An article on the results of Star tank test that appeared in Yachting
in January, 1941
Tank Notes on Star
by Harry A. Hofmann
SEVERAL years ago,
after Pimm had provoked a revolution in the Star Class, the shape of her
hull started a fountain of conjectures on the advantages to be gained
by refining the standard Star hull. Pimm's claim to fame, it is true,
lies in introducing the perfect flexible rig but discussions of her inevitably
led to the question of what had been done to perfect her hull. Naval architects
could tell, to some extent, the effect of a few changes in hull form but
it was felt that only after comparing models in a towing tank could accurate
statements be made about Star boat hulls. To provide a sound basis for
judging Stars, work was started, in the University of Michigan Naval Tank.
The problem began in 1938, when the Star Class rules allowed a vertical
tolerance of 1 inch and a half beam tolerance of 3/4 inch. An inch, more
or less, in the family dinghy would mean an appreciable change in design
but it was apparently felt that in a 22-foot boat the utility of a one-inch
tolerance was restricted to allowing amateur-built boats to measure in.
I mean no slur on amateur ability; that is just the fact of the matter.
But, when one attracts as expert a group of racing men as the Star Class
has brought together, another inevitable question arises- what will be
the effect of changing the hull to some extreme of the tolerances allowed
by the rules?
Whenever it is possible, tank men agree, large models should be used because,
in comparison with small models, they have a proportionately greater range
of towing speeds. Models towed at a low speed are liable to have a flow
of water around them that is in a state of transition; it is neither laminar
nor turbulent flow. This condition is nearly impossible to take into account
in computing the results of the tests and usually leads to some delightful
but quite wrong conclusions. Large models also admit of accurate construction.
All the models used in these tests were finished to an accuracy of 1/32-inch,
which made it possible to reproduce small changes in the full sized counterpart.
So these tests were not made to prove that the standard Star hull is a
poor one; rather, they were made to help explain that one boat may easily
be potentially faster than another. However, lest this information founder
all but the die-hards, it should be realized that any saving in resistance
amounts to only a few seconds per mile advantage, and that smart racing
tactics and a well kept boat are still the decisive factors in winning
Three distinctly different types of hull were used in the series. The
standard hull served as the norm; we wanted to know if it was possible
to have a hull that measured in by the 1938 rules and towed with less
resistance than the standard hull form. Pimm was also included in the
series because she was a never-ending source of questions. And, further
than that, a casual inspection of her lines marked her as an improvement
over the norm. The third model was a combination of as many good ideas
on Star boat design as could be brought together in 1938. It has some
of the same characteristics as Pimm, but the designs divaricate when the
transverse sections are laid out. A hull has been built to her lines but
has not yet been sailed enough. St. Mike will denote her in this discussion.
The models used for these tests were one-third the size of actual Stars.
As a premise to setting down the conditions under which this series was
run, it is necessary to differentiate between these tests and the work
done by Prof. Davidson. The Stevens tank is unique, at this time, in having
the only means available of measuring the lift and drag of sail boat hulls.
However, we were confident that, if a consistent set of conditions was
maintained throughout the series, coupled with the fact that these were
purely relative tests, there would be some real value in the results without
measuring exactly the hydrodynamic forces on the hull. Furthermore, a
Star is somewhat like a centerboard boat in that the keel is a flat plate
and, although there a bulb on it, the opportunities, for making that an
efficient hydrofoil are limited. After considering this difference, the
conditions set as standard were: a constant displacement for all three
hulls; a constant amount of trim (except in the trimming tests); and a
standard angle of heel of 20° (this angle was adopted because the
hull depth does not permit a Star to sail well past a heel of 20°).
Figures will be useless if it is not realized that these tests were purely
comparative. All we wanted to know was which model towed the more easily
and if there were any critical speeds at the resistance of the models
changed noticeably. To that end, the models were towed in positions that
approximate the positions assumed while sailing.
