Marine Waterjet Propulsion

SNAME Transactions, Vol. 101, 1993, pp.
275-335
Marine Waterjet Propulsion

John Allison, Member, Band, Lavis, Severna Park, Maryland
U.S. Navy SES 200 With Waterjet Propulsion
1.0 ~TRODUCTION Barges goe against Stronge Winde and Tyde. This was at a time
when there was great interest in using steam to raise water and
Although some of the earliest attempts to apply mechani- to operate fountains, so there is good reason to suppose Ramsey
cal power to the propulsion of ships involved some type of had a type of waterjet in mind (Dickinson, 1938 [ll). Toogood
waterjet, it is only in recent years that marine waterjet propul- and Hayes were granted a patent in England in 1661 for a
sion has begun to gain acceptance and to challenge the long- propulsion system using an Archimedian screw for this purpose.
established dominance of the screw propeller. In the last few The use of the Archimedian screw as a hydrodynamic device
years there has been a remarkable increase, both in the number had been known from ancient times. Although Archimedes
of waterjet-propelled vessels, worldwide, and in the number of (287 - 212 BC) is credited with its invention for pumping out
waterjet manufacturers. These developments seem to fly in the flooded ships he attributed a similar device to {he Egyptians of
face of accepted theory whereby it has long been taught that the 22nd Dynasty (about 945 BC). Such devices are still used in
waterjets are inherently less efficient than screw propellers. Egypt for irrigation purposes. The drawings of Leondardo da
This paper seeks to examine the background and Vinci (1452 - 1519) contain pictures of a moving spiral for
underlying theory of waterjet propulsion which may help to raising water, and the rotor of his famous helicopter is more like
explain the present resurgence of intei'est in this form of the impeller of an inducer pump of a modem waterjet
propulsion. A review of some leading waterjet propulsor (Figure 1). Early applications of an external pumping element
manufacturers' products is presented, with examples of their involved the use of a form of Archimedian screw. Thus, the
application. In the course of the paper, many aspects of marine origins of modem waterjet propulsion and screw propulsion are
waterjet propulsion technology are touched upon. closely related (Allison, 1978 [2]). There were, however, other
concepts of waterjet propulsion which used positive displace-
2.0 HISTORICAL REVIEW ment pumps, and propulsion systems in which the energy of
steam was used directly without the need for an intermediary
Attempts to use waterjet propulsion date from the 17th century. engine and separate pump. Because of the limitations of
In 1631, David Ramsey, a Scot, acquired English Patent No. 50 technology and lack of understanding of the principles of
which included an invention to make Boates, Shippes and propulsion before the mid-nineteenth century, these waterjet
Marine Waterjet Propulsion 275

propulsors were unable to compete with paddle wheels and, to 1987, the Jetfoil passenger hydrofoil craft (Figure 5), which
later, propellers. uses Rocketdyne waterjet propulsors.
-;;il i:il
Figure 2. Yamato I Magneto-Hydrodynamic Waterjet Ship

,u,O'ClffG SPIP~L FOR P,~ISI/'IG
~AT~.R
Figure 1. Leonardo da Vinci's Drawings
During the mid-19th century there was considerable

interest in waterjet propulsion for naval vessels. Comparative
trials were made by the British Admiralty and the Swedish
Government of waterjet and propeller-driven ships. Evidently,
these trials favored propellers, probably because of the pump
technology available at the time. Saunders, 1957 [4] and
Brandau, 1967 [5] give several sources for detailed historical
reviews, some of which are included in the bibliography (see
also Appendix A).
The idea of waterjet propulsion without a mechanical
device in contact with the water has persisted nonetheless.
From time-to-time propulsors such as the hydrorocket have
surfaced and the water piston propulsor (WPP) was under
development until recently for overwater propulsion of tracked J
amphibians (Swanek, et al, 1989 [3]). Even more recently, the
Japanese have released some details of their new magneto-
3hydro-dynamically (MHD) propelled ship. The propulsors on
this vessel are a type of waterjet in which the seawater in the
A
pump is acted upon directly by electromagnetic forces. The
necessary magnetic field and current are only made possible in
electrical machinery of practical size by the use of supercon-
ductivity. So far, the Japanese ship, Yamato I, 185 tons (Figure J
2) has been demonstrated at only 6 kt.
More conventional pumps have been used successfully in
advanced marine vehicles since the 1950's. The hydrofoil ship
Tucumari (PGH 2) 1968 - 1972 was waterjet propelled (Figure
3). This was followed by the PHM which was initiated by a
joint effort of NATO (Figure 4), but only, six of these waterjet-
propelled ships were eventually built, all for the U.S. Navy.
They are presently stationed in Key West, Florida. The PHM
uses an LM 2500 gas turbine, two MTU diesel engines and
waterjet pumps designed by Aerojet Liquid Rocket Company
(ALRC). Following the PHM program, Boeing built, from 1974 Figure 3. Tucumcari
276 Marine Waterjet Propulsion

U.S. Navy's large SES program (2K/3KSES) 1973 to 1978.
These pumps were designed to absorb 40,000 hp at about 25 kt
up to 80+ kt and they remain the highest-powered pumps
designed so far. The Rocketdyne pump, which was finally
selected for production, was in the early stages of manufacture
(Figure 7) when the program was cancelled in 1978.
- !
Figure 6. U.S. Navy's SES-IOOA
\ i ii
Figure 4. PHM
Figure 7. Rocketdyne Powerjet 46
Abroad, KaMeWa in Sweden, and Riva Calzoni in Italy,

developed high-flow, low head pumps for commercial and
military applications at more modest speeds beginning in 1968
and 1932, respectively. KaMeWa has led the world in the
production of a standard series of highly reliable and efficient
waterjet propulsors of ever increasing maximum size. At
present, their largest pump is the 180 SH, which is matched to
the LM 2500 gas turbine engine rated at 30,000 slap. There are
over 200 ships in service with KaMeWa waterjets, including the
U.S. Navy's SES 200 (Figure 8) which was converted to
waterjet propulsion in 1990. Very recently the Destriero
captured the transatlantic speed record with three KaMeWa
waterjets. Riva Calzoni have produced a series of successful
customized designs and many standard pumps, including
Figure 5. Jetfoil Passenger Hydrofoil Craft propulsors for the Azimut Atlantic Challenger.
KaMeWa's success has been an inspiration to other
manufacturers around the world.
Another type of advanced marine craft, the Aerojet-built In the United Kingdom, Parker has manufactured
U.S. Navy SES-100A testcraft (Figure 6) used waterjet waterjets of increasing size since 1972.
propulsion. This vessel, built in 1973, achieved a speed of 78 Hamilton in New Zealand have pursued their own
kt. independent approach since 1954 and are now beginning to
Very large waterjet propulsors were designed, and produce large waterjet propulsors, in addition to their numerous
developed by Rocketdyne and Aerojet at model scale, for the smaller size models.

inevitable conclusion that waterjets cannot compete with
propellers in propulsion efficiency with pumps of realistic size.
Concurrently with the renewed interest in waterjet
propulsion over the last 20 years, more sophisticated treatments
have evolved which show that the elementary theory may be
misleading because it fails to take into account secondary effects
which can have a marked influence on propulsive efficiency and
the comparison with propellers.
Nevertheless, simple momentum theory, judiciously
applied, provides valuable insights concerning practical marine
i; Figure 8. U.S. Navy's SES 200

m waterjet propulsion.
The simple theory will be presented first and will then be
modified to account for various practical effects. Some of the
alternative formulations which have been published will also be
given.
In the smaller sizes, for recreational and fishing boats, 3.1 Gross Thrust
thousands of waterjets are produced annually by Castoldijet in
Italy, Berkeley in the United States, and others elsewhere. This is the force experienced by the propulsor due to the
In Japan, Kawasaki Heavy Industries (KHI) produces the momentum flux from the nozzle. It is identified with the thrust
former Rocketdyne waterjet pumps for the Jetfoil which they which propels marine animals, such as squid, that use intermit-
acquired from Boeing in about 1987. Recently, Mitsubishi tent waterjet propulsion for rapid movement.
Heavy Industries (MHI have introduced two new waterjet
To = m Vj
propulsors.
where T~ = gross thrust
Japanese manufacturers also produce large numbers of
small waterjets for water scooters and similar recreational craft. mn = nozzle mass flow rate
Recent developments in the USA are discussed later in Vj = jet velocity
the paper. Present manufacturers include American Hydrojet
Corporation and North American Marine Jet.
In simple one-dimensional theory, Vj is assumed to be constant
Table 1 summarizes the main advantages of waterjet
propulsion. and uniform across the jet perpendicular to the flow. Of course,
this is not strictly true in practice and although the effects of
Table 1. The Advantages of Waterjet Propulsion non-uniformity on thrust are small, they contribute to the
break-up of the jet after leaving the nozzle.
• Reduced draft (depending on hull type) Figures 9a and 9b show actual waterjet propulsors under
Elimination of appendages test. Notice that the jet, which appears like a solid glass rod
• Absence of appendage drag near the vena contracta quickly assumes a white appearance and
• Improved maneuverability seems to grow in diameter. This is deceptive. Shearing forces
• No reversing gear needed with air assisted by non-uniformity of the flow tear off the outer
• Less wear and tear on engines and transmissions* layers of the jet to form an expanding cone of heavy spray.
• Essentially constant torque over ship speed range at a given Nevertheless, the potential core of the jet persists for some
power distance downstream, bending under the influence of gravity,
• Enables full power to be applied at low ship speed until it plunges back into the sea when operating in a ship.
• Improved braking especially at speed Gross thrust is of great importance to marine propulsor
Reduced stopping distance operation apart from its obvious primary purpose. Deflection of
• Reduced power requirements at high speed (over 25 kt) the jet produces side force for steering
• Reduced fuel consumption for high speed cruise F = T G sin 0 assuming a horizontal jet.
• Reduced vibration
• Reduced inboard noise Small deflections produce very large side forces with
• Greatly reduced underwater noise almost negligible reduction in forward thrust since loss of
Reduced weight in some cases forward thrust is given by:
Technical Challenges A T = T G (1 - cos 0)
• Fuel consumption at low speed (below 20 kt)
For example, if the jet is deflected 11.5 ° horizontally, the ideal
• Air ingestion in a seaway on some hull types
side force is 20% of the gross thrust while the loss of forward
*Providing broaching (air ingestion) does not occur. gross thrust is only about 2%. Since net thrust is typically about
40% of gross thrust, the side force represents 50% of the net
3.0 T H E O R Y AND P E R F O R M A N C E A N A L Y S I S thrust which propels the ship, for a loss of net thrust of 5%.
This is what gives waterjet ships their excellent steering
In the past, treatments of waterjet theory have begun and capabilities. Of course, there is some loss due to the steering
ended with elementary momentum theory leading to the sleeve. This is addressed in Section 5.0.

Usually, m i = mj unless water is bled from the inlet for
some other purpose, so,
TN = m (Vj - V).
The work done in propelling the ship is
WD = T V = mV (Vj-V).
Assuming perfect inlet energy recovery and neglecting

elevation of the nozzle above sea level, the useful energy added
by the pump is equal to the change of energy of the water
passing through the system. This is given by"
AE = 1/2 m Vj2- 1/2 m i V 2
orif I~ = mi ,
AE = 1/2 m (Viz - V,2).
16,000 SHAFT H©RSEPOWER
Then the propulsion efficiency is given by the ratio of useful
NOZZLE DIAMETER 17.52 INCHES
work done on the ship to the useful work done by the pump
JET DIAMETER ATVENACONTRACTA = 14.72 INCHES
1]i = T V / AE, this is known as the jet
JET DIAMETER 0.841 efficiency
KO-L~TKEETKigETEF[
JET DIVERGENCE HALF ANGLE 5 DEGREES or rlj = m(Vj-V)V/1/2m(Vj 2-v,5
2V s
Figure 9a. Large Waterjet Propulsor Test hence Tlj = ~ (3-1)
The ratio of ship speed to jet velocity is of interest:
V / Vj (3-2)
Dividing by Vj:
2IX (3-3)
YlJ = 1 +Ix
Alternatively, dividing (2)by V
2 (3-4)
rlj =
where a = Vj/V = ~-1 (3-5)
The same expressions can be obtained by considering the kinetic

Figure 9b. Model Waterjet Propulsor Test
energy of the jet (based on its velocity relative to the sea) as a
loss. Then:
3.2 M o m e n t u m Drag Useful work done on the ship
TlJ = Useful work on the ship + energy lost in the jet
Water entering the waterjet inlet is accelerated to ship
speed and, consequently, exerts a momentum drag on the ship.
m V s (Vj - Vs)
Dm = m V
TIJ = m V s ( V j - V s ) +½m(Vj-Vs) 2
where D m = momentum drag
mi = inlet mass flow rate
Vs = ship speed Vs
T~j =
Consequently, the net thrust on the ship is given by
Vs+½ (vj-Vs)
TN = TG- Dm or
TN = m n Vi - m i V s

2 Vs = 2 ix (1 - iX) (3-6)
Tii = and Tij 1 - IX2 ( i - ~)
It will be found that Tij now reaches a maximum depending on

Tij = ~ 2IX as before.
the value of ~ as shown in Figure 10.
The ideal jet efficiency can also be expressed in thrust 1.0

coefficient form. For a propeller:.
2
Tit = )'8
1 +'~ +C T
T
where CT = (3-5A) 3-6
1 V2
~-pAp s
i
For a waterjet: )'4
4
~j =
3 +'~fl+2C T
) . 2 /
where CT = 1
~pAj
T
V2
s
(3-5B) \
0
0 0.2 0-4 ,H 0-6 O'B 1.0
AI, is the propeller disc area and Aj is the jet area at the
vana contracta, usually coincident with the nozzle area Aw Figure 10. Jet Efficiency With Losses
Equation (3-3) is the starting point for less simple
expressions which include various losses and other effects. It
has been derived here in more detail than is usually given to In the event that the losses are equal to the ideal inlet
avoid possible subsequent confusion. energy, the maximum efficiency attainable is 50%.
In the ideal case, this is equivalent to taking water into
3.3 Inclusion of Losses the ship through simple side or bottom openings incapable of
energy recovery, so that:
It can be seen that the jet efficiency increases monotoni- TIj = 2 v, (Vj - V,) / Vj 2
cally as IX tends to unity. When the ship speed matches the jet
velocity, the ideal jet efficiency is 100% but no thrust is or TIj = 2IX(I-IX) (3-7)
developed, therefore, the ship would have to have zero drag
which is not possible. Here TIj is a maximum when IX = 0.5 and Tiim~ = 0.5
Many treatments now introduce a loss coefficient which also.
is defined as ~ = total energy loss as a proportion of the ideal This is as far as the analysis of waterjets is taken in many
inlet energy. treatments, including the latest edition of Principles of Naval
Accordingly, the energy which must be added to the Architecture (1988), where it is stated that according to Kim [6],
water is now: the loss factor (~) excluding the pump efficiency will vary
between 0.25 and 0.5 for well-designed systems, and 0.5 to 1.00
Z~ = 1/2 m (vjZ- V 2) + ~ 1/2 m V = for poorly designed ones. A table is presented showing
examples of jet propulsion efficiencies and propulsive coeffi-
= ,/2 mtV?- V,'+ V,'] cient according to Kim, where propulsive coefficient:
PC = TIj x Tip
= 1/2 m [Vj 2- V 2 (1- ;)1
and Tip = pump efficiency.
and the jet efficiency
The table is reproduced in Appendix B for reference, but
Tij = T V /AE' it must be emphasized that it no longer represents the state-of-
the-art.
therefore:
m (Vj- V) V, / 1/2 m (Vi2- V, 2 (1- ~)) 3.4 More Detailed Expressions for Jet Efficiency
Tij
For the following treatment, reference should be made to
or Tij 2 V (Vj- V=) / (Vj2 - V 2 (1- ~)) the def'mltion sketch shown in Figure 11.

3.5 Inlet Losses
v-NOZZLE EXIT OR . . . . . . .
~VENA CON'[RACTA ~t The inlet losses can be expressed, as previously, as a

\ AsL ------. Ap
~ _ PUMP SHAFT ~ p fraction of the available inlet energy, so the energy recovered is
given by:
h ] I~ [ -L hp'"-,~J-.t""k~i
E. = (38)
Here the loss factor ~ includes only the inlet recovery losses.
<'> "> The inlet efficiency may be def'med as:
L_NOZZLE STREAHLINE
-JET 1]i = (1- ~) (3-9)
V = ship speed, ft.s (m/s) - Station (0)

3.6 Nozzle Losses
Vw = effective velocity of approach, V, = (1 - w) V -
(Station (1) The nozzle losses can be taken into account by a nozzle
Vi = inlet velocity at inlet plane - Station (2) efficiency,
Vp = velocity at pump inlet plane - Station (3) • Kinetic energy in~et
V,t = velocity at stator inlet - Station (4) TI~ = Energy supplied by pump at the nozzle inlet
V = velocity at nozzle exit plane - Station (5)
Energy supplied by pump = 1/2 m Vj2/ll. and nozzle loss
Vj = velocity at jet vena contracta - Station (6)
Ac = capture area - Station 0
= (1- ¢1.) 1/2 m vj 2 (3-10)
A1 = flow area at beginning of inlet ramp - Station (1)
Ai = inlet area - Station (2)
= V 1 / 2 ~ i v "J (3-11)
Ap = pump face inlet area - Station (3)
A,t = stator inlet area - Station (4) hence, energy supplied to the nozzle:
Ani = nozzle inlet area - Station (5)
E = 1/2mVj 2 + ~ l / 2 m v j 2
An = nozzle exit area - Station (6)
Aj = jet csa at the vena contzacta - Station (7)
or E = (1 + ~/) 1/2 m Vjz (3-12)
hp = height of pump inlet centerline above ambient sea
level - Station (3)
3.7 Effect of Nozzle Elevation
hi = height of jet centerline above sea level at the vena
contracta - Station (7) In addition to the kinetic energy of the jet and the various
0C = pump center inclination to horizontal losses, the pump must provide the energy to lift the water to the
~) = p u m p centerline horizontal inclination to ship center- height of the jet centerline. This additional work is given by:
line
= inlet duct vertical inclination to pump centerline W. = m g hi (3-13)
shp = shaft horsepower delivered to the pump where g = gravitational accel, 32.174 ft/s2
bhp = brake horsepower developed by the engine
3.8 Combined Losses and Elevation Term
Figure 11. General Waterjet Propulsor Definition Sketch
Useful work done on the ship:
WD = TV,
For the following analysis the simplifying assumptions
made initially are: = (vj-v)v,
• Pump vertical and horizontal inclinations are zero ( O~ =
= ~ (vj-(I- w) v,)v
¢=0)
• The vena contracta occurs at the nozzle exit plane, (Aj =
Useful energy supplied by the pump:
An)
E" = m [1/2 Vj2 (1 + ~ f ) - 1/2 ~ i (V, 2) + g hj]
• The effective velocity of approach is uniform and is given
by V° = (1 - w) V where (1 - w) is the effective Taylor = m/2 [Vj2 (1 + ~ ) - T l i ( 1 - w ) 2 V , 2 + 2 g hj]
wake factor at station (0) (3-14)
• Pump losses include stator losses
• The external inlet drag is neglected Actual energy supplied to the pump:
• Pressure changes on the hull due to the presence of the E .... = rn/2TIp[Vi2(l+'t[O-~i(1 -w)2V,2+2ghj]
inlet are neglected.
(3-15)
where TI. = pump efficiency.

The jet efficiency is given by:
W^~E
~
F^C10n
Tii = T V / E .... olo
o.g3s
9 C - O.~S
=
Vs) Vs e
c - 0.9~
rh r~ "1
(l-w) 2 V2+ 2ghJl
" / . " / "/ " " / ~ ~ c - o 90
JET
-2- t v j - ( 1 + wit) - r l i s ~r r l o ~ l c v
ql s
c - 0"9
C-LO
.............
Divide top and bottom by Vj2, m/2 and let the jet velocity ratio
I.t v/vj.
=
5 0,5S 0,6 0 es o.l o.rs
2 (0. (1 w) 0'2)
- -
aET vttOCttV nmto. it
~J = 2g hj (3-16)
l+~-rli ( l - w ) 2 0'2 + W qj = 2~, {1 -(1 -w) ~.t} 2g77i VCHErIEjt = Vs/V j
1 4 ~l~-It2- (1 - z ) ( l - w ) 2 4. . . . . .
v/-
OI'C = llj X l ] p x q i . R x (1 - t)
2g(1-(1-w) lt)
1 + W- Th ( 1 - w) 2 la2 + 2g hj
Figure 12. The Effect on Jet Efficiency of Wake Ingestion
TIj = 2 0' (1 - (1 - w) 0') 2g hj (3-17) Usually, the predicted reststance of a ship is based on
I+V-(1-~) ( l - w ) 2 p . 2 + V---~. scaled model test data or analysis for the bare hull which is used
to obtain the effective horsepower.
J
ehp = R T V/550 (3-21)
where R T = the towed resistance or equivalent.
If the jet velocity ratio is defined as the ratio of wake
velocity to jet velocity instead of ship speed to jet velocity then: The actual thrust required will generally differ, as is the
= V / V j = ( 1 - w ) V /Vj (3-18) case with propellers.
T = ILr/(1 - t) (3-22)
Substituting ~/(1 - w) for I.L in equation (3-17) gives: where T = required thrust
1 2 0. (1 - 0') and t = thrust deduction fraction
rlj (3-19) or (1 - t) = Taylor thrust deduction factor.
I+V-(1-~) 0'2+
V." Substituting in equation (3-20) for T gives:
J
shp = RT V J55011o (1 - t) (3-23)
where ~ = (1 - w) VdtVj. Conversely, if the shaft horsepower available is known,
the waterjet thrust can be calculated:
This form is preferred by some authors since when combined T = TID 550 shp/V (3-24)
with the thrust deduction factor it leads to a form of the but the effective thrust for comparison with the predicted drag of
propulsive efficiency analogous to that used for propellers, in the ship is:
which TIj is somewhat analogous to the open water efficiency of
To = RT
a propeller. but R T = (1 - t) T (3-25)
Also TID = TIj Tip So, T© = T (1 - t) 11o 550 shp/V
where TID = TIj ° Tip (3-26)
Where "riD is the quasi-propulsive coefficient.
The pump shaft horsepower will then be given by: Here, Tip is the pump efficiency as installed in the waterjet
propulsor. However, the waterjet inlet produces flow distortions
shp = TV/550TID (3-20)
at the pump face so that the pump performance differs from that
Figure 12 shows the results of plotting Equation 3-17 (or which is obtained in uniform flow. If the pump efficiency is
3-19) for a range of wake factors for small fixed values of lq/ known from a uniform flow performance test, then:
and hj. Tip = Tip. 11, (3-27)
It will be appreciated that high values of jet efficiency are where Tip = installed pump efficiency
possible with wake factors approaching 0.9 or lower. However, and TI, = relative rotational efficiency.
it must be remembered that a significant wake fraction, w, Usually the relative rotational efficiency is close to unity.
represents drag on the ship which contributes to the required
thrust in the first place.

