the frees tream velocity good time for crew training Under high wind condi A is a correlation constant that is tions be sure to record data both at the peak surfing speeds dependent upon the position of and at the lower speed lulls between surfing spurts Mark the wind measuring device the data somehow so that you will remember to use only CL is the sail lift coefficient the average of the two readings For windward work CL is a maximum at low apparent 4 Data Reduction wind velocities where the sails are quite full and the boat usually footed off and is at a minimum at the high The basic purpose of the data reduction process is to apparent wind speeds where the sails are trimmed quite convert the data from apparent wind conditions to true flat and the boat pinched to keep it upright wind conditions Some of the data such as the windward performance will eventually be plotted again in terms of To arrive at a basic form for the upwash correction apparent wind conditions for use during races However equation I used the old classical Gimcrack test data This even the windward data must at first be converted to true data is shown in Figure 18 along with the equation for CL wind conditions in order to establish the best VMG CL C 1 cos C 2 VA C 3 conditions The basic relationships used in the data reduction process are discussed below followed by a The constants required to match the Gimcrack data description of the computer program itself were as follows 41 Upwash Correction C 1 065 The upwash correction is probably the least under C 2 60 stood by the average sailor The important thing however C3 105 is to realize that no single upwash correction equation is The actual value of the CL is not too important We are going to work for all boats My approach has been to select a form for the correction equation that has the necessary only interested in getting a rough idea as to how it changes empirical constants to allow me to match the characteris with apparent wind speed It is important however to tics of a given boat Although I have used several different remember that these constants will not match your forms for this equation it is instructive to follow the particular boat You must arrive at the proper constants correction as I first derived it in the early 1970s through repeated trial calculations with different assumed values The objective is to find the right set of constants First lets look at a very simplified picture and aerody that will give the correct tacking angle as measured on the namic representation of the lift created by the sails The compass for the various apparent wind conditions drawing in Figure 17 is a representation of the vortex systems that can be used to approximate the effect of the The relationships used to convert the upwash velocity sails on the surrounding flow This is the bound vortex w to an upwash angle are as follows system that represents the lift of the sails and the trailing Ve tip vortex Detailed aerodynamic analysis also requires an e w upwash e tan1 wV image system below the water but I wont go into these V wV A C L details here Upwash angle e tan1 A C L From this idealized representation of the sails we can The above relationships are quite flexible providing readily see that the flow direction and speed at the mast you are only working with windward data However we head measuring unit are influenced by the vortex system really need a single equation that can also be applied to the used to represent the sail lift By looking at this simplified reaching and running conditions The general equation 7 that I have used for these purposes is as follows V V V x cos b Upwash e S1 COS S 0 VA COS bA V y A 0 V cos f sinb when VA S 2 use VA S 2 Vz 0 V sin f sinb Again the constants SO S1 and S2 must be deter The final components of the apparent wind vector in mined empirically from the test data For the larger boats terms of the boat coordinate system are therefore given the following ranges have been used depending upon the by boat Vx V cos b S0 2 to 3 Vy V cos f sin b S1 14 to 16 Vz V sin f sin b S2 25 to 30 The sign conventions used in the above equations The plot shown in Figure 19 gives some idea as to how assume a yaw to the left and a roll to heel the boat the upwash changes with different values for the S1 constant For small boats the S1 parameter may be small From the above we see that the wind vane at the or even zero for no upwash correction masthead is exposed to a wind vector V that is composed of three components 42 Heel Angle Corrections V is the component parallel to the boat x We know that the apparent wind angle would be centerline and is negative when the wind correct if the boat were in the upright condition At a 90 is forward of the beam degree heel angle the wind angle indicator would be V is the component from the side useless in measuring the apparent wind angle the reading y would approach zero These two extreme conditions tell V is the component along the mast axis and us that we will have to have some correction equation that z is negative when the boat is heeled must be applied to the measured values The masthead wind vane only rotates about the mast Complex diagrams and geometry can be used to arrive line Zaxis and therefore does not respond to the Vz at the heel angle correction equation However a much component of the velocity The angle of the wind simpler way is to use matrix notation and conventional measured by the wind vane is therefore given by rotation relationships To start this analysis the boat is assumed to be