Each model was towed upright, in the down wind position, and heeled, the
usual position when working to windward. In addition to these tests, the
effect of trimming the boat by the head and by the stern was investigated,
since many skippers have decided opinions on the latter. Another comparison
was the effect of rounding the chines. And here it should be understood
that, when rounded chines are mentioned, the corners at the intersection
of the sides and bottom with the transom were rounded to the same radius
as the chines proper. The set-up of the comparisons has been given above.
Figures 1 to 4 are diagrammatic comparisons of the resistances, and later
in the text the important details of each comparison will be discussed.
But, before going into remarks on the results of the towing, there are
several features of design that should be noted. For instance, Pimm and
St. Mike represent Stars with the ultimate in flat fore and aft contours;
much attention has been paid, in them, to securing a long straight run.
In Pimm, this run starts almost at the after end of the keel; in St. Mike,
an equally long run has been designed but a gradual curve was used to
smooth out the buttocks. By comparison, the standard Star is average in
all respects; the fore and aft section is neither as flat nor as round
as is possible, and the transverse sections are characteristically the
same. Pimm has nearly normal sections throughout but in St. Mike the sections
are flat forward and grow progressively rounder as they go aft (the important
difference in the design of the two boats). Since the tolerance vertically
was plus or minus an inch, it would be possible to find two boats, both
of them measuring in, with two inches difference in fore and aft contour.
For this series, however, that fact was never taken advantage of completely
as it would require that one of the two boats have as much rocker as possible,
within the rule, and it was felt that deeply rockered boats have been
outmoded in the Star Class. Stars come nearly into the realm of planing
vessels and should, therefore, benefit from flat runs.
Aside from learning how near to ideal the standard Star boat comes, we
were interested in trying to produce a faster hull than Pimm's. Figure
1 shows that, for all speeds except those between 4.5 and 5.75 knots,
Pimm and St. Mike are better than the standard Star, in the upright position.
And, happily, above 5.75 knots, St. Mike shows less resistance than Pimm.
This may be considered all advantage since it is in the upright position
(running) that high speeds can be reached, and it is at high speeds that
a fast boat can open a winning lead over another boat.
The reduced resistance of Pimm and St. Mike can most easily be credited
to the long straight runs. This is noticed more pointedly in Figure 2
(model heeled), where the two trial boats hold a distinct advantage. The
divergence in the resistance curves of Pimm and St. Mike quite plausibly
comes front the abnormal sections used in the latter. Pimm had sections
much like the standard Star; St. Mikc's sections had more round as they
moved aft. The advantage of reduced resistance certainly lies in favor
of Pimm, when considering the model results, because she is the easiest
towing boat in the range of speeds that is most likely to be attained
when working to windward. But, as can be noted by comparing Figure 1 with
Figure 2, there is, distinctly less power required to drive an upright
boat and Pimm, in the easiest driving heeled position, tows harder than
either Pimm or St. Mike upright. Which seems to indicate that Stars should
be sailed on all even keel. Although whether or not to let a Star boat
heel is partly discretionary with the skipper, and is certainly dependent
on the force of the wind, as long as the sails can be held full, a Star
should be sailed upright. The increase in resistance shown in Figure 2
is due, for the most part, to increased waves created by the hull. But,
to add to that disadvantage, the sails are less efficient, the boat steers
badly and the general racing efficiency of those aboard is impaired.
The trim tests (Figure
3, model upright) were made the importance of sailing boats of this type
a trim by the head. When the tests were set up, account was taken of the
maximum trim that could be effected by moving the live weights (skipper
and crew) into positions that a possible while sailing the boat, and sailing
her effectively. Under these conditions, Pimm's potential trim range was
4°. Trim forward was accomplished by placing the crew at the forward
end of the cockpit and by keeping the skipper at least 4 feet forward
of the after end of the cockpit. The extreme condition of trim aft had
the skipper and crew both in the after end of the cockpit- a not unusual
place for them Within this range, as the trim forward increased, the resistance
decreased in the order of 12 per cent. The better hull shape of buttocks,
and the consequent better release of water, assumed under the favorable
condition of loading largely accounts for the lower resistance, although
there is a large difference in wetted surface between the extreme conditions
of trim. The value of the increase in initial stability, because of the
increase in the area of the load water line (from a trim aft), is often
propounded by crews who can keep drier by sitting aft, or by skippers
whose backs just fit into the, after end of the cockpit. But the increase
initial stability is only academic and the same boat, driven faster, would
increase her dynamical stability through larger lee bow wave and a change
in the whole wave profile.
