6 t4V. 1976
UNIVERSITY
OF
SO UTHAM PT
ON
department of
aeronautics
and astronautics
ARCHIEF
ADVISORY COMMITTEE FOR YACHT RESEARCH
Lab.
y.
Scheepsbouwkunde
Technische Hogeschool
De!ft
0
S.U.Y.R. REPORT NO.15
A METHOD OF DETERMINING THE EFFECT OF SAIL CHARACTERISTICS ON A
YACHTS CLOSE-HAULED PERFORMANCE
AND OF COMPARING THE MERITS OF DIFFERENT RIGS
SUMMARY
This report describes a procedure whereby, given the hydrodynamic characteristics of a yacht hull, plots may be prepared which enable the windward performance of the yacht to be deduced directly when the lift
and drag coefficients of the rig are known. This method of performance prediction ¡s likely to be useful ¡n cases where it ¡s desired to compare the performance of a given hull under a number of different rigs. Fairly extensive computations are required to present the hull characteristics
in the required form, but these can be carried out quite easily using a digital computer.
Part II of the report presents a set of plots of this kind for a
12.-metre yacht and discusses briefly the inferences which may be drawn from them.
V (knots) mg k (deg) (deg) 9 (deg) y (deg) cp (deg) 5m (deg) NOMENCLATURE
Fore-and-aft Axis Intersection of Centreplane and
Waterplane of the yacht.
Track Line along which the yacht
moves, (fixed relative to the water").
VT (knots) True wind velocity, relative to the water. VA (knots) Apparent wind velocity, relative to the yacht.
V (knots)
Velocity of the yacht, along the track (relative
to the waterS')
Speed made good, component of the yacht's velocity (relative to the water"), in the direction opposite
to VT
Leeway Angle, Angle between track and fore-and-aft
ax ¡ s.
Apparent Wind Angle, Angle between the apparent wind vector and the fore-and-aft axis.
Relative Wind Angle, Angle between the apparent wind vector
(VA) and the track (V5)
Angle between the True wind vector (V1) and the track (V5)
Heel Angle of the yacht.
Angle between main Loom and yacht's centreplane, measured in a plane perpendicular to the
centre-plane.
If these words
are included, this nomenclature and the work in this report are appi ¡cable even when the water has a velocity relative to the earth due to tide or current.
C1 , C0 Lift and drag coefficients of the rig, including
above water part of the hull. (Lift and drag forces measured in trie horizontal plane, perpen-dicular and parallel to VA).
C C C Lift and drag coefficients as above, measured with Lb L20 L30 Lcp
CD10 CO20 CD CD
F F (lb) Components, respectively parallel and perpendicular
to the fore and aft axis, of the air and water
Fxa Fya (lb) forces on the yacht measured in the horizontal plane.
C C Coefficients of F and F
X y xa ya
R , H (lb) Components, respectively parallel and perpendicular
to the track, of the water force on the yacht, measured in the horizontal plane.
Mxa (lb ft) Heel ing moment about the fore-and-aft axis due to air forces on the rig including above-water part of the hull
M (lb ft) Righting moment about the fore-and-aft axis due to water and gravity forces on the yacht.
M (lb ft)
xp Value of M when H = O
, V
O
xw
M (lb ft) Value of M , when H O and V = Û
Xo xw S
i.e. Static Righting Moment
SA (sq ft) Sail Area, taken as area of Mainsail (excluding roach) + 100% of Area of Foretriangle.
CEH (ft) Center of Effort Height, defined as M cos
Coefficients of Air Forces =
Force/pA
. VA2 . SAwhere VA ¡s in ft per sec, and VA = 1.688 VA
= air density, slugs per cu ft.
o o o
PART
NTRODUCT ON
The recent development of methods of measuring aerodynamic characteristics of sails n the wind tunnel has considerably increased the importance of the
quest on:
''How can one decide from their aerodynamic characteristics whether one rig or another will give the better performance to
windward, on a hull of known hydrodynamic characteristics?.
Clearly this is not the only criterion to be considered when designing a yacht rig. A compromise must be achieved, taking into account performance on other points of sailing, convenience of handling, behaviour in gale con-ditions etc. None-the-less it is generally agreed that windward performance is very important, so much so that the use of tank testing in yacht design has so far been directed mainly to improvement in this respect.
Methods of determining yacht performance to windward for given hull and
sail characteristics were developed by Davidsonh23. Procedures more
appropriate to the use of coefficients for a variety of sails have been
sug-5 6,7
gested by Sainsbury and Crewe . Fundamentally all these methods and the
one described in the present paper are based on three equations expressing the conditions of equilibrium of the yacht, namely:
Air drivirìq force
r-water resistance (in fore and aft direction) Air side force r- water sideforce (in transverse direction)
Air heeling moment = Righting moment due to water and gravity forces. Davidson's procedure involves certain special assumptions concerning the
behaviour of sail forces, while that of Sainsbury is the most general. Both
wind speed for chosen values of V (ship speed through the water), and c.p
(heel angle). These calculations are repeated for several values of V at each of three or more values of cç
Vmg
VT 2 = 3020'
=10'
The results are then plotted as in the sketch and the envelope to the several curves for different values of p represents the best obtainable
per-formance of the hull and rig configuration under consideration. To determine the effect of any change of hull or rig (for example a change of boom angle) it is necessary to repeat the calculations and determine a new envelope.