Unless the pump is direct-coupled to the engine, there rh(Vj-(1-w) Vs-CDi½Vs/IVR(1-w)) Vs
will be power transmission losses. Even if direct-coupled, there
may be additional bearing and seal losses where the pump shaft
penetrates the inlet duct, which are not already included in the '2- V ( l + ~ g ) - r l i ( 1 - w ) 2 V2s + 2 g h j ]
pump efficiency.
A transmission efficiency is deemed as:
lit = shp / bhp (3-28) Divide top and bottom by r~/'2, VJz and let ~ = V/Vj as before,
then:
Then T. = ( 1 - t) TIj • Tlp" TI," Tlt550bhp/V,.
2 (Ix- (1- w) Ix2) _ CDi IX2/IVR (1- w)
The overall propulsive coefficient may be deemed as: TIi = 2g hj
l+R/-'qi(1-w) 2Ix2 + V--~. '
OPC = TcV/550bhp whereT,=R r
J
SO, OPC = ( 1 - t ) ~ j - l i p . 1 1 - Tit (3-29).
It is found that t can be positive or negative depending on Ix (2 (1 - (1 - w) g)) - CDi Ix / IVR (1 - w)

the ship speed and the type of hull. Generally, t is positive at Tlj = 2g hj (3-34)
speeds below about 25 kt but may be negative at higher speeds. l+xg-(1-~) ( l - w ) 2 Ix2+ V--~.
This accounts for the apparently optimistic values of propulsive J
efficiency in some installations. The subject of propulsor-hull
interactions is discussed further in a later paragraph.
Before leaving the expression for jet efficiency, given in orif = (1 - w) V / Vj
equation (3-17), the effect of removing some of the simplifying
assumptions initially made will be examined. 1 u (2(1-ix) -cDi Ix/rVR)
With the pump shaft inclination angles O~ and 1~ included = " 2g hj (3-35)
the gross thrust becomes: ~1~ ~ l+v-O-OIx2+ v--7-.
To = m Vj cos Ct cos (~ J
2 I.t (cos ct cos 0 - (1 - w) It) (3-30)

and TIj =
1 + V - ( 1 - ~ ) ( l - w ) 2 I.t + [email protected] Finally, from Equations 3-29 and 3-35 the overall propulsive
coefficient becomes:
J
OPC = (1- t) TIp. TI • Tlt'TIj
If the vena contracta occurs downstream from the nozzle I.t (2 (1-1a.) - C D i g / I V R )
hj will not be identical to h n but the difference will be negligible =
2g hj
for practical purposes. 1 + V - ( 1 - ~ ) IX2 + [email protected]
A term may be included to account for inlet drag. Inlet J
drag is a function of the drop fraction of the inlet lip below the
ship's baseline, the inlet area A i and the local velocity. Inlet
drag:
The quantity ( l_-l_-27@wt)-wt is identified with the hull efficiency of
Di = CDi A 1 1/2 p Vw2 (3-3 l)
where Cni = inlet drag coefficient propeller theory, although this can be misleading. Thus,
and Vw = wake velocity, (1 - w) V (3-32) oPc = ~..~.rl,.rl,.
An inlet velocity ratio is defined as: Ix ( 2 (1 - Ix) - CDi Ix / I V R ) (3-35A)
2g hj
IVR = Vi/V. (3-33) I+V-(1-~) IX2+ V--V--2-.-
'
J
so, Vi = IVR • V = IVR (1 - w) V
therefore V, = V i / IVR (1 - w)
whereT1H = so calledhull efficiency. ( ~ ) •
and Di = CDi 1/2 p A I V i V, / IVR (1 - w)
or Di = CDi 1/2 m V / IVR (1 - w) If the inlet is truly flush with the hull and there is no
external inlet faking or protruding sole plate, the inlet drag tends
to zero and is frequently disregarded in which case equations
Hence, Tlj = (T - Di) V / E"
3-34 and 3-35 revert to the previous forms of equations 3-17 and
3-19.

An alternative formulation for rij is given by Barr and
Etter (1975) which combines the inlet losses and elevation term
but recognizes the effect of non-uniform velocity on the
,(-)
'-W
rij (3-40)
integrated flow, momentum and energy terms. Initially,
riJ =
2 [~I-K+H*-
H*
1]
(3-36)
E' -"-" (vo'l
where:
where K = k + 2 g A h / V. 2 (3-37) V = average velocity of the ingested flow
and K = k - 2g Ah./V *z (should be + not -2g Ahj/V .2)
H* = 2 g H / V. 2 Vm = momentum velocity of the ingested flow
V* = energy velocity of the ingested flow.
where the pump head,
The overall efficiency is then given as:
H = ~
v - (1 - k)
vo
~ + Ah (3-38) rio,,, = rlp-r i , - ri, r i . - rlj
(in the present notation)
and (1 - k) = (1 - 4) = TIt = inlet recovery efficiency
1-t
where rin - 1 - w ' the so-called hull efficiency.
V, = V, = free stream velocity, i.e., ship speed, ft/s
= hj = height of the jet at the vena contracta. 1

As previously demonstrated, ~ is part of the jet
They then substitute the energy velocity and the efficiency expression, so strictly:
momentum velocity of the ingested flow which includes part of
the boundary layer on the hull, i.e., wake, and re-inu-oduce the
1 ~i 1 - --~j
vo)
nozzle efficiency to give: (3-41)
-T~-
: FI - ( 1 - K ) ( v , V ]j
~,--~-j)
'~J ---- H* L
(3-39)
where:
V* = energy velocity of the incoming flow
momentum velocity of the incoming flow However, the value of the overall efficiency will not be affected
g m =
by this distinction.
CD| ~ inlet drag coefficient from
Of more importance is the fact that:
Di = CDt A i 1/'2 p Vo2 K = k + 2 g A h i / V .2 [8]
ivr = inlet velocity ratio V/Vo, i.e., V.r/Vs
3.9 Waterjet-Hull Interaction Effects
Vi = inlet flow vdocity through the inlet area A i (Figure
11). The influence of the hull boundary layer (wake) ingested
by the inlet has already been considered. There are, however,
The above expression for Ylj may be shown to be other effects.
identical to the previous expression of equations 3-17 and 3-19 Pressure measurements made on the hull in the vicinity of
if~)* = 13= = (1 - w) V a n d U . = V,. the waterjet inlet have shown that there is a net lifting force on
Expressions such as equation 3-39 are of academic the stem of the vessel [9]. The opposite is the case for a
interest because they illustrate the effects of including actual propeller, where the pressure near the stern is reduced by the
physical effects of the boundary layer or wake which constitute influence of the propeller in-flow field. Svensson [9] has
non-uniformity of flow velocity. However, data for these proposed a pressure coefficient, Cp. which appears on the
effects are seldom available to the naval architect and the bottom line of the efficiency expression as a negative quantity,
simplified expr~sions such as equations 3-17 and 3-19 are to be thus reducing the hydrodynamic power required and hence, the
preferred for parametric design analysis. required shaft power. The pressure coefficient is det-med as:
Ultimately, the propulsor performance must be provided Cp = (P, - pg hl) / 1/2 p V, z (3-42)
by the waterjet manufacturer when the design conditions, where P, = static pressure in front of the irflet
hullform and pump installation and prime mover characteristics
hi = depth of the inlet
are known.
At the 18th International Towing Tank Conference V = ship speed, as before.
(lq'q'C) 1987, the theoretical model of Etter et al. [8] was Cp is said to be on the order of 0.1.
recommended from which it was deduced that the jet efficiency It might reasonably be expected that P would be lower
is given by: than the static pressure at the inlet depth (pghi) due to the
influence of the inlet potential flow field and indeed this is the
case when the ship is not underway, for instance, in a bollard

pull test, and at very low ship speed if high enough power is
applied to the pump. The inlet may be regarded as a patch of
sinks [9] on the bottom of the hull. The flow will enter from all
directions as shown by Figure 13a. However, the reduction in
static pressure on the hull around the inlet diminishes very
rapidly indeed, as anyone who has investigated the pressure
field on a surface surrounding a strong sink will confirm.
-- ~-/_.,~ -~,o~ - / \
-t
Figure 14. Transom-Mounted Waterjet Propulsor
At usual speeds the pressure around a waterjet is found to

increase relative to the value it would have if there were no
R<I.O
waterjet inlet (as shown by Figure 15. When the term for Cp is
IVE ~ o / - , . , IVR • 1.0
included the propulsion efficiency becomes, according to
Vs - O LOW VS HIGH Vs Svensson:
11| (hi (el
11D=rlp'rl,"
(,tT=-~ / "
2 IS (1 - p) (3-43)
1+¥-(1-0 p2+~_~/_~
V." (1 - w)"
EOUIPOTEPtTIAL SURFACES VELOCITY PROFILE
J
PS < pgh
Figure 13. Notional Waterjet Inlet Flow Patterns
B " 1.5D
As ship speed increases, the flow enters the inlet
predominantly from the forward direction as indicated in Figure
13b. The inlet velocity ratio is greater than unity.
At high speed the capture area of the stream tube entering
the inlet may be less than that of the inlet itself. Some external
diffusion may occur in front of the inlet so that there is a rise of
static pressure at the inlet eptranee, and the inlet velocity ratio is
less than one. BOTTO~'-~
When a vessel starts to move forward under the influence PLATING ~
of its propulsor, there is an increase of draft at the stem due to B - D
the reduction of smile pressure caused by the acceleration of the
boundary layer. This is very noticeable in the ease of planing or /
semi-planing hulls. The so-called phenomenon of suck-down is INLET / INLET VELOCITY RATIO (Ivr)
/
complex and is discussed at length by Payne [19]. The action of / D - NOMINAL INLET DIAMETER ( S )
a propeller in-flow field tends to increase the effect, but a
/s B = BREADTH OF BOTTOM PLATING
waterjet inlet removes part of the boundary layer which is
replaced behind the inlet with relatively undisturbed water at a
static pressure more nearly corresponding to its depth. This is
why the inlet opening should not be too close to the transom,
and why an extension of planing surface aft has been recom- Figure 15. Lifting Force Due to Pressure on the Bottom
mended in some eases. Plating and Inlet for a KaMeWa Waterjet
A recent transom-mounted waterjet propulsor, Figure 14, Installation
has virtually no bottom surface aft of the inlet opening. Since
an inlet close to the transom means a short inlet duct and less
carried water, there is a trade-off here. It is implicit in this expression that the area acted on by
Once a craft reaches planing speed, the hull immersion at the pressure difference P - pgh i is proportioned to the inlet
the stern is drastically reduced and the trim angle returns to a area. Cp is analogous to CDt and the form of the two terms is
value approaching zero. However, many waterjet propelled
similar when Cvl is referenced to ship speed instead of V,.
vessels do not even approach planing speed.
Note that here ( l l ~ - w t ) is not designated as t h e h u l l
efficiency, also, no inlet drag term is included, and ~D differs

from OPC in that the transmission efficiency is not included.
In considering inlet drag it must be remembered that the
draft of any modification made to the hull due the waterjet ~ T ~
i l
installation, such as inlet fences or kflet fairings, must be A~ea ? C~9c~e ;,-~'aa
debited against thrust and hence, reduce OPC. Aj Pz
,
A
C&O
A general expression for overall propulsive coefficient
including all the terms discussed so far, is given by:
Z~c~e~se k~ i
P=essure a: , ~.-"'~
~cogeLLer O ~ s c Ip
I.I. (2 (cos Ct cos O - I-t) - CD, IMIVR) j 9 I

(3-44)
2g hj _ vj = va (L ~ b) I'~. = va(L - a)
1+V-(I-;) gt2 + - - ~ j (l_w)Z
Figure 16. Simple Momentum Theory for an Open Propeller
(Actuator Disk) [2]
Other interaction effects include changes to the flow
pattern around the hull which may influence wave-making and A figure of merit may be det'med as:
frictional resistance. Changes to the lift forces on the hull
influence trim angle and draft. In the inlet duct, change of ideal power
FM = actual power
direction of the flow produces a downward force on the forward
lower part of the duct and an approximately equal upward force 1
m -
on the upper aft par~ of the duct. The net result is a couple actual e f f i c i e n c y
which acts in the bow-down sense. It has been suggested that or FM = 550 shp = ideal e f f i c i e n c y
trim may be adjusted by angling the jets slighdy upwards.
The weight of on-board water may be regarded as a loss
of buoyancy of the ship, or as an additional weight included This expression is useful because it holds when the ship velocity
with the weight of the waterjet propulsors. The combined is zero, i.e., V = 0, in which case:
weight may be more or less than that of the equivalent propeller FM = 1/2 m Vj2/550 shp
system, but generally more, with a corresponding increase of
FM = 1/2 [3 Aj Vj3 / 550 shp.
equivalent draft.
Another effect is the aspect ratio (width to height) of the Useful tables and plots of propeller performance for preliminary
inlet. A wide inlet is to be preferred as more of the boundary sizing may be made with this approach (Allison, [2]). For a
layer is ingested (Purnell, 1976) [17], (Terwisga, 1991) [20]. propeller, Aj varies with V and is equal to half the propeller
Terwisga, 1991, has attempted a rigorous analysis of
waterjet-hull interaction effects on thrust and propulsive disc area in the static case, i.e.:
efficiency. He states that decreases in propulsive efficiency in Aj = 0.5 A v when V = 0
excess of 20% have been measured due to interaction. or Aj = 0.5 Ap (1 + V/Vj) when V ~: 0
i.e., Aj = 0 5 Ap (1 + It) where IJ. = V/Vj as before.
3.10 Comparison With Propellers
Also, the upstream capture area is given by:
Comparison of waterjet pmpulsors with propellers must A, = Aj V / V = A¢/~.
commence with application of the simple momentum theory.
Figure 16 is a definitive sketch for an open propeller in uniform For a given operating condition, an ideal propeller (in
f low. inviscid flow) could be surrounded by a duct which followed the
As for the waterjet propulsor, ideally, dividing stream lines. Such a duct would constitute the
equivalent of a waterjet propulsor.
T = m (V i - V ) where V = V (no wake) For a realistic set of conditions it would be found that the
nozzle area was far larger as a proportion of the inlet area and
The power addeA to the slipstream is: impeller (in this case, propeller) disc area than is usually the
E = 1/2 m (Vj 2 - V z) case for practical waterjets. This is apparent in the following
example.
Hence, the ideal efficiency is:
Consider a thrust requirement of 10,000 lbf at a speed of
m (Vj - Vs) V s 35 kt (59 ft/s). Compare the power required and the physical
1"]~ = TV/P=
1 size of a propeller and a waterjet if the propeller has a figure of
merit of 0.7 and the waterjet has an inlet recovery of 0.83 with a
2 Vs pump efficiency of 0.9. Neglect wake effects and other losses.
or lql,t = ~ as before. Assume p = 2 slug/fr3.

To avoid cavitation, the propeller loading at this speed 1-t
should not exceed 1000 lbf/ft2 (approximately).

Tip TipTI,T - w "
Disc area required = 10 ft2 (c. 3.5 ft dia) 21~ ( 1 - b t )
Required jet velocity = 67 ft/s 1+v-(1-0.2 2ghj Cp.2
Jet velocity ratio, ~L = 0.88
+ ~ (1 - w) ~
Jet area, Aj = 9.4 ft2 J
Jet diameter = 3.46 ft
Ideal efficiency, Tiia = 2 x 59/(67 + 59) = 0.937
1.02
Actual efficiency, Tid = 0.7 X 0.937 = 0.656 TiD ---- 0.9 • 0.99 • ~ .
HP required = TVJ550 Tip
2 x 0.545 (1 - 0.545)
= 10,000 x 59/550 x 0.656
= 1635. 1.01 - 0.83 x 0.5452 + 2 x 32.174 x 2.16 0.07 x 0.5452
102.82 0.952
For the waterjet, assuming a maximum (flange) diameter
for a mixed flow pump not exceeding the diameter of the
propeller, i.e., 3.5 ft, a typical nozzle size would be 13.5 in. dia. 1.02 0.496
TIP = 0.9 • 0.99 • ~ ° 1.01 -- 0.247 + 0.013 -- 0.023
Nozzle area (jet area) = 1.0 ft2
Tit) = 0.630 and TIj = Tij1]p Ti, (1 - t) = 0.6933
Thrust required = p Aj Vi2- p Aj Vj V
(compare Table 2)
10,000 = 2 x 1.0 V. 2 - 2 x 1.0 x 59 V. New hp required is 9500 x 59/550 x 0.63 = 1617.
J J
V. = 106 ft/s Thus, the waterjet ship now requires slightly less power
J
than the propeller ship based on the assumptions made. This is
1.1. = 59/106 = 0.556 only a hypothetical example and no attempt has been made to
2 Ix (1 - i t ) optimize either propeller or waterjet as would be done in a real
TIj = design.
1 - (1 - {) I*2
Several comparisons of waterjet and propeller-powered
2 x 0.556 (1 - 0.556) ships have been published (Svensson, 1989) [9].
TIj = Figure 17 is taken from [9], Figure 18 is taken from [10]
1 - 0.83 x 0.5562
and Figure 19 is taken from [24].
Tlj = 0.664
TO~Oi POw,~e
PO { ~ w l
TIP = TIj Tip = 0.9 x 0.664 = 0.598 goOO
2~t 2 x 0.556
'~id = l+lx = ~ = 0.715
3000
hp = 1794
Although, as expected, the horsepower required by the propeller 7000
is less than that for the waterjet in this simple example, the
difference will be much less when installation effects are taken
into account, including reduction of thrust required due to the 6000
absence of appendage drag for the waterjet ship, change of
displacement including on-beard water weight (even though this
is a penalty for the waterjet), and wake ingestion, etc. as shown 5000
by the following example.
Assume a wake factor (1 - w) = 0.95 and a thrust
deduction factor (1 - t) = 1.02. ~000
Assume a nozzle efficiency of 0.99 and a relative
rotational efficiency of 0.99, with a nozzle elevation of 26 in.,
and hull pressure coefficient of 0.07. 3000
Without appendages and with the waterjet installed, the
required thrust might be 9500 lbf.
ZOO0
Vj = 102.8 ft/s 20 30 ~0 Knots
= (1 - w) V f q j = 0.95 x 59/102.8 = 0.545 Figure 17. Power Required for a 210 Tonnes Triple Unit
Naval Vessel. Comparison Between Waterjet
and Propeller Propulsion [ 15 ]

It is provided here to show how close the comparison of
IOT^L SHAFT propellers and waterjets can be when all effects are considered.
POWE~ ~ - kW
Recent trials of the U.S. Navy's SES-200 (Figure 8) are
SOOq F ~ " expected to provide useful comparisons of waterjet and
PROP . . propeller operation as the ship was converted in 1989/90 from
~000 - - - -
propeller propulsion to waterjet propulsion, with larger engines.
~000 ~ ~ ~ Ultimately, the fuel consumption at various craft speeds and
equal displacement, corrected for differences in specific fuel
2000 -- consumption of the old and new engines will provide a measure
WAT~
of the relative waterjet effectiveness. Figure 20 taken from
~00o
Moore, 1992, [11] shows the resistance-speed curve of the
J
0 i .' ----~ SES-200 with a notional waterjet thrust curve added. It is stated
~0 Z0 30 ~0 50 that the maximum speed of the craft is now approximately 46 kt,
SPEED
FJ~Ot~ "verifying beth the resistance predictions based on towing tank
tests and propulsive efficiency claimed by KaMeWa".
Figure 18. "Gentry Eagle" Performance
1
y.o~¢W 0 )¢f°Pr?,-'uh,on ~v;tem~ DRAG
Comport~on w,t~ F~rooelter$1o¢ o 2~n~ mo;or~¢:tlt
P(on,n(J hull L~/I.=22.$6m
T~,n sorer o = 0.76Z Hewton gader SC progeller~
•PPI propellers: It$$ (2 , 956 kwl
gPH Jp,jnd$: 1|80(2=956 kWt
l'01ol 0¢0~e 00~¢r

l;(wl
2100 ~/."
2000 /
-/
1800 /' / "
1600 ~ J , ~ ~L.~=.
It00 prg_pe
ii ~..rS /~'~
SPEED
1oo0 J \~a,ev, JPS Figure 20. SES 200 Resistance Curve With Notional Waterjet
Thrust Curve
~O0
F L00 An approximate measure of propulsive efficiency is
hp/tonne kt. It is tempting to compare values of this parameter
J Powe# (FPP} : Qc¢ocdlnq tO 1flair
using the full-load displacement for the SES-200 (given in
Jane's 1991) of 207 tomes before conversion, and estimating
IS 10 15 JO JJ knott the increase in displacement with the waterjets installed.
Hp/tonne kt before conversion:
Figure 19. Comparison of WJ and Propellers [24]
= 3200 / (207 x 30)
= 0.515.
In the written discussion of [15] (1982), from which
Figure 17 is taken, the author provided the following If this value of hp/tonne kt is unchanged, the new displacement:
information: = 6800 / (0.515 x 46)
= 287 tormes.
'The efficiencies used in the power comparisons If the actual new displacement is less, which is probable, the
shown at 40 knots in Figure 17 can be split up as new hp/tonne kt is higher than before, but the efficiency is not
follows: necessarily lower.
A report on the performance of the SES-200 waterjet ship
Propellers Waterjets
is listed in the bibliography.
1"1o= 0.663 T~j=tx T~pmp= 0.606
11,' = 0.97 11," = 1 4.0 PUMPS AND INLETS
l l H = 0.985 (w = 0, l l H = 1.032 (w = 0.06, 4.1 Introduction
t = 0.015) t = 0.03) Many types of pumps have been used for waterjet
1"1o= 0.633 T~D= 0.625 propulsion including reciprocating, centrifugal, mixed flow-
mainly radial, mixed flow-largely axial, and purely axial pumps.
For reasons stated in the paper, the required thrust
When it was realized that high flow rates with moderate
for the waterjets is 2% lower than for the propel-
head rise (pressure) gave the best overall propulsive efficiency,
lers resulting in the same power requirement."
interest centered on mixed flow and axial pumps of high specific

speed. High specific speed enables smaller, lighter, higher
speed pumps to be used.
The majority of waterjet propulsors available today use //v / •
some type of mixed flow pumping element (impeller), although
there are some highly successful purely axial flow designs. t
Axial flow pumps tend to be smaller in diameter and
potentially lighter than mixed flow pumps but may not ,,:-I ; \ .,',/~>~ \~.-" /
necessarily achieve quite the efficiency (as a pump) that is • \ \ ' o wl
achieved by the best mixed flow designs (l~p _> 90%).

Inducer pumps usually employ high-solidity long chord
vanes with low angle of attack. They have superior cavitation
performance which means they show little reduction of head
with cavitation inception and minimum cavitation erosion at
very high suction specific speeds. These terms will be defined
later. Accordingly, inducer pumps have been used as the first
stage of high-performance axial pumps, or with or without, a
kicker stage (a row of short axial blades), by themselves.
Inducer pumps have been much maligned on account of their
supposed low efficiency and high power requirements. While it
is true their pump efficiency (89%) is lower than that of the best
mixed flow pumps (91%), they may make up for this by their
potentially higher flow capacity for their size and weight, Figure 21. Runner Profile and Notations for the Derivation of
especially in very high speed applications. the Euler Turbo-Machinery Equation [12]
4.2 Introduction to Pump Performance
The actual head, H = T~H H i where 1~H = hydraulic
It is only possible to include a brief summary of pump efficiency of the pump.
theory and performance analysis in this paper. The reader is The Euler equation enables an ideal head-flow (or
referred to the many excellent works on this subject, some of pressure-flow) characteristic to be drawn. The predicted
which are listed in the references and bibliography. However, "actual" head flow curve may be developed by taking into
some material not readily available is included here. account various losses and departures from the ideal conditions,
Although the principles are the same for all types of for a particular pump geometry. The actual pump
rotating machinery handling fluid flow, the nomenclature, hydrodynamic design must take into account cavitation criteria
conventions and units differ depending on the fluid and the designed to avoid vapor bubble formation and collapse which
application. Additional insights are possible when, for instance, leads to performance loss, vibration, erosion and, if unchecked,
compressor and pump technologies are reconciled. destruction of the pump.
Traditionally, pump technology uses the concept of head Figure 22 is a generic pump head-flow curve with
of fluid rather than pressure and so employs a head coefficient efficiency and power curves added.
rather than a pressure coefficient in presenting pump non-
f
dimensional performance. ~.t \
H = P/pg
or P = pg H
where H = head, or height of a liquid column
p = fluid density I
~,~ERATING POINT
g = acceleration due to gravity.
In customary units, head is measured in feet, pressure in pounds
t "\
per square foot, density in slugs per cubic foot (pounds per
cubic foot divided by go = 32.174 and g is the standard
acceleration due to gravity 32.174 feet per second squared).
Flow may be in cubic feet per second (cusecs), (cfs) or
gallons per minute (gpm) or gallons per hour (gph). HEAO
Pump speed is usually given in revolutions per minute
(rpm) or revolution per second (rps).
~ POWER
In common with that of fans and compressors, pump
design analysis starts with the Euler equation:
Vu2 U 2 - V u l U1 PUMP SPEED IS CONSTANT
Hi = g (4-1)
FLOW
where the impeller blade velocity components for the inlet and
outlet velocity diagrams are shown in Figure 21 taken from Figure 22. Generic Pump Characteristics
Wislicenus, 1968 [12].

