AN ABSTRACT OF THE THESIS OF

DANIEL M. LADD for the degree of MASTER OF SCIENCE

in Mechanical Engineering presented on refiri7r

Title: POWER PERFORMANCE OF PLANING BOATS WITH THE
EFFECT OF PROPELLER SELECTION AND PROPELLER
GUARD DESIGN

Abstract approved: r
Redacted for Privacy
J. XillyKinney

One of the unique problems associated with salmon gillnet fish-

ing is of preventing the fouling of the boat's propeller in the net and
associated fishing gear. Presently, the solution to this problem has
been the addition of a basket type propeller guard which may produce
a drag increase of 50 percent on small planing boats.
Recognizing the need for a greater understanding of the inter-
relationship of boat drag and engine power to the effect of propeller
selection and propeller guard design, this study analytically models
planing boat power performance including the experimentally deter-
mined effects of nozzle type propeller guard design. The model
predicts the drag forces acting on a prismatic planing hull and existing
data on marine propeller performance is incorporated in the model
such that the propeller shaft speed and shaft power for a particular
boat propeller combination can be predicted. The model will also
optimize propeller diameter.
 Three variations of nozzle type propeller guards were tested
on a 16 ft. planing boat. The guards were a commercially produced
outboard motor propeller weed guard, a short (1/D = 0.275) flow
accelerating nozzle and the nozzle equipped with prerotation vanes.
Computer modeling the effect of the three nozzle guards on the per-
formance of a hypothetical 11500 lb. planing boat with a 265 Hp
engine predicted that the top speed of the boat (with an optimum
diameter propeller) would be 34.6 knots, with the weed guard 25.5
knots (a speed loss of 26 percent), with the flow accelerating nozzle
32.3 knots (seven percent loss) and the nozzle with vanes 27.4 knots
(21 percent loss). The validity of the model is demonstrated by a
comparison of the predicted values of propeller shaft speed and boat
top speed, with an open propeller and the weed guard, to the actual
values observed in operation of the 16 ft. test boat. Shaft speed
predictions were within five percent and boat top speed predictions
were within about ten percent of actual observations.
Power Performance of Planing Boats with the
Effect of Propeller Selection and
Propeller Guard Design
by

Daniel M. Ladd

A THESIS

submitted to
Oregon State University

in partial fulfillment of
the requirements for the
degree of
Master of Science
June 1976
APPROVED:

Redacted for Privacy

ofes4or of hanical Engineering
in charge of major

Redacted for Privacy

Head of Depart ent of Mechanical Elgineering

Redacted for Privacy

Dean of Graduate School/

Date thesis is presented Am sr it7) 400.'

Typed by Susie Kozlik for Daniel M. Ladd

 TABLE OF CONTENTS

Chapter Page
I INTRODUCTION 1

Gillnetter's and Propeller Guards 1

Purpose of the Study 5

II PLANING BOAT HYDRODYNAMICS 8

Planing Boats 8
Planing Surface Hydrodynamics 10

III PROPELLER PERFORMANCE 22

Momentum Theory 22
Propeller Performance Curves 29
Interaction of Propeller and Hull 35
Use of the K-J Performance Curves 37

IV PROPELLER GUARD PERFORMANCE TESTING 48

Methodology 48
Test Boat 51
Torque Measurement 52
Shaft Speed Measurement 57
Boat Velocity 57
Thrust Measurement 59
Run Procedure 61

V NOZZLE GUARD TEST RESULTS 64

Test Articles 64
Results 69
Application of Results 74

VI CLOSURE 84

Validity of the Model 84

Summary and Comments 91

BIBLIOGRAPHY 95

NOMENCLATURE 96

APPENDICES 101
 LIST OF ILLUSTRATIONS

Figure Page
1.1 Typical gillnet fishing boat 2

1.2 Wire rope basket guard 2

2.1 Simplified diagram of a hard chine planing boat

hull 9

2.2 Forces on a planing hull 11

2.3 Resistance of Boat A 16

2.4 Resistance of Boat B 17

2.5 Trim angle of Boat B 18

3.1 Fluid flow through an actuator disk propeller 23

3.2 Kort nozzle 27

3.3 Thrust coefficient for a typical three bladed

propeller series 32

3.4 Torque coefficient for a typical three bladed

propeller series 33

3.5 Performance map of Boat C 42

3.6 Thrustline drag of Boat C 43

4.1 Test boat 51

4.2 Outboard motor static test set up 54

4.3 Outboard motor performance map 56

4.4 Test instrumentation 58

4.5 Thrust carriage and outboard 60

5.1 Weed guard and nozzle Al 65

Figure Page
5. 2 Installed weed guard 65

5. 3 Profile of nozzle Al 66

5. 4 Nozzle Al with vanes, three quarter view 68

5. 5 Nozzle Al with vanes, front view 68

5. 6 Thrust coefficient versus advance ratio, PR20

propeller 71

5. 7 Thrust coefficient versus advance ratio, PR 24

propeller 72

6. 1 Test boat thrustline drag versus velocity 85

6. 2 Test boat trim angle versus velocity 87

6. 3 Propeller shaft speed versus test boat speed,

open propeller 89

6. 4 Propeller shaft speed versus test boat speed,

weed guard 89
 LIST OF TABLES

Table Page
2.1 Definition of forces and moments on a planing hull 12

2.2 Specifications of Boat A and B 19

3.1 Ideal shrouded propeller efficiency and the

load coefficient 26

5.1 Performance specifications for propeller nozzles

tested 79

5.2 Optimum performance of Boat C, with various

propeller guards 81

6.1 Comparison of experimental and predicted

maximum test boat speeds 90
 POWER PERFORMANCE OF PLANING BOATS WITH THE
EFFECT OF PROPELLER SELECTION AND
PROPELLER GUARD DESIGN

I. INTRODUCTION

Gillnetters and Propeller Guards

One of the unique problems associated with salmon gillnet

fishing, is that of preventing the fouling of the boat's propeller in the
net and associated fishing gear. A photograph of a typical modern
gillnet fishing boat (hereafter referred to as gillnetter) is shown in
Figure 1.1. Most of the modern gillnetters are relatively small,
medium speed planing boats, 25 to 40 ft. in length, displacing about
10,000 lbs. Propeller protection is almost essential, as the typical
fishing mishap occurs during the short, but busy gillnetting season,
when a boat will accidently run over the surfaced net of another
fisherman, often at high speed or in darkness. Experienced fisher-
men seldom let their boats get entangled in their own nets. Not only
does the fouling of a gillnetter's propeller lead to considerable ex-
pense due to lost fishing time, damage to an expensive net, and
diver's fees for propeller untangling, but this can lead to a rather
dangerous situation as during this time the boat is without maneuver-
ing power.
Figure 1.2 is a photograph of one of the more recent propeller
guard designs. This particular design uses a wire rope "basket" to
 2

Figure 1.1. Typical gillnet fishing boat (drum is for reeling in net).

Figure 1. 2. Wire rope basket guard.

 3

protect the propeller. The more typical basket is constructed of

welded solid steel rod of about 3/8 in. diameter. As one might
suspect, the solid rod basket often suffers from excessive flow in-
duced vibration which can cause premature failure of the welds. The
advantage of the wire rope basket is the flexibility of the wire rope
which tends to dampen out propeller and flow induced vibrations.
This basket has the disadvantage that the breakage of one of the ropes
will cause catastrophic failure when the loose end of the rope and the
propeller become entangled.
Most gillnetters are powered by the conventional inboard engine,
inclined propeller shaft propulsion system. Also, in use is the
inboard-outboard drive combination (also called I/O drive, Z-drive
and stern drive) in which the engine is located inboard, but the pro-
peller drive and steering mechanism are located outboard and is
almost identical in appearance with the lower unit of a standard out-
board motor. Currently no satisfactory propeller guard has been
designed for I/O drive use. Also in limited use is the jet drive, which
has no propeller at all. The boat is propelled by pumping a stream
of water through a nozzle located at the stern of the boat. At first
glance this would appear to be an ideal solution to the propeller foul-
ing problem, however the propulsive efficiency of a jet drive, for
reasons to be discussed in Chapter III, is substantially less than that
of an equivalent open propeller. Also, installation of a jet drive on
 4

an existing conventional boat is a major undertaking. This has limited

the application of the jet drive to uses where propeller damage would
be almost inevitable.
The major drawback to the basket type propeller guard is the
tremendous increase in boat drag, caused by the maze of rods and
supports beneath the hull. A typical 10,000 lb. planing boat will have
a bare hull resistance of about 1200 to 1500 lbs. at 20 knots (Ref. 8).
A rough estimate of the drag of a basket guard for this size boat,
indicates that the drag of the basket at this speed would be some-
where around 700 lbs. , an increase of drag of about 50 percent.
Translated to the effect on the average gillnetter, the addition of a
basket guard will decrease top speed by about one -third, while re-
quiring the same installed horsepower and fuel consumption. To the
fisherman, whose income depends on the sea area he can cover during
the gillnetting season, the performance loss due to a high drag pro-
peller guard configuration can represent a serious loss of possible
income. Also, the propulsive inefficiency associated with the use of

a basket guard, when multiplied by the hundreds of gillnetters along

the Oregon, Washington and Alaskan coastlines, represents a wastage
of thousands of gallons of fuel resources each year. However, the
alternative of not using a guard, represents such a risk that a majority
of the gillnetters have some type of propeller protection.
 5

Purpose of the Study

The lack of an efficient yet net proof propulsion system, is not

due to a lack of effort or inventiveness on the part of the fishermen,
but the lack of an informed coordinated effort. The creation of an
overall efficient boat, propeller and propeller guard system, re-
quires a knowledge of the interrelationship of each component on the
performance of the system as a whole. This study is an attempt to
reduce the hitherto "black magic art" of hull design, propeller selec-
tion and propeller guard design to quantified variables, whose effect
on boat performance can be systematically studied. Previously, any
performance improvements have come about as a result of a trial
and error procedure, which is an expensive and time consuming
process.
It is very possible that a certain propeller guard design may
prove to be entirely acceptable for operation on one type of boat,
while entirely unacceptable for operation on another because of the
way the hull, engine, propeller and guard performance are inter-
related. Therefore, it is desirable to develop a method where any
number of possible variations of planing boat parameters can be
cheaply and easily varied to study the effects on performance. The
work described here combines published models for predicting planing
hull hydrodynamic drag and propeller power/thrust into a single
 6

computer model which can predict boat power performance with a

specified propeller(s) or can optimize the propeller diameter and
pitch to achieve highest efficiency. Also described is the measure-
ment and comparison of the drag and power performance of three
variations of nozzle type propeller guards tested on a small boat.
The results are expressed in nondimensional form and included in the
computer model. An example compares predicted maximum speeds
and efficiencies for a hypothetical gillnetter at constant engine power
and propeller shaft speed equipped with various propeller guards.
This computer model will be a valuable tool to those who require a
specific knowledge of the performance characteristics of planing
boats and to those who wish to optimize a boat and propulsion system
for a specific function.
The development of any effective model requires a large amount
of either analytical or experimental study to establish the functional
relationship between variables. This study will rely on existing in-
formation on planing boat hydrodynamics and propeller performance
characteristics as presented in Chapter II and III. Chapter IV will
outline the experimental procedure devised to collect data on pro-
peller guard performance and Chapter V will present some prelimin-
ary results and will show the effect on performance when the hull,
propeller and propeller guard are modeled as a system. The model
 7

developed by this study should lead to a better understanding of what

can be done to improve propeller guard and overall boat performance.
 8

II. PLANING BOAT HYDRODYNAMICS

Planing Boats

For the purpose of this study, the definition of a planing boat

can be considered as any boat which was designed to be supported in
a major part by the dynamic reaction between the water and the bot-
tom of the boat. This is in contrast with the typical displacement type
craft, which is supported entirely by the buoyant lift of the water
displaced by the hull.
Figure 2.1 is a diagram of a "hard" chine planing hull. The
hard, or sharp intersection, of hull bottom and sides produces an
efficient planing surface (high lift to drag ratio). The "soft" or
rounded chine produces a more seaworthy craft in rough seas and at
low speeds, but is somewhat inefficient at higher speeds. A majority
of planing boat designs (including those gillnetters that may be classi-
fied as planing boats) have a hard chine.
The hull of Figure 2.1 has easily definable dimensions, such as
beam, deadrise angle and propeller shaft inclination, which are im-
portant parameters when discussing boat performance. Most hull
forms are more complex then this simplified diagram, however it will
be shown that for some, and probably most hull forms, a prismatic
representative figure may be arrived upon which will to a first order
approximation represent the performance characteristics of that hull.
 sheer

transom

deadris chine

shaft CL

propeller Lp
disk

Figure 2.1. Simplified diagram of hard chine planing boat hull.

 10

Planing Surface Hydrodynamics

The state of the art limits a rigorous treatment of planing hydro-

dynamics to prismatic planing surfaces assumed to have constant dead-
1
rise, constant beam and a steady running attitude. Murray (5),
presented empirical equations used to describe the lift and drag acting
on a prismatic planing surface. Savitsky (6) summarized the work of
Murray and devised a computational procedure to solve for the running
trim and thus determine the forces on a planing surface. Had ler (3)
refined Savitsky's method to include keel, appendage wind, rudder drag
and the effect of induced propeller forces.
Neglecting appendage, wind drag and induced propeller forces,
a free body diagram of a planing boat can be drawn as in Figure 2,2,
where the various dimensions and forces are defined in Table 2.1.
It is assumed that the frictional drag Df acts only on the hull bottom,
parallel to the keel, midway between the keel and chine lines, the
propeller thrust T acts along the propeller shaft, and the pressure
forces N, act perpendicular to the boat hull. Then for vertical equi-
librium of forces:

'Much of the experimental work on prismatic planing surfaces

was done under the sponsorship of the National Advisory Committee for
Aeronautics (NACA) and related research organizations, as planing
hydrodynamics is directly related to seaplane as well as planing boat
performance.
F.ugure Z.2, Forces on a planing hull.
 12

Table 2.1. Definition of forces and moments on a planing hull.