pointed directly into the apparent wind tan u cos f tan b The boat is then yawed about the vertical Zaxis to the b arctan tan u cos f correct boat apparent wind angle b and then rolled about where b is the angle of the boat centerline the hull centerline Xaxis by the heel angle f to the apparent wind The yaw rotation about the vertical Zaxis is given by f is the heel angle the matrix u is the apparent wind angle measured cos b sin b 0 by the masthead wind vane b sin b cos b 0 The wind speed measured by the rotating cups also 0 0 1 may be in error because of the heel angle However the rotating cups themselves may have a fundamental The roll to the required heel angle is given by calibration error when the flow is not perpendicular to the 1 0 0 rotation shaft It therefore becomes somewhat arbitrary f 0 cos f sin f in applying a geometric heel angle correction to the wind speed The correction equation used by David Pedrick and 0 sin f cos f Richard McCurdy 3 is as follows cos b uncorrected The complete rotation matrix is obtained as follows VA corrected VA indicated cos bcorrected cos b sinb 0 I have used this same equation in my own data reduction A b f cos f sin b cos f cos b sin f programs so that my answers can be compared directly with sin f sin b sin f cos b cos f Pedrick and McCurdys results 43 Leeway Correction The free stream apparent wind vector V may be expressed in terms of its components in the boat Since the leeway angle is not measured directly during coordinate system V the testing we will make use of an empirical relationship x along the hull centerline V y out as part of the data reduction process The basic relationship to the side and Vz up the mast used was derived by David Pedrick 3 8 f stored at location 104 in the program Start with a value of l K VS 2 about 10 and adjust it as you gain experience with the boat where l is the leeway angle in degrees K is the leeway correlation constant The program prompts for each of the input parameters f is the heel angle in degrees The output values are also identified A sample output VS is the boatspeed in knots from the program is shown in the lower right corner of Figure 21 Values for the constant K vary between 9 and about 16 depending on the windward efficiency of the boat This 5 Data Analysis constant will have to be empirically adjusted as experience The purpose of this whole exercise is to be able to is gained with a given boat prepare plots of the boats performance under different 44 Data Reduction Computer Program sailing conditions These plots take two different forms Hand reduction of the data is difficult and time 51 Windward Performance consuming and not worth the effort since excellent The optimum windward conditions are determined by programmable calculators are available The necessary plotting the VMG data against the indicated apparent equations can be programmed on an HP65 or an HP41C wind angle A sample set of data at two different wind With these little handheld computers the data reduction speeds is given in Figure 22 can be performed on deck during the actual boat testing runs as I did on Kialoa IV The optimum VMG correlations are then used to create a complete picture of the windward performance of the Subsequent analysis and plotting of the data can be boat This data will be for the optimum windward accomplished ashore Minicomputers such as the Data conditions in smooth water and will be plotted as a General Nova Minicomputer provide the much needed function of indicated apparent wind speed A typical set of capacity for the storage and analysis of hundreds of data plots is shown in Figures 23 and 24 Figure 23 gives the points and for the computer generation of speed polar optimum windward boat speed Figure 24 shows the plots Home microcomputers such as the Apple computer tacking angle heel angle and indicated apparent wind also do an excellent job angle Whether done on an Apple computer or an HP41C the As experience with the boat increases you will be able to basic data reduction process is the same The basic wind prepare similar plots for different sea states triangle relationships are shown in Figure 20 The 52 Boatspeed Polar Diagram following data reduction steps are required 1 Set upwash and leeway correction constants The next major task is to prepare the complete speed 2 Correct apparent wind angle for upwash polar for the boat This will require the use of the 3 Correct wind angle for heel angle windward performance data discussed above plus the 4 Correct apparent wind speed for heel angle reaching and running data and a lot of guess work 5 Solve the wind triangle to get true wind speed and The usual form of the speed polar is shown in Figure 25 angle The angle parameter is the true wind angle plus the 6 Calculate tacking or jibing angle leeway angle The heartshaped boat speed lines are 7 Calculate VMG prepared at constant true wind speeds This plot should 8 Calculate leeway angle really have several in slope at the points 9 Display wind angle Change sign CHS if on port where the sails are changed However the available test 10 Calculate wind direction data is usually so scarce and contains so much scatter that it An HP41C program for this basic data reduction is the usual practice to just draw smooth polar curves process is given in Figure 21 The name of the boat should I find that a Cartesian form of the speed polar is easier be stored in location 02 Estimates for the upwash to work with on the boat than the polar coordinate plot A constants for your boat must be stored at