Waterjet propulsors operate with fixed-area nozzles in
nearly all cases. Nozzles of several sizes are often available
enabling the pump operating point to be optimized by the
manufacturer for a given application.
T
Figure 23 shows three partial nozzle characteristics which
intersect the pump characteristics at points A, B and C, note that =1
all three points are in the vicinity of the peak pump efficiency.
Nozzle A is the smallest and nozzle C the largest. In general, A
~d
would be selected for a higher-speed vessel and C for a PO,NT
lower-speed vessel.
+J
_w
_o
o
u
I
I Z\
o
%
tel
I
i
I
0. I
i
FLOW COEFFICIENT ~OR QIND 3
Figure 24. Generic Non-Dimensional Pump Characteristics
BASELINE DESIGN POINT JET VELOCITY RATIO IS 0 . 6 2 3
I
1.0 ÷1.5% O P T I M U M FOR
I A S S U M E D LOSSES
[ O o.o
m / n ~ O i "E%
I o.6 / / ~ - 3 % 100% (DESIGN POIN~ ~10 WAKE
FLOW
...~_~.=/__,.,% oPc EF~EC+S
o.4
~ 0.2
Figure 23. Generic Pump Map for 3 Nozzle Sizes
0.5 0.6 0.7
For a given nozzle size, the pump operates at a fixed JET VELOCITY RATIO It (It = Vc/Vj)
operating point on the non-dimensional head-flow curve, or

along a fixed line on a dimensional pump map as illustrated in Figure 25. Relationship of Pump Weight and Propulsive
Figures 24 and 23. Coefficient for Typical Operating Conditions
A fixed operating point specifies a fixed efficiency and
power coefficient. Hence, for a given pump speed, power and This is best illustrated by an example, based on the case
torque are fixed. At low ship speed, the pump speed and power previously discussed. (The simplified version will be used.)
may be limited by cavitation considerations, but at higher ship A thrust of 10,000 lb is required at 35 kt. Based on the
speeds the power and torque remain essentially constant. data provided for the simple case, the maximum jet efficiency
Typically, pump speed may increase by no more than 3% while will be obtained with a jet velocity ratio about IJ. = 0.71 (Table
power is held constant, over the speed range of the ship. 2).
As previously derived in Section 3.0, propulsive
V = 59 ft/s (35 kt nom)
efficiency is primarily a function of the product of pump
efficiency and jet efficiency. At the ship performance design V.J = 59/0.71 = 83.1 ft/s
point a value of jet velocity ratio, ILl.= Vj/V, is selected which
will give a good jet efficiency without resulting in a pump size Thrust T = m (Vj - V )
and weight which is unacceptably large. Figure 25 shows the
relationship of relative pump weight and propulsive coefficient. T = 12 Aj Vj2 - 19 Aj Vj V

Nozzle Area (Jet Area) A = T/( D Vj2 - [3 Vj V ) 6. High pump speed (rpm) to enable a lighter gearbox to be
used
Aj = 1 0 , 0 0 0 / 2 x 83.1 (83.1 -59) 7. Tolerant of flow distortion when matched to an inlet duct
8. Superior mechanical design of bearings, lubrication
Aj = 2.50 ftz system, shaft seals, etc.
9. Lightweight corrosion resistant materials for pump
Nozzle Diameter for 1"ljm," D N = 1.784 ft or 21.4 in. housing and associated parts.
The required hydraulic characteristics favor pumps of
Nozzle Diameter Selected D s = 13.5 in. relatively high specific speed, N s and high suction specific
speed, Nu:
Thus the pump for maximum jet efficiency is about 59%
larger than the maximum size dictated by the application. The N = n Qu2/(gH)3/4 (4-2)
weight of the selected pump (dry) was about one ton, whereas,
the weight of the pump for maximum jet efficiency would be In this form, N is dimensionless and will have the same value in
about five tons. The weight difference is even larger if the
any consistent set of units. In the customary units used in the
weight of on-board water is considered. USA; Pump speed, n, should be in revolutions per second (rps);
The reduction in jet efficiency with the smaller pump is
Flow rate, Q, should be in cubic feet per second (cfs); Head, H,
only about 3.5%. Clearly, there is a trade-off of involving pump
should be in feet (ft); and Acceleration due to gravity, g, in feet
size and power required.
per second squared (ft/sZ).
During the design of a new waterjet ship the influence of
In the dimensionless system, pumps of high specific
pump size and weight on hull structure, engine size, and
speed approach the value of unity for N .
resistance can be examined to determine the optimum pump size
which will enable the ship to meet its mission performance Unfortunately, it is traditional in the USA to use a form
requirements (speed, range, payload) with minimum fuel. of specific speed which is not strictly dimensionless.
Figures 26 and 27 show the results of a pump optimiza- N = N Q1/2/H3/4 (4-3)
tion study for a patrol boat. The results were obtained with the
where N = pump speed, rpm
use of a whole-ship design synthesis model.
Q = flow rate, gpm
H = head, ft.
4.3 Desirable Pump Characteristics for Waterjet
Propulsion The numerical conversion factor for this system of units,
as pointed out by Wislicenus, [12] is as follows:
A superior pump for waterjet propulsion has the
N (US) = 17,170 x N dimensionless (4-4)
following characteristics. (The inlet, duct, elbow, and steering
and reversing gear are considered separately.) Current high-quality waterjet pumps operate at a specific
1. High hydraulic efficiency at a high flow coefficient speed of about 8000 in the U.S. system of units.
2. Minimum outside diameter for a given nozzle size Clearly, for a given head and flow, a pump with a high
3. Low weight for a given nozzle size specific speed will have a high shaft speed and a low torque for
4. Free from cavitation at maximum pump speed down to a the corresponding power. This means a smaller, lighter and less
low inlet head (corresponding to a low ship speed)
expensive pump.
5. Capable of sustained operation with some cavitation
without noticeable cavitation erosion of blades, stator
nozzle, etc.
Table 2. Sample Page From Jet Efficiency Tables
WAKE FACTOR= 1.00000

F2= (l/NOZZLE EFFICIENCY + ELEVATION TERM + PRESSURE TERM) 1.00000
INLET EFFICIENCY= 0.83000
JET PROPULSIVE EFFICIENCY WITH

EFFICIENCY VCVJ VJVC A PUMP EFFICIENCY OF
(0.91) (0.90) (0.89) (088) (0.87) (0.86) (0.85) {0.84) (0.83) (0.82) (0.81) (0.80)
0.55351 0.40000 2 500 0 50369 0.49815 0.49262 0.48708 0.48155 0.47601 0.47048 0.46494 0.45941 0.45387 0.44834 0.44280
0.59501 0 . 4 5 0 0 0 2222 0.54146 0.53550 0.52955 0.52360 0.51765 051170 0.50575 0.49980 0.49385 0.487.90 0.48195 0.47600
0.63091 0.50000 2.000 0.57413 056782 0.56151 0.55521 0.54890 0.54259 0.53628 0.52997 0.52366 051735 0.51104 0.50473
0.66095 0.55000 1.818 0.60146 0.59485 0.58824 0,58163 0.57502 0.56841 0.56181 0.55520 0.54859 0.54198 0.53537 0.52876
0 68454 0.60000 1.667 0.62293 0.61609 0.60924 0.60240 0.59555 0.58871 0.58186 0.57501 0.56817 0.56132 0.55448 0.54763
0.70073 0.65000 1.538 0.63766 0.63065 0,62365 0.61664 0.60963 0.60263 0.59562 0.58861 0.58160 0.57460 0.56759 0,56058
0.70790 0.70000 1.429 0.64419 0.63711 0.63004 0.62296 0.61588 0.60880 0.60172 0.59464 058756 0.58048 0.57340 0.56632
0.70340 0.75000 1.333 0,64009 0.63306 0.62603 0.61899 0.61196 0.60492 0.59789 0.59086 0.58382 0.57679 0.56975 0.56272
0.68259 0.80000 1.250 0.62116 0.61433 0.60751 0.60068 0.59386 0.58703 0.58020 0.57338 056655 0.55973 0.55290 0.54608
MAX. JET EFF = 0.70806 AT VCVJ = 0.70806

The Balje diagram is hard to reproduce as it contains so much
o
information. The regions marked mixed flow and axial are of
x 0"71 LINES OF CONSTANT GROSS WEIGHT most interest for waterjet pumps.
m 6 I41N. I x W O l
0. DESIGN POINT
o4 , , ',. ',, ~ ',
' 0.4 \ \ \ \ \
). 0.2- " ~ ~
--- ---.
"~ ~
-..
~ ~
-.
~ ~ ~ ~ . 2o. 7 ~ " ~ 2 50
,
I I I I I
-~ o.,. ,2~ ,$7~ ~ ~.,:;s ,.T6 .°'
PUMP INLET DIAMETER, FT
~'OZ~ (42901
57 58 5~ 60 61 -~-- S-L3Z ----"--~--

PROPULSIVE COEFFICIENT, % (10000, ~ [~0.000,~ i- |
Figure 26. Pump Size Optimization for a Patrol Boat Versus

PropuLsive Coefficient
(I~ I i i I l T'-"
~-&SZ
, |$0.:()! Oi' ('1(I ,$0,- -- --
i
!, ~-.-7-¢-T.,:--T- ~ I-+- I
z_
T I I t O~ i l
OIIMtlEn . ~i+oo pUMP IP+LET
NO~'ZLE RATIO OtAMETEn. i SO
- o s~s
eU~lI lilLE T
Figure 28. Rtmner Profiles as Funcdons of Specific Speed,
°= o ~ , OIA~ r e n + z.oo
n and Sucdon Specific Speed. S [12]
latH. r t w p l 0~51GN POIHr
i
IIINUSl pOWIR, imp
1 (27-7 to ~ a l m ~ / h r l (227 ml/hr)
$ 1.0OO to J,OQQ G P , ~ ~ -- . . . . ~ C v - , '0 COO G ~ * d
Figure 27. Pump Size Optimization for a Patrol Boat Versus
+°F, \\ "- ~ - - - "......
Power at Fixed Gross Weight
According to Wislicenus, it may be shown by some

70!- z G." Z3 ;c -
simple similarity considerations that the weight of machines
such as pumps is roughly proportional to torque. Thus,
doubling the specific speed might be expected to halve the ~)- ~,'T~ ",~-~.~ ~
weight. Although this rationale has not been checked in detail
for currently available pumps, there is clearly a strong incentive
to use the pump with the highest specific speed for given
required head and flow. However, there are practical limits to
specific speed which include stress in rotating parts and, most
important, cavitation. I V a l u e s oi soecahc soeed (U.S,I N.~" ,~P.u G ~ , ~ .
I
Specific speed varies with the operating point of the
pump and is always quoted for the peak efficiency operating
point. The actual operating point, dictated by the chosen nozzle
size should be at or close to the peak efficiency point. The C~ter cf +r- ~ JZ
CtntrtiuqQI ~ .~IXf]O ~ ptcoe.l~¢
convention of quoting specific speed at the peak efficiency point
enables different designs of pump to be compared. Figure 28
taken from [12] and Figure 29 taken from [13] show how pump
type and impeller design tend to vary with specific speed. Figure 29. Approximate Relative Impeller Shapes and
Figure 30 is the BalJe diagram [14] relating diameter coefficient Efficiency Variations With Specific Speed [13]
to specific speed for the general class of fluid flow machines.

Ioo
STATIC EXHAUST PRESSURE AND
~.o ~ TOTAL INLET PRESSURE. ASSUMING
.....
' ....,o //1......::?,_o;J- Ore. I = Cm. a
N,/~,
Ns = ~ N = rpm
30 PllOl"
n;3,o~ V = Ila/s
zo, ,o., ,,
Ds = ~ H a d = It Ib/Ib
y V l
! D = II
O4
H_ = C A V I T A T I O N S U P P R ~ : S S I O N HEAD
° Z
- - - F - -
_<
3 .O.i I I
1- - , I,,
£3 %: 70 7
0
"~ l \ I~DISPLACEMI~'N
. . . .
$/ . . 2 t b_
L/,_0 ,ooo__
3 6 8 IO ~o 60 O0 I 0 0 300 600 000,000 3000 6000 IOOOO
SPECIFIC SPEED, NS
Figure 30. Balje Diagram - Approximate N D Diagram for Single-Disk Pumps and Low-Pressure-Ratio Compressors
Before proceeding to the subject of cavitation, the other Customarily, g is omitted and plots are made of H/N2D 2 or just
principal pump dimensionless parameters will be briefly H/N 2 for a given pump, versus Q/ND 3 or Q/N, respectively.
reviewed. Many relationships can be derived from the dimension-
less pressure and flow coefficients ~ and ~p as shown in
Flow Coefficient:
Appendix C.
~p = Um/U t
where U m = meridional (axial) velocity entering the

4.4 Cavitation Considerations
impeller
So, Um = Q/Apl where Api = pump fact inlet area A pump requires a minimum absolute pressure at the
Ut = impeller tip speed, U 2 entrance to the impeller if cavitation is to be avoided. If the
Since Api = ~[4 Di2 where D i = equivalent inlet diameter pressure at the pump face falls below this value, vapor bubbles
will form whose subsequent collapse when they reach a
at the pump face
higher-pressure region of the pump will cause erosion, vibration
and Ut = U 2 = /I~D N/60 where D is the maximum
and if severe enough, destruction of the impeller and/or other
impeller diameter parts of the pump.
then (~ oc Q/ND 3 Since the pressure cannot fall below the vapor pressure of
Frequendy plots are made using either Q/ND 3 or just Q/N the fluid (fresh or salt water in the case of a waterjet pump), the
for a given pump, as the x-axis, where Q is in gpm or gph. required pressure which depends solely on the design of the
Pressure Coefficient: pump, is reckoned from the vapor pressure, and is called the net
= P/pUt 2 (or sometimes 2Pip Ut2) positive suction head, NPSH.
where P = pressure rise through the pump
NPSH = HAT + hpi + Vvi2/2g - H i - H. (4-20)
Since P = pg H
~l/ = gH/Ut2 (or sometimes 2g H/Ut2) where:
where H = head rise through the pump. HAT = atmospheric head, ft
Also, since Ut2 oc N 2 D 2 h In. = static head at pump inlet
~/~ P/I~ N 2 D 2 or gH/N 2 D 2 Vpi = velocity at the pump entrance (Q/Apl)

Hi = elevation of pump shaft centerline above SL corresponds very nearly to a line of constant power as previ-
Hv = vapor pressure expressed in feet of water ously mentioned.
Some manufacturers including KaMeWa, indicate zones
The waterjet inlet must be capable of supplying the NPSH of operation on their pump maps. Zone I, which is the majority
demanded by the pump to avoid cavitation. Usually, this is no of the area on the map, is unrestricted. Zone I is separated from
problem at the design point, but at low ship speed it is generally Zone II by a line which closely follows a line of constant
necessary to reduce power (pump speed) to match the pump specific speed usually about N = 10,000. Zone II, which is for
required NPSH to that available from the inlet for the existing reserve operation in rough weather, for instance, or at overload
ship speed. The available NPSH will depend on the efficiency displacement, is separated from Zone HI by another line of
of the inlet. approximately constant N a, about 12,000. Zone HI is for
Total head in the free stream = H^T + VwZ/2g (at SL). intermittent use only. Zone I11 is bounded by a third line which
is of the same general shape. Operation above this line is
Total head available at the entrance to the pump: prohibited. These lines and zones are illusla'ated in Figure 31
which is a generic pump map for a mixed-flow pump and Figure
Hpi = H^T + (1 - ~) Vw2/2g - H i 32 which is for a mixed-flow pump similar to those produced by
KaMeWa. It will be noted that the thrust tends to fall off at
Therefore, NPSH available =
constant power (and speed, rpm) as the ship speed is reduced
Hpi - Hv = H^T + (1 - ~) VwZ/2g - H i - H v (4-21) and the operating point crosses from Zone I into Zones II and
III. Beyond Zone I11 the thrust would collapse as the pump
NPSH may be non-dimensionalized by dividing it by the pump would no longer be able to absorb the horsepower and develop
head, H. The resulting ratio is known as the Thoma parameter useful head on account of the vapor cavities in the flow. Unless
[12]: the engine was shut-down by its governor, it would be in danger
O H = NPSH/H of overspeeding.
Cavitation inception is hard to detect during a pump test,
so NPSH is gradually reduced at constant pump speed, until the
pump head rise H begins to fall off. The amount by which the I I ! I I I I
head is allowed to decline due to cavitation inception varies with
..... I I
the application. Up to 3% is allowed by the Hydraulics
Institute, but 2% is usual. The limiting value of NPSH is then m ...... [ I
defined as the value at which the head is reduced by 2% or other
selected value.
A corresponding specific speed is defined as the suction
specific speed, N :
co
~ SHP
N,, = nQU2/(g NPSH) 3/4 (dimensionless) (4-22) f

I
10 20 30 40 50 60
N is constant for all values of pump speed at which cavitation
CRAFT SPEED, KNOTS
inception occurs (within practical limits).
Like pump specific speed, the dimensionless form of N B
is seldom used. Instead, the same units are used as for N and g Figure 31. Pump Map for a Generic Mixed-Flow Pump
(Computer Generated)
is omitted.
N,, = N QIr2/NPSH3/4 (U.S. custom) (4-23) NSS - 13,400 NSS - 12,000
where N = pump speed, rpm

Q = flow rate, rpm
and NPSH = net positive suction head, ft.
Combining equations 4-3 and 4-23 gives:
N = N u (NPSH/H) 3/4
or N = N (H/NPSH) 3/4 (4-24) z
The upper bound of N= for high-performance non-

inducer pumps is about 10,000 (about 0.6 dimensionless). For
inducer pumps, the limit may be as high as 30,000 which means
0 20 30 40 50 60
that full shaft speed (and power) can be maintained down to a SPEED, KTS
lower ship speed.
Lines of constant suction specific speed may be Figure 32. Mixed-Flow Pump Performance Map Showing
superimposed on a pump map showing thrust versus ship speed Operating Zones
for various pump speeds. A line of constant pump speed

A local cavitation number, analogous to the blade
cavitation number of propellers is of importance in pump
design. /
(~p = (Pi- Pv)/1/2 I9 V,12

where Pi = inlet static pressure at an impeller blade
Pv = vapor pressure
Vri = blade relative inlet velocity.
According to Wislicenus [12], it is impossible to operate

free from cavitation at values of (I P much less than 0.4.
However, it is important to remember that pumps can be
designed to tolerate some cavitation without damage. Experi-
ence with KaMeWa pumps has shown that these pumps can
operate for a considerable portion of the time in Zone 11 without
noticeable cavitation erosion.
5.0 TYPES OF PUMP AND OTHER WJ FEATURES
5.1 Centrifugal and Radial-Flow Pumps
Figure 34. Volute Propulsion Pump [12]
Centrifugal pumps were used on early waterjet-propelled
craft (Table 3) because they were readily available and, perhaps, Current applications include the Scbottel pump jet
because the need for higher flow rate, lower head pumps was thruster, Figures 35 and 36. This device has a rotatable volute,
not fully appreciated. For very high speed hydrofoil craft, somewhat like that shown in Figure 37. Similar in form to a
centrifugal volute pumps were a valid choice, and the TUCUM- centrifugal pump, on which the design is based according to
CARI (PGH-2) launched in 1968 has this type of pump (Figures Schottel. Note that the flow leaving the impeller is not
33 and 34). completely radial, however.
Table 3. Comparison of 19th Century and Modern ~,

Waterjets of Comparable Power [15] ~ ,
Pump Power, kW 560 600

Pump Type Centrifugal Mixed-Flow ~ . ~ [ ~ ~ " " -~
Pump Housing Dia, m 5.79 0.85 ~ ~ - t
Impeller Dia, m 4.27 0.85
Impeller Weight, kgs 8000 110
Number of Blades 12 6
Intake Area, m z 2.6 0.30
Ship Speed, kt 9.3 10 to 30
Thrust, kN 21.6 40 to 25 )il , ...........
.5,
j -.
the dimensional characteristics of a

% 2~' ~ ~ . o centrifugal pump for two pump speeds. Figure 39 shows the
" ~ ~ ~ ~ published performance of the Schottel thruster. This pump
gives excellent low-speed performance with freedom from
~ ~ ~ ~ . j , cavitation. The thrust relationships based on the dimensionless
•~ characteristics of centrifugal fans can be applied to such
thrusters, see for example, equations 4-15 through 4-17
(Appendix C).
Figure 40 also from [12], shows a radial-flow pump with
axial discharge suitable for a hydrofoil application. This pump
could also be regarded as a mixed-flow pump with a high degree
Figure 33. Propulsion System for TUCUMCARI
of radial flow.

S SCHOlq'EL Pump-Jet
(kN) N
,oo ~ 3o0 ,too 5c~ ~ o 700 ~ o ¢,0o iooo Noo,~.col3oo (h.p.}

' I ' 'I ' ' ' ' "
I [ P • nqo and
I __Lz!'~" 'D°'?'=::.....
, ~ uc ~o 250 k w (3aO t~ ~.1

' :57
"- *uptol.C-.COro.m.
uo to 400 kW (:550 tl o )
up {o 650 k W (.QC~ h D.I

Up ~0 I . a 3 0 t ~>m.
r
uo (e I .c::w{~~ (1,,3C--0tl p.)
...... "
...... - '°'" .......... ........... :;,"/"~'
,,,,, ~#
r,, To,°........................... "
N . in~t
(
S o e c ~ units a r e
- I avaliat~le o n f ~ e s t
N
0 100 200 300 4130 500 ~00 T00 ~]00 e00 1000 (kW}
" 7.,~, ,i •., 7~ , . , ~ r , . . ~ . < + : ~ , < ,,:1,:,

Figure 39. Schottel Pump-Jet Performance
Figure 36. Principle of Schottel Pump-Jet
Ill:-
tt i t ~ /~"-...U/
' - IN h'
........ ~~A' "~ .,~lL4t l
Figure 40. Radial-Flow Propulsion Pump With Axial

Discharge [12]
Figure 37. Vertical Propulsion Pump With Rotatable Volute
and Stationary Casing 5.2 Mixed-Flow P u m p s
~ "" , . I
Many of the pumps currently in use in waterjet propulsors
may be classified as mixed-flow machines although the degree
of radial-to-axial flow in the impeller may vary [12].
Figure 41 shows the velocity diagrams and blade profile
for a mixed-flow impeller. This figure may be compared with
the actual geometry of a KaMeWa pump impeller shown in
Figures 42 and 43, or that of the PHM foilborne propulsor-
derived second stage shown in Figure 44 or the MJP pump in
Figure 45.
Manufacturers of waterjet propulsors rarely publish the
head-flow curves for their pumps. Only the head and capacity
coefficients at the operating point, i.e., for a given nozzle size,
may be deduced from the thrust performance.
Figure 46 shows the test-derived characteristics of the
axial waterjet pump for which the jet is shown in Figure 9a.
Figure 47 shows the full-scale characteristics of an axial pump
with an inducer stage based on the model test results, and the
~A~E OF VOLUM~ FLOW
full-scale test results for three nozzle sizes. It will be noted that
while the predicted head was not obtained, the peak efficiency
Figure 38. Characteristic Curves of a Centrifugal Pump at of the pump was higher than predicted from the model tests.
Two Different Speeds of Rotation n 1 and n 2 [12] Full-scale tests of the pump without an inducer were not
possible.

--' ---'-7 I
,==~ 0.6 k_ vs Q/N
0.4
Figure 41. Velocity Diagrams and Blade Profile for a Mixed-
ca_
Impeller O. [ I_~_ L___ I i
~b
I IMPELLER t~_ × 6 I
cx_
,::,:: 5
uJ ~"
-- 43[ Q/N
cL
L__ oo:~ I I , I , ,
5
Figure 42. KaMeWa Mixed-Flow Waterjet
t~
4
L~J
N X
3
AII/N2 vs Qlrl
2
O
]
O,a -
LO
0 I I I t I 1 l
Figure 43. KaMeWa Mixed-Flow Waterjet Cut-Away to l 2 3 4 5 6 7

Show Impeller Q/N, NORMALIZED FLOW RATE, GPt'I/RPH
$TATOR-
Figure 46. Mixed-Flow Waterjet Model Pump Characteristics
$.3 Axial Pumps
Single-stage axial flow pumps are high-specific speed,

high-flow, low-head machines, and by themselves are unsuited
to very high-speed craft waterjet applications. Therefore, axial
stages are used in series with inducers or another axial stage
when a very high head is required. Single-stage axial pumps are
used successfully for moderately high speeds, however.
Figure 48 shows the dimensionless head-capacity curve
NOZZLE of an actual axial flow pump with data points for four pump
speeds. This plot is insU'uctive because it shows the extent to
which the data collapse into a single line when
non-dimensionalized.
Axial pumps tend to be smaller in diameter than most
mixed flow pumps. Figure 49, from Wislicenus, compares an
Figure 44. PHM-Dedved Mixed-Row Second-Stage Pump axial and a mixed-flow pump with a large radial component.