T propeller thrust
A - gross weight of boat
Df - frictional component of drag
T - running trim angle
LCG distance of center of gravity from transom, measured along
keel
CG center of gravity
inclination of thrust line relative to keel
N - resultant of pressure forces normal to hull
a distance between line of action of Df and CG
f distance between line of action of T and CG
deadrise angle
B beam (width) over chines
Lk wetted keel length
PH horizontal distance from transom to propeller hub
PV vertical distance from keel to propeller hub
VCG distance from CG to keel
M moment of forces about CG
R horizontal resistance
planing speed
c - distance between line of action of N and CG
 13

A = N cos T + T sin(T +E ) - Df sin T (2. 1)

For horizontal equilibrium:

T cos (T ) = Df cos T + N sin T (2.2)

For equilibrium of pitching moments M, about the center of gravity:

M = Nc + D a - Tf =0 (2. 3)

For a given boat design and loading the displacement A, center

of gravity, beam B, shaft angle E, deadrise angle 3 and the distances
LCG, a and f are specified. The unknowns (N, c, Df and T) in the
above equations are evaluated by a simultaneous solution of these
equations together with the planing formulas for lift, drag and center
of pressure as described in Savitsky. Summing forces parallel to the
keel results in:
T cos = A sin T +Df (2.4)

Rearranging equations (2.1), (2.2) and (2.4) and substituting into

equation (2. 3) results in:

M=A
[c cos (T + E) f sin T
COS E COS E

f
+ Df c tan E-
[a COS E

Except for the correction of an inadvertent omission (-Df tan E), this
is the same as Savitsky's equation (35). In the above equation, the
only unknowns are frictional drag Df, center of pressure moment arm
 14

c, and the running trim T. In general the computational procedure

can be described as follows:
1) A trim angle 7 is assumed
2) The frictional drag Df and moment arm c are calculated using
the drag and center of pressure relationships described in
Savitsky's and Hadler's papers.
3) The moment M is calculated (equation 2. 5)
4) A new trim angle is assumed and steps 2) and 3) are repeated
until the moment approaches zero, using successively better
approximations of the trim angle. Any of several root finding
routines may be used to calculate the trim angle approxima-
tions.
Once the equilibrium trim is found, the horizontal resistance
can be calculated from equation 2.2 and the calculated normal force,
or:
cos ('T + c)
R = T cos (T + e) A sin T + Df (2. 6)
Coss
from which the effective power is:
ehp = RV (2.7)
where V is the speed of the boat.
Due to the complexity of the relationships in step 2) of the above
procedure and the iterative process to find the running trim, this
procedure has been programmed in the FORTRAN extended language
for use on the CDC CYBER 70, Model 73 high speed digital computer
 15

at Oregon State University. The first portion of program HULLS,

listed in Appendix C, is the FORTRAN equivalent of the procedure
described above.
To check the validity of this procedure and program HULLS,
on predicting resistances of real boats, the results of several com-
puter predictions were checked against the published results of scale
model tests for two planing hull forms. Figure 2.3 shows the re-
sistance of Boat A, scaled up (refer to Reference (8) for scaling
method) to a boat displacement of 10,000 lbs., as a function of boat
velocity and compares this with two predictions using program HULLS.
Figure 2.4 and Figure 2.5 compare the predicted and model resist-
ances and trim angles for Boat B, scaled to 10,000 lb. boat displace-
ment. The trim angle is included here only to make a more complete
comparison between model and actual boat behavior. Model test data
for Boat A were taken from Clement and Blount (1), Figure 15b (LCG
four percent aft of centroid of Apl). Test data for Boat B were taken
from data sheet No. 1 of Reference (8), Form characteristics of Boat
A and B together with the computer input data are tabulated in Table
2.2.
A comparison of the relative magnitudes of the forces and
moments acting on a planing hull will show that the most important
variables concerning boat resistance are longitudinal location of cen-
ter of gravity LCG, deadrise angle p, and the beam or breadth over
1600 Calculated using Maximum Beam

Model Test
Calculated using
1200 - Average Max.
Beam and Beam
at Transom

800

400

10 20 30
Velocity (knots)

Figure 2.3. Resistance of Boat A.

1600 Calculated using Maximum Bum

Model Test Calculated using

1200
.. Average Maximum
Beam and Beam
at Transom

800

400

0 10 20 3
Velocity (knots)

Figure 2.4. Resistance of Boat B.

 Model est Calculated using
4 Average of
Maximum Beam and
Beam at Transom

Calculated using
Maximum Beam
2

0 10 20 30
Velocity (knots)
Figure 2.5. Trim angle of Boat B.
 19

Table 2.2. Specifications of Boat A and B (scaled to 10,000 lb.

displacement).
Boat A Boat B
A, Displacement (lbs) 10,000 10,000
Lp, Chine length (ft. ) 27.6 31.86
Bpa, Mean Beam (ft. ) 7.48 6.37
Bpx, Max Beam (ft. ) 9.00 7.78
Bpt, Transom Beam (ft. ) 6.38 4.97
Apl, Planing Area (ft. 2) 206.5 202.7
Centroid of Apl (% Lp) 48.2 48.8
LCG (% Lp) 44.2 42.8
pT Deadrise Angle at Transom 12.5 12.5
'
[30.6, Deadrise Angle at 0.6 Lp 14.0 14.0
Program HULLS input data
A, Displacement (lbs. ) 10,000 10,000
13, Deadrise Angle 12.50 12.50
Bpx, Max Beam 9.0 7.78
(Bpt + Bpx) /2 7.69 6.37
LCG, long CG (ft. ) 12.2 13.63
VCG, vert. CG (ft. ) 1.38 1.5
E , shaft angle (degrees) 13. 10.0
PH, horizontal position of propeller
hub (ft. ) 0. 0.0
PV, vertical position of propeller
hub (ft. ) 0.8 0.7
ACV roughness allowance 0. 0.
2)
CDA, Air Drag parasite area (ft. O. 0.
 20

the chines B. When using a computer prediction technique, it is

important to use a properly representative value of both deadrise
angle and beam, as in most planing hull designs these do not have a
unique value. As can be seen from Figure 2.3 and Figure 2.4, using
the maximum beam Bpx, tends to somewhat over predict the high
speed resistance (up to ten percent for Boat A), while using the aver-
age of the maximum beam and the beam at the transom Bpt, tends to
underpredict the resistance (about eight percent for Boat B). For use
in program HULLS it is recommended that a weighted average be
used such that
B = 0.75 Bpx + 0.25 Bpt (2. 8)

as this will bring the computed high speed resistance of both Boat A
and Boat B to within five percent of the model test resistance. It is
recommended that the deadrise angle at a position 60 percent of the
projected chine length Lp (Figure 2. 1) forward of the transom be used
as the representative value, as this appeared to give the best results
when compared with the model tests of Reference (8).
The computer model predicts a lower resistance at lower speeds
(below 17 knots in Figure 2. 3) because side wetting and excessive
wave drag are not included in the model. Both effects commonly
occur on planing boats at lower speeds. At these speeds, planing
boats behave more like displacement craft and the previously dis-
cussed planing equations do not adequately model the situation. If
 21

one's main interest is in these lower boat speeds, as would be the

case for a gillnetter also used as a troller or a rather low powered
highly loaded boat, a correction should be made to account for the
low predicted resistance of program HULLS. Unfortunately, little
research has been done on planing boats at low speeds, as planing
boats by their very nature are designed for operation at planing
speeds.
The current version of program HULLS has no provision for
including appendage and rudder drag nor the effect of propeller in-
cluded forces, however work is continuing to include these effects in
future versions of program HULLS. Also an effort is underway to
model the drag characteristics of the basket type propeller guard to
gain a better understanding of the actual effects such a device has on
boat performance.
The important point of this chapter is that the complex drag
characteristics of planing boats may be adequately modeled with a
rather simple prismatic model. In modeling boat power performance
it is necessary to know the drag of the boat so that the power needed
for propulsion, by a conventional marine propeller (Chapter III), can
be predicted.
 22

III. PROPELLER PERFORMANCE

Momentum Theory

To aid in the understanding of propeller performance some

simple momentum principles will be reviewed here. This theory
assumes that the fluid flow is uniform, inviscid and irrotational and
that the propeller can be represented by an ideal actuator disk.
Figure 3.1 is a diagram of the flow into an actuator disk propeller,
where Va is the flow velocity around the propeller, u is the final
velocity added to the fluid by the action of the propeller, T is the
propeller thrust and Ap and As are the propeller and propeller
slipstream cross sectional areas normal to the flow. Then from
momentum:
T=m (3. 1)

where m is the mass flow rate of fluid through the propeller or:
p As (Va u) (3. 2)

where p is the fluid density. The rate that kinetic energy is lost in
the slipstream is (reference frame is with respect to the propeller):
.
K. E. = 1/2 m u2 (3. 3)

The efficiency of the propeller is then the useful work done by the
propeller divided by the power input, where the power input will be
the useful work plus the loss, or:
 Va

T Va + u
Ap
>. As

slipstream
Propeller disk

Figure 3.1. Fluid flow through an actuator disk propeller.

 24

TVa mu Va 2
e 2 2 + u/Va (3.4)
TVa + 1/2 mu m uVa + 1/2 m. u2

Since:
T (3. 5)
u =
m p As (Va + u)

solving for u results in:

1 2
1/2 (va + 2 4T
u (Va + (3. 6)
p As

Substituting (3.6) into (3.4) results in:

4
e= 4T
(3. 7)
3 +GAsVa2 + 1)1/2

For an open propeller, by use of Bernoulli's equation, it can

be shown that the added velocity at the disk is exactly half the final
added velocity u in the slipstream. Then equation (3. 5) and equation
(3. 4) reduce to:
2
e (3. 8)
1 + (C + 1) 1/2
T

where:
C = (3. 9)
T 1/2 p ApV a4

where CT is called the load coefficient. Since the fluid at the pro-
peller disk is moving at only half its final induced velocity, it can be
seen that the slipstream behind an open propeller must contract. A
measure of this contraction is the ratio u/Va. From equation (3. 8)
and equation (3. 4) it can be shown that:
u/Va = (CT + 1)1 /2 - 1 (3. 9a)
 25

Equation (3. 7) partly explains the relative inefficiency of the jet

drive, mentioned in Chapter I, as for practical reasons the exit area
of the jet must be made much smaller than the equivalent open pro-
peller. Gongwer (2) presents an equation for efficiency which takes
into account the duct losses of a jet system. This equation is
2
e (3. 10)
Va
2+ +K
Va

where K is the so called loss coefficient:

AP (3. 11)
K
1/2 p Va2

and AP is the pressure or head loss due to flow through the jet duct.
Most conceivable jet drive configurations will have a loss coefficient
in the range of 0. 6 to 1.0, which for an optimum value of Va , places
an upper limit on the value of ideal efficiency at around 0.5 to 0. 6.
Open marine propellers can have real efficiencies approaching 0.80.
Equation (3. 7) raises the question that perhaps it would be
possible to alter the slipstream of a propeller in order to increase
the ideal efficiency. Such an increase can be expressed as

2(1 + (C + 1)1/2)
e' e - 1 (3. 12)
e 4T
3
4(p AsVaL.
+11/2

where e' is the ideal efficiency of the propeller with an altered slip-
stream and e is the standard ideal propeller efficiency from equation
(3. 8). Shrouding the propeller with a nozzle (also called a duct) is
 26

one way of altering the propeller slipstream. A "Kort" or flow

accelerating type nozzle will cause an increase in the propeller slip-
stream area, and thus an increase in ideal propeller efficiency.
Figure 3.2 is a cross section of a typical Kort nozzle. Assuming
that the exit area of a nozzle is the same as the propeller disk area,
equation (3. 12) can be written:
2(1 + (CT + 1)1/2)
e' e
-1 (3. 13)
e 3 + (2C + 1)1/2
T

Table 3.1 is a tabulation of some values of (e' - e)/e and the load
coefficient. It can be seen that the efficiency of a highly loaded pro-
peller (large load coefficient) may be increased considerably by the
addition of an accelerating nozzle. Generally, tugboats and similar
vessels, on which Kort nozzles are often used, operate at a value of
load coefficient approaching ten and some of the large super tankers
may have load coefficients of around four. Most planing boats would
operate with load coefficients less than one.

Table 3.1. Ideal shrouded propeller efficiency and the load coefficient.
CT (e' - e)/e
0.1 0.0006
0.5 0,008
1.0 0.020
2.0 0.044
4.0 0.078
10.0 0.138
 27

Direction
of
Fluid
Flow

Figure 3.2. Kort nozzle.

 28

Equation (3.13) neglects the loss of efficiency due to the drag of

the duct. Van Manen (4) presents an equation for an efficiency factor
due to nozzle drag:
T -DN 41 CDN
eDN =1 _ (3. 14)
T D CT

where DN is the nozzle drag, 1 is the length (chord) of the duct paral-
lel to the propeller shaft axis, D is the propeller diameter and CDN
2
is the sectional drag coefficient of the nozzle. From equation (3. 14),
it can be seen that the nozzle with the largest aspect ratio (D/1), will
have the lowest frictional drag. The minimum length of a nozzle
is only limited by the ability of the nozzle to achieve the desired
effect of suppressing the contraction of the slipstream without flow
separation from the nozzle surface. Generally the larger the design
load coefficient, the longer the nozzle needs to be.
Another type of nozzle, less common than the Kort nozzle, is
the flow decelerating nozzle. This nozzle slows down the flow into
the propeller disk which has the advantage that it increases the local
pressure near the propeller blades which will tend to delay the onset

2
An equivalent way of analyzing nozzle effect, other than
momentum analysis, is to consider the nozzle as an annular airfoil,
with the slipstream contraction (a function of the load coefficient)
providing an effective angle of attack such that a component of the
sectional lift will be in the direction of the thrust line. In this case
CDN is analogous to the drag coefficient of a wing at a certain angle
of attack.
 29
3
of blade cavitation. This type of nozzle will produce a net decrease
of thrust. The decelerating nozzle has found use on very fast
vessels, where cavitation is a major problem and on warships where
the shrouding of the propeller with its coincident decrease of cavita-
tion produce a much quieter running propeller. For a much more
detailed discussion of propeller nozzle theory, performance, and
design, see the additional references listed in Appendix A.
The interest in the propeller nozzle for planing and other small
craft, is not the possible improvement in propeller efficiency, but
the propeller protection provided by the nozzle. Although it is un-
likely that the addition of a nozzle will increase the speed of a small
lightly loaded planing boat (with a small propeller load coefficient),
the possibility that a nozzle will act as an effective propeller guard
while doing little damage to the efficiency of boat operation, should
certainly be investigated.

Propeller Performance Curves

To analyze the effects of propeller selection, one must know

something of the various performance characteristics of propellers.

3
Cavitation is the phenomenon where the fluid pressure near a
propeller blade is so reduced, that the fluid forms vapor bubbles
(boils), which can seriously effect propeller performance and en-
durance.
 30

The most common method of quantifying propeller performance char-

acteristics is to perform tests with a systematic series of model
propellers. The results of these tests can be reduced to a set of non-
dimensional variables from which full size propeller performance
can be predicted.
There are a great number of possible variables that can be used
to describe propeller performance, however those that will be used
here are:
Thrust Coefficient
T
KT = (3. 15)
p n2D4

Torque Coefficient

KQ = (3. 16)
p n2D5

Advance Ratio
Va
(3. 17)
nD

where T is the propeller thrust, Q is the propeller torque, p is the

fluid density, n is the number of propeller revolutions per unit time,
D is the propeller diameter, and Va is the fluid velocity around the
propeller. These coefficients are nondimensional and any consistent
set of units may be used. From the definition of efficiency, it may
easily be shown that:
 31

KT J
e
P 2 KQ (3.18)
Tr

where e is the actual propeller efficiency.

Figure 3.3 is a typical open water performance diagram for a
systematic series of propellers. "Open water" refers to the fact that
the propellers were tested free from the interference of hull and
appendages and "systematic" means that each model propeller tested
is geometrically similar to all the others with the only difference
being the "pitch" of the propeller.
In Figure 3.3, the thrust coefficient and efficiency are plotted
against the independent variables advance ratio J and pitch ratio P/D
where P is the pitch of the propeller. A working definition of pitch
is the axial distance the propeller would advance, in one revolution,
in an unyielding medium. This definition is somewhat arbitrary as
many screw propellers have a pitch distribution which varies with the
blade radius. Figure 3.4 is a diagram of the corresponding torque
coefficient for Figure 3.3, which may be calculated from equation
(3. 18).