Crewe's work includes a study of the way in which sail characteristics affect the performance calculation and he has shown that many, but not all, non-optimum rig configurations may be eliminated at an early stage in the calculations. Even so, with any of these methods, comparison of several Sail configurations on the same hull is liable to involve considerable calculation
for each sail
condition. To perform many such calculations by handwould
bean extremely laborious task. On the other hand it is now possible to carry out the calculations on a digital computer, and to arrange for most of the plotting to be done automatically. Thus, if facilities of this kind are
avail-able, calculation of V - V1 envelopes for a number of configurations may in some cases be regarded as a feasible alternative to the procedure described
in the present report.
It will be realised that much of the earlier work on performance calcu-lation was directed at the problem of comparing the merits of different hulls on the basis of tank test results. This purpose can be achieved by manual calculation of V - V envelopes without excessive labour. Fortunately
mg T
used, since any reasonable system will usually give much the same assessment of the relative merits of various hulls of simi lar type. Thus the use of a
simplified representation of sail characteristics, such as Davidson's Gimcrack sail coefficients, is usually acceptable for this purpose. However, when the purpose is to study the effect of various rig configurations and to elucidate
the optimum rig characteristics for sailing to windward, then it is valuable to have at hand a method whereby the performance of a given hull under many different rigs can quickly be examined.
lt is the purpose of Part I of the present report to show how hul I
charac-teristics may be presented in a form which enables the speed made good to
wind-ward to be deduced directly when the lift and drag characteristics of the sails are known. Fairly extensive computations are required for any hull, but these can be performed quite easily using a digital computer. Thereafter the wind-ward performance of this hull under any rig can be deterniined with very little
calcul at ion.
The procedure to be described makes use of the same basic principles and the same equilibrium equations as the other methods mentioned above.
Accord-ingly no substantial difference is to be expected in the final results.
However, it is important always to bear in mind that any performance predictiun based on tank and wind tunnel results is subject to various errors and
uncer-tainties. These arise both from difficulties in scal ing-up model results and from differences between test and fullsize conditions, such as the gustiness of real wind and the waves on real water.
A set of plots representing the hull characteristics of a 12-metre yacht has been prepared in accordance with the method discussed and is presented in
Part II of this report.
PRINCIPLES OF THE METHOD
A basic requirement for any performance prediction is knowledge of the hydrodynamic characteristics of the hull, which may be obtained from tank tests, or otherwise (e.g. the characteristics of a notional hull might be arbitrarily specified, or devised by modifying those of a known hull). For
the present purpose it is necessary that this data shall be available in a
form from which the resistance, sideforce and righting moment can be computed for any values of ship speed
(V),
heel angle (cp) and leeway angle (X) within appropriate limits.Not all tank test results are in the required form, since tests are often made for only a few values of p and V , and results cannot readily be
derived for intermediate values of these variables. Fortunately, however, Crewe6 has given a set of figures for a 12-metre which can be used for the present purpose with a little approximation. Furthermore, it is now the
prac-tice at M.I.T. to present yacht test results in the form of polynomial expressions for resistance and sideforce in terms of V , p and X
. Such
data ¡s well suited for our present purpose, and data for a 12-metre tested at M.I.T. have been used to illustrate the present discussion. More extensive
performance calculations for the same yrht are given in Part I of this report.
Together with the hull characteristics we must specify two parameters con-cerning the rig, namely the Sail Area and Centre of Effort Height (CEH).
Methods of investigating the effects of changes in Sail Area and CEH are des-cribed later. Some knowledge is also needed of the way in which sail lift and drag vary with heel angle; this requirement is also discussed ¡ri greater
detail later.
We now consider Figure 1, which consists of three separate plots, each for a different value of the apparent wind angle,(). The axes represent values of lift and drag coefficients, C and C , for the rig at a
L0
D2Ochosen heel angle, 20 in this case. For convenience the plots were prepared for those o' values for which wind tunnel sail data were usually obtained.
The actual o' values in the tunnel were 17.5, 20 and 22.5 degrees.
However, corrections were introduced to allow for the effect of the tunnel walls, and the results were considered to apply to o' values of approximately 18, 20.5
and 23 degrees ¡n open air. More recently the procedure for calculating wind tunnel corrections has been revised, and results are now obtained for somewhat different values of o' . In any event it is usually desirable to plot wind
tunnel test values of CL and CD against o' and faired values can then be
picked off for any desired values of o'.
The lines in Figure I are computed from the hull data, etc., specified
above, and characterise the behaviour of the hull. Their significance is defined thus:
If at the specified heel angle, 200 in this case, and at the o'
value of the plot lift and drag coefficient values C and C
L2O D20
for the rig (including the above-water part of the hull) correspond to any point on one of these hull characteristic lines, then, in a
true wind of 13 knots the yacht will make good to windward the speed indicated alongside this line, when sail ing in calm water at the apparent wind angle specified for the plot.
If the CL2O and CD values for the rig correspond to a point falling between the lines of constant V , the appropriate V
mg mg
value may be estimated by interpolation.
It is important to note that the above statement does not imply that the yacht will actually sail at a heel angle of 20°. As will be seen below, the plot is arranged so that Vmg values at the heel angles which actually occur are obtained from sail tests at a fixed heel angle.