There are very few truly axial pumps, but Figure 50 shows an
American Hydro Jet (AMJET) pump which could be considered
to be axial. The Ultra (formerly Dowty) pumps are truly axial
and might be considered propeller pumps (Figure 51). there are
one and two-stage types available to suit different head
requirements. Other waterjet pumps described as axial-flow
include Castoldi, Hamilton, Nomera and PPjet (Parker). Figure
52 shows a cut-away view of Castoldi JET 05 waterjet unit. ,. ~ . . . . . .-~ ~,~
,~ , r~-\ t /i it /! il II i # -
/ / """ ,,, ,'. -.-~-,
Figure 49. Comparison of Axial-Flow & Radial-Flow

Pumps [12]
~ i[
x
F.ED,°TED
~.OM
~c . _
~LA.V~eW
5 L
90 PREDICTED FROM
~ ~ - . D E L DATA
)-
zo so
u. Figure 50. American Hydro Jet Axial Flow Pump

O 1 0 0 % N O Z Z L E . N P S H 36 F T
~ . g 0 % N O Z Z L E , NPSH 36 F T 5.4 Inducer Pumps
~ ' 1 1 0 % N O Z Z L E . NPSH 3 6 F T
AH PUMP H E A D RISE. FT
N E N G I N E S P E E D . RPM Inducer pumps were originally developed for rocket motors.
Q F L O W R A T E . GPH
One of the first inducer impellers, made by Bell Aircraft, now
25 310 35 4O Bell Aerospace Textron, was about 3 in. diameter. Based on
Q#N
personal recollection of the author, it resembled a small marine
propeller with very fine pitch and very high area ratio
Figure 47. Full-Scale Mixed-Flow WJ Pump Performance (overlapping blades). Its purpose was to raise the pressure of
the fluid sufficiently to avoid cavitation in the main pump
element.
Computer programs were developed by NASA for the
I I
design of inducers, versions of which were used by Rocketdyne
I ,
I and ALRC to design inducer stages for axial and mixed-flow
I waterjet propulsors which required very high head at both low
speed (to overcome hump drag) and at very high speed. These
o 4o pumps were used in hydrofoil and SES applications where their
compact shape and size were essential. The ALRC pump
,% (Figure 53) is used in the U.S. Navy's PHM craft and the
Rocketdyne pump (Figure 54) was eventually selected for the
3KSES program which was cancelled before the manufacture of
o 3o
a 1000RpM the first pump was completed. A photograph of the inducer for
1400RPM this pump (PJ 46) is shown in Figure 55. The inducer is
O IS00RP~
approximately 46 in. diameter. Rocketdyne also developed a
range of inducer pumps consisting of an inducer with a kicker
stage. The kicker stage is a row of short axial blades integral
0'I,0 000 0~0 O.lO with the inducer. Most of the head rise occurs in the kicker
V~U
stage, but the inducer allows the pumps .to operate at very high
suction specific speeds ( N > 20,000). The PJ20, whose
Figure 48. Dimensionless Head-Capacity Curve of an Axial- inducer was the model for the PJ 46, is the propulsor of the
Flow Pump [12]

Boeing Jetfoil passenger hydrofoil craft. Both the PJ20 and the
Jetfoil itself are now built by Kawasaki Heavy Industries in
Japan. A cut-away view of the PJ20 and a dimensional drawing
are shown in Figure 56. The kicker-stage blades are clearly
visible in the cut-away view.
Figure 53. PHM Propulsor
~. '?i, ?o',,
Figure 54. Rocketdyne Powerjet 46
, .~,,, ~,
~:~ ~
Figure 51. Ultra Hydrojet 300 Axial-Flow Waterjet Pump Figure 55. Rocketdyne PJ46 Inducer (Full-Scale Pattern)
I Rocketdyne also developed a larger inducer pump suited

to lower head requirements than the PJ20, the PJ24 a photo-
graph of which is shown in Figure 57. Figure 58 shows the
~, dimensions. The PJ24 has no kicker stage as can be seen from
the photograph of the impeller in Figure 57. The performance
of the PJ24 is shown in Figure 59.
Inducer pumps, such as the range formerly made by
Rocketdyne (PJ16, PJ20, PJ24), are suited to high-speed
applications where small size and total weight are at a premium.
They cannot compete with current mixed-flow waterjet
propulsors in propulsive efficiency or installed power because
they depend on low flow and high head to achieve their thrust.
They have a low flow coefficient at optimum pump efficiency
and therefore a relatively low specific speed.
Figure 52. Cut-Away View of a Castoldi JET 05 Waterjet The PJ24 is roughly comparable in size of impeller inlet
Unit diameter with the KaMeWa 63S. At 40 kt, the PJ24 requires

3000 shp to produce 12,500 lbf of thrust (1"1o = 0.51) whereas,
the KaMeWa 63S only requires about 2400 shp (lql) = 0.64), or
less, depending on the application. Thus, the power required by
the PJ24 is 25% more. KaMeWa have often said that inducer
pumps require up to 30% more power than their pumps,
Haglund et al. (1982) [15]. However, this can be misleading as
it is possible to design inducer pumps for much higher specific
speeds than those of the Powerjet range of pumps.
? .~E,,.ces
~NGINEEnlNG NOTES
I l.ul~¢oil~{eXleln~l~t,r+olyl
I
f-/ L
A~sl B97 ~ mlo
P,lch di.~ele, 4 O 0 ,~
Numee, of ~ln 32
................ ~[ I:: it ~- ' \,!,0

~OT~ I
• -_.i ~, ~ .... ~- ~1I
Figure 58. PJ24 Dimensions
Figure 56. Powerjet 20
0 )o 1o
c~fr s~uD. ~,0rs
!
~o o
Figure 57. Photograph of PJ24 Showing Inducer Figure 59. PJ24 Performance
Figure 60 shows actual test results for an inducer pump The most recent inducer pump design was performed by
small-scale model. The model efficiency is low due to scale DTRC, Stricker, et al., 1992, [16], from which the following
effects. Figure 61 shows two view of the model inducer which information and figures are taken. This pump design, which had
was about 7 in. diameter. to meet specified thrust, craft speed and envelope characteris-
tics, posed particularly challenging problems. Four pumps were

required to fit within the planing flap of an advanced am-
t_, ,,gn~ ,¢ ~*or on :F,,, . . . . .
phibious assault vehicle (AAAV). This meant that the overall
diameter was limited to 16.5 in. and the inlet duct configuration J /
/ / . /
was necessarily far from optimum giving a low effective inlet
efficiency. Furthermore, the pumps had to provide adequate /
/ /
J I /
hump drag thrust margin at 18 mph, with a top speed about 28
mph. . . . . !
', 0 ,.~,~'.,..---~J/ !~ , ,i
90 F / -L"/N?
r •
• " l,'! -
,ll6 I" ./41- I
.821- .701- 11 H u 1::{4 z_v, ~ . . _ . . , ~ ,

Figure 62. Cross Section of Propulsor System
o
× .18 t-'-. 6fd- .90'
%
:1:
66L 54t- ' 60 L- I I _ _ [ . . . . L___ L__---3--

O.q8 0,50 0.52 0.54 0.56 O SB 13.60
Q/N, GAL/REV
Test Conditions
No Distortion
5000 rpm
15 psia Suction Pressure
Figure 60. Inducer Waterjet Pump Model Test Results
80r~tl0~
L~!,-
/ ....;--~ .~
First P^Rr~t v^~

\--2_,
Figure 61. Model Inducer Figure 63. Stack-Up of Precision Cast Waterjet Components
Figure 62 shows the propulsor configuration and Figure 5.5 The Propeller as a Pump
63 shows a stack-up of the precision-cast pump components.
Figure 64 shows the trade-off for power versus diameter at the Because of its fixed geometry, a waterjet propulsor is
stator cavitation limit. Figure 65 shows the inducer suction constrained to operate at an operating point on the pump
performance. The selected diameter of 16.1 in. was the largest characteristics which remains essentially constant over the
that could be accommodated. Table 4 shows the design point normal range of performance conditions, i.e., pump speed and
performance. The design flow coefficient, ~), is about 0.44, power, ship speed, sea state, displacement, ship trim angle, etc.
which is exceptionally high. A great deal can be deduced from Accordingly, the shaft torque and power remain almost constant
the published performance data, which cannot be included in for a given pump speed, regardless of the other conditions.
this paper for space reasons.

If a ship loses speed due to increased resistance in a high
4so
I sea state for instance, propeller torque will increase rapidly if
r~usT = 3757 LBS at 18 MPI4 propeller speed is maintained, and it is easy to overload the
~ J V R = 2.6
410 ~ * OES!GNPOINT engine in the case of diesel propulsion. Furthermore, reducing
propeller speed will lead to a further increase in torque
,--,~ I'", JVR = 214
coefficient possibly necessitating further speed reduction, and so
on. This is why careful matching of propeller and engine
c ~ """] jvR = 16 characteristics is so important at the design stage particularly if
-. JVR = 2.0 .-"
---~ 370 ..... ~ jv~ = IS ..... fixed-pitch propellers and diesel engines are to be used.
t__ Allowance must be made for increased resistance in waves, hull
~ ........ STaTOR CnVIT~ITTON LZI'I,IT fouling, displacement growth throughout the ship's life, and
other factors which may shift the resistance curve closer to the
T
T
maximum allowable thrust curve based on the engine
i characteristics.
330 ,I , : : : : : E ; i
IS.O 17.0 19.0 21,0 23.0 The propeller may be regarded as a pump with a variable
25,0
nozzle area if the stream tube and propeller shown in Figure 16
DIRMETER ( ±riches )
is surrounded by an imaginary thin duct of identical shape. If
the flow is inviscid nothing will be changed for a given set of
Figure 64. Design Point Map for 3 Degree Incidence Angle conditions. Then the downstream jet area becomes the nozzle
area of an equivalent waterjet propulsor with an inlet area equal
T a b l e 4. W a t e r j e t D e s i g n P o i n t C h a r a c t e r i s t i c s to the capture area of the propeller.
If the propeller performance is characterized by curves of
Thrust 3767 ibs K.r, KQ and 1']o versus J, where:
Craft Speed 18 mph =
Headrise 57.5 ft KT thrust coefficient, T / p n z D 4
Flowrate 22994 GPM KQ = torque coefficient, Q/O n 2 D 5
NPSH 37.2 ft open water efficiency (ICr,/KQ) (J/2~)
11o =
RPM 1251
Diameter 16.1 inches J = advance ratio V/n D
Uti p 88 fps n = propeller speed (rps),
Vax 41.6 fps
Then it can be shown that:
Tor~de 1633 ft-lb
Horsepower 389
Pump Efficiency 88% j r /8KT+I
+ 1t (flow coefficient)
3.00 "t'.,- ,. • , , ., , .... , . . . . .. , , ,...
~. /~ r i o J[nclder~ ~ q t ~ - 3 d~r~s
2"lSt iii • Oe,lgn PoLnc
2 •50 -~
Illl ..--2 q ~s~ and V = 4 K.r,/~ 3 (pressure coefficient)
z25 l/\ '.." also Aj (----A) = Ap ~)/(2~1/"+ j2/~2)u2
F-
= Ap ~ / 2 ~ when J = 0
and
f Ca=l,i
r /8KT+I 1 KT J
1.25 T C=-t.O J
1.00 ! . . . . ' , ', . . . . ; . . . . : ' ', ', " , . . . . ', " ', 'i . . . . ~ . . . .
0.00 0.20 0.40 0.60 0.80 1.00 1.20
Note that Tli = Tlp~TIp=p
Figure 66 shows the characteristics of a Gawn propeller

as a propeller and as a pump. As expected, the head coefficient
Figure 65. Inducer Suction Performance
is very low and the flow coefficient is high giving a high
specific speed, ranging from 0.426 (7300) at J = 0 to 2.15
(37,000) at J = 1.55 corresponding to maximum efficiency. For
The opposite is true of a propeller. Because it has no
comparison, the DTRC inducer pump has a specific speed of
fixed set of geometric restraints on the shape of the flow path
0.529 (9100) at its design point (J = 0.94). At J = 0.94 the
through the disc of the propeller, the operating point can vary
propeller specific speed is 0.9 (15,500).
widely with consequent wide variation of torque and power at a
given shaft speed. Similar comparisons can be made with ducted propellers
where the duct exit area is less than that of the propeller disc.

the required flow at the craft design speed and power. At this
1.01 GAWN ADIA 0 " 0.8 condition the inlet has the optimum inlet velocity ratio and is
o PlO " 1.8 entirely free from cavitation. Strictly, all other conditions
REF. DUCANE 1972 EFFICIENCY AS A require a different inlet area to maintain the optimum inlet
o.81 PROPELLER rro
velocity ratio and recovery. In practice, satisfactory inlet
KT
w performance can be obtained over a wide range of inlet velocity
o.0! O.24 ~ ratio and craft speed with fixed area inlets. The optimum inlet
0.2O ~ design for a given application requires model testing and
iterative adjustments of geometry guided by theory and
~ 0.4 0.16
experience. This is only economic for large projects, in most
0 0.12 ~ cases, especially in the smaller sizes a standard inlet geometry is
~ 0.2 0.08 used which has been found to give good performance for the
usual types of installation.
oo.o o
In addition to good head (pressure) recovery, an inlet
20 should provide a low level of flow distortion at the pump face
ADVANCE RATIO J S0% since this will give a higher value of relative rotative efficiency
EFFICIENCY AS A PUMP, / ' - ' ~ / ~ T0 (1"1,). Some level of distortion is inevitable because the flow has
I r)PUMP /F /" ~"~--EFFICIENCY AS
Lu Z
=. o.12! /
/ A PROPELLER
, ~° . - to be turned into the correct direction to enter the pump which is
necessarily at some height above the inlet in nearly all ship
~
A u. 0.191.0 p ~ All 50 =.
applications. Usually an elbow duct is required 1 through which
0 0.01 '0.8 / ~ . , . 40 =
the shaft driving the pump passes. The shaft creates a wake
I- AIIAp / /
which adds to the distortion. In the past, shafts have been
enclosed in a streamline fairing to minimize their wake, and
O 0.0 0.2 / / 1.0- 10
turning vanes have been used in the elbow to reduce turning
/ o _-,.4--~8 J 1.2 losses and distortion. Most current waterjets have flush inlets
I/ , , 7 T "7 o
'0 0.1 0.2 0.3 0.4 0.5 with a low elbow angle and use neither shaft fairings nor turning
FLOW COEFFICIENT
vanes.
Hydrofoils with waterjet propulsion have the most severe
Figure 66. The Free Propeller as a Pump flow-turning problems, see for instance, Figure 33 which shows
the TUCUMCARI inlets and ducts leading to the waterjet
5.6 WaterJet Inlets pumps. Figure 67 shows the corresponding arrangement for the
Jetfoil and Figure 68 for the PHM. Vaned elbows are illustrated
The provision of a high-efficiency waterjet inlet is very in Figures 34 and 40. Figure 69 shows the SES-IOOA, which
important. Equation 3-44 reduces to the following if Tli is the was one of the two 100-ton test craft for the 2K/3KSES
only variable. program, with its original pod inlets. Figure 70 shows the pod
inlet with its auxiliary openings which could be controlled to
OPC = a constant/(a constant - Tli I.J.2) give the right inlet velocity ratio according to the speed of the
craft. In 1974, the pod inlets were replaced by variable area
where T~i = (1 - ~) as before.
flush inlets, shown in Figure 71. Although the flush inlets had
lower drag and allowed the craft to attain a higher speed, they
In the example on page 28:
were subject to air ingestion in a seaway. An extensive program
OPC = 0.4745/(1 - ~ i x 0.297) was successfully completed to keep air out of the inlets by
means of inlet fences (Lavis, 1976) [21].
If Tli = 0.83 OPC = 0.630 (as before)
If ~li = 0.80 OPC = 0.622 (-1.2%)
If Tli = 0.77 OPC = 0.615 (-2.4%)
Thus, each percentage point reduction of inlet recovery costs

0.4% reduction of OPC relative to its value initially with Tli =
0.83 which is about the state-of-the-art for current waterjets.
5.7 Types of Waterjet Inlet
Waterjet inlets can be divided into two main categories;
pod inlets and flush inlets. Pod inlets are used on hydrofoils
such as the PHM and Jeffoil. Flush inlets are used on all other
craft including conventional monohulls, planing craft,
catamarans and surface effect ships.
Figure 67. Waterjet Inlet and Flow Path for the Jet foil
Waterjet inlets may be of variable or fixed inlet area.
Theoretically, the optimum inlet should be designed to match
1 An exception would be a waterjet pump driven by a shaftless motor.

Fo*lbo¢.e
;-&.,,...7°, . LW.o,~.M".":.
. . . . .
Figure 70. SES-IOOA Pod Inlet
Figure 68. PHM Waterjet Inlet and Ducting Arrangement

Figure 71. SES-100A With Flush Inlets
Figure 72 illustrates how cavitation arises on a waterjet

inlet, in this case a pod inlet, if the inlet velocity ratio Vi]V" is
,~ ~ ~ - ~ ~
either too high or too low. High inlet velocity ratio occurs at
low ship speed and increases with power, so there is an upper
limit to the power which can be applied when the ship is
stationary or at low speed without causing the inlet (or the
pump) to cavitate. Low inlet velocity ratio occurs at high ship
speed and is lowered further if power is reduced for any reason.
Very low inlet velocity ratio occurs when a pump is idling while
the ship is underway, for instance, or when a wing pump of a
three-pump configuration is shut-down and the ship proceeds on
the other two pumps.
On a flush inlet, cavitation arises inside or outside the lip
as described for a pod inlet. It was to avoid undesirable inlet
velocity ratios that the variable geometry flush inlet was devel-
oped (Figure 73). This type of inlet was demonstrated on the
SES-10OA. The intent was to keep the inlet operating within the
cavitation bucket, Figure 74, which shows the boundaries for the
different types of cavitation which may occur on this type of
inlet. Note that there is a wide latitude on inlet velocity ratio at
low ship speeds up to about 50 kt. Beyond 50 kt the bucket be-
comes narrow. This is a very important point because very few
waterjet ships exceed 50 kt, and it helps to explain why fixed
area inlets are used. The variable area inlet was developed for
the 3KSES which had a projected top speed approaching 100 kt.
However, the variable geometry inlet was mechanically
complex and very expensive. Further model testing showed that
Figure 69. SES-100A With Pod Inlets

the flow did not behave entirely as predicted based on the
extensive modified potential-flow computer simulations which
were conducted using the Douglas-Neuman method adapted to
non-axisymmetric internal flow (Figure 75). It was found that a
reasonable compromise inlet performance could be obtained
with a round fixed-area inlet and this was selected for the
3KSES prior to cancellation of the program. The fixed inlet
a.o F
Z.O~
OUTSIOEL[P CAVITATIONUSED IN
..A, CO.T,OL,V,TEM
- Z~.~ ~ L[P OR T~ROAT

//
would have saved a great deal of cost and some weight
depending on the weight of on-board water. 1.0,
I
j f = r
=-=-osL - aO,~ ~ ~ X O~ [ 13:2,:,6
<
Inlet'no[ (A) Design V i / V o -~0.;
CoYilofion
. ~ ~ O.05~
l POINTS OF NO "DETECTABLE"
~VITATION
O IOl PORT ~ 85 STBO
~ ~[SSION85
SEVERELY °EG~OEO PUMP [~LET
O101 STBO o 97 PORT -- STATICRECOVERY
(B) V i / V o Too High
Exlemal
I +IO2 PORT & 98 PORT STATIC RECOVERYIr4PROVIMG
Co-,,i)o);o. ~;..~: .... XlOZ STBO o 99 PORT

O.Z 0.4 016
.......
O.t8 I'.O
INLET VELOCITYRATIO
\
Figure 74. SES-100A Inlet Cavitation Bucket
IIII Jill I[[lli

(C} V i / V o Too Low
),o
Figure 72. Effect of Inlet Velocity Ratio on Cavitation
LOWElSl
l OEtkn(,~t~ LI.(5
[:~o,tea " ~rooA.o
o
i'i
i 4.o/i,
:j.
SECTI~~-4
................
I
I
t
L I I! I
I
10.10. ~ SID{ PLAT(PROFIL[
I -, [ L
-12 -e -( I izZ 111 20 24 211 ' 31 ~6
x DISTAHCE FT
~ - - " le.~l '
; ~s.o,'- V--'-- III I I III

tttwrt0a
Figure 73. Design for a Large Variable-Area Waterjet Inlet

[--
8
The shape of some current waterjet inlets may be seen ~ 1 1 1 1
from the pump illustrations, Figures 42, 43, 45, 50, 51 and 62.
Very little information has been published concerning the
performance of inlets, but it is known that KaMeWa and Riva -e -4 o 4 ex ~2 le 20 14 2i 32 36
OISTANCEFT
Calzoni have performed extensive painstaking experimental
investigations. Published pictures of KaMeWa inlets are
generic only. KaMeWa will provide an actual inlet design for Figure 75. Potential Flow Simuiation of Flush Inlet
each application. The inlet duct is shipyard provided to
KaMeWa's design. Figure 76 is a photograph of the inlet duct Performance analysis indicates that inlet efficiencies
for the SES-200 taken from [11]. Figure 77 is a photograph greater than 0.8 are achieved. Further performance gains may
looking into the inlet. be possible using wider inlets to capture more of the boundary

layer. This has been studied by Purnell, 1976, [17] from which KaMeWa has published some qualitative information on
Figure 78 is taken. Here K= is the inlet momentum velocity inlet design and performance (Svensson, 1991) [18]. These
ratio V J V corresponding closely to the wake factor (1 - w), reports are of great interest to the naval architect who wishes to
and jet velocity ratio is 1/~.L. understand waterjet propulsion as they are based on extensive
experience of a wide range of applications.
5.8 Nozzles
Nozzles may be of the Pelton type in which the outer and

inner peripheral walls are straight and parallel as shown in
Figure 44, in which case, the vena contracta occurs downstream
of the nozzle exit plane as in Figures 9 and 10, or the nozzle
may have a well-rounded entrance to a parallel throat as in
Figure 50, in which case, there is no external vena contracta and
the nozzle exit area is coincident with the jet area (prior to
break-up of the jet surface).
Well-designed stators and nozzles result in very low
nozzle losses ~ ~ 0) so the nozzle efficiency is very high lqn
- ~ 100%. A value of 99% can be used for high-efficiency
waterjet pumps, or 1 + ~l/= 1.01 in the jet efficiency expression.
No matter how well-designed a pump may be, the
velocity in the jet will not be quite uniform. Figure 79
illustrates this point. In a rigorous analysis the momentum flux
and kinetic energy of the flow should take this into account for a
given volume flow rate.
__/- vj
Figure 76. Waterjet Inlet Duct Before Installation in
SES-200 [11]
" i085 ~1 x
Figure 77. View Looking Into SES-200 Inlet [11]

/
~ . W I D E BOUNOAR~'LAYERINLET
d 0,55 Figure 79. Velocity Distribution in a Model Waterjet Nozzle
>0'.~ Jet [9]
0~_0
0.50 ~CONVENTIONAL INLETIK.=0.95)
1.6 L8 2.0 It is possible for a parallel-throat nozzle to experience

DESIGNJET VELOCITYRATIO
cavitation on the nozzle walls if not correctly designed. The
SES-100A pump nozzles had parallel throats but the pumps
Figure 78. Predicted Effect of Wide Boundary Layer Inlet designed for the 3KSES program had Pelton-type nozzles, as
on OPC [17] was the case for the Rocketdyne Powerjet range of pumps.

Current waterjet pumps, including KaMeWa pumps have the illuslrations of pumps in Figures 43, 51, 52, 53, 57, 58 and
parallel throat nozzles in most cases. The DTRC inducer pump 69.
has only a vestigal nozzle and a hollow jet (see Figure 62).
5.9 Steering and Reversing Gear
Steering in a waterjet-propelled craft is accomplished by

FORWARD
deflecting the waterjet with a steering sleeve (or a pair of vanes
called buckets). This is a very powerful effect as it operates on
the gross thrust as explained in Section 3.0. Steering sleeves are
usually hydraulically operated although some manufacturers use
mechanical linkages in the smaller sizes. .... j
Waterjet-propelled ships are highly maneuverable as
shown in the photographs of Figures 80 and 81. NEUTRAL
REVERSE
Figure 82. Flow Re-Direction on MJP Waterjet
Figure 83 shows a generic thrust vector diagram for a

steering and reversing gear at zero forward speed. A similar
diagram can be drawn for any pump with steering and reversing
Figure 80. The Marintechnik Hamidah Crew Boat Turning at gear. The available steering angle and permissible powers will
Zero Speed vary. Generally, the ahead bollard pull of a waterjet propulsor
will be greater than that of an equivalent fixed-pitch propeller,
but less than that of a controllable pitch propeller, while the
astern bollard pull will be somewhat less than that of an
equivalent fixed-pitch propeller, but maneuvering is easier on
the engine and transmission with a waterjet.
THRUST AHEAD
\ /
\\ //1/
\/
B
+c~O STEERING ANGLE
MAX~ MAX
PMAX
I
Figure 81. Passenger Vessel Kasumi Jet With KaMeWa

Waterjet System Executing Turning Maneuver
THRUST ASTERN
at Over 29 Knots
Reverse thrust is obtained by use of a reversing bucket of Figure 83. Thrust Vectors Available at Zero Speed
some type, which deflects the jet downwards and forwards
usually. Partial deployment of the reversing bucket catches only 5.10 Braking
part of the jet, giving variable net thrust either forward or astern,
with a neutral, zero net thrust position. A clutch is desirable to Unlike propellers, waterjets may allow full power to be
avoid residual craft motion in the neutral position. This applied in reverse at full forward speed which may be main-
capability gives the ship excellent maneuverability alongside tained until the ship has slowed considerably after which power
when docking or undocking. Figure 82 shows the principle as must be reduced progressively to the maximum allowed at zero
applied to the MJP waterjets. Other mechanisms may be seen in speed.