The effect of fluid viscosity on propeller performance can be

described using the Reynolds number:
Ux
Re = (3.19)
 0. 40

a)
U
0. 30

Ci)

0. 20

0. 10

J Advance Ratio
Figure 3. 3. Thrust coefficient for a typical three bladed propeller series.
 0.06
a)

4-1
4-I
0
a) 0.04 1.0
0'
;-4

a 0.8

0.2 0.4 0.6 0.8 1.0 1.2 1.4

J Advance Ratio

Figure 3.4. Torque coefficient for a typical three bladed propeller series.
 34

where v is the kinematic viscosity of the fluid. The U and the x are
a significant velocity and distance, and for propellers, vary as to
which effect(s) are to be studied. The two Reynolds numbers used
in this work are the Reynolds number based on propeller diameter
and shaft speed:
nDC
Re D = (3. 19a)
v

where n and D have the previously defined meanings and C is the

mean chord of a propeller blade, and the Reynolds number based on
airfoil chord:
Ref= Va 1 (3. 19b)

where 1 is the chord of a duct or vane. Fortunately, above a certain

critical Reynolds number (based on propeller diameter and shaft
speed) the effect of this parameter is small and can be effectively
ignored as is indicated by the successful practice of scaling up eight
inch diameter model propeller tests to 20 foot full size, with minimal
accounting for the effect of the vastly different Reynolds number
(Schoenherr (7)).
The effect of blade cavitation on performance can be described
using the cavitation number,
p pv
CT (3. 20)
0 1/2 p Vat
where p is the absolute static pressure at the propeller hub and pv
is the vapor pressure of the fluid. Any type of blade cavitation will
 35

adversely affect propeller efficiency. Generally cavitation becomes

a major effect in propeller performance with a combination of small
advance ratios and small cavitation numbers. Planing boats normally
operate at large advance ratios but small cavitation numbers, how-
ever cavitation can be a serious effect on propeller performance of

high speed planing boats. Of course, the other undesirable feature

of cavitation is the blade erosion which may effectively destroy a
propeller.

Interaction of Propeller and Hull

The K-J performance diagram of Figure 3.3 is called an open

water test diagram because it represents the performance of a pro-
peller in a uniform velocity field. When a propeller is used to propel
a ship or boat, certain corrections must be made to account for the
variations in flow field around the hull, near the propeller.
When a boat or ship travels through the water a certain portion
of the vessel's velocity is imparted to the water behind the vessel.
This motion is known as the wake. Although the mechanism of wake
production is extremely complex, for most purposes the average
velocity of the wake can be fairly represented as a fraction of the
velocity of the vessel, thus
V Va (3.21)
w
V
 36

where w is the Taylor wake fraction and Va and V are the velocity of
the water near the propeller (relative to the ship) and the velocity of
the ship respectively. It can be seen that the advance ratio of equa-
tion (3. 17) can be expressed as
Va V(1 w)
(3. 22)
J nD nD

A quantity somewhat related to the wake is the thrust deduction.

This factor takes into account the apparent increase in hull resistance
due to the induced pressure field of the propeller. The fluid pres-
sure ahead of the propeller is reduced, which superimposed on the
pressure field of the hull, acts on the hull in such a way that the
required propeller thrust is greater than the "thrustline drag" of the
hull. 4 This relation can be expressed as
T
R (3.23)
T
t)

where TR is the thrustline drag and t is the thrust deduction fraction.

From equation (3. 15) it follows that
TR
KT = - 2
(3. 24)
p n2D4 p n D4(1 - t)
Schoenherr (7) indicates that a hull with a large wake fraction
will also have a large thrust deduction. The exact relationship

4
Distinction must be made between boat resistance, thrustline
drag and propeller thrust. Resistance is the horizontal boat drag,
thrustline drag is the projection of the horizontal resistance onto the
propeller shaft line and propeller thrust is the actual thrust developed
by the propeller.
 37

between these two coefficients is impossible to tell, however

Schoenherr indicates that for small fast ships it is sufficiently accur-
ate to set the thrust deduction fraction equal to the wake fraction.
Merchant ships can have rather large wakes, with wake fractions on
the order of 0.30, which if ignored in the performance analysis can
lead to considerable error. Fortunately, Had ler (3) indicates that
the wake fraction and thrust deduction for planing boats is small,
varying somewhere between 0.0 to 0.10 for both. It is suggested
here that the wake and thrust deduction be considered equal and that
the larger value of 0.1 be used only for full bodied boats or those
with a large shaft angle. The best values to be used can only be
determined from experience.

Use of the K-J Performance Curves

The use of the K-J performance curves as in Figure 3.3 and

Figure 3.4 is well documented (refer to Scheonherr (7) or various
references in Appendix A). The most common ship design problems
are as follows:
1. Given the open water characteristics of the propeller series,
the design speed and resistance of the ship, and the diameter of the
propeller, find the propeller pitch and shaft speed for best efficiency.
2. Given the open water characteristics of the propeller series,
design speed of ship, and engine speed find the propeller pitch and
 38

diameter for best efficiency.

These two types of problems may be easily solved by defining two new
coefficients,
KT TR
KTd = (3.25)
2 2
pVa D2 p V2 (1 - w)2 (1 - t) D2

and
KT T n2 TR n2
KTn = (3. 26)
4 4
J p Vat p V4 (1 - w) (1 - t)
For the first problem, KTd is a known quantity and the parabola
KT = KTd J2 drawn on a KT - J diagram represents the locus of

possible solutions. It is then an easy matter to pick the solution of

J and P/D with the highest efficiency, and thus from the definition of
advance ratio:
Va (3. 27)
n=
DJ

The second problem is similarly solved using the coefficient K Tn.

Although the above two problems are common enough in ship
design, different problems arise when considering planing boat
performance. On an ordinary planing boat, the hull and powerplant
are rarely, if ever designed together. Indeed it is a rather simple
procedure to entirely exchange engines on most boats. Given a boat
and powerplant combination it has been mostly a trial and error
procedure to find a propeller which is suited to both the performance
characteristics of the boat and the engine.
 39

Large ships are usually powered by either diesel engines or

steam turbines. The diesel powerplant can be characterized as a
constant torque device, limited only by a maximum permissible engine
speed and a steam turbine can be characterized as a constant power
device. Neither device will be exceptionally affected by a poor choice
of propeller which will not allow the powerplant to achieve either
rated power or speed. However most gillnetters are powered by a
spark ignition engine(s), which tend to have very definitely humped
torque and power curves, where a mismatching of propeller and engine
may permit the engine to deliver only a portion of its rated power.
Given an existing boat and engine combination, it would be de-
sirable to be able to predict the power performance characteristics of
a given propeller, that is the propeller horsepower and shaft speed
required to produce a certain boat velocity. These performance data
can then be compared with the engine power characteristics to de-
termine whether the engine is being most effectively utilized. A

propeller-engine combination may be optimized for either maximum

speed or maximum economy.
One method (readily adaptable to a computer program) of solving
the above problem may be described as follows. Given the thrust
required to produce a given velocity (Chapter II) for a specific boat
and the open water characteristics of a propeller of given diameter
and pitch, one can compute from equation (3.24)
 40

R
(3. 28)
4(1
pD - t)KT

where the only unknown on the right hand side of the equation is the
thrust coefficient (KT). Assuming one starts with a preliminary guess
for the thrust coefficient K TN, shaft speed can be calculated from
equation (3.28) and,
V(1 - w)
J nD
(3. 29)

However from Figure 3.3, the thrust coefficient is a function of

advance ratio J and the pitch ratio P/D only, of which the pitch ratio
is known. Then from Figure 3.3
K = f(J) (3. 30)

This new value of thrust coefficient may be substituted into equation

(3.28) from which a new shaft speed may be calculated. For com-
puter application, this successive substitution routine may be slightly
altered to obtain convergence for extreme operating conditions (see
program HULLS, Appendix cy. Once the shaft speed is found, the
advance ratio may be calculated from which the torque coefficient can
be solved for using the efficiency of Figure 3.3 or taken directly from
Figure 3.4. Then
pK n2 D5 (3. 31)

and

SHP = 2 TrnQ (3.32)

 41

where SHP is the power delivered to the propeller. Repeating this

procedure for several boat speeds a complete performance curve of
the propeller may be obtained.
This iterative routine has been included in the second part of
program HULLS described in Chapter II. Propeller performance data
were taken from four K-J diagrams presented in Schoenherr (7) for
four different series of propellers. The four propeller series con-
sisted of three and four bladed propellers with either wide or narrow
blades. Input data for program HULLS must specify which propeller
series is being used (see Appendix C). The curves in the K-J dia-
grams were incorporated into the computer program using polynomial
curve fits computed with the aid of the Statistical Interactive Pro-
gramming System (SIPS) of the Oregon State University computer
center. A tabulation of the propeller series used and the polynomial
coefficients is presented in Appendix B.
Figure 3.5 is an example of the above calculations carried out
for Boat C, with computed (using program HULLS) thrust require-
ments as in Figure 3.6. Figure 3.5 was prepared using as input
data five different 1.75 ft. diameter, three bladed, wide blade pro-
pellers (propeller series four of Appendix B) with pitches ranging
from 1.4 ft. to 2.8 ft. (13/D from O. 8 to 1.6). As can be seen from
Figure 3.5, at each boat speed there is a propeller pitch and cor-
responding propeller speed which requires the least power. This
 42

D= 1.75 ft.
Thrust data from
Figure 3.6
w=0
t =0

Propeller Shaft Speed (rev/sec)

Figure 3. 5. Performance map of Boat C.
 A = 11500 lbs.
2000 9.7°
B = 8.0 ft.
LC G == 13.0 ft.
6.96°
PH = 0.0
1500 -,- PV = 0.93
ACf = .0004
Air Resistance = 0

1000

500

0 ,
0 10 20 3

Velocity (knots)
Figure 3. 6. Thrustline drag of Boat C.
 44

point corresponds to the solution to the first type of ship design prob-
lem discussed previously. The pitch required for best efficiency may
or may not remain fairly constant throughout the speed range of the
boat, as this depends on the drag characteristics of the boat.
Superimposed on Figure 3. 5 is the propeller shaft power output
of a hypothetical engine which produces a maximum of 265 Hp at
40 rps shaft speed (4800 rpm engine speed assuming a 2:1 gear re-
duction between engine and propeller). From Figure 3. 5 this engine-
propeller diameter combination is not very well suited to this boat
as a selection of a propeller with a pitch of 1. 4 ft. (P/D = 0.8) allows
the engine to produce maximum rated horsepower, but will only
propel the boat at a speed of 31. 5 knots. Selection of a propeller with
a pitch of 2.1 ft. (P/D = 1.2) will propel the boat at a speed of 32. 5
knots, but will not allow the engine- to produce maximum power. The
solution to this problem is to either change the engine-propeller gear
ratio, such that maximum horsepower is produced at a shaft speed of
28 rps, or change the propeller diameter. Since the latter is usually
much easier to do than the former, the third section of program
HULLS is designed to calculate the propeller diameter and pitch for
best efficiency given a certain shaft horsepower and shaft speed.
The computer routine to solve this problem is described as
follows. The input data needed are thrustline drag as a function of
boat velocity, shaft horsepower, shaft speed and some maximum
 45

allowable diameter which serves as a starting point in finding the

optimum diameter. Referring to the K-J diagrams of Figure 3.3
and 3.4, and given a starting diameter then:

K
Q
- SHP
(3. 33)
Q p n2 D5 27r p n3D

The unknowns are thrust coefficient, advance ratio and pitch ratio.
The next step is to assume a value of the advance ratio (J) then,
JnD
V
- w) (3. 34)

from which
TR
T= f(v)
(1 - t) (1 t) (3.35)

where f(v) is a relationship such as shown in Figure 3. 6. Then,

T
KT (3.15)
p n2D4

From Figure 3.3,

P/D = f(KT, J) (3. 36)

but then from Figure 3.4

[1]
K = f(J, P/D) (3. 37)

This new value of torque coefficient is compared with the known torque
coefficient from which a new approximation for advance ratio may be
calculated using any one of a number of equation root finding routines.
Program HULLS uses the rapidly converging secant method. Once
 46
[.,
1]
the value of the advance ratio is found such that KQ = KQ' then
JnD
V =(1 - w) (3. 34)

Using this procedure, several values of propeller diameter may be

chosen and a curve drawn through the corresponding velocities, such
that a maximum velocity and corresponding diameter may be approxi-
mated. Since input power is constant, this maximum velocity also
corresponds to the highest propeller efficiency. It is interesting to
note that given the additional freedom of allowing the shaft speed to
vary, at constant power, the optimum open propeller diameter will be
the largest that can be made to fit on the boat as this will produce the
smallest load coefficient, CT and thus the highest efficiency (equation
3. 8).

For the boat in the previous example, with an input of 265 Hp

and shaft speed of 28 rps, the optimum diameter becomes 1.74 ft.
with a pitch of 2.46 ft. (P/D = 1.41) and a corresponding boat velocity
of 34.7 knots, exactly as indicated by Figure 3.5. With an input of
265 Hp and a shaft speed of 40 rps as for the hypothetical engine of
Figure 3.5, the optimum diameter becomes 1.46 ft. with a pitch of
1.76 ft. and a boat velocity of 34.2 knots, an increase of about five
percent in boat speed over what is obtainable with a 1.75 ft. propeller
with the same engine and gear ratio. Of course, in the selection of
 47

the actual propeller one is limited to a commercially available dia-

meter and pitch. Generally, a wide range of combinations of pro-
peller diameter and pitch are readily available.
 48

IV. PROPELLER GUARD PERFORMANCE TESTING

Methodology

The previous chapters have discussed how existing data can be

used to build a model of planing boat performance. This chapter
discusses the method devised to measure propeller guard performance
such that it can be included in the planing boat model. Although this
method was used only to test nozzle type propeller guards, it is felt
that this method could have general application to performance testing
of all types of propeller guards. Performance data on propeller
guard designs will be helpful in evaluating the effectiveness of a de-
sign and what steps should be taken to optimize that design.
Intermediate scale testing of guard performance on a small
planing boat was selected as the best method of obtaining the desired
results. Full scale testing on actual fishing boats was eliminated as
too costly an option. The lack of facilities adaptable to small scale
boat performance testing and the inherent error involved in scaling up
test results excluded this possibility also.
When evaluating propeller guard design based on the criterion of
minimum drag (i. e. minimum performance loss), one would desire
a guard with a streamlined shape to eliminate pressure (form) drag
and a minimum of wetted area to minimize frictional drag. This
 49

criterion immediately suggests a propeller nozzle, as a nozzle is

streamlined and has the minimum wetted area that will still com-
pletely surround the propeller. Therefore it was decided that a
preliminary test program should concentrate on some form of nozzle
guard as this type of guard is essentially the ultimate as far as pro-
peller performance is concerned. Additionally the nozzle guard would
eliminate most of the flow induced vibration problems of current
guards. The problems associated with a nozzle type guard are the
relatively complex fabrication techniques required and the open inlet
of the nozzle which may require additional structure for adequate
propeller protection. Hopefully the improved efficiency of a nozzle
guard would more than offset the above disadvantages. However,
should it be decided that a simpler design is in order, in this case,
the opinion of the author is that it is better to modify an efficient de-
sign to make it simpler, rather than to modify a simple design and
attempt to make it more efficient.
Due to the difficulties in quantitatively evaluating the effective-
ness of a propeller guard design in actually preventing propeller
5
fouling, for this study, it was deemed necessary that this question be

5
There are literally hundreds of different ways a net could get
caught in a propeller. A comprehensive experimental study to de-
termine the probabilities of a particular propeller guard preventing
fouling under any number of operating conditions was deemed a too
costly and time consuming undertaking.
 50

left open to the judgement of the designer and the particular prefer-
ence of any users.
The evaluation of nozzle type propeller guard performance was
accomplished by a test procedure somewhat similar to that used in
testing of propellers. In this way guard performance was directly
related to its effect on propeller performance characteristics. The
results obtained were propeller torque, propeller (and guard) net
thrust, velocity of advance and propeller shaft speed. The per-
formance of a propeller and guard combination were compared to
the performance of the standard propeller and some interpretations
made as to the effect of the guard on the overall performance.
The methodology can be described as follows; the test boat with
a standard propeller, was operated through a range of conditions
while recording measurements of thrust, carburetor pressure drop,
boat velocity and shaft speed. Torque was computed from an empiri-
cal relationship between shaft speed and carburetor pressure drop
derived from data taken during a static engine test. These measure-
ments were reduced to apparent propeller performance coefficients,
that is, the thrust coefficient, torque coefficient and advance ratio as
described in Chapter III. These coefficients are not the actual open
water test coefficients of the propeller, as the water flow field under
a hull is not uniform, however for the purpose of comparisons only,
this distinction was felt not to be important. Propeller performance
 51

characteristics calculated from actual operation on the test boat will

be referred to as the apparent performance characteristics. Pro-
ceeding in this way, a propeller and guard combination was tested in
the same manner as a basic propeller, such that a comparison could
be made between performance with and without the propeller guard.