Plots are needed for several o' values, and on each plot it is
conve-nient to mark the point corresponding to the CL2O and CD2O values for the rig at the value of o' appropriate to the plot. In Figure I such
points are shown for two different rig settings, with the main boom at angles of 10 and 30
to the centre line of the hull.
It is now apparent that the greatest speed made good is obtained with
the 10 boom angle at an apparent wind angle (o') near 20.50.
A simple plot of V against o' , as shown in Figure 8, may be used
to arrive at a good estimate of the optimum V and of the o' value at
mg which this occurs.
Since points on the curves of constant V represent the behaviour mg
of the yacht when sail ing under a rig having the indicated values of
CL and C0 , it is possible to indicate as parameters on the curves the
values of any quantities associated with sailing conditions, such as the heel angle (cp) , true wind angle (y) , leeway angle (x) , ship speed
(V)
, etc. Some of these quantities are indicated in Figure1 and on the
Although a later section describes n some detail the procedure by which
the hull characteristic lines are computed, a brief explanation is now given, since it may faci i ¡tate understanding of their use.
For any characteristic line, SA, CEH, V , and V
T mg
are fixed. We choose a series of values of the leeway angle X
covering the range in which sailing conditions will lie, and for each value of X we carry out the following procedure.
Since - c' + X
, V , VT
and are known so that
V and
VA may be calculated from the geometry of the vector
triangles (see sketch).
With V X and CEH known, the heel angle (tp) , water
sideforce and resistance are determined from the known hydrodynamic and stability characteristics of the hull.
The air force on the sails is of course equal and opposite
to the water force on the hull, and the apparent wind velocity and direction (VA and ) being known the air force is resolved
into lift and drag components. These are converted to coefficients
CL and CD based on the known sail area of the rig (SA).
These
CL and CD values clearly refer to the rig heeled
at the angle q, just determined, which is probably not equal to
the heel angle, say 20° , for which the sail characteristics are
known. The next step is to use the known (or assumed) variation
of C and C with q, to find the values, C and C
L D L20 D20
at q, = 20 to which the
CL and CD values just determined
would change if the heel angle were changed from q, to 200
without any other change in the rig. These values, CL and CD2 , are plotted to define a point on the hull characteristic
1 ¡ne.
Other points on this line are found by repeating the calculation for different values of . The whole procedure has now to be
repeated for the various values of V and mg
Accordingly, ¡f at cp = 200 and at the value of one of the plots, the rig gave C and C values corresponding to a point on one of the hull
L20
characteristic lines, the yacht, with this rig and at this , might well
sail in a 13 knot wind at a heel angle somewhat different from 200. However,
the CL and CD values given by the rig at the actual heel angle would be such as to provide the sail forces necessary to give the indicated V
It is evidently the case that the solid lines in Figure 1, whose func-tion is to represent the hull behaviour, are dependent to some extent on a property of the sails, namely the variation of CL and CD with p.
For-tunately the experimental evidence at present available suggests that this variation is likely to be fairly consistent so that a single plot, as ¡n Figure 1, can probably be used with adequate accuracy for a variety of sail configurations. However, this is certainly a point to which attention should be paid in future sail testing.
The importance of accurate knowledge of the variation of CL and CD with p may be minimized by choosing the values of p and V1 for the
plot so that optimum performance is likely to occur at an actual heel angle which does not differ greatly from the heel angle for which the plot is
pre-pared and for which, of course, sail data must be available. Then the necessary corrections to CL and CD are likely to be for relatively small changes of p, in the part of the plot relevant to optimum performance.
In order to study yacht performance under different wind conditions it is desirable to prepare similar plots for several values of VT , choosing
these values so that the optimum performance occurs at heel angles near
those for which the sail data are available. Suitable values for most displacement yachts are about:
p lO 20 30 degrees
VT 7.5 13 20 knots
For yachts of different types, e.g. Catamarans, quite different values of p may be required. However, much the same values of VT should be considered in order to demonstrate the effect of varying sail characteristics when sailing in light, medium and strong winds.
CHANGES OF STABILITY
t is now appropriate to consider how we may examine the effect of
changes of stabil ¡ty on our performance estimates. This is obviously impor-tant if we are interested in the effects of, say, changes in ballast ratio. Furthermore changes of stability have much the same effect in the calculations as changes in centre of effort height (GEH). The latter may be of interest either because we wish to consider performance under a different rig with
different CEH, or because we find, for example, that CEH may change somewhat as we vary boom angle.
The significance of a change of either stability or CEH is that it alters the relation between the sideforce H and heel angle p . Davidson1 adopted
the reasonable approximation that the transverse component of the sail force acts perpendicular to the centre-plane of the yacht. If CEH is the height, above the waterline, at which this force intersects the centreplane, the rela-tion between the air heeling moment M and sideforce F
is:
xa ya
M = CEH . F sec cp
xa ya
Fyw secç
We may with advantage go a step further and define CEFI as
Ma
cosIf this relation is used to calculate GEH from sail test results, the
ques-tion whether the sail force is exactly perpendicular to the centre-plane becomes of little importance in the present work.