The reverse thrust will greatly exceed forward thrust The impeller diameter is 337 ram, dry weight 888 lbf and the
initially as it is now the sum of the gross jet thrust reverse power rating is 550 hp.
component and the momentum drag. The reverse thrust will Dowry Hydraulic Units Ltd (United Kingdom) was part
rapidly decrease as the ship slows down until at zero speed it is of the Dowry Group of Companies and is now Ultra Hydraulics
only a fraction of the forward thrust available (Figure 83). Ltd. Beginning with the Dowry turbo-craft jet boats of the
Actual numbers must be obtained from the waterjet 1950's. Dowty/Ultra axial waterjets are among the most
manufacturer at the time of contract, but it will be obvious that numerous in the world. Waterjets produced for mil!tary vehicles
waterjets have excellent braking capability, especially at speed. and craft have high thrust at low speed without suffering from
cavitation. A special feature of some units is a 360 degree
6.0 AVAILABLE W J PROPULSORS rotating nozzle.
Some of the principal manufacturers of waterjet

propulsors are listed in Appendix D in alphabetical order,
showing the size and power range of their products. Further
details for many of these companies are given in "Jane's High
Speed Marine Craft".
The following brief comments may be helpful to persons
unfamiliar with the industry.
Aerojet Liquid Rocket Company (ALRC) later known as
Aerojet Techsys~ms Company, a division of Aerojet Gener~,
built, in 1970, the two-stage, two-speed pumps for the SES-
100A (4000 hp) and the very successful 18,000 hp-rated pumps
for the PHM. They are no longer in the waterjet business apart
from supporting the PHM program. Until recently, the PHM
pumps were the highest-p6wered Waterjets in the world.
American Hydro Jet Corporation (AMJET) is a fairly new
company founded in 1989 and manufacturing waterjets in the
range 400 to 5000 hp. They are the only U.S. manufacturer of
larger pumps and recently tested a 56-ft fishing boat with their
AMJET J-3580 pump driven by a Textron Lycoming TF35
marine gas turbine engine developing 3000 shp, at speeds close
to 50 kt. They ~ e currently installing three AMJET -1700 units,
1100 hp each in a 90-ft motor yacht designed to make 40 kt.
The AMJET steering and reversing gear is of unique
design with two coordinated steering vanes, or buckets, which 10 20 30 40 50
B O A T SPEEO, K T
act as thrust reversers when opposed. The result is a smaller and
lighter arrangement than is found with most steering and
reversing gears. Figure 84. Castoldi Jet TD 337 Pump Map
Berkeley introduced their jet drive in 1960. They are
well known manufacturers of high-performance transom- Hamilton (New Zealand) are true pioneers of small-to-
mounted waterjet propulsors for pleasure boats, fishing boats medium-powered waterjets and are now developing larger units
and commercial applications. Prime movers are principally up to 4000 hp. Beginning in 1954, Hamilton popularized
automotive-type engines. ~,aterjets for boats following their conquest of the Colorado
Castoldi Jet, located near Milan, Italy and represented in River rapids in 1960. Hamilton waterjets are used in over 120
the USA by the Bird-Johnson Company at the time of writing, countries around the world and their customers include the U.S.
produce some 2500 waterjet units annually, in the power range Deparanent of Defense. An interesting side line on public
60 to 1200 maximum continuous horsepower. Their units awareness of waterjet propulsion until recently is the fact that
feature hard anodized aluminum alloys and stainless steel for the only reference to the subject in the bicentennial edition of
full seawater corrosion protection and built-in gearboxes with a the Encyclopedia Britannica (1976) is to Hamilton in connection
range of possible gear ratios to give optimum matching with the with boating.
engine. A clutch is provided for true neutral, i.e., zero thrust Jacuzzi (USA) waterjets were used on a large number of
operation. A remotely operated movable inlet screen is military craft in the Vietnam War era and are still in use on
provided. Castoldi jet pump maps are of interest because small military craft in many countries.
Castoldi measure actual net thrust in a laboratory boat. 2 Figure The Jacuzzi waterjet interest was bought by North
84 shows actual measured performance for a TD 337 waterjet. American Marine Jet, Inc. known as Nomera. They manufac-
ture a range of axial/mixed flow waterjets up to 700 hp.
2This method is also used by North American Marine.

KaMeWa (Sweden) is the giant of the waterjet world both
EXHAUST
in size of pumps and in technical prowess. KaMeWa, famous OUTLET
AIR iNTAKE
for their controllable-pitch propellers, delivered their first EXHAUST
SILENCER /
waterjet to the Swedish Government in 1968 and have been \ /i
\
marketing their $62/6 series of waterjet pumps since 1980. \ //
About 225 vessels with KaMeWa waterjets have been listed in E~ECTOR
COOLING
Jane's, many of them with two or even three pumps. On Figure AIR
104 KaMeWa shows a total of 326 installations. The KaMeWa \
$62/6 pumps follow a systematic series beginning with size 40

and presently ending with size 180 (30,000 hp). Designs are
listed for larger sizes. More details of KaMeWa pumps are
given later in the paper. KaMeWa has the advantage of WATER JET
possessing a superb marine laboratory and research department. \

\
Not only are they able to use these resources to develop their \\
products, but are able to assist their customers to solve technical
problems which sometimes arise through the application of a
comparatively new technology in the conservative shipping
industry. Numerous technical papers have been published by
KaMeWa's technical staff many of which are referenced in this !J ~\ \\
paper. KaMeWa has been an inspiration and a challenge to /i ,\ FLEXIBLE
WATER JET
DRIVE SHAFT '\ COUPLING
others also trying to understand the intricacies of waterjet
GAS TURBINE :
propulsion and its merits relative to propellers. A monohull L M 500
boat, the Destriero, recently broke the transatlantic speed record

using KaMeWa waterjets, cutting 31 hours nearly 40% off the
previous best time for a west-to-east crossing, at an average
speed of about 50 mph. A plot, specially provided by KaMeWa
for this paper, showing the propulsive efficiency of the model Figure 86. Kvaerner-Eureka Total Waterjet Propulsion
and full-scale craft will be found in Figure 85. These results are Package
astounding when compared with those for the best propeller-
driven craft. MJP (Sweden) has been making the larger sizes of
waterjets since 1986. A new type of mixed-flow pump was
DES3RIERO - 67 m HARD CIhlNE MONOHULL developed as well as an advanced maneuvering control system
I R I A L RESULTS FOR "/5% FULL LOAD
with an autopilot feature to make harbor maneuvers easier.
PPjet (Parker) (United Kingdom) has 20 years experience
in supplying waterjet propulsors, in the range of powers up to
4000 hp. The larger sizes are built with advanced composite
materials using glass fiber, kevlar, or carbon fiber as appropriate
for static parts. These materials are totally corrosion-resistant
and have excellent structural properties combined with light
eo ~c - - - -
weight. The method of construction enables the pump inlet
SHIP S P E E D , K T S
housing to be molded to the hull contour with little change to
WATER JETS: K 6 M e W a 2 • 125 SII • 1 • 1 2 5 B
R E S I S T A N C E DATA F R O M 1:9.6 S C A L E M O D E L the basic mold. The pump uses a single-stage mixed-flow
F U L L - S C A L E p O W E R F R O M T o n G u E A N D R P M M E A S U R E M E N T S AT I N P U T
SHAFT TO WATER JET impeller and has an unusual feature whereby the vertical
inclination of the nozzle can be varied to change boat trim to
Figure 85. Propulsive Efficiency for Destriero - Supplied by suit load conditions, for instance, in getting over hump, or to
KaMeWa account for the lift due to the waterjet inlets at high speed.
Riva Calzoni (Riva Lips) (Italy). If KaMeWa is the giant
Kawasaki Heavy Industries (Japan) continue to make the of the waterjet industry, Riva Calzoni is its patriarch. Riva
former Rocketdyne PJ20 pump for the former Boeing Jet Foil Calzoni delivered their first waterjet in 1932 and today build
hydrofoil passenger craft. The PJ20 is discussed in more detail large custom-designed fabricated pumps as well as a more
later in the paper. standardized range of medium powered waterjet propulsors.
Kvaerner Eureka (Norway) gained their first order for Designs are offered up to 50,000 kW (67,000 hp). Pumps are
waterjets in 1990 from the Royal Norwegian Navy. Units up to fabricated in stainless steel and optimized for each customer's
4000 kW per shaft are offered. Kvaerna announced in 1991 that requirements. Like KaMeWa, IRC, as Riva Calzoni has been
they had developed a complete integrated gas turbine powered known, benefits greatly from an excellent research organization.
propulsion system for high-speed craft (Figure 86). The Riva Calzoni hydraulic laboratory continuously conducts
Marinepower (Turbodrive) (USA) is a small company studies on pumps, turbines and waterjets for the group's many
that supplies jet-drives for the boating industry, using activities and products. IRC 115 DX waterjet units are used in
automative-type engines. the 74 m wave-piercing catamarans built by International
Mitsubishi Heavy Industries (MHI) have introduced two Catamaran with an input power of 3860 kW (5200 hp) each.
waterjets with inducer-style double-row blading. The IRC 190 DL unit absorbs 15,000 kW (20,000 hp).

Recently, Riva Calzoni and Lips Propellers (the equations in 2 and 3 dimensions at the Norwegian Institute of
Netherlands) have joined forces to produce optimized waterjet Technology with the object of improving efficiency and
propulsion systems, and are now known as Riva Lips. Figure reducing noise and vibration.
86A shows a 3-D CAD drawing of a Riva Lips Type 115
1 !
waterjet inlet.
Figure 86A. 3-D CAD Drawing of a 115 Type Waterjet

Intake
Rocketdyne (USA) sold their waterjet interests to

Kawasaki Heavy Industries but it is believed that they reserved
the right to build large waterjets for the U.S. Department of
Defense should the need arise. Rocketdyne had brought their
PJ46, 40,000 hp, waterjet propulsor design to readiness for Figure 87. Schottel Pump Jet
manufacture and had started to make the first unit when the
3KSES program was cancelled. This pump was optimized for
speeds of 80 kt and above. The 3KSES was expected to achieve
speeds up to 100 kt at light weight. In this speed regime, the
PJ46 remains unchallenged in efficiency combined with low
size and weight for its power and thrust. It also had unsurpassed
suction performance at low speed. It is, however, totally
unsuited to current commercial high speeds in the 30 to 50 kt
region.
Schottel (Germany), Schottel of America (USA), builds
pump-jets intended principally for very low speed applications
! 650turn ]
and as maneuvering thrusters on account of their outstanding
bollard pull characteristics (Figure 87). However, they have
found favor as main propulsion units also in craft with speeds up Figure 88. SMP 165 Transom-Mounted Waterjet Propulsor
to 35 km/h (18.9 kt), including river cruise boats, patrol craft
and amphibious support vehicles. Power up to 1000 kW (1340
hp) are offered with thrusts up to 16,0130 lbf at bollard pull. A
unit resembles a centrifugal or mixed-flow fan with the rotatable . .iii.iii:iiiiiiiii
. .s< i i,iiiii'i'iiiiiii!ii!i
jii!:iii iiiiiiii!iif!
i!i::iiiiii;i!i!!iiiii::
::iiii::;;!i:i:::
iiiiiiiiiiiiiiiiii J i!iiiii;iiiii;ii:
volute and central inlet mounted horizontally in the bottom of
the ship. The volute discharges through a nozzle inclined
downward at 18 degrees to clear the bottom of the boat.
Because of its 360 degree rotation, exceptionally good
maneuverability is offered by the thruster. Several units can be
mounted in the hull forward and aft. The centrifugal-type pump
is resistant to cavitation and the unit can operate in very shallow
water. A grill helps to exclude weeds and debris. This
fascinating device has endless applications and occupies an
important niche in the panoply of waterjet propulsors.
Super Marine Power Ltd (United Kingdom), is a new
comer to the jet-drive scene. Their product, the SMP 165
transom-mounted waterjet unit is unique in that only part of the :i: i ili{
inlet duct actually penetrates the hull (Figure 88). The unit is i:!ii{::
designed for 170 hp with gasoline engines or 165 hp with diesel
engines. A special transmission allows two units to be used
side-by-side with a single engine of 350 hp, either gasoline or
diesel.
Ulstein Propeller A/S (Norway), offers waterjets in the
power range 300 to 5000 kW (400 to 6700 hp). They have an
unusual steering and reversing arrangement (Figure 89). Their Figure 89. External Features of the Ulstein-Liaaen Waterjet
inlet design has been analyzed by solving the Navier-Stokes Unit

Ultra Hydraulics Ltd (United Kingdom), has been inlet diameter squared. Obviously, this implies a certain nozzle
discussed under Dowty Hydraulic Units. It is noted that Ultra size but nozzle size is rarely available and the chart works
now has a different address from Dowty ("Jane's 1991" shows reasonably well for currently available large waterjets for which
the Dowty address). inlet diameter can usually be estimated if it is not known.
Scales for inches and centimeters are provided. The cavitation
7.0 W A T E R J E T P E R F O R M A N C E AND limit shown corresponds to a suction specific speed of N E =
APPLICATIONS 10,~100 and roughly corresponds to the continuous operation
limit previously described. For non-continuous operation it may
It is desirable that the naval architect should be able to be permissible to cross this line depending on the manufac-
make preliminary waterjet selections and performance turer's recommendation and past experience.
predictions without resorting to manufacturers. This is
particularly true for parametric performance studies of new or "tI
proposed vessels, both for naval (military) applications or
commercial use. Obviously, it may not be desirable to reveal
that such design studies are being undertaken. However, once a
preliminary design has reached a mature stage it is essential to ::I \ .................. s........
engage the waterjet manufacturer in the t'mal waterjet selection
process and ship performance predictions before the hull design
and prime mover selection are cast in concrete. These remarks
are generally true for all applications but are of the utmost
importance where large vessels are concerned. For smaller
craft, the situation is more relaxed as manufacturers of small
waterjets usually provide conservative waterjet performance
maps and have a range of impeller trim sizes available to match
the speed of any suitable engine-gearbox combinations. This
enables a fairly firm selection to be made for boat applications,
but it is still a good idea to consult the manufacturer at the
earliest possible time. 4S ~
4 [ soSs ~ ~r
7.1 PreliminaryWaterjet Performance Maps SHIP SPEED, KT I
$ - PUMP FACE INLET DIAMETER, CM
Some preliminary waterjet performance maps have been
developed for large mixed-flow waterjet propulsors, which will o.1 o.2 o.= a.4 0.5 O.e 0.t 0.8 0.9 1.0 1.1 1.2 1.3 H P I C M 2
enable the naval architect to make reasonable thrust and power 3 4 5 6 7 '. R,',i,'
predictions. It must be emphasized that these maps are generic POWER DENSITY, HPIE 2
in nature and are not for any particular manufacturer. In

selecting prime movers to provide the horsepower indicated by Figure 90. Generic Computer-Simulated Pump Map
the chart, the usual installed power margins should be applied.
For smaller waterjets, a thrust margin of at least ten percent Figure 91 is another generic performance map with
should be allowed as well as a power margin. Small vessels estimated operating zones superimposed. Figures 90 and 91
may not benefit to the same extent from some of the phenomena were independently constructed and are not necessarily
observed in large vessels with custom-designed inlets. consistent especially at the speed extremes. They are provided
Figure 90, which is based on a computer simulation, here for general guidance. Actual values must be obtained from
shows thrust per horsepower for craft speeds from 5 to 55 kt as a manufacturers.
function of power density det'med as horsepower divided by
4.0
_ ~.-~ ~ - ZONE t
3,0
T/S 2
Ibf/cm 2
2.0
~0.45
HP/S2 ~0.40
1.0
0"i
Figure 91. Generic Mixed-Flow Pump Map

KaMeWa has published a preliminary waterjet pump
selection guide, Figure 92. The KaMeWa range of waterjets
form a geometric series in which the model number indicates the
,!
approximate inlet diameter at the pump face, in centimeters. I
i/
==,ODD
BASEO ON PUBLISHEDKIMIWI S6~= DATA
Beginning with size 40, each succeeding size is about 1.122 /
30,000
times the previous size rounded to the nearest centimeter in most
cases. An exception is size 60 which is not offered as a standard
size. Table 5 from [22] shows a recent list of sizes published by
NOMINAL PUMP SPEED N - K HPV3 RPM
KaMeWa. Sizes as low as 32 have been provided.
SIZE
50" 50 63 71 80 90 100 112 125 1 =o,ooo
~ 140 NOMINAl. Iii,oo 0

NOmSePowE~
le,ooo
14,000
~ 20 ' . , . v. size 112.ooo

/
I to.ooo
10
,Io ]II,ooo
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 18 HP X 10- 3
30 +,ODD
'i ' 2' I '; ' ~' 6 ''7 ' ~I' 9 '1'0' 1'1' 1'2kWXIO -3
POWER
:o ~ +,ooo
Figure 92. KaMeWa Preliminary Size Selection Chart

t
o / .
40 SO
. .
SQ 70
.
80
.
90
. . . m. . . .
100 1tO 1~0 130 t40 15Q 160
Io
170 180
'
Table 8. Size Options for KaMeWaType 62 Waterjet S ~MP SlZlr.,C*C
Units [22]
Figure 93. Mixed-Flow Pumps - Speed/Power Relation
+ t t~ M~ 7
, , :~, . .-= +=
Table 6. Riva Calzoni (Riva Lips) Medium Power Range
. ~ '= :+.,+ m . '"-
Pumps
':~ ~ ,,'~ -~ :,': :1
i~s0,
\
",,\
Rlvl C81t0111IRC ~'yoe w81erle: UNIt
M e d i u m sized, standardiled p u m p units

Power range 200 to 1200k%t~
Model Max power Dry mass including
input integral air duct
kW ho kg
Figure 93 which was constructed only from data IRC 28 D ]50 -;75 200
published by KaMeWa, indicates the trends of horsepower and IRC 32 DL
IRC 36 DL
=50
660
600
900
240
350
pump speed with pump size. 1RC 38 D 600 800 500
IRC41DL 900 1200 760 °
Riva Calzoni (Riva Lips) have a more limited range of IRC 43 DL 950 1300 630
IRC47 DL 1200 1600 730
standard sizes but tend to build more custom designs. Table 6 "includes dl~onnectln= clutch
from [23] shows some model sizes. Riva Calzuni (Riva Lips)
has also published a list of applications from 1932 to 1988
giving both power and pump speed, rpm. Other manufacturers will provide pump maps on request
Hamilton provides pump maps m their brochures of or for specific projects.
which Figure 94 is an example. In estimating speed it is useful to remember that impeller
American Hydro Jet, the only current U.S. manufacturer tip speeds are usually limited to maximum values not exceeding
of pumps in the larger sizes, will provide pump maps on request. 185 ft/s although many pumps run at considerably lower t i p
Figure 95 is a pump map for their waterjet pump and Figure 96 speeds. In contrast, inducer pumps have a maximum tip speed
shows their Hydrojet Series 700 waterjet propulsor. of 200 ft/s. For preliminary design purposes .assume a

maximum tip speed of 150 ft/s at rated power, or use Figure 93,
for the actual power required.
13OO _ L_
[-28oo
I[I I I,, 4~
~o~
• 2z~ Figure 96. American Hydro Jet Series 1700 Waterjet
Propulsor
I • ,, I
)i°-
~-ieoog
7.2 Pump Selection and Engine Matching
Besides the desired thrust at the design operating point, it

is essential to have resistance curves for the vessel at maximum
~lllllllll
II IIII
I l i l l l l l l l I displacement including projected lifetime growth, and in high
o "r~" < sea states. These resistance curves are then superimposed on
candidate pump maps. If pump maps are not available, they can
Figure 94. Hamilton 291 Pump for Utility Craft be constructed from Figures 90 or 91 or the required thrust
computed from the resistance can be plotted directly on Figure
91 for different pump sizes. The best candidate pump is usually
i I the smallest size which will meet the design point thrust
,o ! cAviT..,o, u,,T i requirement without crossing the continuous operation limit at
9 i,'! i I! lower ship speeds. In high sea states operation in Zone II may
----4 be permissible.
----4 J i When the tentative pump size has been selected, the
power required at each ship speed may be read from Figure 91.
The required shaft speed may then be estimated from Figure 92
for a KaMeWa pump, or a rough idea can be obtained for any
other mixed-flow pump assuming a tip speed of 150 ft/s at the
design point and an impeller diameter about 40% larger than the
i v I-.7"-., \s inlet diameter.
I-- 5 ~.. ' ~ ' \1500 HP It will then be possible to make a preliminary engine and
gear ratio selection, and to plot the power and torque required
on the engine map. Since it is possible to apply full power to
the pump above a minimum ship speed which can be determined
by following the appropriate line of hp/s2 back to the limit on
Figure 91, the excess thrust available for acceleration may be
determined.
7.3 Examples of Preliminary Pump Selection and Sizing
Preliminary pump selection and sizing is illustrated in the

0 10 20 30 40 50 80
following examples.
CRAFT SPEED, KT
1. Figure 97 shows the predicted resistance of a semi-

Figure 95. AM JET Waterjet for a Motor Yacht planing landing craft at three displacements. The craft is to
have a top speed of 30 kt at its design gross weight and LCG
position. The pumps selected for waterjet propulsion should be
It is also useful to remember that pump flow rate and free from significant cavitation at the overload displacement in
pump speed (rpm) remain almost constant over the speed range sea-state 3.
above 20 kt,. for a given engine power but that power changes
very nearly as (rpm)3 when pump speed varies.
avoid pump cavitation, full power should not be maintained
below 20 kt except for intermittent operation. On no account
//-/
should full power be maintained below 11 kt.
/ /" OVERLOAD SS3
It will also be seen that the craft can reach its maximum
SO ~
// Z O N E III / ZONE II
J
speed at overload displacement in SS3 without crossing the
~ / / DESIGN
~'- -~__ ./ / / DISPLACEMENT
cavitation limits, using reduced power until 22 kt is attained, but
t ~ " - "-.~-'~> DESIGN 552 with an adequate excess of thrust over drag. The preliminary
design performance predictions may be refined by adding more
thrust curves for other powers and by including values of (1 - t)
30 CALM other than 1.0 to test the effect of this parameter, until an
appropriate value is provided by the manufacturer for this
hull-type.
~ 20
It is now necessary to look at the actual performance of
available pumps in the size range indicated. Strong candidates
would be a K a M e W a 90S, a Riva Calzoni (Riva Lips) IRC 82
/.// DLX, or an American Hydro Jet 2700/3500. However, the
possibility of using a smaller pump should be examined using
°1'o ~rs 2O is 3'o 3's Figure 91 with a higher value of hp/s 2. This would mean
CRAFT EPEED, KT
accepting a lower thrust margin.
Before leaving example 1, it is instructive to deduce some
Figure 97. Thrust and Drag - Example 1 other parameters for the candidate pump.
Assume the following typical values:
Resistance at Design Point = 37,400 lbf
Wake Factor (1 - w) = 0.95
Predicted Thrust Required = T = R/(1 - t)
Pump Efficiency, l~p = 0.90
Assume t = 0 initially
Inlet Recovery Factor, 1~i = 0.82
Thrust Per Pump (Assume 2 Pumps) = 18,700 lbf
Est. OPC (Initial Guess) = 0.56, i.e. f l i t " 1~, • Tip • lqi • (1 - t) Relative Rotative Efficiency 1], = 0.98
Initial power estimate hp = T V J 5 5 0 OPC Height of Nozzles = 2.5 ft

= 18,700 x 30 x 1.6878/550 x 0.56 Nozzle Area/Inlet Area = 0.40
= 3074 per pump
Nozzle Efficiency, l~n = 0.99
Assume l~t = 0.97, then shp/pump = 3074 x 0.97
= 2982 Density of Seawater = 1.989 slug/ft 3
Thrust per hp, T/hp = 18,700/2982 = 6.27 Aj = 0.4 x 7~/4 3.0282 = 2.88 ft 2 (Dj = 1.915 ft or 23 in.)
From Figure 90 hp/s 2 = 2.25 approximately.