Test Boat

Figure 4.1 is a photograph of the small planing boat used as a

test platform. The boat was 16 ft. in length, with an approximate
empty weight of 300 lbs. The hull was constructed of aluminum and
had a rounded or soft chine planing form with external spray strips.

Figure 4.1. Test boat.

 52

This boat is the product of a commercial manufacturer of small

recreational use vessels. Although this boat is powered by a con-
ventional outboard motor which is hydrodynamically similar to the
I/O propulsion system, it is felt that the results will be generally
applicable to inboard engine inclined shaft propulsion as well, as
the scope of this experiment was to measure only differences in
performance characteristics and not their absolute magnitude. In
the case of a nozzle type guard, application of these results may
require some correction to account for the lift and increased drag of
the nozzle in inclined flow.

Torque Measurement

Due to the extreme difficulty in obtaining a direct measurement

of propeller torque on an outboard motor, particularly while in actual
operation on a boat, a static power test was performed on the test
motor. During the actual propeller tests, engine speed and airflow
measurements were compared with the results of the static test to
determine the torque input to the propeller.
The test motor was a standard water cooled, two cycle two
cylinder spark ignition power plant with a displacement of 22.0 cu.
in. and a nominal power rating of 25 Hp at 5500 rpm engine speed.
Ignition voltage was provided by a quick rise magneto and the fuel
was controlled by a single carburetor equipped with a nonadjustable
 51

high speed jet. Power was transmitted to the propeller through a

12:21 right angle gear box. Cooling water was taken in just under-
neath the antiventilation plate and discharged back into the water
through the exhaust system. A set of four geometrically similar
propellers were obtained for the motor, with a common diameter of
8.75 in. and pitches of 9, 10, 11 and 12 in. In actual data analysis
only the 9 in. and 11 in. pitch propellers were used.
In the static test set up, engine torque was measured as a
function of engine speed and carburetor pressure drop. The pressure
difference between the inlet and venturi portions of the carburetor is
approximately proportional to the square of the engine airflow. Engine
power is approximately proportional to the engine airflow.
Figure 4.2 is a photograph of the static test set up. Power was
absorbed with a water cooled Prony brake dynamometer driven by the
propeller shaft through a driveshaft and flexible coupling. Dynamo-
meter reaction was measured with a force scale. Two electric water
pumps were used, one to assist the engine water pump in circulating
water through the cooling passages and the other for circulating water
by the engine water intake and exhaust outlet. A 55 gallon drum was
used as a cooling water reservoir. Carburetor pressure drop was
measured using a differential pressure gauge and shaft speed was
measured with an electronic counter using a photocell pickup. Each
test condition was established by adjusting the brake friction and
 54

Figure 4.2. Outboard motor static test set up.

 55

throttle opening to obtain the desired shaft speed and carburetor

pressure drop.
A set of 400 data points were collected on the static test set up.
Figure 4.3 is a parametric representation of the least squares fit to
6
the data set.
The formula is
Qo = -283.43 + 6.1966 no + 3570.2/no

-0.048408 n o 2 + 1440. 1A/n o

-6953.2 A/n02 (4. 1)

where Qo is the shaft torque in lb. -ft., A is the carburetor pressure

drop in in. H2O, and no is the shaft speed in revolutions per second.
During actual propeller test runs a correction was applied to account
for air density where
B. P. TA° )1/2
Q = Qo ( TA B. P. 0
(4.2)

B. P. and TA are the barometric pressure and absolute air tempera-

ture during the run and B. P. 0
and TA are the averages of the baro-
metric pressure, and air temperatures for the static test. For most
weather conditions this correction will seldom exceed two percent.
Due to the unavoidable differences in operating conditions be-
tween a static test and normal outboard operation, equation (4. 1)

6
Derived using SIPS system of the OSU computer center. Cor-
relation coefficient R2 0.96 (R2 = 1 is a "perfect" fit).
 32 56

28

24
\!
A=
20 --
4.;
4.4

a)

cr 16 --1-
O
H
0
a
12

8
Maximum
Torque

Maximum
4 Speed

A is carburetor pressure drop in in.

H2O

20 30 40 50 60
no, propeller shaft speed (rev/sec)

Figure 4. 3. Outboard motor performance map.

 57

should be used with caution. The equation should be reliable for

discerning differences in magnitude of engine performance rather
than the absolute magnitude itself. It is thought that the major differ-
ence between normal operation and the static test is the higher ex-
haust back pressure necessarily involved in the exhaust system of
the static test.

Shaft Speed Measurement

During the static test of the outboard an electronic counter was

used to measure engine speed. As this method was not convenient
for the propeller performance tests an alternate method was devised
using the ignition pulses of the engine, modified by the electronic
circuitry of an electric tachometer into a DC voltage that varied
linearly with input frequency. The system was calibrated with a
voltage function generator and proved to be quite linear and insensi-
tive to input waveform. The output voltage was measured with a
digital voltmeter as shown in the photograph of Figure 4. 4.

Boat Velocity

Boat velocity was measured using a paddlewheel type flow

transducer inserted through the bottom of the boat hull. The velocity
was indicated on a dual range meter, such that the scale of the meter
could be expanded from 0 to 10 knots to 0 to 20 knots, depending on
Figure 4.4. Test instrumentation (from top left, carburetor differ-
ential pressure gauges, velocity indicator, thrust
readout, hydraulic thrust transducer, paddlewheel
velocity transducer, digital voltmeter).
 59

the range switch setting. The indicator meter and transducer are
shown in the photograph of Figure 4.4.
The system was calibrated in the test boat by timed runs over a
measured course. Upstream and downstream7 speeds were averaged
to account for the current velocity. The paddlewheel frequency out-
put was assumed to vary linearly with boat velocity and the meter
output was calibrated accordingly.

Thrust Measurement

The entire outboard motor was mounted on a carriage such that

the motor was free to travel a short distance parallel to the propeller
shaft axis, while otherwise securely mounted to the boat transom.
Four low friction (rolling coefficient of friction = 0. 0003) linear
motion ball bushings were used to guide the motion of the carriage.
Figure 4.5 is a photograph of the outboard motor and thrust carriage.
No correction for the friction of the bushing was used as it was felt
the vibration of the motor would compensate for any stickiness in the
bushings. No hysteresis was observed during the runs.
The thrust force was converted to hydraulic pressure with a
diaphragm type force transducer mounted low on the thrust carriage

7
All boat runs were made on the Willamette river in the im-
mediate vicinity of Corvallis, Ore. The measured course was two
city blocks long (771.4 ft. ).
 60

Figure 4. 5. Thrust carriage and outboard.

 61

near the propeller thrust line. The hydraulic pressure was indicated
on a conventional Bourdon tube gauge graduated to read in pounds
force. The gauge was equipped with an adjustable damping valve.
The hydraulic system was factory calibrated and no further calibra-
tion was attempted other than a cursory check of the general accuracy
of the system.
During the actual runs a correction was made to account for the
component of the weight of the outboard motor and frame in the
direction of the thrust line caused by the varying running trim of the
boat (Figure 6.2) or:
AT = 109 sin T (4.3)

where the weight of the outboard and frame was 109 lbs. The trim
angle was measured with an inclinometer. The data of Figure 6. 2
were applied to all the propeller test runs, with the assumption that
the variation of trim angle with velocity always behaved in the same
manner.

Run Procedure

A driver and data taker were required to operate the test boat
during a run. The boat driver steered the boat and maintained con-
stant engine speed, while the data taker recorded values of thrust,
boat speed and engine airflow. The runs were made at constant
engine speed, with a minimum engine speed of 3000 rpm (28. 6 rps
 62

propeller speed). This produced a minimum Reynolds number of

5.0 x 105 where,
nDC (4.4)
Re D =
v

D and n have the previously defined meanings, C is the mean chord of

a blade, 8 and V was taken to be the kinematic viscosity of fresh water
at 600 F. This number is well above the critical Reynolds number
for propeller tests defined in Schoenherr (7) Table 7.
Since the test boat had a rather flat resistance curve (Figure
6. 1) a large variation in apparent advance ratio could be produced.
The upper limit of advance ratio was limited by the power available,
and the lower limit by the minimum engine speed required to keep
the engine speed and airflow within the measured performance en-
velope of Figure 4.3.
During the test runs it was discovered that due to wind gusts,
water current variations and other uncontrollable factors the four
measured variables were subject to random variations and usually
cycled about some averages value. To remove this variation, a set

8
The three bladed propellers tested had a diameter of 8.75 in.
with the blades having a mean chord of about 3.5 in.
91t is apparent from chapter III and the definition of thrust
coefficient and advance ratio, that a vessel with drag varying exactly
as the square of the speed will always operate at the same advance
ratio, no matter what the forward speed. With a more linear drag
variation the faster the boat goes, the larger the advance ratio.
 63

of about 15 data points, at each test condition, were recorded at

approximately five second intervals. These values were averaged
to obtain mean values of thrust, engine speed, boat speed and engine
airflow and these means were used as a single data set. The esti-
mated error for the measurements after averaging are as follows:
Thrust + 1. 5%

Engine Speed + 1. 0%

Boat Velocity + 2. 5%

Airflow (carburetor pressure drop) + 3. 0%

The estimated error for the implied measurements of torque and trim
angle are:
Torque (repeatability between runs) + 7. 5%

Torque (absolute value) + 15.%

Trim angle + 10.%

 64

V. NOZZLE GUARD TEST RESULTS

Test Articles

Figure 5.1 shows the two nozzle guards tested. Figure 5.2 is
a photograph of the "weed" guard as installed on the outboard motor.
This guard is commercially produced for use on small outboard
motors and is primarily intended for shielding the propeller from
underwater obstacles and large underwater growth. Currently the
guard is available only for outboards of 25 Hp and smaller. This
guard was tested to provide a standard of comparison for developments
of future nozzle guards.
The weed guard is not a true nozzle. Its design is such that it
can be easily fabricated as an aluminum sand casting with a rather
poor surface finish. The guard surface is not streamlined and the tip
clearance between propeller and guard is rather large, about 0.4 in.
or 4.5 percent of the propeller diameter. Figure 5. 1 also shows the
guard designated as nozzle Al. This guard was constructed in the
Mechanical Engineering shop at OSU. The ring portion of the guard
was machined from an aluminum casting on an ordinary lathe. The
inside of the nozzle was step machined and contoured with a machinist's
scraper.
The profile of the nozzle is shown in Figure 5.3. This nozzle
profile was taken from van Manen (4) (his nozzle 19a) and is a standard
 65

Figure 5.1. Weed guard (left), Nozzle Al, and Al vane assembly.

Figure 5.2. Installed weed guard.

 R+C+Xu R R+C +Xl

Shaft axis

L
Ordinates of Nozzle (in percent)
Y/L 0 1.25 2.5 5.0 7.5 10 15 20 25 30

X 1 /1_. , 18.25 14.66 12.80 9. 87 8.00 6.34 3.87 2.17 1.10 0.48
Xu /L 18.25 20.72 21,07 20.80 straight line
Y/L 40 50 60 70 80 90 95 100
Xi /I- 0 0 0 0.29 0.82 1.45 1.86 2.36
Xu/L 6.36

Figure 5.3. Profile of nozzle. Al,

 67

flow accelerating type nozzle. It was felt that for a first attempt a
standard nozzle profile should be used even though for the purpose
under study an optimum nozzle profile would probably be "neutral"
rather than flow accelerating. To minimize the frictional drag, as
discussed in Chapter III, the nozzle was made as short as possible,
yet long enough to enclose the propeller. The nozzle length was 2.4
in. (1/D = 0.275). This value of 1/D is much smaller than is com-
monly used in ordinary applications of accelerating nozzles. The
blade tip clearance, for the two propellers tested, averaged to about
0.04 in. or about 0.5 percent of the propeller diameter.
Figure 5.1, Figure 5.4 and Figure 5.5 show nozzle Al and the
prerotation vanes designed for it. These were installed to provide
protection for the inlet of the nozzle while perhaps slightly assisting
propeller efficiency. The theory behind prerotation vanes is that the
vanes will turn the flow into the propeller such that the net flow rota-
tion at the exit of the propeller will be zero. In normal propeller
operation the fluid in the propeller slipstream will have a nonzero
angular velocity. This flow rotation represents a loss of energy. To
obtain a net flow rotation of zero, the torque on the prerotation vanes
must be equal and opposite to the propeller torque. Generally, unless
the rotational losses are substantial, the extra drag of the vanes more
than cancels out the gain due to zero flow rotation.
 68

Figure 5. 4. Nozzle Al with vanes, three quarter view.

Figure 5.5. Nozzle Al with vanes, front view.

 69

A GOttinger 625 profile (similar to a Clark Y) with a thickness-

chord ratio of 0.2 was chosen for the shape of the vanes. The four
vanes each had a chord of 1.25 in. , a length of 3.125 in. and were
spaced at 60 degree intervals. Generally it is better that stationary
element spacing not coincide with propeller blade spacing (as with
this design). Coincident spacing may aggrevate vibration problems,
however this spacing was chosen as a convenience as vibration prob-
lems were not considered important for this preliminary design.
The four vanes did not have enough area to completely counteract the
propeller torque, under most operating conditions, however for this
design the opinion was that a minimum wetted surface area was more
important. In this case, the vanes had the area required (at maximum
lift drag ratio) to counteract the torque of the high pitch (PR24) pro-
peller at an advance ratio of about one. At advance ratios less than
this the propeller torque would not be fully counteracted. It can be
shown that the vane torque will counteract the propeller torque only
2
at a unique value of K Q /J2 (in this case K Q 0.04).