The air heeling moment is, of course, equal to the water righting moment,
M . Since the water sideforce acts some
distance below the axis of moments which is taken in the water-plane, increase of sideforce reduces M If
as an approximation we take this distance d as constant, we may write
M - d.H sec p - M = M H sec cp. CEH,
x xw xa whe re H = sideforce and H F y M cos p M cos p
H=
CEH (CEH + d)Here M depends principally on the heel angle but varies somewhat with the
Xp
yacht's speed through the water.
The values of CEH and d depend on the geometry of hull and rig. As
a rough rule, the centre of effort usually lies near the centre of area of the sailpian, and d is of the order of half the draft. Thus a 25% increase of
stability, i.e. of M , has the same effect, on the important relationship Xçp
between H and p , as a roughly equal percentage decrease of CEH.
The procedure we adopt to cope with variations of stability or CEH is
simply to carry out the computations for two or more different values of the relation between H and p to plot both sets of results. This is done in Figure 2, where the lighter lines correspond to 25% greater sideforce at any heel angle than the heavy lines. Interpolation between the lines for
differ-ent CEH makes it possible to determine V for intermediate values of CEH.
mg
The above discussion is intended to clarify the effects of stability and
CEH. lt is not necessary to assume, in calculating the hull characteristic
lines, that d is constant, since the value of M can usually be taken
directly from test results.
CHANGES OF SAIL AREA
We may readily examine the effect of changes of sail area, since say, a 10% increase of sail area has much the same effect on the sail forces as a 10% increase of the CL and
CD values for the sails. Approximate
allow-ance can be made for the fact that a part of the total CD due to the hull, of the order of .02 for a 12-metre rig at r 2U°, will not increase. Thus
the increase of CL would be 10% and that of C0 , 10% of (C0 - .02). It
may also be necessary to allow for an increase in the height of the rig, for unless all dimensions are increased in proportion it will probably be incorrect to work from the original sail coefficients. Thus it would be appropriate to estimate the performance for the increased sail area by relating the new sail CL - C0 point to hull curves adjusted for 5% increase of CEH. (See also
pp 23-4)
USE 0F C AND C
If it is desired to examine the effect of sail changes expressed in terms of sail driving and sideforce coefficients, C< and C,, , the hull
charac-teristic lines can readily be plotted on C - C axes. In fact these
characteristic lines are identical in form with those on the CL - CD plot,
and the change only requires drawing C - C axes at an angle to the CL - C0 axes. Figure 3 shows hull characteristic lines identical to those of Figure 1 plotted on C - C areas.
There have been various discussions in the past as to the relative effects of changes in sail driving force (Cg) and sail sideforce (Cv)
We may now easily obtain some quantitative information on this point. For
instance Figure 3 shows that, in the vicinity of the optimum point (10 boom angle), a rig change will improve Vmg provided that the resulting increase
of C exceeds approximately 1/6 of the increase of C . This assumes that
the rig change does not involve a significant change of CEH. If such a change does occur it must be allowed for by the procedure described in the previous section. It is also assumed that there is no change in the optimum
value of . Any such change appears likely to have only a small effect but
¡t could ¡f necessary be allowed for. It does not by any means follow that the same relationship between changes of C and C appi ies in other con-ditions or to other yachts. However, more information on this point may be obtained if the present procedure is applied to yachts of other types.
We may note that, given the present procedure for determining Vmg
plots of CL and CD are no less convenient to use than plots involving C
and C
. CL
and CD plots may often be preferred since CL and CD are
fundamental aerodynamic quantities whose behaviour can more readily be fore-seen and understood from general principles than ¡s the case for C and C,,.
METHOD OF COMPUTATION
When we set out to compute CL and CD for a point on one of the hull characteristic lines, Sail Area, CEH, V , V and ' are given, and we
choose arbitrarily a value of X . Different X's will give different points
on the line.
Now - ' + X and it can be shown from the geometry of the velocity
triangles that
Sin (2y - = (1 + 2 V "y ) sin
mg! T
y
=v
/cosS mg
and VA = VT sin
y/sin
Thus VA and V are found.
For any value of , sail tests give M and F . The relation
xa ya
between Fya and Mxa is conveniently expressed by defining the Centre of Effort Heiqht, CEH , by the equation
M CEH . F
xa ya
It is found in sail tests that CEH thus defined varies only slightly with and cp . Accordingly it is convenient to treat GEH as a parameter
in the present computations.
Now M M and F
=F
xw xa yw ya
so that M = F . CEH . sec cp
xw yw
Also the hydrodynamic characteristics of the hull determine M xw
and F in terms of p , V and X. CEH, V and
X are known, and so
yw s s
we have 3 relations involving the three unknowns p, M and F
.
Unfor-xw yw
tunately M and F are generally complicated functions of V , p
and
X , so that an analytical solution is not possible.
However, no great diffi-culty arises in obtaining the required solutions for p and F by an
iterative process using a digital computer. R can then be computed from
the hydrodynamic characteristics of the hull.
Hull hydrodynamic characteristics are usually given in terms of
R and H rather than F . Since R and H , like F
and F
yw xw yw
are components of the resultant horizontal water force
F R2 +
H2 cos(tanR/H
-yw
or very nearly
F = H - R.X (where X is in radians)
yw
Further, since R. ). is of the order of 2% of H little error results from
tak i ng
F
=H
yw
and this approximation has been used in the computations in the present report. The resultant horizontal force on the yacht having now been found, the lift and drag components of the air force are determined and these are con-verted to coefficient form using the apparent wind angle , the apparent
wind velocity VA , and the sail area of the yacht rig.