T = m(Vj-Vw)
Pump inlet size for first trial S = 4 ~ = 36.4 in. or 92.5 cm T = V. 2 - V wVj ,.~
pAj J -
or S = 3.03 ft
The corresponding power density, hp/s 2 is 2982/92.3 = 0.35 V w = (1 - w) V = 0.95 x 30 x 1.6878 = 48.1 ft/s
hp/cm 2. Also, the thrust density T/s 2 is 18,700/92.32 = 2.2
lbf/cm 2. 2Vj = V w + (Vw2 + 4T/O Aj)
It is now possible to plot the thrust curve through the
design point, using values o f T / S 2 obtained from Figure 91. = 48.1 + (48.12 + 4 x 18,700/1.989 x 2.88) 1/2
Jet Velocity, Vj = 86 ft/s

Vk (kt) 15 20 25 30 35
Flow Rate, Qj = 247.8 ft3/s or 11,200 gpm
T/S 2 (lb/ft 2) 2.59 2.48 2.37 2.2 2.02
Mass Flow Rate, m = 15,589 lb/s or 492.7 slug/s
T (lbf) 22,100 21,100 20,200 18,700 17,200
(1 pump) Jet Velocity Ratio (Wake), ILl.w = V j~/j = 0.559 or a = ~-1 =
1.788
2T (lb) 44,200 42,200 40,400 37,400 34,400
(2 pumps) Jet Velocity Ratio, ~1. = 50.63/86 = 0.5888
Ideal Jet Efficiency, 1"11= 21.1./(1 + ~.1) = 0.741

It will be seen that the thrust curve passes through the Zone I
limit line at about 22 kt and the Zone 11 lhnit line at 20 kt. To

1 2~t (] - r*) Absolute Specific Speed, N = 6790/17,170 = 0.395 _Z-0.4
Actual Jet Eft., TIj = ~ " 1 + g t - (1 - 4) t-t2 + 72g hj
Suction Specific Speed, N = NQI:Z/NPSH3/4
J
Net Positive Suction Head, NPSH = Ill VwT/2g - h i - H v
where [.1.= 1.1.w since Cp and CDI are unknown at this stage. Assume h i = hj, then NPSH = 29.5 - 2.5 - 0.5 = 27.5 ft
1 2 x 0.559 (1 - 0.559) and N , = 600 x 11,2001:z/27.53/4 = 0.308

Therefore, Tlj = ~ ° 1.01 - 0.82 x 0.5592 + 2g x 2.5/862
Also, (Jp = NPSH/H = 27.5/91 = 0.302
TIj = 0.669
Thus, the pump is operating at a very conservative
If the assumptions made are correct, the estimated OPC condition relative to cavitation when at its design point. This
should agree with the value calculated as follows: situation changes drastically if full power is maintained down to
20 kt.
OPC = Tit ° Tip" 1"1, ° (1 - t) Tlj (1 - t) assumed to be 1 To a first approximation, H remains constant at constant
power but NPSH is reduced at lower speed. At 20 kt NPSH is
OPC = 0.9 x 0.98 x 1.0 x 0.669 only about 11.5 ft so N~, increases to 10,150 which is the limit
OPC = 0.57 (0.56 was estimated). for continuous operation, or close to it. Also, (~p decreases to
The calculations can be repeated for the revised value of 0.126, a value at which some cavitation is unavoidable.
OPC if desired.
For the application in Example 1, a suitable engine would 7.4 Selection of Engine and Gearbox
be the M T U 16V 396 TB94 (two required) with ZF or Reimjes
gearboxes with 3.43:1 reduction ratio. A suitable ZF gearbox is Most waterjet propulsors are powered by high-speed
the BUK755 coaxial box with integral clutch. The approximate marine diesel engines, although in some applications marine gas
weight of this box is 1530 lbf. Alternatively, the Paxman turbines are preferred because of their low weight and small size
Valenta 18RP200-0-CM might be chosen with a 2.5:1 reduction especially at very high powers. Presently available marine gas
ratio. Many other selections are possible. turbines cannot match the specific fuel consumption of diesel
engines, so once again there is a trade-off involving weight,
O t h e r Parameters size, initial cost and fuel cost.
Approximately impeller diameter = 1.4 x S = 1.5 x 36.34 = 50.9 Work Boat World publishes an annual table of diesel
in. = 4.24 ft. engines, in order of power, showing the number of cylinders,
shaft power, speed, torque, specific fuel consumption, dimen-
If maximum tip speed is to be 150 ft/s, N = 150 x 60/71; x 4.24. sions and weight. It is not difficult to find an engine to match
Pump speed should not exceed N = 675 rpm. any likely waterjet requirements. Incidentally, the table also has
an attachment showing some leading waterjet manufacturers
Alternatively, for a KaMeWa 90S pump, K = 41 (Figure 93). with details of a few of their pumps giving minimum-maximum
Therefore, N = K shp 1/3 = 41 x 29821/3 = 590 rpm. power range, recommended power, dimensions & weight.
Diesel engine makers which spring to mind in connection
The first answer is appropriate to a pump average power with larger high:performance waterjet vessels include Detroit
for this size, but the actual power is considerably less than Diesel Allison (DDA), MTU, and Paxman. For marine gas
average in this case. For instance, according to Figure 92, the turbines, Textron Lycoming, Stewart & Stevenson (Allison Gas
rated power is about 6000 hp for a size 90 pump which would Turbines) and, of course, General Electric are among the best
then have a speed of 745 rpm with a tip speed of about 166 ft/s. known.
For preliminary gearbox and engine selection assume 600 Waterjet pumps do not usually require reversing
rpm initially. gearboxes*, therefore, the transmissions are lighter, less costly
and smaller than those for propellers. Well-known manufac-
Specific Speed, N = NQlrZ/H3/4
turers include Cincinnati Gear, ZF and Reintjes who have
developed a new gear design for waterjets. In the smaller sizes,
Pump Developed Head = 3.)j:/2 g Tin - Tii Vw2/2g + hj
Twin Disc, MPM and Nigata are just a few. Transmissions are
H = 86:/2g x 0.99 - 0.82 x 48.12/2g + 2.5 listed in "Jane's High-Speed Marine Craft".
H = 116.1 -29.5 +2.5 7.5 Example 2

H = 89.2 ft
Determine the maximum braking force at 30 kt, for the
or, based on actual power, H = 550 hp x Tlp x Tle/(m lb/s) 92.8 ft craft in Example 1.
Average, N, = 600 x 111,2001/a/91 TM = 6790 _-6800
*Some operators reverse the pumps to try to back-flush the inlet grilles to remove debris such as plastic bags or seaweed.

According to the hypothetical diagram in Figure 83, the The question of provision, or not, of an inlet screen must
astern nozzle thrust is about 43% of the ahead gross thrust at the be addressed early on. A screen may exclude some types of
same power, therefore, the astern nozzle thrust at 30 kt would debris which could damage the pump. On the other hand, small
be: objects pass through the pump without harm and collision with
objects too large to pass through the pump may break the screen
T n = 0.43 x m Vj = 0.43 x 492.7 x 86 x 2 = 36,440 lbf which is then ingested by the pump possibly with serious
consequences. This happened to the American Enterprise when
Add to this the inlet momentum drag: a piece of the screen fractured a pump casing causing the
machinery space to be instantly flooded. Inlet screens can also
D m = m V , = 492.7 x 48.1 x 2 = 47,400 lbf.
significantly reduce inlet performance. In cases where the pump
rotation can be reversed, the screens can be flushed when the
Total Braking Force = T + D m = 36,440 + 47,400 = 83,840 lbf
boat is stationary. On balance, screens are probably desirable
If the gross weight of the craft is 250 LT (560,000 lb) the initial for boat applications but are not necessary for large waterjet
rate of deceleration, pumps.
Seawater is the enemy of all things mechanical on marine
a = F/m = 83,840 x 32.174/560,000 = 4.82 ft/s 2 craft. Corrosion can be minimized by the use of stainless steel
for pump casings, or by the use of composite non-metallic
i.e., a = 0.15 g (nearly 3 kt per second)
materials. All exposed mechanisms for steering and reversing
By integration the stopping time and distance may easily be including hydraulic cylinders, hoses and linkages must be
calculated. Remember, power must be decreased once the ship suitably protected from the corrosion and abrasive effects of
has slowed to 20 kt. seawater, sand and marine growth. Particularly since these
systems often remain close to the water surface with the craft at
7.6 Example 3 rest.
Figure 95 shows a pump map for an American Hydro Jet

8.0 NOISE
AMJET 1700 Series pump with an inlet diameter of 17 in. and a
nozzle diameter of 9 in. This type of pump is shown in Figure
Waterjets are quieter than propellers and therefore
96. Superimposed on the map is the estimated resistance curve
waterjet-propelled vessels are more difficult to detect by sonar
for a 12 meter motor yacht in calm water at its design displace-
than propeller-driven ships or small craft. This fact has obvious
ment approximately 12 long tons. It requires 1060 horsepower
military advantages. Figure 98, published by Schottel,
to reach its design speed of 40 kt. The resistance at higher
compares the measured underwater noise of two push-boats with
d!splacements and in head seas is not known but this pump
the same thrust; one propeller driven, the other with the Schottel
provides an exceptionally good cavitation margin, as full power
pump jet. Figure 99, published by KaMeWa shows sound
can be maintained down to about 14 kt in this application. The
measurements for the JET CAT waterjet-propelled catamaran.
pump has a maximum speed of 2300 rpm so the design speed
Unfortunately, very few data are available for propellers. Figure
will be about 2050 rpm, suitable for direct drive with one of
100 shows an analytical comparison of the underwater noise
several engines including an MTU 8V396TE74 and a Stewart
predictions for waterjets and subcavitating propellers for the
and Stevenson V16. Many other engines are available in this
same ship.
power range if a gear is used.
aft 140
7.7 Example 4 li!flll I IIIlill I I!1[[111 ! fFlili!l
130 I IIIIjJ I ~t ill ~ I I I
• i~ i!!i
Figure 94 shows a pump map for a Hamilton 291 with the ( }. t II I II I Iit
120
resistance curve for a new type of utility craft superimposed. iJ"l~ I likitl I I IIIlill iailJ
I~ ~ q llK iTYprcALPROPELLERI~
110 IPUIIK i i 1 ~ SPECTRUM LEVEL nl
Hamilton offers a wide range of impellers so that it is not
100
difficult to match the pump to available gearboxes which can be i lilt "~\ ift'~,L,-'I-4J N
90 illliil "~A~Z4111 I t ilill!l ;'T"4J.]lidi
used with a specially-rated automotive type of engine. In this i111111 It1,~ I I il!t!ll I l i I~
case, a Crusader 454 engine was selected with a Borg Warner 80
Ill;Ill t III%".~L I II I Ij lull
[!!! I Ill!ill "~J~ 1 I
transmission. 70 'IH, l' Illl ,~"H.~
t r~ i ,ll ,,1
i III ! l illlii I I I IlIlt'k'~."r:h'.qJ III
60 ~l~llt~ l!!!!llt I111!11!!~J'~tu"l
I_[.~JYY~CAL
I SCHOTTEL PUMP-JET"
7.8 WaterJet Installation Considerations I lJ]!t SPECTRUM LE~/EL ) I II!,FII
50
j III II P tllllil I IHItll
40 i llill II I I i Iq
Many waterjet pumps, including KaMeWa, are transom- I IIIh I t llJl li II
30
mounted and the thrust is transmitted directly to the transom 100 tO00 tO000 100000
FREQUENCYIN HERTZ
which of course must be strong enough and stiff enough to take
both the thrust and the weight of the unit. Others, including Measured spectrum of underwater sound generated by
Hamilton, are hull-bottom-mounted with a load-bearing integral, two push-boats wtth the same thrust, wRhout any special
or bolted, inlet duct. Some pumps are structurally connected to snienong measures.
the hull at beth transom and bottom. The type of mounting and Propeller
the thrust path are design considerations when a new waterjet- SCHOTTELPump-Jet
propelled ship or boat concept is being developed, as the local
hull shape and scantlings are affected. Figure 98. Comparison of Waterjet and Propeller Noise
316 M a r i n e Waterjet Propulsion

We can expect to see:
3C 1
Larger numbers of waterjet-propelled ships world-wide,
particularly catamaran and SES ferries
~0 ..~_,,o~ 30k°;q. " ~--L ~'~-'~ °-r . . . . . . _ )
, "t~" ~%1 ' . . " ~ ]0 knots Larger powers with smaller pumps for the same or higher
efficiency than at present
Lighter weights at all power levels
Higher efficiency particularly in the smaller sizes of pump
" ~ i
Improved steering and reversing gear of smaller size and
I "" " L., .': 5 ~n~ts ---~, r
-io less weight
I .,,. ,,
Renewed development of inducer pumps for special

-20
20 :O ~O 160 31S 5 ::0 !.25 23 5 10 20 i'0 applications initially, expanding later into other market
t6 ]l,S $3 125 250 500 1KHt 2 /- $ ;6 31.5
125 ~ 5C lOO 200 LO0 riO0 ~8 3~S ~.3 125 25 areas. Figure 101 compares large inducer and mixed flow
pumps at very high speeds.
Figure 99. Hydro-Acoustic Performance of the JET CAT Shaffless pumps as an out-growth of the Westinghouse
Installation. Sound Pressure in 1/3-Octave Band shaftless propeller development
at a Hydrophone Depth of 10 m
Renewed public confidence in waterjet propulsion for
CONCEPTUAL SES UNDERWATER NOISE
pleasure craft and fishing boats
Wilh Sub-Cavilltln9 Prop elle¢$
A big increase in the U.S. industry as waterborne mass
transportation around urban areas expands
Motion control via active waterjet deflection
Continued growth of military applications in the U.S.
following the success of the SES 200 and other current
o waterjet applications.
=o
MIXED FI.OW PIJMP
,6 ~. s ~ tls :so soo i, al ,~ s~ i~,

O¢lave C;In(ec F r l ~ u e n c y . HZ
CONCEPTUAL SES UNDERWATER NOISE

With WATER JETS
F ~ = , a Sa*,,l
t
We (MIX.FL)DRY( * 4 = 'Np (INO.)ORY APPROX.
I
Wp (MIX,Ft..) WET - 4.1 I Wp (INO,) WET APPROX. I
IO ~o 30 40 so ~o 7o Io Jo
SHIP SPEEO KT
Figure 101. Comparison of Large Inducer and Mixed Flow

Octave Ceclle¢ Frequency. HZ
Pumps at Very High Speed
Figure 100. Analytical Comparison of Propeller and Wterjet Figures 102 through 104 specially provided for this paper
Noise [BLA] by Mr. Roll Svensson, show a statistical analysis of the growth
of the waterjet industry reflected by KaMeWa's experience as
For passenger ferry service, the noise environment in the the world's leading manufacturer of large waterjet propulsors.
cabin is important, and has been found to be significantly lower
on waterjet-propelled passenger vessels, especially at the lower 10.0 CONCLUSIONS
frequencies.
A broad review of marine waterjet propulsion including
9.0 FUTURE TRENDS the underlying theory and latest results has been presented.
Currently available waterjet propulsors have been reviewed..
There is every indication that the tremendous growth of Examples of their application have been presented. Practical
interest in waterjet propulsion will continue as waterjets replace design considerations, advantages and challenges have been
propellers in more and more applications.
discussed. Future trends have been forecast which include
continuing growth in the number and capability of marine Total number o/ihstctlar;ons : ~bt. 320
waterjet propulsors. O/splccemenc : u~ to 2000t
~./ze o / P/J : uo :o /dO SZ[
~c'wer : u.o bo ]OOO0 H~./znc/b
,~ e'lax. Po~er I w ' ; - u n , t
I dehvered c~ur~nq ctcrua/yectt" Speed : uo to ~3 K n o t ;
[kwI rzoe ~.#j___L

%
Ii. ZOO00
I
57%
---..~.
50
tSO00
36%
I
---7-
25°/°
lO000
20
I
-.4 i
/o% /
5000 I
i Figure 104. Breakdown of Waterjet Applications

(KaMeWa Experience)
It is hoped that this paper will be of assistance to naval

/ea~ o# d e l i v e t ' y architects and designers in becoming more familiar with the
field of waterjet propulsion and in making preliminary
Figure 102. Growth of Waterjet Industry (KaMeWa performance estimates. Enough references and bibliography
Experience) have been provided to enable almost any aspect of this subject
to be pursued in more detail.
•~ax. lie# c~ '~J.u~t
dfhvcrl¢ ~or~nq fttefltye~r 11.0 ACKNOWLEDGMENTS
i 180
The author wishes to thank all those friends and

colleagues whose valuable contributions and help have made
this paper possible. He is indebted to all those whose work has
been drawn upon in compiling the paper. References have been
given wherever possible.
Thanks are due to those waterjet manufacturers who
kindly provided technical information about their products not
otherwise available, in particular, Mr. Roll Svensson of
KaMeWa, Mr. Phil Brodie of Hamilton, Mr. Paul Roos of
l~ 125 AMJET, Mr. Malcom Breeze of PPjet, Mr. Ole Midttun of
Bird-Johnson (Castoldi), also Mr. Phil Wassinger of MTU, and
t 111 all other companies who sent material or whose brochures were
used.
Special thanks are due to Mr. Jack Offutt, Mr. John
. I 100
Stricker and Mr. John Purnell of CDNSWC, for help and
encouragement
Lastly, Mr. David Lavis without whose help this paper
would never have been undertaken, and Ms. Linda Peters for the
technical publication tasks which she has so excellently
J 71
accomplished.
I t t ~ l - #l [ - ~l,'
r'iar o/CebrIP¢~ 12.0 REFERENCES
Figure 103. Growth of Waterjet Pump Size (KaMeWa 1. Dickinson, H.W., "A short History of the Steam Engine,"
Experience) Cambridge, 1938.

2. Allison, J.L., "propellers for High-Performance Craft," 17. Pumell, John G., "The Performance Gains of Using
Marine Technology, Volume 15, No. 5, October 1978. Wide, Flush Boundary Layer Inlets on Waterjet Propelled
Craft," DTNSRDC, Report PAS-75-45, March 1976.
3. Swanek, Richard A., Johnson, Virgil E., Jr., Altmarm,
R.J., and Kobayashi, Sukeyuki, "Water Piston Propulsor 18. Svensson, Rolf, "A Description of the Waterjets Selected
for an Advanced Marine Corps Amphibian, Paper No. for "Destriero"," Fast 91.
89-1537 CP, 1989.
19. Payne, Peter R., "Design of High-Speed Boats: Volume I
4. Saunders, Harold E., "Hydrodynamics in Ship Design, - Planing," Published by Fishergate, Inc., Annapolis,
Volumes I, II and HI," SNAME, 1957. Maryland, 1988.
5. Brandau, John H., "Aspects of Performance Evaluation of 20. Terwisga, Tom van, "I'he Effect of Waterjet-Hull
Waterjet Propulsion Systems and a Critical Review of the Interaction on Thrust and Propulsive Efficiency," Fast 91,
State-of-the-Art," NSRDC Report 2550, October 1967. Trondheim, Norway, June 1991.
6. Kim, H.C., "Hydrodynamic Aspects of lmemal Pump Jet 21. Lavis, David R., and Benton, H. Schaub, Jr., "SES-100A
Propulsion," University of Michigan, Ann Arbor, 1964. Waterjet Inlet Fence Model Test and Evaluation Report,"
Payne, Inc. Report No. 159-10, March 1976.
7. Barr, Roderick A. and Etter, Robert J., "Selection of
Propulsion Systems for High Speed Advanced Marine 22. Trillo, Robert L., "Jane's High-Speed Marine Craft
Vehicles," AIAA, Paper No. 74-334, San Diego, 1991," 24th Edition, 1991.
California, February 25-27, 1974.
23. Trillo, Robert L., "Jane's High-Speed Marine Craft and
8. Etter, R.J. et al., "Model Testing of Waterjet Propelled Air Cushion Vehicles," 20th Edition, 1987.
Craft," Proceedings of the 19th Meeting of the A'Iq'C,
1980. 24. Svensson, Roll', "Waterjets - Propulsion for Naval
Vessels," Navy International, July/August 1987.
9. Svensson, Rolf, "Experience with the KaMeWa Waterjet
Propulsion System," AIAA, Paper No. 89-1440-CP, 25. Lalangas, P.A. and Yarmoulis, P.L., "Design and
1989. Construction of a 25 m High Speed Aluminum Motor
Yacht," SNAME Transactions, Volume 91, 1983.
10. Dorey, A.L., "High Speed Small Craft," The 54th Parsons
Memorial Lecture, RINA, 1989 13.0 NOMENCLATURE
11. Moore, Robert C. and Bender, Gregory, L., "The acceleration

Evolution of the U.S. Navy SES-200," HPMV 1992. a = inverse of jet velocity ratio, B-1
a -- axial induced velocity factor (Figure)
12. Wislicenus, George F., "Hydrodynamic Design Principles A = area
of Pumps and Ducting for Waterjet Propulsion," NSRDC, A¢ = capture area for waterjet
Report 3990, June 1973. A cap = capture area for propeller
Aj = jet area at vena contracta
13. Karassik, Igor J. and Krutzsch, William C., "Centrifugal
and Axial Pumps," Marks' Standard Handbook for Aj = minimum slipstream area for propeller
Mechanical Engineers, 8th Edition, McGraw-Hill Book A n = nozzle exit plane area
Company. Ani = nozzle inlet area
Ap = impeller disc area, propeller disc area
14. Balje, O.E., "Axial Cascade Technology and Application
b = velocity increase fraction behind propeller
to Flow Path Designs, Part H - Application of Data to
bhp = brake horsepower developed by engine
Flow Path Designs," Journal of Engineering for Power,
B = breadth of bottom plating
ASME, October 1968.
CDi -- inlet drag coefficient
15. Haglund, K., Svensson, R. & Bjorheden, O., "Design and C o -- Ratio of NPSH to NPSH at head breakdown
Testing of a High-Performance Waterjet Propulsion Cp = pressure coefficient
Unit," Second Symposium on Small Fast Warship, RINA, CT -- thrust coefficient
London, May 1982. D = nominal inlet diameter, propeller diameter
Di = inlet drag
16. Stricker, John G., Becnel, Alan J. and PumeU, John G.,
Dm = momentum drag
"Development of a Waterjet Propulsor for the Marine
Corps High Water Speed Landing Craft Application," Dp = pump diameter
23rd American Towing Tank Conference, New Orleans, D s -- specific impeller diameter based on flow
June 1992. D = specific impeller diameter based on power

ehp = effective horsepower (of hull) T = effective thrust
E = energy rate, E", E", E .... S = pump size
Er = energy recovered by inlet (rate) T = thrust
FM = figure of merit Te = gross thrust of jet
Fs = steering side force TN = net thrust on ship
g = acceleration due to gravity u2 = impeller tip speed
go = gravitational constant Um =
meridional (axial) velocity
hi = depth of the inlet, elevation of pump shaft
Ut = tip speed, also Utiv
eenterline above SL Ut =
blade tip speed at inlet (Figure 21)
h. = height of jet at vena contracta
J
blade tip speed at exit (Figure 21)
= static head at the pump inlet U2 =
V = velocity
H = pump head v = velocity of approach
HAT = head due to atmospheric pressure v. =
slipstream velocity of propeller
J
H i = ideal head from Euler equation v =
p velocity through propeller disc area
Hv = head due to vapor pressure vw =
mean wake velocity approaching inlet
H* = pump head divided by free stream velocity head
V = velocity
HP = horsepower V =
velocity of approach
ivr = inlet velocity ratio, also IVR •
J = advance ratio of propeller V = axial velocity
JVR = jet velocity ratio, same as a = ~1,"1 Vi = velocity at inlet plane
k = inlet loss fraction, ~ Vi = jet velocity (at vena contracta)
K = factor defined in 3-37, pump speed coefficient Vk = ship speed, kt
(Figure 93) V = momentum velocity of incoming flow
KAN = nozzle area coefficient m
Vn = velocity at nozzle exit plane
Kr~ = speed coefficient Vo =
free stream velocity ( V )
K = power coefficient V =
P velocity at pump inlet plane, also V
i% = thrust coefficient for propeller, T/On2D 4 P
ship speed
thrust coefficient based on flow V =
Vri = blade relative inlet velocity
thrust coefficient based on power
Yst = velocity at stator inlet plane
mass flow rate V = tangential velocity at r I (Figure 21)
U1
mi = inlet mass flow rate V = tangential velocity at r 2 (Figure 21)