Results

Each guard was tested with two propellers, the 8.75 in. dia-
meter by 9.0 in. pitch propeller (designated as PR20) and the 8.75
in. diameter by 11.0 in. pitch propeller (designated as PR24).
 70

These dimensions correspond to pitch ratios P/D 1.03 and 1.26

respectively.
Figure 5. 6 and Figure 5.7 show the apparent thrust coefficient
10
as a function of apparent advance ratio, for the PR20 and PR24
propellers. The solid lines in Figure 5.6 and Figure 5.7 represent
the least squares curve fit through the standard propeller test data
points. Data from the propeller with guard test were compared to
these base lines.
Both Figure 5. 6 and Figure 5.7 show that the weed guard has a
serious drag problem, as the addition of the weed guard leads to a
serious loss of thrust (a 25 percent smaller thrust coefficient for the
PR20 propeller at advance ratio of 0. 75). Comparison of the torque
coefficients for the open propellers and the propellers with the weed
guard, despite some data scatter, showed the result that the differ-
ence in torque coefficients was negligible. The average of 19 obser-
vations (refer to Table 5. 1) showed the percentage difference between
the open propeller and the propeller with weed guard torque coeffi-
cients to be a nominal two percent (standard deviation 0. 11).
Both Figure 5.6 and 5.7 show that the plain nozzle Al has
excellent drag characteristics as the thrust coefficient for the open

10
In this chapter the word apparent will be dropped, with the
understanding that the coefficient being discussed is the apparent
rather than the actual coefficient.
 0.32

0.28

'41
4.1
O
O
U 0.20
4-1

0
X 0.16 X No Guard 0
O
O Weed Guard O 6
es
0.12 O Nozzle Al w/o vanes 0
O
a. A Nozzle Al vanes as = o
44 0.08
0.04
0
0 0.4 0.5 0.6 0.7 0. 8 0.9 1.0
J Apparent Advance Ratio
Figure 5. 6. Thrust coefficient versus advance ratio PR20 propeller (P/D = 1.03),

-4
 0.32

0.28 0
0

a)
0
0
0
a)
o
0.20
X No Guard 0
0
0.16 0
pq
0 Weed Guard
E1

0.12 0 Nozzle Al w/o vanes

a)

ci 0.08
E-4

0.04

0 0.4 0.5 0. 6 0.7 0.8 0.9 1.0

J Apparent Advance Ratio
Figure 5.7. Thrust coefficient versus advance ratio PR24 propeller (P/D = 1.26).
 73

propeller and the propeller in the nozzle are very nearly the same
over the entire range of advance ratios. Additionally it appears that
the nozzle actually adds thrust at smaller advance ratios, which
should be expected as a small advance ratio corresponds to a large
value of the load coefficient. However the difference in torque co-
efficients is substantial (refer to Table 5. 1). An average of 21
observations showed that the propeller in the nozzle torque coefficient
was about 11.7 percent greater than the open propeller torque co-
efficient, for corresponding advance ratios (standard deviation 0. 076).
However even with this drawback the nozzle is a substantial improve-
ment over the weed guard.
Due to time limitations and poor weather, nozzle Al with the
guide vanes was only tested with the PR20 propeller. Figure 5. 6
shows the effect of the guide vanes on the thrust characteristics of
the nozzle. For the data shown, the vanes were set at an angle of
attack of zero degrees, which approximately corresponds to the maxi-
mum lift drag ratio of the vanes. Data were also taken for an angle
of attack of nine degrees, however the values of thrust coefficient
behaved erratically so that no valid conclusions could be reached. It
is believed that this behavior was due to premature flow separation
(stalling) over the vanes such that the drag of the nozzle system took
drastic changes. The low Reynolds number (from 1.5 to 3.0 x 105)

experienced by such small vanes at the speeds the tests were

 74

conducted, and the coincident slow transition from laminar to turbu-

lent flow probably would contribute to this problem. In this case,
an angle of attack of nine degrees was probably too close to the stall
angle, for the type of flow encountered in the test set up.
Figure 5.6 shows that the vanes add a significant drag to the
nozzle system. Averaging the percentage difference in torque co-
efficients for the six data points indicates that the vane-nozzle system
requires about a 23 percent greater torque than the open propeller
for equivalent advance ratios (standard deviation 0. 106). The com-
bination of these effects lead to considerable performance degradation
of the vane-nozzle system compared to the open propeller. However
the performance of this system is still better than that of the weed
guard (refer to Table 5.1 for comparison of performance specifica-
tions for nozzles tested).

Application of Results

Of course the ultimate goal of these propeller guard tests is the

application to performance prediction of full size boats. To do this
with the preceding data several assumptions are made. First it is
assumed that the percentage change in torque coefficient with the
addition of a propeller guard geometrically similar to any of those
tested is the same for any type of propeller and is independent of
advance ratio such that the propeller and guard torque coefficient can
 75

be expressed as AK
Q
KQg = 1 + KQo KQo (5. 1)

where KQg is the propeller and guard torque coefficient, KQo is the
open propeller torque at the corresponding advance ratio and
AK
Q
/KQo is the percentage change in torque coefficient (assumed to
be a constant). Secondly, some of the extra drag of any nozzle sys-
tern is due to the interference effects caused by the mounting of the
nozzle. It is assumed that the magnitude of the extra drag will be

approximately the same for any type of mounting systems, thus this
effect need not be accounted for. Finally, it is assumed that the
change in thrust coefficient caused by the addition of a propeller
guard will be independent of the type of propeller used and can be
broken up into three separate components. That is,
KTg = K + AK + AK + AK (5. 2)
To Tf TA TD

where KTg is the thrust coefficient for the nozzle and guard system,
K
To is the thrust coefficient of the open propeller, AK Tf is the change
in thrust coefficient due to the frictional drag of the nozzle, AK TA is
the change caused by the flow altering characteristics of the nozzle and
AK
TD
is the change caused by the pressure (form) drag of the nozzle.
AK
Tf
and AK
TD
will always be negative and AK TA should be positive
for an accelerating nozzle and negative for a decelerating nozzle.
 76

In the reduction of data from Figure 5.6 and Figure 5.7 it fol-
lows that
AKTR = AKTA + AKTD = KTg - KTo AKTf (5. 3)

where AKTR is an abbreviation for AKTA + AKTD' KT is the

measured value of thrust coefficient, KTo is the base line performance
of the open propeller from the propeller test and AK Tf is a computed
value of frictional drag, which is calculated as follows. Since,
Ff = 1/2 pVa 2 Cf A w (5. 4)

where Ff is the frictional drag, A w is the wetted area of the nozzle

and Cf is the frictional drag coefficient, then
Ff 1/2 pC f A w Va 2
AK
Tf
- --
p n2 D4 p n2 D4

Cf A
w
J2
__ (5. 5)
2 D2

where J is the advance ratio and D is the propeller diameter. Taking

the frictional drag coefficient to be equal to the turbulent flow flat
plate friction factor (i. e. the Schoenherr mean line, equation [5. 7] )
a value of .AKTf may be calculated. Since a nozzle profile is not a
flat plate over which the flow may not be fully turbulent, the
Schoenherr friction coefficient may not be a particularly good estimate
for the true frictional drag, however for the purpose of this study,
this factor amounts to merely a small correction to account for the
 77

relatively small difference in Reynolds number for flow over the ex-
perimental model propeller guard and a full size guard.
The definition of the Reynolds number for this purpose is

Va 1
Re = (5.6)

where 1 is taken to be the mean chord of the nozzle surface and V is

the kinematic viscosity of water. The Scheonherr mean line is de-
fined by the following equation
0. 242
1/2 log10 (Re Cf) (5. 7)

When scaling up the model test to full scale, it is assumed that the
ratios A w /D2 and 1/D remain constant.
Assuming that the pressure drag can be expressed as follows,
FD = 1/2 pAD CD Va2 (5. 8)

then
2
FD 1/2 p AD CD Va
AK =
TD
p n2 D4 p n2 D4

AD CD
2
J = Y J2 (5. 9)
2 D2

where AD is the projected frontal area of the nozzle, CD is the cor-

responding drag coefficient (assumed constant) then Y will be a
constant for a particular nozzle design (assuming A D /D2 is constant)
which must be determined from the experimental data.
 78

From the discussion in Chapter III, it is apparent that AKTA

(the change caused by the flow altering of the nozzle) will primarily
be a function of the load coefficient CT. Reducing the data from
Figure 5.6 and Figure 5.7 it was found, 11 that the AK TR of equation
(5. 3) could be best expressed as
AKTR =XCT +YJ2 (5. 10)

For convenience, the load coefficient was taken to be the open pro-
peller load coefficient at the equivalent advance ratio, or
8 KTo
CT = (5. 11)
Tr J2

From equation (5. 10) it is clear that

AK = X CT (5. 12)
TA

and

AK = Y J2 (5. 13)
TD

Equation (5.12) will not be valid as the advance ratio approaches zero,
because the load coefficient will then approach infinity, however for
the somewhat limited range of advance ratios (about 0.6 to 1.2 for
most boats at moderate speeds) that planing boats operate, the approx-
imation of equation (5. 12) is felt to be valid.

11
Again by use of the SIPS system of the OSU computer center.
 79

Reducing the data of Figure 5.6 and Figure 5.7 in this way and
using the SIPS subsystem of the OSU computer center to fit a least
squares curve approximation of the form of equation (5. 10) through
the data points, a very good model of the effect of the nozzle on the
thrust coefficient can be obtained. The specifications for the three
nozzle configurations tested are tabulated in Table 5.1.

Table 5.1. Performance specifications for propeller nozzles tested.

Nozzle Weed Al w/o Al vanes
vanes
1 X = AK TA/CT 1.34 x 10-3 3.82 x 10-3 2.52 x 10-3
2 Y = AKTD/J2 -8.07 x 10-2 +0.06 x 10-2 -2.06 x 10-2
3 1/D 0.47 0.275 0.275
4 Aw/D2 3.42 1.92 2.73
5 AK
Q Qo
-0.02 +0.117 +0.227

6 mean chord 1,
ft 0.344 0.2 0.2
7 wetted area
A
w, ft2 1.82 1.02 1.45
8 frontal area
AD, ft2 0.0875 0.102 0.145
9 A
D
/D2 0.164 0.192 0.273
10 CD 0.98 0.15
11 No. of
data points 19 21 6
 80

From the previous discussion it should be clear that an efficient

nozzle guard should have a small wetted area ratio A w /D2, a small
form drag parameter Y (a small drag coefficient), a non-negative
load coefficient parameter X (a nozzle profile which does not cause a
net loss of thrust) and a small torque coefficient parameter AK /K
Q Qo.
From this criteria, nozzle Al without prerotation vanes is the best
propeller guard of the three variations tested. From just this data
alone it is difficult to tell whether the weed guard or nozzle Al with
guide vanes will affect boat performance the most. However when
incorporating the data of Table 5.1 into the power performance model
of program HULLS, some interesting comparisons can be made.
Referring to the example of Chapter III, to the boat with dimensions
and thrust characteristics as in Figure 3.6, and incorporating the
effect of various propeller guards leads to the results tabulated in
Table 5.2. In Table 5.2, propulsive efficiency is defined as
T RV
p. e. = (5. 14)
SHP

where TR is the thrustline drag at the corresponding velocity V and

SHP is the propeller shaft horsepower, in this case 265 Hp.
From Table 5.2, it is clear that the plain nozzle would act as
a very good propeller guard, if it is capable of fully protecting the
propeller. How effective it is cannot be determined from this study,
but the fact that a plain nozzle with the right choice of propeller, has
 81

the potential to hardly affect boat performance should be enough in-

centive to spur further interest into this question.

Table 5.2. Optimum performance of Boat C (Figure 3. 6) with the

effect of various propeller guards, 265 Hp at a propeller
shaft speed of 28 rps.
Open weed Nozzle Al Nozzle Basket
propeller guard w/o vanes Al vanes guard (est.)
Maximum
speed (knots) 34.65 25.53 32.28 27.40 20.69

Optimum
dia. (ft.) 1.74 1.53 1.71 1.62 1.49

Optimum
pitch (ft.) 2.47 2.36 2.33 2.16 2.22
P/D 1.41 1.55 1.36 1.33 1.47

Propulsive
efficiency .757 .451 .660 501 . 336

The dismal performance of the weed guard is directly attribut-

able to its lack of streamlining. The value of the drag coefficient in
Table 5.1 (line 10), calculated from equation (5. 9), shows how im-
portant streamlining can be. The only positive characteristic of the
weed guard is that it is somewhat stronger and less fragile than the
Al nozzle. However, it too has the drawback that the inlet to the
propeller is unprotected.
Adding prerotation vanes to nozzle Al should not have affected
the nozzle performance as badly as it did. The results show that
 82

work needs to be done on designing the prerotation vanes to comple-

ment the nozzle performance rather than detract from it. The major
problem with the existing design would appear to be the unstreamlined
vane supports. It is the considered opinion of the author that a nozzle,
with a good vane design would be the ultimate propeller guard design.
For the purpose of comparison, estimated performance char-
acteristics of a basket type guard were input into the propeller size
optimizing routine, with the results as shown in the last column of
Table 5.2. In estimating the performance characteristics of the
basket guard, it was assumed that the basket only added drag to the
propulsion system, but otherwise didn't effect propeller performance.
In computing the drag of the basket, conservative estimates of the
ratios Aw/D 2, 1/D and CD AD /D2 were used (4.0, 1. 5 and 0. 36
respectively). These parameters are not strictly applicable to a
basket guard as the propeller diameter on a boat can be changed
while leaving the basket size the same, thus the ratios discussed
above are not strictly constant, however the analogy between the
performance characteristics of a basket guard and a propeller nozzle
is good enough to show how much improvement can be made over bas-
ket guard performance. The small optimum diameter of the propeller
with the basket guard (also with the weed guard) is due to the assump-
tion that the surface area of the guard varies directly with the square
of the propeller diameter.
 83

The poor performance of the basket guard is due to the high drag
characteristics of the cylindrical basket members. Any performance
improvement for a basket guard design must necessarily come from
a reduction of drag area and/or a streamlining of the basket mem-
bers. This would suggest that a major improvement in basket design
(and probably the least expensive) would be to streamline and reduce
the size of the basket members. Still, the resulting basket guard
would not be as efficient as an effective nozzle guard.
 '84

VI. CLOSURE

Validity of the Model

The worth of an analytical model is its ability to predict the

behavior of the real phenomena. As a test of the model, data were
entered into program HULLS to simulate the performance of the 16 ft.
test boat used in the experimental data collection. The test boat had
a maximum planing surface beam of 56 in. and a beam (chine to chine)
at the transom of 54 in. for a weighted average beam (equation 2. 8)
of 55.5 in. (4. 625 ft.). The deadrise of the test boat was a constant
five degrees. The gross weight and center of gravity location were
approximated by estimating the weights and center of gravities of the
empty boat, test equipment and occupants. The skin friction rough-
ness allowance AC used was the standard 0.0004.
f
Figure 6. 1 is a comparison of the test boat predicted thrust and
the thrust as recorded during the propeller runs. In the calculations
a correction was made to account for the wind resistance of the boat
hull and occupants, where the parasite area is defined as the product
of the frontal area and its associated drag coefficient. The value of
parasite area in the calculations was approximated by noting the
difference in boat drag between upstream and downstream runs and
using an approximate current velocity in the test run area of about
 A = 1100 lbs.
P = 5°
B =4.62 ft.
LCG = 6.04 ft.
160 VCG = 1.0 ft.
c = 0.
PH = -1. 25 ft.
PV = 0.4
AC
f
= 0.0004
120

Air Resistance Parasite Area = 15 ft. 2

80
Upstream
A Downstream
Calculated
40

ti

0 5 10 15
Velocity (knots)
Figure 6.1. Test boat thrust versus velocity.
 86

two knots. 12 For simplicity, it was assumed that the air drag force
acted through the center of gravity of the boat.
Figure 6.1 shows the same trend as in Figure 2.3 and Figure
2.4, that is the model tends to underpredict boat thrust at lower
speeds which is due to the inadequacy of the planing formulas, dis-
cussed inChapter II, at these low speeds. For this reason, no
calculations were performed for speeds less than ten knots. However,
in the higher speed range (above about 12.5 knots) the model per-
forms rather well. As a planing boat is primarily intended for higher
speeds, this low speed limitation should not pose a serious problem
in use of the model, as long as the user is aware of it.
Figure 6. 2 shows the calculated and experimental trim angle
for the test boat. The calculated values of trim angle show the
correct trend, although their values are somewhat displaced from the
experimental values. This difference is unimportant as the calcu
lated values refer to the actual angle between the keel of the boat and
the horizontal plane, while the experimental values refer to an arbi-
trary reference angle set when the boat was motionless in the water.