(There are various different methods which may be used in calculating sail area as a basis for coefficients, and ¡t ¡s important that the area used here is consistent with the area used in calculating the sail data)
We have now determined the values of sail CL and CD and the heel angle p corresponding to the specified values of Vmg VT
and X
So far no assumptions have been made other than those commonly involved in performance calculations, including the assumption of constant CEFI
However, it is necessary to convert the values of CL and C0
which we have
calculated to those which the same sails would have at the fixed heel angle at which the Sail tests have been performed, say 200 . This conversion
requires knowledge of the variation of sail CL and
CD with cp and,
strictly speaking, the results are only applicable to those sails for which the variation assumed is correct.
The results of some 12-metre sail tests were found to agree well with the
follo.iing relations:
2 2
C
=C
.cos
20/cos r.pL20
2
CD = CD . cos 20/cos2
+ .ûûo(
- 20)( - 20)(' and cp being measured in degrees)
The last terni is chosen empirically to fit an existing set of measurements,
and i t may iel be possible to f md better expressions for general use. If
di ffernt expressions are found to be appropriate to a particular rig under consideration, it is not difficult to make the necessary corrections since only the very last stage of the computation is affected, and this may readily be altered. Furthermore, the results can often be made somewhat insensitive
to the variation of CL and CD with cp , by choosing V1 so that optimum
performance for the hull and rig concerned occurs at a heel angle near to that for which the sail data is available.
The calculations described above are repeated for the same values of V1 and Vmg with several values of X , giving pairs of values of
CL and CD
all on one hull characteristic line. Then by varying Vmg and
the
various hull characteristic lines shown in Figure 1 are obtained.PART II
APPLICATION TO A 12-METRE YACHT
INTRODUCTION
This part of the report presents a set of plots prepared in the form described in Part I , and representing approximately the characteristics of a
12-metre yacht. When values of sail CL and CD are known, for a heel angle of 100, 20° or 300, it is possible to read off directly from the plots the
speed that the yacht would make good to windward, sailing in a true wind of 7.5, 13 or 20 knots. Values of heel angle, true wind angle, and leeway angle may be obtained at the same time. The effect of varying the yacht's stability or the sail centre of effort height can also be studied.
To illustrate the use of the plots sail CL and CD values from wind tunnel tests of a 12-metre rig are indicated.
As explained in Part I, the preparation of these plots requires hull data in a form from
which
the resistance and other important characteristics can be derived for any value of the heel angle. The procedures at the M.I.T. tankdiffer from those used elsewhere and regularly produce results in this form. Therefore the author has welcomed the opportunity to use data on a 12-metre tested at M.I.T. for a first application of the procedures described in this
report.
It should be noted that the yacht to which the present work refers dif-fers somewhat from other recent yachts of this class and is not entirely
representative of the characteristics of modern 12-metres. Thus, while the plots indicate broadly the characteristics of 12-metres, and to some extent of yachts in general, it should not be supposed that they are exactly
applic-able to all modern 12-metres.
HULL CHARACTERISTICS
Stability
The data concerning the stability of this yacht are in less complete form than is usual, since the effect of speed on the righting moment was not investigated. However, Crewe6 showed that, for another 12-metre, this effect was small.
The information given on the stability of the M. 1.1. 12-metre is that
Static Righting Moment = M0
= 57J
-
2 in degrees)= H sec p . (CEH + d)
= H sec cp .
36.5
(for general discussion of this relationship see Part I, page 9.)
Crev'e also showed that an expression of this kind well represented the data for the 12-metre that he studied.
The present report includes plots for the yacht with the stability given
by the above relation, and also for the yacht with the static righting moment, and consequently the sideforce at any heel angle, arbitrarily increased by 25/.. These two cases are referred to as 'Normal Stability and '25% Increased
Stability'
Since changes of CEH have the same effect in the calculations as changes
of stability, we may if we wish regard the normal and increased stability cases as representing hulls both having the original stab lity hut with differing centre of effort heights. If we assume the same value for d as that obtained
by Crewe for his 12-metre, namely 3 ft. approximately, we arrive at the
follow-ing values:
Normal Stability CEH =
33.5
ft.25% Increased Stability CEH 26.2 ft.
The significant effect of changes of either hull stability or CEH on
the calculations is to vary the relationship between the sideforce (H) and
heel angle (p) , which for the two cases is given in round figures n the
following table.
o
p=
10 20° 30°Normal Stability H 1L+S0 2500 3200 lb
25% Increased Stability H = 1800 3153 +000
By interpolation between the plots for the two conditions the performance of the yacht may be obtained for any intermediate values of hull stability or
CEH.
In accordance with the usual M.I.T. practice the value taken for GEH in the stability relation includes allowance for the effect of the velocity gradient of the wind. Wind-Tunnel tests indicate that GEH for a 12-metre rig is approximately 28 ft. without velocity gradient, so that the value of 33.5 ft. taken for the normal stability case implies that GEH ¡s increased by some 20% as a result of the velocity gradient. Although this increase
appears to be a large one rough calculations suggest that an increase of this order is not impossible. However, no reliable data on the appropriate figure
is available and for this reason the velocity gradient is often disregarded in performance calculations.