%
jet mass flow rate Vw = effective velocity of approach (with wake)
m. = nozzle mass flow, Vu1 velocity vector (Figure 21)
n -- rotational speed, rps V = velocity vector (Figure 21)
%
N = rotational speed, rpm --
specific speed V average velocity of ingested flow (3-40)
=
N s =
energy velocity of incoming flow
N u = suction specific speed V* =
W = Taylor wake fraction
NPSH = net positive suction head (1 - w) = Taylor wake factor
OPC -- overall propulsive coefficient Wo = energy (work) to lift water to height h i (rate)
inlet static pressure at an impeller blade
Wp = pump weight
Pv -- vapor pressure
P = pressure, p u m p pressure Greek Letters
PC = propulsive coefficient
Po = free stream pressure far ahead of propeller O~ = pump shaft vertical inclination
Ps = static pressure in front of inlet (Xo = steering angle
Pl -- pressure in front of propeller disc ~ = inlet duct vertical inclination to pump shaft
P2 -- pressure behind propeller disc AE = change of energy (rate) also A E "
q = dynamic pressure Ah = height of jet at vena contracta (hi)
Q = flow rate ~ = loss coefficient, inlet loss coefficient
R = resistance (drag) 1"1 = efficiency
Rt = towed resistance 1"1o = propulsive efficiency, 1]J "TIp
t = thrust deduction fraction 1]H = hull efficiency, (1 - w) / (1 - t)
( 1 - t) = thrust deduction factor 1]i = inlet efficiency
shp = shaft horsepower at pump 1lid = ideal propeller efficiency

T~j = jet efficiency of propulsor pump horizontal inclination
~. = nozzle efficiency ¢ = flow coefficient. Q/V=U, or Q/nD 3
31o = open water efficiency of propeller V = nozzle loss coefficient
= pressure coefficient, PIpU, 2 or H/[3nZD =
11oA = overall efficiency
~p = pump hydraulic efficiency, also T~pump
Metric Conversion Factors
11, = relative rotative efficiency, also 1"1,,
Tit = Izansmission efficiency, shp/bhp lft = 0.3048m
0 = jet deflection for steering 1 ft3 = 0.02832 m 3
~L = jet velocity ratio VJVj or Vw/V j (VJVj) 1 in. = 2.54 cm
X = ratio of circumference to diameter of a circle 1 long ton = 1.016 tormes
[3 = mass density ! mile = 1.609 km
O = cavitation number 1 shp = 0.7457 kW
OH = Thoma cavitation parameter, NPSH/H 1 lbf/ft2 = 47.88 Pa
Op = pump cavitation number 1 lbf = 4.448 N
'U = cavitation parameter

APPENDIX A
SOME HISTORICAL EVENTS PERTINENT TO MARINE WATERJET DEVELOPMENT
287-212 BC Archimedes Axial Water Pump

1452-1519 Leonardi Da Vinci Axial Water Pump, Screw Propulsion
1631 Ramsey, David Patented Steam (WJ) Ship Propulsion
1642-1727 Newton, Isaac Principia 1687
1661 Toogood and Hayes Patent for Archimedian Screw WJ
1680 Hooke, Robert Archimedian Screw Propeller
1706-1790 Franklin, Benjamin Proposal for WJ Propelled Boat, 1775
1736-1819 Watt, James Reciprocating & Rotary Steam Engines
1782 Rumsey, James Wl Propelled 80 ft Ferry Boat (Potomac)
1787 Kempelen Steam Turbine
1791 Barber Patented Gas Turbine
1802 Symington Charlotte Dundas Steam Tug (Clyde)
1807 Fulton, Robert Clermont Steamship (Hudson)
1812 Bell Steamship Comet (Clyde)
1836-1845 Smith Screw Propellers, England
1840's Ericsson Marine Propellers (Including Contra-Rotating), Sweden/USA
1853 Ruthven, John WJ Ship Enterprise (Not Successful)
1853 Seydell WJ Ship Albert (Successful) (Oder)
1863 (British Admiralty) WJ Ship Nautilus, 10 kt (Thames)
1863 (British Admiralty) WJ Ship Waterwitch Versus Viper
1870 Ramus, C.M. Planing Boat Rocket Propulsion
1878 (Swedish Government) Comparative Trials WJ and Propeller
1880 Thomeyemft J1 Pump-Jet Propulsion (Exterior)
1894 (Royal National Lifeboat Institution) WJ Lifeboat
1932 First Riva Calzoni WJ
1959-1967 Cambell, Donald Bluebird, Gas Turbine Jet Propulsion
1952 (Etablissements Billiez) WJ Ferry (France)
1954 First Hamilton WJ New Zealand Rivers
1968 First KaMeWa WJ Mixed Flow Pump WJ
1968-1972 Tucumcari (PGH-2) Boeing/Centrifugal Pump
1971-1980 SES 100A 74 kt (1978) Inducer Pump WI
1971-1983 SES 100B 92 kt (1977) (Surface-Piercing Props)
1973-1978 2K/3KSES Most Powerful WJ Development
1974 - Present PHM Boeing/ALRC Inducer/Mixed Flow, 2-Speed, Coaxial Shaft Pumps
1974 -Present Jetfoil Boeing/Kawasaki Inducer/Axial Pumps
1989 Riva Calzoni Atlantic Challenger WJ
1990 SES 200 WJ Conversion
1991 KaMeWa 180 Sll Largest Current WJ Built
1992 SEC SES Largest WJ Ship
1992 Yamato First MHD Waterjet Ship
1992 Destriero Atlantic Speed Record with KaMeWa Waterjets

APPENDIX B
(See Paragraph 3.3)
EXAMPLES OF JET PROPULSIVE EFFICIENCY

ACCORDING TO KIM, 1964 [6]
WARNING: This table no longer represents the state-of-the-art.
D = nozzle diameter ~lj = jet efficiency

~e = pump efficiency PC = propulsive coefficient
Nozzle r Loss factor " x
diameter / - - - - - - 0.50 p - ~ f ~ 1.00
Type of ship (m) ~j nj
0.05 0.59 0.53
5.1-m power boat 0.10 0.60 0.54 max
30 knots 0.15 0.53
~lP = 0.90 0.20 0.45 0.41
0.25 0.30 0.27
1.83 0.55 0.47 0.49 0.42
161.0-m cargo ship 2.44 0.59 0.50 0.50 0.43 max
25 knots 3.05 0.60 0.51 max 0.49 ~l-~
~p = 0.85 3.66 0.57 ~ 0.43 0.37
1.22 0.58 0.50 0.50 0.43 max
Destroyer 1.83 0.60 0.51 max 0.49
33 knots 2.44 0.57 ~ 0.43 0.37
~p = 0.85 3.05 0.52 0.44 0.38 0.32
3.66 0.46 0.39 0.32 0.27
1.22 0.38 0.32 0.36 0.31
Cargo ship 1.83 0.48 0.41 0.46 0.39
16 knots 2.44 0.55 0.47 0.49 0.42
~lp = 0.85 3.05 0.59 0.50 0.50 0.43 max
3.66 0.60 0.51 max 0.49
Hydrofoil boat 0.15 0.51 0.46 0.47 0.42

60 knots 0.31 0.59 0.53 max 0.48 0.43 max
0.61 0.42 ~-g 0.30
~lP = 0.90 0.92 0.28 0.25 0.18 0.16

APPENDIX C
OTHER USEFUL RELATIONSHIPS
Many relationships can be derived from the dimensionless pressure and flow coefficients ~q/and ~). The following will be found
useful. The pressure coefficient used here is P/PUt 2.
Specific Diameter:
D= = DP'n/Q u2 (2= pl/'ml/~) V1/'/~ ~ (B ~.,d on flow) (4-11)
Specific Diameter:.
D = D P-V4/Hplrz= (550 Tlp)xtz D (Based on power) (4-12)
Speed Coefficient:
iq, = NQ/DP = 15 ~/PV(N win) (4-13)
Specific Speed:
N, = n Qln/p~'4 = (2/~lrz p3/4)-1 (~u2/~/-3/4(n rps)

(4-14)
Thrust Coefficient:*
K.r = DT,a/Q = (p,a 2~,/4/~1/2)V1/4/¢1/2 (Based on flow) (4-15)

Thrust Coefflcient:*
K.r~ = D HP/T 3a = [2/925 (p/t) 'rz] ~/V4/Tlp (~,n (Based on power) (4-16)
Nozzle Area Coefficient:**
KAN = An/D 2= E/4 ~/(2~1/) u2 (4-17)
Power Coefficient:
Kp = 550 hp/p/U4 D 2 U,3 = ~ ~J/fqp (4-18)
Efficiency:
Tip = P Q/550 hp = ~ ~//Kp (4-19)
* For ideal static thrust based on head rise throug h the pump.
**Gives ideal nozzle area for a given pump operating point defined by (~ and ~tt at static conditions.

APPENDIX D
WATERJET PROPULSOR MANUFACTURERS WORLD-WIDE
Horsepower Range
Company Min - Max
I i
Aerojet (ALRC)/Aerojet Tech Systems Co. See Note in Text

P.O. Box 13222, Sacramento, CA, 95813, USA
TEL: (916)355-3011
American Hydrojet Corporation 200-5000

2033F W. McNab Rd, Pompano Beach, FL, 33069, USA
TEL: (305) 978-6676; FAX: (305) 978-6597
Berkeley Pump Company/Berkeley Jet Drive 50 - 600

2222 5th Street, Berkely, CA, 94710, USA
TEL: (415) 549-1808; FAX: (415) 549-2610
Castoldi S p A 15 - 1300
Viale Mazzini 161 - 20081 Abbiategrasso, Milano. Italy
TEL: (2) 94821; FAX: (9) 4960 800
Doen Marine Pry Ltd

20 Jannali Drive, Dingley, Vic, 3172, Australia
TEL: (03) 551 3048; FAX: (03) 568 3922
Dowry (see Ultra) See Note in Text
FF Jet Ltd AB 65 - 1400

P.O. Box 79, 67101 Kokkola, Finland
TEL: 358 68 21505; FAX: 358 68 21435
Hamilton, CWF & Co. Ltd 300 - 4000

Lunns Road, Christchurch, New Zealand
TEL: 64 (3) 486 969; FAX: 64 (3) 348 6969
Jacuzzi (see North American Marine Jet Inc.) See Note in Text
KaMeWa AB 600 - 32,000

P.O. Box 1010, S-681 Kristinehamn, Sweden
TEL: 46 550 840 00; FAX: 46 550 181 90
Kawasaki Heavy Industries, Ltd 3800

1-1, Higashi Kawasaki-cho 3-chome, Chuo-ku, Kobe, 650-91, Japan
TEL: (078) 682 5321; FAX: (078) 682 5549
KEM Equipment, Inc. 50 - 350

10800 SW Herman Road, Tualatin, OR, 97062, USA
TEL: (503) 692-5102; FAX: (503) 692-1098
Kvaerner - Eureka M S 6OO0 - 32,OOO

P.O. Box 38, N-3401 Lier, Norway
TEL: (03) 859000; FAX: (03) 850475
Marine Jet Power (MJP) 1300 - 12,000

S-740 63 Oste~bybruk, Sweden
TEL: (46) 295 20785; FAX: (46) 295 21383

Horsepower Range
Company Min - Max
Marine Power, Inc./Turbodrive 145 - 575

Ponchatoula, LA, USA; TEL: (504) 386-2081
Mitsubishi Heavy Industries (MHI) 2850 - 5500

U.S. Office 630 Fifth Avenue, New York, New York, 10111
TEL: (212) 969-9000; FAX: (212) 262-2113/3301
North American Marine Jet Inc./Nomera 50 - 950

P.O. Box 1232, Benton, AR, 72015, USA
TEL: (501) 778-4151; FAX: (501) 778-6381
PPjet (R.G. Parker (Engineering) Ltd) 6 5 - 4000

Units 5-7 Ailwin Road, Moreton Hall, Bury Street, Edmunds, Suffolk,
IP32 7DS, England
TEL: (0284) 701 586; FAX: (0284) 750 545
Riva Calzoni SpA (Riva Lips) 650 - 32,000

Via Stendhal, 34 - 20144 Milan" Italy
TEL: 39 2 4146 356; FAX: 39 2 48300458
Rockwell International Corporation See Note in Text

Rocketdyne Division
6633 Canoga Avenue, Canoga Park, CA, 91303, USA
Schottel 150- 1600

5401 Spay/Rhine, Germany
TEL: (2628) 610; FAX: (2628) 61 3000
Super Marine P o w e r Ltd

Boumcmouth, England 165 - 175
Taipan Jets Australia Pty Ltd 40 - 500

Lot 1, Siganto Drive, Oxenford, Qld, 4210, Australia
TEL: (075) 531 98; FAX: (075) 531 963
Ulstein International AS 600 - 7000

N-6065 Ulsteinvik, Norway
TEL: 47 70 10 050; FAX: 47 70 11 442
Ultra 20 - 325
Anson Business Park, Cheltenham Road East,
Staverton, Gloucester, GL2 9QN, England

APPENDIX E
BIBLIOGRAPHY
Allison, John, "Waterjet Performance Modules for SES Gasiunas, A. and Lewis, W.P., "Hydraulic Jet Propulsion: A
Whole-Ship Design Synthesis Model," 1988. Theoretical and Experimental Investigation into the Propulsion
of Seacraft by Waterjets," The Institution of Mechanical
Allison, John, "Waterjet Performance Section for ASSET Engineers, October 1963.
Initialization Program for U.S. Navy," 1989.
Gregory, Douglas L., "Performance Characteristics of a Full
Allison, John, "Waterjet Preliminary Design and Performance Scale Twin Disc Waterjet," NSRDC, Report 400-H-01, August
for Intra-Theater Sealift Ship (ITSL)," for U.S. Navy, 1988. 1970.
Allison, John, "Waterjet Selection and Performance Review for Gregory, Douglas L. and Hale, Malcolm R. Hale, "Performance
a Large SES Sealift Ship (SFS)," for U.S. Navy, 1988. Characteristics of a Full-Scale Jacuzzi Waterjet," NSRDC,
Report 400-H-02, December 1970.
Allison, John, "Waterjet Selection and Performance for an SES
Passenger Ferry," 1989. Holden, K. et al., "On Development and Experience of Waterjet
Propulsion Systems," 2nd International Congress of Interna-
Allison, John, "Waterjet Selection and Performance at Model tional Maritime Association of the East Mediterranean, Trieste,
and Full-Scale for the Advanced Material Transporter (AMT)," 1981.
for U.S. Navy, 1991.
Koops, B., "Waterjet Propulsion System Measurements at
Allison, John, "Waterjet Selection and Performance for a Very Matin," 19th ITI'S, Contribution to the High-Speed Marine
Large SES Car and Passenger Ferry," 1990-1991. Vehicle Committee Report, 1986.
Allison, John, "Comparative Waterjet Performance Studies for a Lakshminarayana, B., "Visualization Study of Flow in Axial
Tracked Amphibian," for U.S. Government, 1988-1991. Flow Inducer," ASME Paper No. 72-FE-33, May 1972.
Bartholomew, R.J., "2000 Ton LM 2500 Engine Overspeed Levy, Joseph, 'q'he Design of Waterjet Propulsion Systems for
Estimate and Thrust Degradation Due to Waterjet Inlet Air Hydrofoil Craft," January 1965.
Ingestion," Aerojet Interoffice Memo, 15 June 1973.
Lewis, Edward V., "Principles of Naval Architecture, Second
Blount, D.L., Grossi, L. and Lauro, G., "Sea Trials and Revision - Volume II - Resistance, Propulsion & Vibration,"
Model-Ship Correlation Analysis of the High Speed Gas SNAME, 1988.
Turbine Vessel "Destriero"," FAST '91.
Lewis, James W., "Calm Water Performance Trials on the 160 ft
Bowden, John O. and Embry, Gerry D., "SFS - The 55 Knot SES-200 With Diesel Driven Waterjets," NSCSES Report No.
Sealift Ship," Naval Engineers Journal, May 1989. 60-248, June 1991.
Carmichael, A. Douglas, "Design Optimization of Waterjet Mavlyudov, M.A., "Waterjets of Dynamically-Supported Ships
Propulsion Systems for Hydrofoils, Part I, An Overview," (DSS) - Some Problems and Methods of Solution," Krylov
Massachusetts Institute of Technology, Report No. 72-13, Shipbuilding Research Institute, St. Petersburg, Russian
October 1972. Federation, Personal Communication, June 1993.
Comstock, John P., "Principles of Naval Architecture," Miller, E.R., Jr., "Waterjet Propulsion System Performance
SNAME, 1967. Analysis," Hydronautics, Inc.
Delao, Martin, "Some Experimental Results of Tests of a Parker, R.G., "Waterjet Drive - Aspects of Design and Use and
Low-Speed, Waterjet Propulsion System," AIAA, Paper No. Its Place Among Propulsion Systems," R.G. Parker (Engineer-
66-718, Los Angeles-Long Beach, CA, August 8-10, 1966. ing) Ltd., Conference on Propulsion for Small Craft - Propellers,
Stem- gear, Engines and Installation, Nov. 9-10, 1982.
Du Cane, Peter, High Speed Small Craft," John de Graft, Inc.,
1972. Pearsall, I.S., "Cavitation," Institution of Mechanical En-
gineers," 1974.
Etter, R.J. et al., "Model Testing of Waterjet Propelled Craft,"
Proceedings of the 19th Meeting of the A'ITC, 1980. Perkins, W.F., Jr., "Discussion Notes for a Review of Progress
in Design of Large SES," Marine Technology, 1974.
Forde, Magnar, et al., "Computational Fluid Dynamics Applied
to High Speed Craft With Special Attention to Water Intake for Roos, Paul W., "Efficient Application of Jet Drives to Small
Waterjets," Fast '91, Trondheim, June 1991. Craft," Small Boats Symposium, ASNE, May 26-17, 1993.

Rossell, Henry E. and Chapman, Lawrence B., "Principles of Svensson, Roll', "Experience with the KaMeWa Waterjet
Naval Architecture, Volume 1I," SNAME, 1939. Propulsion System," American Institute of Aeronautics, Paper
No. 89-21440-CP, 1989.
Savitsky, D., "Waterjet Testing at Stevens Institute," 18th ITI'C
Contribution to the High-Speed Marine Vehicle Committee Svensson, Rolf, "A Description of the Waterjets Selected for
Report, 1986. "Destriero"," Fast 1991.
Schlappi, Herman C., "An Innovative Energy Saving Propulsion Svensson, Rolf, "Waterjets for Luxury Yachts," Sweden.
System for Naval Ships," Naval Engineers Journal, April 1982.
Van Terwisga, Tom, "On the prediction of the Powering
Sherman, Peter M. and Lincoln, Frank W., "Ram Inlet Systems Characteristics of HulI-Waterjet Systems," Marin Jubilee,
for Waterjet Propulsors," AIAA, Paper No. 69-418, Seaule Waginengen, May 11-15, 1992.
Washington, May 21-23, 1969.
Venturini, G.N., "Waterjet Propulsion Dynamics," International
Stark, Nicholas R., Zseleezloy, John and Krishnamoorthy, V., Hydrofoil Society Paper, 1974.
"Water Channel Tests of Rohr Marine 3KSES Waterjet Inlets,"
Hydronautics, Inc., Technical Report 7717-1, July 1978. Venturini, Giovanni, "Waterjet Propulsion in High Speed
Surface Craft," High-Speed Surface Craft Conference, Sussex,
Streeter, Victor L., "Fluid Mechanics," Second Edition, United Kingdom, June 24-27, 1980.
McGraw-Hill Book Company, Inc. 1958.
Verbeek, R., "Application of Waterjets in High-Speed Craft,"
Stricker, John G., PurneU, John G., and Brophy, Michael C., Marin Jubilee, Waginengen, May 11-15, 1992.
%Vaterjet Technology Development Summary (1972 - 1978),"
DTNSRDC-81/019, June 1981. Wilson, R.A., "Tests of Waterjet Inlet Broaching at
DTNSRDC," 18th ITI'C, Contribution to the High-Speed
Stricker, John G., Be,end, Alan J. and Purnell, John G., Marine Vehicle Committee Report, 1986.
"Advanced Waterjet Systems," ASNE Small Boats Symposium,
May 26-27, 1993. "Aspects of Performance Evaluation of Waterjet Propulsion
Systems and a Critical Review of the State-of-the-Art,"
Svensson, Roll', "Experience With Waterjet Propulsion in the NSRDC, Report 2550, October 1967.
Power Range up to 10,000 kW," SNAME Power Boat
Symposium, Miami, Florida, February 1985. "MJP Waterjet Propulsion System," MJP News, Waterjets,
January 1992.

Discussion
Robert Latorre, Member this paper will be a required reference for those working in
The author has provided the Society with a comprehensive the field of marine propulsion.
summary of marine waterjet propulsion. I have some comments to supplement this paper and have
The problem of inlet cavitation and performance loss at one request of the author.
high speeds illustrated by Fig. 74 is difficult to resolve using Experimentally, I have been successfid in integrating flush
the potential flow simulation as the author noted in discussing inlet waterjets with large, hard-chine monohulls; Destriero is
Fig. 75. However, work in the area of computation fluid dy- an example of this application. The findings reported in this
namics has developed computer tools which can model a paper and my experience support the utilization of waterjets
viscous fluid by solving the Navier-Stokes equation for the on vessels which, until now, only propellers would have been
pressure distribution C, and velocity components u, v, w in considered.
the fluid and along the waterjet inlet surface. Comparative studies to evaluate various propulsor concepts
Using the Tokyo University Modified Marker and Cell should consider total hull drag, propulsor characteristics, and
(TUMMAC V) program, we performed at the University of hull-propulsor interaction factors. However, little data have
New Orleans a series of two-dimensional fluid flow calcula- been published regarding hull-waterjet interaction factors,
tions for a 30 deg angle waterjet inlet set at I V R = 1.0. The and often "propeller magnitude" interaction factors are inap-
calculations were compared with the wind tunnel pressure propriately applied to waterjets. For this reason, waterjet
measurements for a waterjet inlet tested at similar conditions studies indicate too low efficiency. In fact, waterjet vessels
(Fig. 105(a) herewith). The resuhs are in nondimensional for- designed for speeds greater than 25 knots are likely to exceed
mat where the pressure coefficient is given by their full-load performance predictions during sea trials.
Destriero and other construction projects afforded me the
P.~ -- erd opportunity to define and manage extensive model experimen-
Cp- I/2pV~ tal programs to study waterjet-hull interaction effects. These
interaction effects make it possible to achieve overall propul-
The test data in Fig. 105(b) show that where I V R = 1.0 sive coefficients (OPC) at high speed, for vessels with flush
there is a sharp peak in the value of C~ in the lip near the inlet waterjets, superior to those for submerged or surface
stagnation location. propellers. The significant performance advantage for flush
The test data show that at I V R = 1.0 this ramp has a low inlet waterjets to a great extent results from hull efficiency,
value of Cp and so cavitation may occur in the ramp section of ~ln = (1 - t)/(1 - W). Table 4 herewith gives a comparison
the inlet. The potential flow calculation provides a qualitative of ~qn for propellers and waterjets for a 500 metric ton hard-
picture of the pressure distribution, but it is not able to esti- chine monohull. Above 25 knots, waterjets can have an OPC
mate the inlet ramp and lip surface Cp. In contrast, the TUM- advantage over propellers of 16 to 19%, due to lqn alone.
MAC V calculation shows good agreement with the experi- The superior OPC for flush inlet waterjets at speeds above
mental measurements. 25 knots is supported by fidl-scale trial data from a number
In conclusion, with the adoption of C F D codes such as
of vessels. Those data have been previously reported in Fig.
TUMMAC V the waterjet surface pressure distribution can
6 of reference [26] below. (The definition of OPC is the ratio
be estimated by the numerical solution of the Navier-Stokes
equation. This will enable the designer to easily perform a of E H P bare hull from model tests divided by total SHP from
cavitation check without extensive prototype testing. The time fidl-scale trials.)
and costs lead to the following question for the author. While I would like to request that the author expand his discussion
we have a reasonable technique of checking propeller cavita- to provide his definition for relative rotative efficiency with
tion based on the propeller's cavitation number and loading, regard to flush inlet waterjets and the likely range of its
can the author recommend a procedure to evahmte the cavita- magnitude.
tion in this waterjet inlet and pump components? Table 4
Donald L. Blount, Member

The author is to be congratulated for this treasure describing
current technology for waterjet propulsion. I am confident 25
35
45
55
TYPICAL HULL EFFICIENCY - ~H
SPEED-KNOTS PROPELLERS
0.88
0.94
0.95
0.95
FLUSH INLET
WATERJETS
1.05
1.07
i.i0
1.13
~H WJ/~H I
1.191.161.161.19PROPS
(a) COMPARISON OF 30 DEG (b) COMPARISONOF INLET LIP Cp