12
A11 velocities are measured relative to the water's surface.
Naturally a boat proceeding downstream will be going faster relative
to the still air, than when proceeding upstream at the same water
velocity, thus a boat going downstream will experience a greater air
drag. The difference between upstream and downstream drag will be
directly proportional to the current velocity, the boat velocity relative
to the water and the parasite area of the boat.
5

4
A

A
3

A
2

Calculated
1 Experimental

0 10 1

Velocity (knots)
Figure 6. 2. Test boat trim angle versus velocity.
 S8

Obviously, the keel of the boat need not be horizontal with the boat
at rest in the water, thus partly accounting for the difference be-
tween calculated and experimental trim angles.
Figure 6.3 and Figure 6.4 show the calculated and experimental
propeller shaft speed as a function of boat velocity, for both the open
propeller and the propeller in the weed guard. For the lower pitch
propeller (PR20, P/D = 1.03) the agreement between experimental
and calculated shaft speeds is excellent in the higher speed range,
while the deviation in the low speed range can be explained by the low
speed inadequacies of the planing model of Figure 6.1. The reason
for the low calculated shaft speeds for the high pitch propeller
(PR24, P/D = 1.26) is unknown, but is probably due to the design
differences between the experimental propeller form and the form
from which the propeller performance curves used in program HULLS
were adapted. Even so, most of the experimental data points fall
within five percent of the calculated values.
To predict the maximum speed of a boat, one can compare the
calculated propeller speed and associated horsepower with the engine
performance curve. The point where the propeller horsepower curve
intersects the maximum engine horsepower curve will be the maxi-
mum shaft speed of the propeller, which corresponds to a maximum
boat speed. Proceeding in this manner, using the outboard motor
performance curve of Figure 4.3, the maximum speed of the test boat
 89
60

50

40

30
10 12 14 16 18 20
Velocity (knots)

Figure 6.3. Propeller shaft speed versus test boat speed,

open propeller.

10 12 14 16 18 20
Velocity (knots)

Figure 6.4. Propeller shaft speed versus test boat speed,

weed guard.
Table 6.1. Comparison of experimental and predicted maximum test boat speeds.
Maximum Maximum Maximum Maximum
speed (knots) speed (knots) shaft speed shaft speed
experimental calculated (rps) exp. (rps) calc.
No
PR20 Guard 18.6 20.0 51.4 55.0
Propeller Weed
Guard 16.2 16.0 50.5 50.3
No
PR24 Guard 17.3 19.5 44.8 46.4
Propeller Weed
Guard 14.4 14.9 42.6 41.6
 91

was predicted with the results as in Table 6.1.

With the exception of the PR24 propeller without a guard, the
results of Table 6.1 show a fairly good agreement between the cal-
culated and experimental test boat speeds. Considering the number
of seemingly innocuous variables, such as minor changes in boat
loading, wind speed, current direction, wave height and period,
which can have a rather dramatic effect on the top speed of a planing
boat, the comparison of Table 6.1 would seem to be as good as could
be expected. In any case, the model would appear to be reliable in
correctly predicting performance trends, in this case, for example,
the lower top speed of the test boat with the high pitch (PR24) pro-
peller.
In conclusion, it would appear that within certain limitations the
model can be expected to yield fairly accurate results. The model
can be particularly useful in predicting trends of boat performance,
which in conjunction with actual boat performance tests may be ex-
pected to yield practical results.

Summary and Comments

In this paper a computer model was discussed by which the drag

of a planing boat may be predicted with a relatively small amount of
input data. Using existing data on marine propeller performance,
a method was shown to predict the performance of a planing boat
 92

powered by a conventional marine propeller. Also a routine to op-

timize the propeller diameter was discussed. An experimental
program to test prototype nozzle propeller guards was described,
and a method to include the effect of the guard on propeller per-
formance was incorporated into the boat performance model. The
computer model was used to make several comparisons between boat
performance with various propeller guard designs and the model was
compared with the performance of a small boat to check its validity.
The practical aspects of the boat performance model is that a
multitude of possible boat, engine, propeller and guard combinations
may be simulated analytically, before any expensive hardware is built.
Also the model can help improve the performance of existing boats
by revealing the cause of certain inefficiencies (the example of Figure
3. 5). The experimental propeller guard test procedure can be used
to evaluate the performance of any number of propeller guard designs,
from which it is possible to make an interpretation as to the effect of
a certain design to any size and displacement planing boat by use of
the performance model.
The model is incomplete. Various aspects of boat performance
have yet to be included in program HULLS, to account for such effects
as the drag of the rudder, the propeller shaft and its supports, the
effect inclined flow has on propeller performance and the induced
propeller forces which may affect the trim of the boat. Trim tabs
 93

(transom flaps) are often used on planing boats to vary the running
trim, most boats have some type of keel, both of which could have a
more or less pronounced effect on boat performance, neither of which
is included in the performance model yet. The low speed deficiencies
of the planing model should be improved and a more complete analysis
of basket guard performance should prove helpful. More inclusive
propeller performance data should be included in the model. High
speed boats often have difficulties with propeller cavitation, which
can have a very pronounced effect on boat performance. Limited
experimental data on partially cavitating propellers is available
(for instance van Lammeran (A8) in appendix A) and could be included
in the model.
Also this study is limited to planing boats, although there are
many small boats (gilinetters) that would more correctly be classified
as displacement boats. Fortunately the propeller performance analy-
sis applies equally well to both displacement and planing boats and it
would be a simple matter to modify program HULLS to accept input
data for a displacement type boat. None of these considerations
should be considered unimportant, however the present model does
work reasonably well and should provide valuable results providing
the user understands the limitations as well as the features of this
performance model.
 94

By use of the model, a fisherman may decide that the perform-

ance loss due to a basket type guard may so adversely affect his
production that the installation cost of a more efficient guard, whose
performance can be reasonably predicted, is justified. Another
fisherman may decide that a moderate performance increase does not
justify the additional expense. In evaluating the cost effectiveness of
a propeller guard design the overall system must be evaluated, that
is, the initial cost of a propeller guard, the cost of fuel, the value of
improved boat performance, the risk of propeller fouling and guard
durability must all be included to decide which system will be more
effective in the long run. Unfortunately, this study could not cover
all these aspects of gillnet fishing economics, however it has laid the
groundwork for evaluation of the one question from which all the others
follow, that is, how much more efficient can a propeller guard be
made.
 95

BIBLIOGRAPHY

1. Clement, Eugene P. and Blount, Donald L. "Resistance Test of a

Systematic Series of Planing Hull Forms," Transactions of the
Society of Naval Architects and Marine Engineers (SNAME),
Vol. 71, 1963. pg. 491.
2. Gongwer, C. A. "Water Jet Propulsion for Surface Craft,"
proceedings Fourth Symposium on Naval Hydrodynamics, Office
of Naval Research, Washington, D. C. 1962.
3. Had ler, J. B. "The Prediction of Power Performance on Planing
Craft," Transactions SNAME, Vol. 74, 1966. pg. 563.
4. van Manen, J. D. "Analysis of Ducted Propeller Design,"
Transactions SNAME, Vol. 74, 1966. pg. 522.
5. Murray, Allan B. "The Hydrodynamics of Planing Hulls,"
Transactions SNAME, Vol. 58, 1950. pg. 658.
6. Savitsky, Daniel. "Hydrodynamic Design of Planing Hulls,"
Marine Technology, Oct. 1964. pg. 71.
7. Schoenherr, Karl E. "Propulsion and Propellers," Principles of
Naval Architecture, Publ. SNAME, New York, N. Y. 1941.
8. . "One Set of 16 SNAME Small Craft Data
Sheets," publication of SNAME Technical and Research Committee,
1967.
 96

NOMENC LA T URE

A test outboard motor carburetor pressure difference

(in. H2O)

a on planing boat, distance between line of action of Df and

CG

AD projected frontal area of nozzle

Ap propeller disk area
Apl planing area of boat, area enclosed by chines and transom
As propeller slipstream cross-sectional area
Aw wetted surface area of nozzle
B beam (width) of planing hull

B. P. atmospheric barometric pressure (in. Hg)

Bpa mean beam of planing hull (Apl/Lp)
Bpt beam over chines of planing hull at transom
Bpx maximum beam over chines of planing hull
C mean chord of propeller blade
c on planing boat, distance between line of action of N and
CG

CD form (pressure) drag coefficient

CDN nozzle (duct) sectional drag coefficient
Cf frictional drag coefficient, usually Schoenherr turbulent
friction coefficient
CC on planing boat, location of center of gravity
CT, propeller load coefficient (eq. 3. 9)
 97

D propeller diameter
Df on planing hull, frictional component of drag
DN nozzle (duct) drag force
e ideal propeller efficiency
eT ideal propeller efficiency with nozzle
eDN nozzle drag efficiency factor
ehp effective horsepower of planing hull (=RV)

e actual propeller efficiency (eq. 3.18)

f on planing boat, distance between line of action of T and
CG

FD pressure drag force

Ff frictional drag force
J propeller advance ratio (eq. 3.17)
K duct friction loss coefficient
K. E. kinetic energy (rate)
KQ propeller torque coefficient (eq. 3.16)
lc Ell new value of propeller torque coefficient
KQo open propeller torque coefficient
KQg propeller torque coefficient with installed guard
KT propeller thrust coefficient (eq. 3.15)
K open propeller thrust coefficient
To
KTd propeller thrust coefficient with n eliminated (eq. 3.25)
KTg propeller thrust coefficient with installed guard
 98

KTn propeller thrust coefficient with D eliminated (eq. 3. 26)

1 nozzle (duct) chord (parallel to shaft)
LCG on a planing boat longitudinal distance of CG from
transom, measured along keel
Lk on planing boat, wetted keel length
Lp on planing boat, projected length of planing bottom,
measured from transom to intersection of chines
M on planing boat, moment of forces about CG
N on planing boat, resultant of pressure forces normal
to hull
n propeller shaft speed
no test outboard motor propeller shaft speed (revolutions/
second)

P propeller pitch
p absolute fluid static pressure at propeller hub
p. e. ship propulsive efficiency (RV/SHP)
PH on planing boat, horizontal position of propeller hub,
measured forward from transom
pv vapor pressure of water at propeller hub
PV on planing boat, vertical position of propeller hub,
measured down from keel
Q propeller torque
Qo test outboard motor propeller shaft torque (lb. -ft.)
R boat resistance (drag) measured in horizontal direction
(parallel to boat velocity)
Re Reynolds number (eq. 3. 19)
 99

Re Reynolds number based on propeller diameter (eq. 3. 19a)

D
Rei Reynolds number based on airfoil chord (eq. 3. 19b)
SHP propeller shaft horsepower
T propeller thrust
t thrust deduction fraction
TA air temperature (absolute)
TR thrustline drag, projection of boat drag, onto propeller
shaft line
U representative velocity
u added fluid velocity in propeller slipstream
V velocity of boat or ship (with respect to the water)
Va velocity of advance of propeller (with respect to water
near propeller)
VCG on a planing boat, vertical distance of CG, measured
up from keel
w Taylor wake fraction
X load coefficient parameter (eq. 5. 10)
x representative length
Y form drag parameter (en. 5. 10)

Greek Letters
deadrise angle of planing hull

PT
deadrise angle at hull transom

6
deadrise angle of hull, at 0.6 Lp forward of transom
 100

A boat displacement, weight of

AC
f
frictional drag roughness factor
AKQ change of open propeller torque coefficient due to
propeller guard
AKTA
change of open propeller thrust coefficient due to
propeller and guard loading
AK
TD
change of open propeller thrust coefficient due to
propeller guard form drag
AK
Tf
change of open propeller thrust coefficient due to
propeller guard frictional drag
AK AK + AK
TR TA TD
AP nozzle (duct) pressure drop
propeller shaft angle, measured from boat keel
V kinematic viscosity
p fluid density
To cavitation number (eq. 3.20)
T planing boat trim angle, angle between horizontal and
boat keel
APPENDICES
 101

APPENDIX A

Additional References

References not directly related to this report but could be

helpful to anyone wishing to pursue this subject further.

1. Chen, C. F. "Shrouded Supercavitating Propellers," Fourth

Symposium on Naval Hydrodynamics, Office of Naval Research,
Washington, D. C., 1962.
2. Clement, E. P. "How to Use the SNAME Small Craft Data
Sheets for Design and Resistance Prediction," Society of Naval
Architects and Marine Engineers (SNAME) Technical and Re-
search Bulletin No. 1-23, 1963 (reprinted 1971)
3. Dyne, G. "An Experimental Verification of a Design Method for
Ducted Propellers," Symposium on Pumping Machinery for
Marine Propulsion. Publ. ASME, New York, N. Y., 1968.
4. Fletcher, Herman S. "Experimental Investigation of Lift, Drag,
and Pitching Moment of Five Annular Airfoils," NACA Tech.
Note 4117, 1957.
5. van Gunsteren, L. A. "Ring Propellers and Their Combination
with a Stator," Marine Technology, Vol. 7, no. 4, Oct. 1970,
pg. 433.
6. Had ler, J. B. and Hubble, E. N. "Prediction of the Power
Performance of the Series 62 Planing Hull Forms," Transactions
SNAME, Vol. 79, 1971, pg. 366.
7. Johnson, Virgil E. "Waterjet Propulsion," Seventh Symposium
on Naval Hydrodynamics, Office of Naval Research, Washington,
D. C. , 1968.
8. van Lammeren, W. P. A., van Manen, J. D. , and Oosterveld,
M. W. C. "The Wageningen B-Screw Series," Transactions
SNAME, Vol. 77, 1969, pg. 269.
 102

9. van Manen, J. D. "The Choice of the Propeller," Marine

Technology, Vol. 3, No. 2, April 1966, pg. 158.
10. Meyerhoff, Leonard, Hill, John G. , and Meyerhoff, Stanley.
"Ducted Propeller Applications for Modern Ships," Transactions
SNAME, Vol. 80, 1972, pg. 136.
11. Morgan, W. B. "Theory of the Annular Airfoil and Ducted
Propeller," Fourth Symposium on Naval Hydrodynamics, Office
of Naval Research, Washington, D. C. , 1962.
12. Morgan, W. B. and Caster, E. B. "Comparison of Theory and
Experiment on Ducted Propellers," Seventh Symposium on Naval
Hydrodynamics, Office of Naval Research, Washington, D. C. ,
1968.