The test on rounded chines (Figure 4, model upright) provided the surprise
that should be a part of any experiment. Although it was popularly considered
that rounded chines are desirable, every increase in bilge radius was
accompanied by an increase in resistance, in the range of speeds normally
sailed by a Star. The answer to the phenomena was found by observing the
models as they pass through the tank.
As the bilge radius increased, there was a noticeable smoothing out of
the wave profile. That is, the waves no longer appeared sharply defined
but looked as if they had been made in a liquid with the viscosity of
heavy oil. A sharp wave profile is formed by the sudden release of the
water as it passes from under the chine. The sudden removal of the restraining
surface, above each portion of the water, results in an immediate acceleration
of the particles in an upward direction, giving rise only to the pure
up and down vibration that is seen as the well-defined wave profile of
a sharp-chined boat. With a round chine, on the contrary, while the acceleration
at corresponding depths is the same, the direction of the acceleration
is constantly changing as the water rises to the surface. And there is
a tendency for this rotating direction of acceleration to create horizontal
vortices whose axes of rotation are parallel to the chine. The effect
of this rotating mass of water is to demand more energy from the hull.
The direct result, of course, is an increase in resistance. So, while
there is an apparent decrease in the waves formed by the hull, the unnoticed
whorling is actually increasing the wave making resistance of the boat.
ROUND chines add another complication to getting speed from a hull. This
bad feature is an extension of the range in which burble will appear.
In the wake of any transom-sterned boat, a severe agitation of the water
may be noticed, an indication that a breakdown is occurring in the flow
of water under the boat. When water is released from the transom properly,
the burble that means an increase in resistance is not present. However,
it is possible to run without burble only above and below definite speeds,
the range depending on the lines of the boat. But, also, the shape of
the chine has some effect. With a square-chined Star, it is safe to say
that the burble will appear at about 5 knots and run clear aft at about
6 1/2 knots (when the boat is upright and on the designed trim). Increasing
the bilge radius to 3 inches, on a Star, increases the burble range by
dropping the lower limit to 4 1/2 knots and by raising the upper limit
to 7 1/2 knots. Also, a model with round chines, when running heeled,
begins to climb out of the water forward as the speed increases, the round
chine acting as a flat surface on which the model tries to plane. To complete
the analysis, it is necessary to go back to sharp-chined heeled models,
in which the burble range is from 4 1/2 to 6 1/2 knots.
With the sharp-chined models, there was no indication that the bow tried
to rise, a further proof that the sharp entrance angles are imperative.
The easiest way to decrease the burble range is to trim the boat by the
head and, conversely, the handicap of a trim by the stern is that it will
cause the water to break away from the transom at speeds lower than normal.
Under all conditions except that of a rounded chine, the burble range
increases by dropping the lower limit; the upper limit is quite constant,
being 6 1/2 knots. Interestingly, that is approximately the speed at which
some styles of Star hulls show the first signs of planing.
In summary, it can be said that many Star boat owners might have had a
faster boat than the standard Star because:
1. The fore and aft contour of the boat had been flattened and a large
decrease in resistance accrued.
2. The boat was sailed down by the head, at least while running, which
would give a 12 per cent decrease in the power required to drive her.
3. Square chines might have been used, which mean a decrease in resistance.
Although this work was done directly on Stars, the information is also
applicable to other types of square-chined boats. But it must always be
remembered that, even though a skipper have all the information in the
world and his boat has an enviable heritage from designer and builder,
the combination must sail hard to be a winner.