If the user of the plots in the present report desires to obtain perform-ance estimates assuming a GEH not corrected for velocity gradient, this may readily be done by interpolating for the desired GEH between the
character-istics given for the two different stabil ities. Resistance, Sideforce and Leeway
The tank test results for this yacht are given in the form of 13 term polynomial expressions for the resistance and sideforce expanded to fullsize, in terms of the yacht's speed, heel angle and leeway angle. These po1ynomias are very convenient for the calculations that have to be made, but in order to indicate the hull characteristics in a more easily comprehensible form Figure L+ has been prepared from them. Here resistance is plotted against speed for
o o o . .
zero heel, zero leeway, and for heel angles 10 , 20 and 30 , with fixed
side-force values at each heel angle corresponding to the normal stability and 25% increased stability conditions. Values of the leeway angle have been indicated as parameters on these curves.
At the time when the study of this hull was carried out it was believed at M.I.T. that the relationship between sideforce and leeway was subject to scale effect between tank model tests and the fulisize yacht. A correction was applied for this which had approximately the effect of making the leeway
angles given for the fulls 12e yacht less by some 30% than those measured on the model at the corresponding speed and sideforce. More recently fulisize tests of the 5+ metre Antiope have indicated that the scale effect on side-force is small. Thus the present plots may well be based on somewhat smaller
leeway angles than actually occur with a 12-metre.
SAIL CHARACTERISTICS
The CL and CD values marked on the ordinate and abscissa scales of the plots really represent the lift and drag components of the air forces necessary to propel the yacht, divided by V . Sail Area. The sail
area assumed in the preparation of the plots was 1900 square feet (this represents the area of foretriangle and mainsail without roach for a repre-sentative 12-metre rig). Thus, if a rig of different area, or Sail coeffici-ents based on a different nominal area for the rig,are used the CL and CD values for the rig must be multiplied by the area used divided by 1900
before being applied to the plots.
As pointed out in Part I of this report, preparation of the plots requires
knowledge of the variation of sail CL and CD with heel angle. The present plots have been prepared on the basis of the empirical relations given on page 13 of Part I
The sail characteristics indicated on the plots are taken from faired results of tests of a 1/9 scale model of a 12-metre rig in the wind-tunnel. They are intended to indicate the approximate values of the sail coefficients that may be obtained with such a rig but they may well not represent the opti-mum characteristics that can be achieved by adjusting sail trim etc. No
allowance for scale effects between model and fullsize has been included.
As is appropriate in this and most other methods of performance
calcula-t i on calcula-the air forces on calcula-the hul 1 are included with the sail forces. n the
the bare hull gave negligible lift, and thit its drag was sufricient to in-crease the drag coefficient of the rig by an amount increasing roughly
o o
linearly from .015 at = 18 to .03 at r 25.7
The CEH values for these sails measured in the wind tunnel showed
little variation with heel angle or apparent wind angle over the ranges re-quired for the present appl ication, and corresponded to approximately 28 ft
fullsize. This ¡s ¡n the absence of wind gradient. To obtan performance
figures from the plots for this CEH we should relate the sail points with interpolated performance lines of the way from the lines for increased stability (CEH 26.2 ft) toward those for normal stabil ity (CEH = 33.5 ft)
Alternatively if we accept the M.I.T. assumption that the wind gradient causes some 20% increase of CEH we may relate the sail points with the
per-formance lnes for normal stability. It should be noted that the wind gradient will probably have some effect on the CL and CD values so that the performance figures thus obtained might not be exactly correct in the presence of a wind gradient.
INTERPRETATION OF THE PLOTS
In order to illustrate the use of the plots we will now consider certain inferences concerning the behaviour of 12-metre yachts which can be drawn
from them. However, it should be borne ¡n mind that we are here more con-cerned to illustrate the use of the plots than to enter into detailed discussion of 12-metre behaviour.
Performance in Light Winds
Consider first the plots (Figure 5) for sailing ¡n a true wind of 7.5 knots, on which have been inserted points representing the behaviour of a
12-metre rig in the wind tunnel at heel angle io', with boom angles 10 and
30
Let us read off Vm values for boom angles 10 and 3, and plot these against the apparent wind angle
()
as in Figure 8.We see that an optimum V of about L.7 Kn is achieved with ô = 1°
mg m
and o about 22 . For boom angle 1 , we can take the following approximate
figures from the plots
19 23° 11° 20.50 Ql O "2 y 40 °
and for optimum performance, which occurs about = 220, we find by inter-polation that c 10 and y = -5 approximately.
Although at o' 23°, V for Orn = 3 is about equal to that for = 10, the wider sail trim ¡s inferior at the lower values of . On the
other hand, ¡f by further reduction of the boom angle, or by other alterations of the sail trim, C could be increased above the value it has for o
= 1,
L m
without the simultaneous increase of CD exceeding about 1/3 of the increase
of CL , then the resultant sail point on Figure 5 would come nearer to the
V = +.3 line and the yacht's performance would be improved. This draws attention to the fact that in light winds sails should be trimmed to give
large CL values provided that the increase of C0 ¡s not excessive. Experience in certain 12-metres has shown that this requirement can be met and optimum performance in light winds achieved by using sails of relatively large camber with the boom approximately amidships.