13. Oosterveld, M. "Model Tests with Decelerating Nozzles,"

Symposium on Pumping Machinery for Marine Propulsion, Publ.
ASME, New York, N. Y. , 1968.

14. Rossell, Henry E. and Chapman, Lawrence B. , Editors,

Principles of Naval Architecture, Society of Naval Architects and
Marine Engineers (SNAME), New York, N. Y. , 1941.
15. Sacks, A. H. and Burnell, J. A. "Ducted Propellers A Critical
Review of the State of the Art," Progress in Aeronautical
Sciences, Vol. 3, Pergamon Press, New York, N. Y. , 1962.
16. Savitsky, Daniel. "On the Seakeeping of Planing Hulls," Marine
Technology, Vol. 5, No. 2, April 1968, pg. 164.
17. Savitsky, Daniel, Roper, John K. and Benen, Lawrence.
"Hydrodynamic Development of a High Speed Planing Hull for
Rough Water," Ninth Symposium on Naval Hydrodynamics,
Office of Naval Research, Vol. 1, Washington, D. C. , 1972.
18. Todd, F. H. "Table of Coefficients for A. T. T. C. Model Ship
Correlation and Kinematic Viscosity and Density of Fresh and
Salt Water," SNAME Technical and Research Bulletin No. 1-25,
1964 (reprinted 1970)
19. Troost, L. "Open Water Test Series with Modern Propeller
Forms," Transactions North East Coast Institution of Engineers
and Shipbuilders in Newcastle upon Tyne (NEC I), England
1950-51, pg. 89.
 103

20. Weissinger, Johanness and Maas, Dieter. "Theory of the Ducted

Propeller - A Review," Seventh Symposium on Naval Hydro-
dynamics, Office of Naval Research, Washington, D. C. , 1968.
21. Wislicenus, G. F. Fluid Mechanics of Turbomachinery, Vols.
1 and 2, Dover Publications, New York, N. Y.
22. Wislicenus, G. "Pumping Machinery for Marine Propulsion,"
Symposium on Pumping Machinery for Marine Propulsion,
Publ. ASME, New York, N. Y., 1968.
23. . "Explanatory Notes for Resistance and Pro-
pulsion Data Sheets," SNAME Technical and Research Bulletin
No. 1-13, 1953 (reprinted 1972).
 104

APPENDIX B

Particulars of Propeller Series

Performance Data Used in Program HULLS
Model test data of four propeller series, presented in Schoenherr (7),
were reduced to empirical polynomial equations. Particulars of the
propeller series are as follows:
All propeller series had elliptical blade form with ogival blade cross-
section. Propeller series 1 and 2 were four bladed propellers with a
model test diameter of nine inches. Propeller series 3 and 4 were
three bladed propellers with a model test diameter of eight inches.
In the following table, these terms have the following definitions.
P/D, Pitch Ratio, Propeller Pitch/Propeller diameter
M. W. R. , Mean Width Ratio, mean expanded width of one blade/
propeller diameter
B. T. F. , Blade thickness fraction, thickness of blade at root/pro-
peller diameter

Propeller Series P/D M. W. R. B. T. F.

1 0.6 - 1.35 0.20 0.05

2 0.6 - 1. 35 0.25 0.05

3 O. 4 1. 6 0.30 0. 05

4 0.4 1.6 0.40 0.05

 105

Table B. 1. Curve fit coefficients.

Thrust Coefficient KT = E (P/1))111JnC.
J = Advance Ratio
Ci
Propeller Series
1 2 3 4

J -.14644 -1.122 -. 35649 -. 36996

P/D . 38958 . 80043 .63395 . 48175

(P/D) J2 -. 74675 . 3062 -.67787 -.0068867

3/2 1.3862 .010251
( (P/D) J) 1.6274 -.097408
3/2 -.27698 1.5795 -.034565
J 0.

(P/D) 3/2 .10381 -.28412 -. 51063 -. 39528

(P/D)2 J -. 86663 0: -.77676 0.

(P/D)2 0. 0. .36144 0.

J2 0. -1.1127 0. 0.
 106

Table B. 2.
Torque Coefficient KQ = (P/D)171Jr1Ki
= Advance Ratio
K.
Propeller Series
1 2 3 4

J .024705 .013151 .01882 .015769

12 -.0099412 -.038731 0. -.01622

(P/D)2 .077502 .071361 .075399 .070074

(P/D)J .0'78489 -.015613 -.097448 -.064269

(P/D)2 J -.17858 -.11443 -.13744 -.022769

((P/D)J)3/2 . 3594 .095306 .30194 .027956

(P/D)2J2 0. 0. -.0094812 0.

((P/D)
23)3/2 0. .024384 0. 0.

((P/D)J2) 3/2 0. -.018276 0. 0.

P/D J2 -.17434 0. -.13748 0.

 107

APPENDIX C

Program HULLS

Figure C. 1. Program HULLS data input flow diagram.

108
 109

Table C. 1. Explanation of program HULLS input flow diagram.

Input Variable Description Units Format
A A Displacement lbs. F10.2
P Deadrise Angle Degrees F10. 2
LCG Long. C G ft. F10. 2
VCG Vert. CG ft. F10.2
E Shaft Angle Degrees F10. 2
PH Prop. Horiz. ft. F10.2
PV Prop. Vert. ft. F10.2
AC
f
Roughness Allowance F10. 5
2
CDA Air Parasite Area ft. F10.2
B IP No. of Propellers 315
IZ No. of Blades/Prop.
M 1 --.. Narrow Blade
2 = Wide Blade
C w Wake Fraction 2F10. 3
t Thrust Deduction
Title Prop. Guard Title 8A10
D X Propeller -- 5E11. 4
Y Guard
1/D Performance
Aw/D2 Coefficients
AK /KQ
Q
(see Chapter V)
E D Prop. Dia. ft. 2F10. 3
P Prop. Pitch ft.
F SHP Shaft Power Hp. 3F10. 3
n Shaft Speed rps.
D max. Max. Dia. ft.
 110

Table C. 1 (cont.)
Decision Description
1 IP = 0 Start new boat input
IP <0 Stop
Title If no title go to decision 3
3 Input decision parameter (13)
1, go to input E
2, go to input F
3, go to decision 4
4 Input decision parameter (13)
1, go to input .A
2, go to input B
3, go to decision 3

Operation Description
1 Calculate and write out boat
thrust, resistance, etc. from
10 to 40 knots
2 Calculate and write correspond-
ing shaft speed and horsepower
from 10 to 40 knots
3 Calculate and write out optimum
propeller diameter, pitch
propulsive efficiency and
corresponding boat speed
 PROGRAM HULLS(INPUT.OUTPUT,TAPE5=INPUT,TAPE6=OUTPUT) TAUR=TAU/57.296
OINENS/ON ii(401.VFS(401.TRI4(40).T(401.EHP(46),PORP(4:1 LAmOA=ALAMOA(TAU,CV.CLOC)
C.DRAG(40) CL11=(TAU**1.1).1.012*SORT(LANDA1-.1065,BETA*TAU..1.66,
DIMENSION GTITLE131 C(1.0129SORT(LAMDA)1**7.6
REAL MOMENT,LAMOA,LCG VFS1=VFSIIVI*SORT(1.-CLB0/(LAMDA*COSITAUR111
REAL J0(71,KT2 OLANDA=-G.4644-1.21549,BETA.3.04898*TAU
CONNON/8LK3/VES,T C+0.19031,BETAITAU-0.004929.(BETA/TAU)**2
COMMON/GLKA/D,RPS.P.ITYP IF(OLAMOA.LT.0.10LAMDA=g.
CONMON/BLKB/WF,TD OF=0.5.1.9393*(1.8*(CF(VFSMANDA*0/1.2083E-05)
COMMON/EILKUDIA/BLKO/AOKTRO0KTR,PSI,WARAT/2LKE/A0K0 C+OCF),(VFSIVFS1.LANDANVSO,DLANIDA)/COS(BETAR)
DATA V/400./.TR/m/40.0.1.EHP/41m0S,PORP/40,0a,DRAG/4P10./ D=OELTATANITAURINDFICOS(TAUR)
KITE(6,11) CP=1.75-1./(5.21CV.42/LANDA.4,242.39)
11 FORMAT(IHISTHIS PROGRAM WILL CCMPUTE BOAT RESISTANCES/X C=LCG-CPLAMCA4B
CS FROM 11 TO 41 KNOTSX/X A=VCG-0. 254*TAN(BETAR)
CWOELTA.ROAT DISPLACEMENT (LBSWX mOMENT=DELTA*(CCOS(TAURNEPSRI-F*SINITAUR)1/COS(EPSR)
CIMETA=00AT DEADRUE ANGLES/X C.OF*1A-F/COS(EPSR)-C.TAN(EPSR))
CalmMEAN BEAM NIOTHIFT)S/X C SECANT METHOD TO FIND NEW APPROX. FOR TAU
CSLCG=LONG. CENTER OF GRAVIFWX STAU=TAU
CWWCG=VERTICAL CENTER OF GRAVITYX/X IF(KOUNT.E(2.1)G0 TO 70
CMEPS=PROP. SHAFT ANGLE MEASURED FROM KEELP/X TAu=(MOMENT*TAuo-TEST.TAu)/(NOMENT-TEST)
CIMBH.POSITION OF PROP. MEASURED FORWARD OF 2/X IF(ABS((TAU-TAU01 /TAU).LT..0021G0 TO 210
CS INTERSECTION OF KEEL AND TRANSOMx/X IF(TAU.LT..15)TAU=.15
CFPV.POSITION OF PROP. MEASURED VERTICALLY DOWNWX 70 TAU1=STAU
CS FROM KEELS/X TEST=MOMENT
CMCF.RCUGHNESS ALLOWANCE (USUALLY .0004)5/X) KOUNT=KOLNT.1
C INPUT BOAT CHARACTERISTICS TF(KOUNT.GT.1001CALL ERROMHULLS),RETURNS(300)
13 READ(5.10010ELTA,BETA,B,LCG,VCG,EPS.PH,PV GO TO 50
100 FORMATIF10.2) 2J0 AIRD=COA*0.5.0.002371VSO
C INPUT ROUGHNESS ALLOWANCE O=ONAIRO
REA0(5,105)0CF T(IV) .I0ELTA'S/N( TAURWIFI/COSCEPSRFAIRO/ICOSCEPSR+TAURI)
105 FORNATIF10.5I EHRt/V)=0.1IFS(IVI/550.
C INPUT AIR RESISTANCE PARASITE AREA PORP(IV)$SORTICELTA/11.9383+vS0,0,8»
READ(5.100)COA TRIM(IV)=TAU
WRITE16.110)DELTA,BETA.9.LCG.VCG,EPS.PH,pv.00F DRAGIIV)=0
110 FORMAT(///XtDELTA=AF11.2/XABETA=$F10.2/XXB=AF10.2/X$LCG=XF10.2/X C TEST FOR STABILITY, SAVITSKY FIG. IA
l*VCG=AF10.2/XIEPS.$F10.2/KAPH=$F10.2/x$PV=AF10.2/x$DCF=XF10.6) STAB=SH
WRITE(6,1111CDA PTAU=2.284+91.502.PORP(IV)..,2,0.083973*SCRT(BETA)/PORP(IV)
111 FORMAT(XIAIR RESISTANCE PARASITE AREA =$F10.2//) 1-0.42997/PoRP(U)
NRITE(6.15) IF(TAU.GT.PTAU)STA9=8HUNSTABLE
15 FORMATI4XtVELOCITy$4X/TRIm$6X50RAG56XIEFF. HP$3X C OUTPUT TRIM ,DRAG.ETC.
C$PORPOSING$2x*PROP.A/4X5(KNOTS)A5X$ANGLE55X5(LBS)* 700 WRITS16,5001Y(IV),TRIPIIV),ORAG(Iv),EHP(IY),PORP(IV)
C15X$COEFF43X*THRUSTILBSW) 1,T(IV),STA8
RETAR=1ETA/57.296 500 FORMAT(1x,4F10.2,F11.3.F10.2,5x,A1)
EPSR=EPS/57.296 WRITE16,20031
F=(VCG,PV)*COS(EPSR1-(LCG-PH)*SIN(EPSR) 2000 FORMAT(1141)
C CALCULATE TRIM, DRAG, ETC. FROM 10 TO 40 KNOTS C INPUT NC. OF PROPELLERS AND PROP. CHARACTERISTICS
00 300 iy=looto 501 REAO(5,120)IP.U.N
VtIV1=IV 121 FORMAT(3I5)
KOUNT=1 TO 13
C ASSUME FIRST TAU IFI/P.LT.VISTOP
TAU=0.5 P=IP
VFS(IV)=V(/11)1.6509 WRITE(6,130)IP.I20
VSQ=VES(IV)**2 130 FORMAT(////X/No. OF PROPELLERS=*I3/XINO. CF BLADES=*I7/x
C CALCULATIONS AS IN SAVITSKY ANO HADLER ISNARRON BLA9E(m=1) OR NICE BLADE(M=2).M=XI3,///)
CV=VFS(IV)/SORT(32.2'8) C INPUT WAKE FRACTION ANI THRUST DEDUCTION
CLOC=CLOISETA,2.+DELTA/(1.9383.VSO,B*81) READ15,121)14F,TD
50 IF(KOUNT.E0.2)TAu=3. 121 FORmATUF10.3)
 WRITE(6,132)WF,TO 7P1 FORMAT(////X1OUTPUT WILL 2E OPTIMUM PROP. CIA ANL °ITCH GIVEN 39pt
132 FORMAT(X*WAKE FRACTION=*F10.3,11X,*ThRUST DEDUCTION=*F10.3/////) 1/X* PER PROP., REV/SEC, ANO MAX. PROP. DIA.*/)
C INPUT NOZZLE GUARD CHARACTERISTICS C INPUT SHP,RPS AND MAX. DIA.
READ(5,125)GTITLE READ(5.703)SHP,RPS.01A4X
125 FORNATIRA1C) 701 FOR4AT(3FIV.3)
IF(GTITLE(1).EQ.134 )G0 TO 134 WR/TE(6,7021SHP,RPS,Dt4MX
WRITE(6,125)GTITLE 707 FORMAT(X$SHP=*F11.2/X$RPS=*F10.2/X*DIAMX=IF10.2/X)
READ(5,1221AOKTR,90KTR.PSI,WARAT,ADKO C ROUTINE TO FIND MAX. VELOCITY GIVEN SHP,RPS, AND MAX. DIA.
122 FOR4AT(5E11.4) C FITS PARABOLA TO 3 POINTS AND ESTIMATES MAX. TO SUCCESSTVLY GETTER
WRITE16,1331PSI,WARAT APPROXIMATICNS
133 FORMAT(1X*PROP. NOZZLE CHARACTERISTICS ARE*/X*L/0=*F1".5/X 0=550..SHP/(6.21118*PPS)
1 /WETTED AREA/D2=*F13.5//) KOUNT=1
GO TO 135 POTAMX=1./DIAMX4
134 PS/=0. IFIVELD(1./POIAMX0.251.GT.VEL0(1./(1.02POIAMX)E.25))G0 TO 125
WARAT=0. HDIA=POIA4X/2.
AO K()=0. 714 ON=PDIA4X+HOIA
135 IF(TZ.EQ.3.AND.M.ER.1)ITYP=3 705 02=04
IF(IZ.E(1.4.AND.N.E0.1)ITYP=1 01=02+HDIA
IF(IZ.ECI.3.AND.4.E0.2)ITYP=4 C3 =02 -HO IA
IF(T1.E0.4.AND.1.E0.2)ITYP=2 F3=VEL0(1./030.25)
C DECISION PARAMETER F2=VEL0(1./021).25)
515 READ(5,140)N IF(ARS(IF2-F3)/E2).LT..04.ANO.HOIA.GT.POIAMX/10.)Go TO 720
140 FORMAT(I3) F1=VELD(1./010.25)
IF(N.E0.1)G0 TO 600 707 IF(KOUNT.GT.1001OALL ERR(SHHULLS),RETURNS(731)
IF(N.E0.2)G0 TO 700 KOUNT=KOUNT+1
IF(N.E0.3)G0 TO 960 IF(F2.GE.F1.ANO.F2.GE.F3)G0 TO 715
GO TO 999 IF(F3.GE.F2)GO TO 712
C INPUT PROP. DIA. AND PITCH F3=F2
600 REA0(5,150)0IA,PITCH F2=F1
150 FORNAT(2F10.4) 03=02
PRAT/O=PITCH/DIA 02=01
WRITE(6,160)DIA,PITCH,PRATIO 01=02.HDIA
160 FORNAT(X*OUTPUT WILL RE SHP PER PROP. AS A FUNCTION OF VELOCITY* F1=VEL0(1./010.25)
1//X*DIAMETER=*F10.2/X*PITCH=*F10.2/X*PRATIO=*F10.3//) GO TO 707
WRITE(6,6601 712 IF(03.LE.POIAMX)G0 TO 719
661 FORMATI3X*VELOCITY*2X*REV/SEC*3X*SHIP HEW) F1=F?
00 675 1=10,40 E2=F1
KOUNT=1 01=02
C ROUTINE TO FIND RPS, GIVEN PROP. DIA. AND PITCH 02=01
KT1=0.1 03=02-HOIA
ITEST=0 F3=VEL0(1./030.25)
625 RPS=0.5 RPS.0.5 SCIRT(T(I)/(1.9353KTIP11.-T0)DIA4)) GO TO 707
J=VES(I)(1.-WF)/(RPSO/A1 715 C4=02+0.5HOIA(FI-F3)/(F1-2.F2+F3)
KT2=XKT(J,PRATIO,ITYP) IF(AIS(1.-(0)0/0213.251.1T..001)G0 TO 731
IF(ABSI(KT1-KT2)/KTI).LT..002)ITEST=ITEST+1 IF(HDIA.GT.PCIA4X/2J0.)HOIA=HOIA/2.
IFFITEST.GT.3)G0 TO 650 GO TO 705
KT1=((T1+1(T2)/2. 720 HOIA=PDIANY/10.01
KOUNF=KOUNT+1 GO TO 705
IF(KOUNT.GT.100)CALL EPRISHHULLSI,RETURNS(675) 715 HDIA=POIAMX/100.
GO TO 625 GO TO 704
650 Q=XKO(J,FRATIO,ITYPP01.9383RPSPPSDIA5 7?9 HDIA=HCIA/2.
SHP=2.3.141599RPS/551. IF(HDIA.LE.PCIAMX/231.)G0 TO 731
675 WRITE(6,600)V(I),RPS,SH0 GO TO 704
680 FORNAT(X,3F10.2) 730 04=POIANX
WRITE(6,20001 731 WRITE16,950)
GO TO 515 950 FORMAT(X*OPTIMUM PROP. CHARACTERISTICS ARE //1
700 WRITE(6,701) WRITE(6,951) CJ
 451 FoRMAT(77.70IA75xsPITCH*67$ SPEED KNCTS*5x7PRopuLx FUNCTION CL0(9ETA,CLEI)
CASIVE EFF.A) C SUCCESSIVE ITERATION TO SOLVE HAULER EO. (F)
VEL=VEL011./DH**1.25) CL0=1.
TX=THRUST(VEL) KOUNT=1
PITCH=OIA,PKTJ(Tx/(1.9383+RPS,FRPS=p*11.-TO) POIA**41,vEL,(1.-wF)/( 5 TEST=CLO
1RPS*0/A) ,ITTP) KOUNT=KOUNT+1
EFF=Tx+VEL/(550..SHP*P) CLO=CL040.0165.1ETA*CL04.0.6
VELK=VEL/1.6889 IF(KOUNT.GT.501CALL ERRt3HCLO),RETURNS(10)
WRITE16,552I0IA,PITCH,VELK,EFF IF(A0S(tIEST-CLO)/CLO).GT..0001)G0 TC 5
952 FORmATI2F10.3,47,F10.2,10X,F7.41 11 RETURN
960 WRITE(6.975) ENO
975 FORMATIMX,OPTION 1, ENTER DIFFERENT 00AT PARAMETERS? /7
C*OPTION 2, ENTER DIFFERENT PROPELLER PARANETERs*/X
C*OPT/ON 3. ENTER DIFFERENT PROPELLER SIZES/X
CtOPTION 4 r END PROGRAM*)
C DECISION PARAMETER
READ(5,1401N
WRITEt6.1100)N
1100 FORMATI/MOPTION=7/3/1H1)
IFIN.E0.1160 TO 13 FUNCTION CF(RE)
IF(N.E0.2)G0 TO 501 C SUCCESSIVE ITERATION TO SOLVE SCHOENHERR MEAN LINE
IF(N.E0.3)G0 TO 515 CF=0.11
999 STOP KOUNT=1
END 5 TEST =CF
CF=(0.242/ALCG10(CF4RE))**2
KOUNT=KOUNT+1
IFIKOUNT.GT.50) CALL ERRI2HCF1,4ETURhS(10)
IFTABS((CE-TEST)/CF).GT..03011G0 TO 5
10 RETURN
ENO