locations 96 98 sample plot is shown in Figure 26 This plot contains a and 100 before running the program Start with an Sl value combination of true wind data and apparent wind data of 00 no upwash Set S0 3 and S2 30 If you have significant upwash at the masthead the calculated tacking The solid lines running from left to right are the boat angle will be larger than the value read on the compass speed curves at constant true wind speed The long between tacks With a trial and error process and the use dashed lines running roughly from the top of the plot to of plots similar to Figure 19 repeat the calculations until the bottom are lines of constant indicated apparent wind the calculated tacking angle is close to the measured value speed The very short dashed lines running from the left As you gather data at different apparent wind speeds you up toward the right are lines of constant apparent wind will have to adjust the S0 and S2 constants and then again speed The optimum windward sailing conditions are search for the proper S1 constant The leeway constant is shown at the left side of the plot Optimum downwind 9 tacking conditions are indicated by the VMGMAX line at the Biography right side of the plot Arvel Gentry is presently a research supervisor in the The generation of this plot is quite an art The basic Aerodynamics Research Department at the Boeing problem is that you never have enough data and what Commercial Airplane Company He has raced his own data you do have seems to always have too much scatter I boats very successfully primarily in Southern California find it helpful to draw a complete polar plot before even and has extensive crewing experience on larger ocean looking at the data obtained on the boat This purely racing yachts He has authored numerous magazine guessed polar can then be used both as a means of articles on sailing aerodynamics and sailboat performance analyzing the test data and as a basis for the subsequent He has conducted research efforts in support of Americas polars based on the experimental data As more test data is Cup projects and designed the mast section shapes used obtained the polar can be shifted and revised so as to on Courageous and Freedom He has also developed better match the test results Owners of MHSrated boats specialized sailboat performance recording equipment can now get a computer generated polar diagram along and served as a sailing performance test engineer on Jim with their rating However this curve should only be Kilroys Kialoa maxiboats considered as the starting point for your own diagram 6 Use of Performance Data Good boat performance data curves have their greatest use as a yardstick for measuring boat performance during races Current boat performance can always be compared with the windward plots or the polar data and adjustments sought that will bring the performance up to or greater than the plots The data can also be used as basic input information for detailed tactical situations such as selecting the optimum current crossings As both a tactical and navigational aid the data can help predict where you will be at a later point in a race and allow studies of possible tactical decisions under different future true wind conditions 7 Conclusions Knowledge of your boats detailed performance can have a significant effect on your chances of winning Experience in the 12meter Americas Cup boats and maxiocean racers clearly illustrate this However even the local sailor can improve his chances if he learns more about his boat The technology required to assess boat performance is available to the average serious sailor The required equipment includes a complete set of sailing instruments a programmable calculator such as the HP41C and a technical understanding of the testing and data analysis process 8 References 1 Arvel E Gentry Optimum Downwind Tacking SEA Magazine February 1970 2 Arvel E Gentry Are You at Optimum Trim SAIL Magazine March 1974 3 David R Pedrick and Richard S McCurdy Yacht Performance Analysis with Computers Chesapeake Sailing Yacht Symposium January 1981 4 Joanne Fishman Kilroy Is Here MOTOR BOATING and SAILING Magazine August 1981 10 Figure 1 Single channel strip recorder for Kittiwake 1972 Figure 2 Sample data record from single channel recorder 11 Figure 3 Strip recorder on Kialoa III 1975 Figure 6 Navigation station on Kialoa IV 1981 Figure 4 Performance testing on Kittiwake 1976 Figure 7 MicroNova Computer on Kialoa IV 1981 Figure 5 Kittiwake dual channel strip recorder Figure 8 Shoreside computer room for Kialoa IV 12 Signet Telcor Brooks Gatehouse Figure 9 Boatspeed sensors Telcor Signet Brooks Gatehouse Figure 10 Combination wind speed wind direction systems 13 Figure 11 Brooks Gatehouse Hercules System 190 Figure 12 Cockpit of New Kialoa IV 14 Figure 13 Rochester Microcomputer based system Figure 14 Sample output from Kialoa III and Kittiwake dual channel strip recorders 15 16 Figure 15 Sailboat performance data recording form Figure 16 Typical set of hand recorded test data Figure 17 Vortex representation of sail lifting system 17 Figure 18 Lift coefficient correlation for Gimcrack Figure 19 Influence of S1 parameter on upwash angle 18 Figure 20 Apparent wind triangle parameters 19 Figure 21 Performance data reduction program for HP41C 20 Figure 22 Typical VMG optimization plots Figure 23 Typical boatspeed plot for optimum VMG conditions 21 Figure 24 Optimum windward conditions 22 Figure 25 Typical boatspeed polar plot 23 Figure 26 Typical boatspeed polar plot in cartesian coordinate form 24