Another method by which performance ¡n the 7.5 knot wind might be
im-proved is by increasing the sail area. This would result in an increase of lift force proportional to the increase of area, and a somewhat less than proportionate increase of drag, since part of the drag is due to the windage of the
hull.
(see pp. 11 and 17-18) We might estimate the effect of such changes by moving the sail point upward from the o = I point along a linethrough
this
point from CL = 0, CD = .02 (.02 being the part of the drag coefficient corresponding approximately to the air drag of the hull.It
will
be noticed that for the 7.5 knot wind no plots are given for the25% increased stability. This ¡s because the change of stability ¡s found
to have a barely significant effect on the position of the constant V9
lines. The only effect of this change is to reduce the heel angle by 20%. Values of Vmg y and . do not change significantly.
Performance in Moderate Winds
For the 13 knot wind (Figure 6) two plots are superimposed, the heavy one for the normal stability case, and the lighter one for the 25% increased stability. Sail points have been inserted for wind-tunnel tests of a 12-metre
rig at heel angie 20°, boom angles 1° and 3.
When we relate the sail points to the normal stability plot we see that
the lo boom angle ¡s rather better than 30 and optimum Vmg is close to 6.2 knots and occurs with near and slightly above 20.50. t is possible to
read off from the plots values of y and cp for the sailing condition e.g. at = 20.5, ô = 370 and cp = 22°
For this condition it appears that increase of CL would only ¡rnprove performance if the associated increase of CD was less than about 1/14.5 of
the increase of CL
For the increased stability case, optimum V9 is nearly 6.14 knots and ¡s achieved with = 20.5°,
5m
= 10, y
370
and cp = 18°. In this con-dition it seems that some advantage might be gained by increase of sail area,
though, after allowing for the increased centre of effort height, the benefit obtained is likely to be small.
Performance ¡n Strong Winds
For the 18 knot wind, (Figure 7) plots are again shown for normal and increased stability and sail points are given for heel angle 300, boom angles
o o o .
1 , 3 and 5 . The improvement of optimum performance resulting from the
increased stability is now about 0.14 knots compared with about 0.15 knots ¡ri
the 13 knot wind and zero ¡n the 7.5 knot wind.
In the 18 knot wind it appears for the normal stability case that the largest boom angle, 5°, gives the best V , just over 6.4 knots, with
= 21.5 , y = 35 and cp = 30 . It is clear that neither increase of CL
nor of sail area would be beneficial. Indeed some reduction might be advan-tageous if the reduction of drag achieved at the same time exceeded about 1/7 of the reduction of lift. For the increased stability case, it appears that
the boom angle 30 gives the best Vmg nearly 6.8 knots, with = 210
y 350 and cp = 270. In fact slightly better performance may perhaps be
obtained ¡n the normal stability case with a boom angle somewhat greater
than 50, and ¡n the increased stability case with a boom angle between 30
o and 5
CONCLUDING REMARKS
With the aid of the above examples, and of the discussion ¡n Part I of
the report, it should be apparent that the effect of different sail charac-teristics on the performance of a yacht can readily be deduced from plots of the kind described. However, as there are uncertainties in the expan-sion to fuilsize of model test results for both sails and hulls, it should be recognized that too great reliance must not be placed on the exact results obtained from performance predictions by this method, or indeed by any other method. Attention must also be paid to the fact that the present
calculations refer to yacht behaviour in calm water whereas yachts commonly
sail in a seaway which ¡s likely to have a considerable effect on their performance. Subject to those reservations it is believed that plots of the type presented in the present report will be of value in indicating the general trends to be expected in yacht behaviour and the directions ¡n which improvement may be sought by changes in rig characteristics.
In order that the general behaviour of yachts and the desirable characteristics of sails may be better understood it would be useful to prepare hull characteristic plots in the form described here for a number of representative yachts.
In future experimental and theoretical work on sails, regard should be had to the need for expressions giving the variation of CL and CD with heel angle. Such data will evidently be of value in connection with the present method of performance estimation and is likely also to be helpful in other studies of yacht performance.
REFERENCES
DAVIDSON, K. S. M., "Some Experimental Studies of the Sailing Yacht'', T. Soc. Nay. Arch. and Mar. Eng., New York, 1936.
A Review of the Gimcrack Sail Coefficients, DL TM 16, Stevens Institute of Technology, 1935.
Sailing Yacht Testing Technique, DL TM 17, Stevens Institute of Technology, 1936.
+. Investigation of Yacht Sail and Rig Characteristics, DL TM 55, Stevens
Institute of Technology, l9+O.
SAINSBURY, J. C., "4indward Performance Prediction", Southampton University Yacht Research Report 8, 1961.
CREWE, P. R., "Estimation of Effect of Sail Performance on Yacht Closehauled Behaviour", T. R.I.N.A., 196k.
CREWE, P. R., "Methods of Estimating Yacht Performance". Read at a meeting of Southern Joint Branch RINA & I Mar. E, 196k.
HERRESHOFF, H. C., "Hydrodynamics & Aerodynamics of the Sailing Yacht", T. Soc. Nay. Arch & Mar. Eng., New York, 1964.
r
1.CL2
1.i
1 0.o
0.
o
Viz,-'
u 3Oe.10
o
o
01
0.2
D20o
Fig I
Plots of 12 metre performance
against CL & CD
Rig
at heel angle 200
rrue wind 13 knots
0.2 CD
03
Vmg:6.4
/
6.0
56
/
¡
i
I
A.4z37T
As71
y
/
.226 A .34A.2
25A-3
A .23 2. J/
I
.Au2/
IVs668
( .331
6'lo-4iji
iliiir
3)
2Vu 763
Vmg64 6.0
5.6
A.3!