FUNCTION ALAMDAITAU.CV,CLO)
C NEWTON RAPHSON METHOD TO SOLVE HAULER E0.(4)
TAU11=TAU4=1.1
/TMAx=i
ALAMDA=2.0
5 ROOT=SORTIALANDA1
F=TAU11*(0.012,ROOT*0.00554ROOT*45/(CV=CV0)-CLO FUNCTION THRUST(VEL)
FP=TAU114q0.006/ROOT+0.01375,ROOT.,03/(CVITV)) C LAGRANGIAK INTERPOLATION OF THRUST AS A FUNCTION OF VELOCITY
TEST=ALAMDA-F/Fp CONNON/81K3/VES(40),T(41)
IFIABSI( ALAMDA-TEST)/ALAMDA).LT..0001)G0 TO 10 I=10
IF(ITMAX.GT.100)CALL ERRI6HALANDA),RETURNS(iO) 3 IFIVES(I).GT.VEL.OR.I.E0.40)G0 TO 5
IFITEST.L.T.0.160 TO 6 I=I+1
ALANDA=TEST GO TO 3
ITMAx=ITMAx+1 IF(I.LT.12)I=12
GO TO 5 THRUST=T(I)*(VEL-WFS(I-1))*(VEL-VES(I-2))
6 ALAMOA=ALANDA/2. C/((l/FS(I)-VES(I-111.(VESII)-VES(I-2)))
GO TO 5 C+T(I-1).(VEL-VES(I1)*(VEL-VES(I-2))
10 ALANDA=TEST C/t(VESII-1)-VFS(I))*(VFS(I-11-VES(I-21))
RETURN C+T(I-2)*(VEL-VES(I))*(VEL-VESII-11)
END C/(( VES1/-2)-VFS(I1)*(VESII-2)-11FS(I-1))1
RETURN
ENO
 FUNCTION XKUJI,POQIITYP)
CURVE FIT OF KO VS. J AND P/0 BLOCK DATA
C
REAL J C KT,KQ CURVE FIT DATA
COMNON/9LK2/A101(4)0K0(4),CK0(4),OK0(4),EK0(4) CIMENSION VES(400,T(NO)
1,F014),GKO(4),NK014),AAK0(4),BAK014) COMMON/9LK1/AKT(4),BKT(41,CKT14),OKT(4),EKT(4),FKT(4)
1,GKT(4),NKT(4).AAKT(4)
CONNON/BLKE/ADKO
XKO.AKO( ITYP) J4.BKO(ITYP).J+J+CKO(ITVP).FCCPOD COMMON/9LK2/AK0(4),9K0(4).CK0(4),OK0(41,EK0(4)
2,FKO(4),GKQ(4)0110(4),AAK0(4),OBKO(NI
14.0KO(ITYPJ*P00*J+EKUITTP)P0000*JATKO(ITYP), COMMON/ILK3/VES,T
2(POCMJ)"1.5+GICI(ITTP)*(P00*J)..24.MKOI/TYPI* DATA VES/40.P./.T/40*1.,
3(POW200*J)"1.5.AAKOIITTPI(POD4J*J)"1.5 DATA AKT/.14644,1.122,.35649,....165,,
461K0(ITYPIPOD*JJ 1OKTI.39958,.900431.63395,.49175/,
XKO=XKO(1..NAOKO)
RETURN 1CKT/.74075,.3052,-.67747,-.0CE9967/,
ENO
10KT/1.6274,.097409g1.3862,.010251/,
1EKT/..27698.1.5795,0..-.034565/.
1FICT/.10101,....28412,.51063,...039529/.
iGKT/.06063gr!.,...77576,1./.
IHKT/0.,0.g.36144,1./,
1AAKT/0.,1.1127,0..1./
FUNCTION KKTCJIIPOD,ITYPI DATA AK0/.024705..013151,.a1882,.01576W.
C CURVE FIT OF KT VS. J ANO DID IBK0/...0099412,46039131,0...91622/,
REAL J 1CK12/.077502,.071161..075399,.070074/,
CONMON/OLK1/AKT(4),BKT(4),CKT(4),CKT(4),EKT(4) 10K0/-.070489,.015613,...097448,-.054269/,
1,EKT(4),GKT(4),HKT(4),AAKT(4) IEK0/.17058,-,114431.-.13744,-.022769/,
XKT=AKT(ITTP)JNOKT(ITYP)*P00+CKTII/TYP)* 1FK0/.3594,.095396,.3019411.027956/,
1PO0PJA.OKT(ITYP)*(POO*J1"1.5.NEKT(ITYP)*J"1.5 iGKO/0.0.....1994912,0./,
2.FKT(ITYPI4P00"1.5GKT(ITYP1*POD*P00J iHK0/0.,.024384,0.,0./,
3NHKTIITYPIPPOO'POONAAKT(ITYP)JJ IAAKQ/0.0-.019276,0.,0.1,
KKT=XKTODELKT(2.5465.KKT/(JJ),J) 1BBK0/-.17434.0.1-..13749,0./
RETURN ENO
ENO

FUNCTION DELKT(CT.J) SUBROUTINE ERR(FUNC),RETURNS(M)

C NOZZLE GUARD KT CORRECTION C ERROR MESSAGE SUBROUTINE
REAL J NRITE(6,10)FUNC
COMMON/BLKC/DIA/BLKO/AOKTROOKTR,PSI,WARAT/BLKA/OgRPS.P.ITVP 19 FORMAT(IXO,ITERATION MAXIMUM ERROR, CALLED FROM 8A1C8 PROGRAM
IFIPSI.LE.0..OR.WARAT.LE.00G0 TO 10 18UNIT8,///)
OKTR=ADKIR'CTA.BOKTRJ*J RETURN M
OKTF=JJ*WARAT'CF(RPS*OIA*DIAPSI+J/1.208E-.05)/2. ENO
DELKT=OKTRA.DKTF
RETURN
10 OELKT=0.
RETURN
ENO
 FUNCTION PKTJ(KT.J,ITYP/
C
FUNCTION VELC(DIA)
FALSE POSITION 0ETNOD TO SOLVE FOR P/D GIVE KT AND J C ROUTINE IC FIND BOAT VELOCITY GIVEN PROP. DIA.
REAL KT.J
DP=1.3 CONMON/EILKA/O,RPS.P,ITYP
P01=0. CONNON/OLKEWWF.TO
KOUNT=1 COMMON/OLKC/C
5 XL=F00 REAL J.J2.101.02.KO
KOUNT=KOUNT+1 C=IIA
XR=P01.10F KOUNT=1
FL=KKT(J.XL.ITYP)-KT J=16.9/(RPS*DIAI
FR=KKT(JO(R,ITYP)-KT J2=1.6
DENOM=1.13834RPS+RPS4P+DIA+.4
IF(FL.FR.L1.0.)G1 TO 11 KO=P*O/(DENOW.DIA)
POO=P004.0P
POO=PKTJITHRUST(J+RPS*DIA/(1.-WF))/(DENON,(1.-T011).J,ITYPI
IF(KOUNT.GT.12)CALL ERR(4HPKIJI,RETURNS(30) K111=XKC(J.POO,ITYP)-K0
GO TO 5 5 P002=PKTJ(THRUST(J2+RPS*DIA/(1.-WF))/(0ENOW41.-111)).J2,ITYP/
19 X2=IXL.FR-KR,F0/(FR-FL) KOUNT=KOUNT4.1
KOUNT=K1UNT+1
IF(KOUNT.GT.50)CALL ERRI4NVELD),RETURNSt101
IF(KOUNT.GT.100)CALL ERRI4NPKIJI.RETURNS(30) KO2=XKIVJ2,POD2,ITYP)-KO
F2=XKT(J.X2,ITYP)-KT
SJ=J2
IFIA8S(F2/KTI.LT..03002)G0 TO 31 J2=(.12,KG1-J*K02)/(K01-K02)
IFIF2*FR.LT.O.IGO TO 21 IFIABS((J2-J)1J2).LT..000021G0 TO 10
05=02
J=SJ
ER=F2 KO1=KO2
GO TO 10
20 01=02 GO TO 5
FL=F? 10 VELD=J2*RPS*DIA/11.-WFI
RETURN
GO TO 19
10 PKTJ=X2 END
RETURN
END