/
/
1V1667
Az4j1 :32.9
/th//'
e =
1 7i
I
*O
9 :221
D 9 A 28
Ï//
A .26
/
I
/
A.15AWA:o218°
/Az2
1
vas'.»
jiii1iii
14
CL20
13
1.2 1 .1i
o
V1s37Y 40
0. 0. 0.Figicont. GJ
12 metre rig CL & CD at e:20,
Çm:1°&3°
Vmgz6A
-V-..-06e
56
V3=5 1
=4i'21A
e:25o
2IV,:74$
A:5z415
e237
:.
Az3 2. A=25/
7 A:151/
rV=7.17A:23e.6
Le:123
II
I01
D20
1. 1. 1.
0
o
o
0.
o
01
o2
0.3
D20
o
01
0.2
C
D200.3
Fig. 2
Plots of 12 metre performance against
CL & CD
Rig at heel angle 20°.
True wind 13 knots.
Vmg64t
6.0
156
7g
/./
7/,
tì/
///AWA20.5O
11H
IHl
r
Vmg 64 6.05.
11/
AWA:18°
Fig.2 cont.
Normal stability
Stability increased 25°/e
02
01
02
01
0.3
02
Cx20
06
08
I0$
10
cx20
10
10
12
AWA
o.18
:6.4
.-
6.0
AWA :
20.5
cY20
.6.0
Fig.3
Plots of
12 metre performance against C & Cy
True wind 13 knots.
Enter with sail C & C
for heel angle 20°
® 12 metre rig C & C
at O20°, Çm° & 1°
Vrng :6.4
5.6
60
AWA:0.
23°
5.6
Cv20
5.6
Cr20
!I1
III
I Pf8
0.1
1100
1000
900
800f
u, m -J700
600
500
400
ÇH:
1800-e-1o1
z 1450
300F-8:30°
60
8:20°
SI D E FORCEz H z 4,000
Hz 3200
70
'z
H:3,150
/
H: 2,500
Z
Fig.4
12 metre hull characteristics.
Normal stability
Stability increased
25/e
AJ
ti
I
lilt
80
Vs(KNOTS)
/
ÇHEL
ANGLE :e0
[LEEWAY ANGLE:A:0
i
111
o
o
o
06
Vmg4.8
/
4.4
/14.0
/
CL,0
Y:
/.
'/1/
I
/
i
35//e= 7.5
= 34
/
2 I4'
',
II
/
/
7'I
If
/
/
/
I
25
y33
/
I
1.I/
/
/
,/
m
AWA18°
ABIIC010
-Vmg48
4.4
/.,/
4.c
k'!
15¡
AWA
205
Azl5
C010
-i ¡ i i Iii r
o
01
02
0.3
o
01
02
03
Fig.5
Performance plots
for 12 metre
True wind 75 knots
Erter w;th sad characteristics for
10° heel.
1.
I
i
i
i
o
o
aUi
z 15Fig. 5 con t.
Normal stability
25e1
change of
stability
or CEH changes
e ony
:56
e :13.1
Vrrs .4
/
e:IoK
I
I
I
I
I-I
iv-51.
A :2
2.6/
Aaii
AWA ::25.7'
/
Colo
111I1H
4
3 2Vmg:4.8
/
Q,
/
4.4
CL,0
i
/
/
/
/
2.5/
20.
eio
/
/
//
/
/
s,
/46
I
/
/
/
/
/
,//
//
1
/
/
/
/
/
/
/
/
.9.8
/
/
/
/
/
/
/
I
/
/
/
/
/
/
/
/
/ r AWA:o:23°
ACD,0
IHI
Liir
0.102
0.3
Fig.6
Performance plots for 12 metre.
True wind 13 knots.
14
13
12
i 1
10
09
08
O.o
o
12 metre rig
CL&CD at
e2o,
1°&3°
Fig.6 cont.
Normal stability, CEH 33.5ft
-
Stability increased
25°/e or CEH reduced to 262ft.
Vm.
6.4
Q
6.0
.5.6
CL2O
e2o4
A À. 15.r-40
r
/
/
i
e-io
/
'2
/1.5AWA:o:23°
CD20
»iii itir
0.1 .204
o
A2
-LIItIJ11d
01
02,-'
03
'- D0
V
Eß
64
6.0
Fig.7 Frformance plots for
12 metre
True wind 18 knots.
I
GB
64
60
Vmg:1
¡168
64
6.
f
1'11
0.0
o
o
o
o
12 metre rig CL & CD
at O3O Çm
1°, 3°& 50
0
0.10.2 CD30 0.3
Fig. 7 cont.
Norma! stability1
CEH 33-5ft.
Stability increased 25°/e or CEH reduced to 262 ft.
)
\ \
I te:31L
¡/
:39
e=3c t-j
\\\
e:30
IH
:39
e25
:37
-P
\
e
=2t
=37e:
25)... 5 2 1 3AWA:o
230
f
Az2 - -I i i i I6.5