W A T E R JET INLET R A M P Cp Cp
Cp RAMP 0 / ' ~ b / / I LIP Rn : 10 7 1
o:,(.~_-2s4 . . . ~ IVR= I.o
02 0.8
06
o l\ Y" I . 04
-0.2 0.2
-0.4
\ I EXP
0
~ _~.,¢ 2 POTENTIALCAL -0.2
3 TUMMAC V
-06
-4 -3 -2 -1 0 1 2 S/b 4 -3 -2 q 0 1 2 s/o
Fig. 105

Again, the author is to be congratulated for an excellent vehicle. The size and weight can become a serious design
paper. problem. Furthermore, most of the control actuators are ex-
posed outside the hull and they become vulnerable to damage.
Additional reference The Propulsor Technology Branch of the David Taylor
26 Blount, D. L., "Achievements with Advanced Craft," ASNE Model Basin has been conducting active development work
Small Boats Symposium, May 1993. in the inlet shaping and, in collaboration with Band Lavis
Associates, developing an innovative steering concept for
John Stricker, Visitor, Naval Surface Warfare Center, more reliable steering and bucking performance, especially
Carderock, Maryland for larger vehicles.
[The views expressed herein are the opinions of the discusser and not J. B. Hadler, Member
necessarily those of the Department of Defense or the Department of
the Navy.] [Oral.] Along with others I would like to compliment the
authors on a very comprehensive paper on waterjet propul-
The author has done an accurate and complete job of de- sion. I will find it most usefid for my classroom work on ship
scribing the current status ofwaterjet technology and its prac- propulsion.
titioners worldwide. This paper provides a broad range of I would like to clarify a point ofconfilsion when we attempt
practical and theoretical information and will be used as a to compare the propulsive performance of the waterjets with
vital reference by those involved in the development and that of the screw propeller. We have developed a method for
application of high-performance propulsion systems. relating the performance of the screw propeller when it is
Because the waterjet propulsor becomes an integral part tested in open water and then tested behind the model to
of a hull form, it is difficult to separate propulsor and hull account for the differences. We have developed the concept
forces. Therefore, characterizing a given propulsor as installed of the "wake fraction" to account for the different velocity of
in a particular hull in terms of propulsive coefficient is a the ship in relation to the integrated average velocity behind
challenge and leads to the use of terms such as thrust deduc- the ship and have called this the "effective" wake fraction.
tion (t), inlet drag coefficient (Co,), and static pressure coeffi- We also note that the propeller must produce more thrust
cient (Cp) all of which appear in the general overall propulsive than the actual resistance of the ship due to the hydrodynamic
coefficient equation (3-44). Where hull forms and proptdsors interaction of the propeller with the hull. This we have given
are designed for minimum power, which will be the ride the name of "thrust deduction." We also note that the torque
rather than the exception in the filture, quantifying these from the open water test is usually not the same as the torque
parameters may be made more difficult. Baseline bare hull behind the model due to the fact that the velocity field behind
resistance (Rr) as defined in equation (3-21) may itself be the model is not uniform. This difference has been given the
impossible to quantify precisely since hull and propulsor ge- name of "relative rotative efficiency.'" These three factors
ometries will undoubtedly be developed simultaneously. are called the "propulsion factors" and go to make up the
Weight and volume of the propulsion system become impor- hydrodynamic propulsive efficiency of the ship. This gives us
tant factors in this interactive design process, as the attthor an easy method for relating the power requirements to propel
covers in Section 4.2. the ship with the resistance measured on the model of the
ship. It must be noted that these are convenient factors that
Charles Dai, Visitor, David Taylor Model Basin, have evolved as a result of the methods used in model testing
Bethesda, Maryland of the ship and the screw propeller as both separate and
I would like to compliment Mr. Allison on a very thorough combined entities.
discussion on the subject of waterjet propulsion. The paper There is a tendency on the part of the naval architect and
contains a good introduction to the basic principle of waterjet marine engineer to carry these concepts over to other methods
propulsion through a simple momentum analysis. Examples of marine propulsion. In the case of waterjet propulsion these
in the paper are very useful in demonstrating some of the concepts are questionable. We will have to develop new con-
design considerations for selecting a waterjet system. cepts largely dependent on how we test or predict the perform-
I would like to make some comments on the subject matter. ance of the waterjet system as a unit and how we relate the
There are many benefits of a waterjet propulsion system as interaction of the system with the hull. With the growing
compared with those of a conventional propeller. The author application of waterjet propulsion to ships and various craft,
has already mentioned those benefits in his paper. For the the solution to this problem is now receiving the attention
waterjet propulsion system to be more competitive as com- of the International Towing Tank Conference (I'll'C). The
pared with the conventional propeller, the losses have to be ultimate answer from the propulsive efficiency point of view
kept to a minimum. is the comparison of the amount of power that has to be
Two major areas of loss are the inlet loss and the loss in installed in the ship to propel it at the design speed. It is
the internal flow passage. Carefid hydrodynamic shaping of primarily a question of how we can determine these values
the inlet is required in order to minimize inlet head loss. The reliably for each system.
inlet shaping can also greatly influence the hull hydrodynam-
ics. The interaction between the inlet and hull should be Michael Schmiechen, Member
taken into consideration in the inlet design. The internal losses [Oral.] I would like to stress the point made by Prof.
can be minimized by better impeller design and reduced Hadler concerning the performance evaluation of waterjet
length of the ducting. Flow turning in the passage should also propulsion. Evidently the conceptual framework traditionally
be minimized to reduce secondary flow loss. used for the evaluation of ship hull-propeller interaction can-
Another challenge of a waterjet propulsion system is its not be applied without adequate modifications. The approach
poor performance in the low-speed region (speed less than developed by the present discusser 25 years ago aims at a
25 knots). The inlet lip should be designed to reduce excessive rational reinterpretation of the established concepts in terms
angle of attack at low speed or be more receptive to angle of of first principles, the balance of mass and energy fluxes in
attack at low speed. The mechanical design of the steering particular, i.e., at an adequate operational interpretation
bucket and its control can also be very challenging for a large [27,28].
This approach is even necessary for the evahmtion of screw predicted from the model resistance and calculated jet effi-
propeller propulsion, e.g., in the most interesting case of fidl- ciency. It is used by waterjet manufacturers because it is
scale ships under service conditions, where the operational something people are familiar with from propeller usage and
interpretation in terms of results of hull towing and propeller is easier to handle than changes of pressure distribution on
open water tests is not possible. The work on this problem the hull which are credited with additional lift at the stern.
since 1980 [29] has found its climax in fidl-scale tests with Equations (3-43) and (3-44) contain the terms for both thrust
the German research vessel Meteor [30]. As has been shown deduction factor and hull pressure coefficient. Presumably,
in a recent project this approach can be used in the design thrnst deduction factor tends to unity when the pressure
of propulsors as well [31,32]. Its advantage is that wake and coefficient term has its correct value.
thrust deduction are treated implicitly, in the whole process no The choice of the correct value of inlet drag coefficient Co,
explicit assumptions on the interactions of hull and propeller is another thorny problem. CD~ is a fimction of inlet velocity
being necessary. ratio, i.e., ship speed and inlet flow. Modifications to the hull
to accommodate the inlet may increase drag. A less than trnly
Additional references flush inlet (dropped lip or semi-flush inlet) will also have an
27. Schmieehen, M., "Perfornmnce Criteria tot Pulse-Jet Propel- associated drag.
lers," Proceedings, 7th ONR Symposium on Naval Hydrodynamics, During inlet testing for the large SES program, the force
Rome, 1968, pp. 1085-1104. accmmting system yielded an apparent negative inlet drag
28. Schmiechen, M., "Uber die Bewertung hydromechanischer
Propulsionssysteme (on the evahmtion of propulsion systems)," Schiff- over part of the inlet velocity, ratio range explored. This was
steehnik, Vol. 17, No. 89, 1970, pp. 91-94• originally thought to be due to experimental error, but is now
29. Schmiechen, M., "Eine axiomatische Theorie der Wechsel- believed to be due to pressure changes around the inlet which
wirkamgen zwischen Schiffsrnmpf und -propeller, (An axiomatic the- were not known at the time.
or,/of the interactions between ship hull and propeller)," Fritz Horn For flush inlets, inlet drag per se is usually neglected if
zum 100, Geburtstag gewidmet, Schiffstechnik, Vol. 27, No. 2, 1980. only for lack of knowledge. Inlet drag effects, if any, are
pp. 67-99• lumped in with the other interaction effects and covered by
30. Schmiechen, M., Proceedings, 2nd INTERACTION Berlin
'91, 2nd International Workshop on the Rational Theory of Ship the so-called thrust deduction factor.
Hull-Propeller Interaction and Its Applications, VWS, the Berlin I would also like to thank Mr. Stricker for the excellent
Model Basin, June 13-14, 1991, Mitteihmgen der Versnchsanstalt photograph of his model inducer pump components exhibited
fiir Wasserban und Schift'ban, I-left 56, 1991• in the registration area.
31. Schmiechen, M., "Entwurfnnd Bewertung von Diisenpropel- Dr. Dai makes some important points concerning the neces-
lern als Pnmpen. (Design and evahmtion of dncted propellers as sity tbr minimizing inlet duct losses and optimizing the inlet/
pumps)," FDS Report No. 245/1993, based on VWS Reports Nos. hull integration. Some very innovative design and testing is
1209/92 and 1221/93.
32. Sehmiechen, M., "'Entwurft, nd Bewertnng yon Diisenpropel- being conducted at CDNSWC (formerly David Taylor Re-
lern als Pumpen (Design and evaluation of dncted propellers as search Center, Carderock) in this area.
pumps)," VWS Report No. 1229/93, short version to be published With regard to the ongoing work on steering and reversing
in Trans• STG 87 (1993), presented at the STG Symposium Potsdam, gear in which I and my colleagues at BLA are participating
Sept. 4, 1993. with CDNSWC, we have developed some new concepts espe-
cially suitable for the next generation of large waterjets for
Author's Closure which a U.S. patent application is being processed.
I would like to thank all the discussers for their kind and In reply to Prof. I.zttorre's interesting discussion and ques-
helpfid comments. tion concerning a procedure to evahmte cavitation in the inlet
Mr. Stricker draws attention to an important point concern- and pump components, I would say there is no simple criterion
ing weight and volume in the interactive design process. comparable with propeller cavitation number.
We have tbund that the most effective way of optimizing Each design must be evahmted on its merits, invoking the
the design of a waterjet ship is through the use of a whole- various techniques available for inlet analysis. The propensity
ship design synthesis computer model in which the different to inlet cavitation is minimized when the inlet is operating
types of waterjet pumps are characterized. The results are at its design inlet velocity ratio.
extremely illuminating. In the simplest terms, it is found With regard to the pump, its suction specific speed should
that the pump with the highest propulsive efficiency is not be a guide, depending on the type of impeller. A local blade
necessarilv that which will result in the lowest filel consump- cavitation number, ~rp analogous to blade cavitation number
tion to fid'fill the stated mission of the ship. of propellers, is used in pumps, as discussed in the paper.
If we assume that pumps of equal, or nearly equal, hydraulic Inlet flow distortion at the pump face is also an important
efficiency are available in different sizes, then it may well
factor.
turn out that a smaller, lighter ship with a smaller, lighter
With regard to the potential flow simulation of a flush
pump of lower propulsive efficiency will be able to deliver
waterjet inlet presented in Fig. 75, the boundary layer defect
the required payload at the required speed over the required
distance while consuming less filel than a larger ship with a was taken into account by first calculating the displacement
more efficient but heavier pump. This results in part from thickness from preliminary results and then moving the physi-
the weight magnification effect whereby a pound of weight cal boundaries the appropriate amount, after which the poten-
in Group 200 translates into several pounds of hull weight. tial flow simulation was repeated.
This trade-off can only be made by the use of a whole-ship I would especially like to thank Mr. Blount for the excellent
design synthesis model containing accurate characterizations photographs of Destriero, one of which is exhibited in the
of strncture, propulsors and other machinery as well as of the registration area, and also for the video of Destriero's perform-
other groups' equipment. ance trials, which we have viewed during this presentation.
Concerning the thrust deduction factor, I am afraid that Mr. Blount's discussion highlights the most important area
this is often just an expedient to express the observed fact of research and past controversy, which is still little under-
that the power required by a waterjet-propelled vessel at stood. It should be mentioned that the hull-waterjet interac-
design conditions is less (or more in some cases) than that tion is not always favorable to waterjet efficiency.

WARNING
THE FOLLOWING PROGRAM CONTAINS MATERIAL THAT

MAY NOT BE SUITABLE FOR MATURE AUDIENCES
pL• KT HT OE NF RPIAERNTDILCYL E ART 18 DISPONTING HIMSELF WITH HI8

LITTLE FINEND8 IN THE 8ARGASSO SEA ... AN IRRESISTIBLE FORCE
HE 8Wilts TOO NEAR THE SURFACE SWEEPS ART ALONG
A SHADOWAPPROACHE$1
VI Vz
ART IS A F I N E FELLOW!
HEIGHT: O F'T 0 Ot X10-3 (2ram) ..... V.o-0
WFJGHI": :DOL8 XI0-S (1 rag)
B U O Y ~ MEUTRN. HE TRAVELS AT A STEADY 2 KT
THE SHADOW IS OVERTAKING HIM DOING 40 KT
HE DOES NOT SEE IT!
ART 18 SUDDENLY ACCELERATED TO 15 KT

ART ENTERS A DARK TUNNEL
AND iS LIFTED UP
MrS ~
LITTLEIIOOYCOlIFI TOA llUOI)ENSTOP
(RELAlrWETO THE SEABED}
* AND THENB ACCELERATEDV~OLENTLyBACKWARD8
HE IS CUT OFF FROM HELP BY THE DREADED
WHEN HE LEVEL8 OUT, HE 18 STILL" " " .~ t . Vp • 18 KT FORWAi~D8
DIVIDING STREAMLINE ! TRAVELING FORWARD AT 18 KT "~ • 2. Vp • 0 MOMENTARILY
3. Vp - BACKWARD8
BUT SOMETHING HORRIBLE
WHAT A DRAG ! IS ABOUT TO HAPPEN;. ALL THIS IS ACCOMPANIED BY
AWFUL SWIRLING SENSATIONS
A R T FINDS H I M S E L F B A C K A M O N G S T
FINALLY ART EMERGES IN THE HIS FRIENDS O N L Y A L I T T L E W A Y
JET WITH A VELOCITY FROMWHERE HE STARTEDI
Vp : Vj - Vs
HE IS HALF DEAD.l_
.J~ -I L~,~ .f:,.~.~. ~, ~.~

WAKE U Z.1 7.1 V C =o,<
INLET 16TO18 O.S 14.3
THE J E T P L U N G E S B A C K INTO THE pure. I~ o,~4 sJ .o.s
JET -34 0.4 -17.3
S E A B E H I N D THE SHIP UNDER THE JETWAKE -24,O 1.6 4.0
INFLUENCE OF G R A V I T Y TOT~ 4.e .tA A R T IN O L D A G E
Fig. 106 A light-hearted look at what really happens in a waterjet

H/STORY OF A PARTICLE PASSING THROUGH A WJ
....... /-\--\
j . vJ _
. . . . . . . . . . . . _~ IICLE
3 rr- I---3o r,~--~-~ 12~ rv . . . . .
SHIP SPEED, V s 40 KT
TAYLOR WAKE FRACTION 0.05
JET VELOCITY RATIO, Vi/V s 1.6
JET VELOCITY 64 KT (107 FT/S)
HEIGHT OF NOZZLE ABOVE SL 3 FT
HISTORY OF A PARTICLE - SUMMARY
APPROX APPROX
VELOCITY~f VELOCITY DISTANCE DISTANCE
RELATIVE RELATIVE APPROX RELATIVE RELATIVE TO
TO SEABED SHIP DWELL TO SEABED SHIP
(KT) (KT) TIME (FT) (Fr)
LOCATION (FWD 4-) (BACKWARDS ÷) (SEC) (FWD +) (BACKWARDS +)
WAKE 2 38 2.1 7.1 130
INLET 15 -~ 18 25 -p 22 0.5 14.3 20'
PUMP 18 -) 0 -~ 24 22 -* 64 0.1 -0.5 7
JET -24 64 0.4 -17.3 43.2
JET WAKE -24 -} 0 64 -~ 40 1.0 -5 72.5
TOTALS 4.t -1.4 272,7
• ALONG INLET CENTERLINE
"1" ROTATIONAL VELOCITIES NOT coNSIDERED; MEAN BLADE VELOCITY IS ABOUT 140 FT/S
Fig. 107
With regard to relative rotative efficiency, factors involved In the case of a surface effect ship, the nature of the flow
in the comparison of model or fidl-scale pump tests and effi- entering the inlet duct is affected by the outflow from the
ciency of the pump installed in the ship include the hull cushion. This complex wake flow is difficult to simulate in
boundary layer and details of the constnlction of the inlet lip the laboratory and, hence, is another source of difference
and duct, including the elbow and shaft fairing if fitted. between model and full-scale velocity distribution at the pump
In the absence of test data concerning inlet flow distortion face giving rise to small differences in torque and efficiency
at the pump face and its effect on torque and efficiency, a between model and full-scale pumps when the latter operate
value of% = 1.0 is usually assumed, but this is not necessarily in a ship.
correct. From the foregoing, it will be clear that the value of relative
Strict specifications for inlet flow distortion were imposed rotative efficiency is dependent, among other things, on the
for the 40 000 shp pumps developed for the large SES program hull type. It is not possible to assign a magnitude to it with
in the 1970s, with the object of minimizing cavitation damage any certainty, but if I had to choose one, I would say it was
and loss of efficiency. on the order of 0.01 below unit, i.e., 0.99 typically.
WATERJET PUMP POWER CAPACITY
AT CRAFT SPEEDS OF 40 KT TO 45 KT
1OOOOO 100,000HPPUMPGOALS - ~
(EP =065-07)
NOTE:SOMEPUMPSCAN
9OOOO ABSORBMORE POWER
AT HIGHERSHIPSPEEDS
BUT IT IS POWER AT 40-45KT
,,=, 8oooo WHICH COUNTSPLUSPROPULSIVE
EFFICIENCY
6OOOO- INDUCER PUMPS

IJJ ,,I,
50oo RIV/VLIPS P U M P S 3KSES ~ I
o •
KAMEWA PUMPS
o s.. 11
(EP = 0.5) _
.=
== DESTRIERO I
SES-100A • I= ~ ~=*
O"
1930 1940 1950 1960 1970 1980 1990 2000
YEAR
Fig. 108 Trend of waterjet power with current goals
I thank Prof. Hadler for his kind comments and am pleased I regret that I have not had the opportunity of reading Prof.
to hear that he will find the paper usefid in the classroom. If Schmiechen's numerous papers, most of which are inaccessi-
this proves to be the case, then one of the objectives of ble to me.
preparing the paper will have been achieved, namely to make I am pleased Dr. Barr (comments from the floor by Roderick
the subject understandable to the uninitiated. Barr, Member) enjoyed our light-hearted look at what really
I entirely agree with Prof. Hadler and Prof. Schmiechen happens in a waterjet (Figs. 106 and 107)). This was done
that the concepts of propulsion factors which have evolved with a serious purpose to help visualize the hull interaction
for propeller-driven ships cannot be carried over into waterjet and flow through a waterjet propulsor.
propulsion without proper definition and appropriate modifi- With regard to the trend in waterjet unit power shown
cation.
in Fig. 108, the goal of developing an operational 100 000
Nevertheless, the same fimdamental principles apply to
both propellers and waterjets. The existence of a hull bound- hp waterjet propulsor by the year 2000 is not deduced
ary-layer affects the propulsor performance in both cases, and from the shape of the curve. Rather, this goal is the result
the presence of the propulsor affects the flow around the hull of a requirement from shipping interests for a high-speed
near the stern in both cases. cargo ship for which there is an urgent demand to fill
In the case of waterjets, models may be tested with the the gap between present large containerships and aircraft.
inlet opening closed and faired with the hull, and with the Whether or not this goal is achieved depends greatly on
inlet flowing but with no nozzle thnlst. Self-propulsion tests the success achieved in obtaining financial support for the
can also be performed. Hull pressures may be measured with ship program.
and without inlet flow. Comparable hull pressure measure- I agree that a quantum leap, is involved, but we have all
ments during propeller self-propulsion tests are complicated the tools and we have made quantum leaps in this country
by the presence of appendages and by the propeller blade- before. The 40 000 hp pumps for the 3KSES, which was
induced pressure pulses. beginning manufacture when the program was canceled, rep-
It is hoped that the IT/'C will be able to resolve, in a resented a quantum leap in power as great as that now contem-
straightforward manner, the problem of dealing with the ef- plated, as shown Table 5.
fects ofhull/propulsor interactions when estimating the power I agree with Dr. Barr that variable area inlets could help the
required to drive a waterjet ship. low-speed performance of waterjets by maintaining optimum
Table 5
Propulsor PHM 3KSES KaMeWa 180 WJ 2000

Unit Power, hp 16,000 40,000 40,000* 100,000
Speed, kt 60 90 50 40
Year 1974 1978 1991 2000
Currently 30,000, limited only by non-availability of a higher powered prime mover.

inlet velocity ratio (IVR) and, hence, good inlet recovery. The trated in Fig. 74. Figure 72 shows the qualitative effects of
inlet efficiency of current fixed inlets is quite low at low ship incorrect inlet velocity ratio.
speed. Hydronantics Inc., of which company the present Hy-
The SES 100A variable area inlet is depicted in the cutaway dronautics Research Inc. is the successor, did an outstanding
view of Fig. 71. The geometry is generally similar to that job in developing the variable geometry inlet for the SES
shown in Fig. 73 and the hydrodynamic performance is illus- 100A and the 3KSES program.

Marine Waterjet Propulsion - SNAME 1993

Uploaded by

Copyright:

Available Formats

Marine Waterjet Propulsion - SNAME 1993

Uploaded by

Document Information

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Marine Waterjet Propulsion - SNAME 1993

Uploaded by

Copyright:

Available Formats

SNAME Transactions, Vol. 101, 1993, pp.

U.S. Navy SES 200 With Waterjet Propulsion

Marine Waterjet Propulsion 275

Figure 2. Yamato I Magneto-Hydrodynamic Waterjet Ship

Figure 1. Leonardo da Vinci's Drawings

During the mid-19th century there was considerable

276 Marine Waterjet Propulsion

Figure 6. U.S. Navy's SES-IOOA

Figure 7. Rocketdyne Powerjet 46

Abroad, KaMeWa in Sweden, and Riva Calzoni in Italy,

Marine Waterjet Propulsion 277

i; Figure 8. U.S. Navy's SES 200

278 Marine Waterjet Propulsion

The work done in propelling the ship is

Assuming perfect inlet energy recovery and neglecting

AE = 1/2 m Vj2- 1/2 m i V 2

JET DIVERGENCE HALF ANGLE 5 DEGREES or rlj = m(Vj-V)V/1/2m(Vj 2-v,5

The ratio of ship speed to jet velocity is of interest:

Alternatively, dividing (2)by V

where a = Vj/V = ~-1 (3-5)

The same expressions can be obtained by considering the kinetic

Marine Waterjet Propulsion 279

It will be found that Tij now reaches a maximum depending on

The ideal jet efficiency can also be expressed in thrust 1.0

280 Marine Waterjet Propulsion

~VENA CON'[RACTA ~t The inlet losses can be expressed, as previously, as a

V = ship speed, ft.s (m/s) - Station (0)

Marine Waterjet Propulsion 281

5 0,5S 0,6 0 es o.l o.rs

282 Marine Waterjet Propulsion

It is found that t can be positive or negative depending on Ix (2 (1 - (1 - w) g)) - CDi Ix / IVR (1 - w)

2 I.t (cos ct cos 0 - (1 - w) It) (3-30)

Marine Waterjet Propulsion 283

= hj = height of the jet at the vena contracta. 1

284 Marine Waterjet Propulsion

At usual speeds the pressure around a waterjet is found to

Figure 13. Notional Waterjet Inlet Flow Patterns

efficiency, also, no inlet drag term is included, and ~D differs

I.I. (2 (cos Ct cos O - I-t) - CD, IMIVR) j 9 I

286 Marine Waterjet Propulsion

should not exceed 1000 lbf/ft2 (approximately).

Marine Waterjet Propulsion 287

l'01ol 0¢0~e 00~¢r

288 Marine Waterjet Propulsion

achieved by the best mixed flow designs (l~p _> 90%).

Vu2 U 2 - V u l U1 PUMP SPEED IS CONSTANT

Marine Waterjet Propulsion 289

Figure 24. Generic Non-Dimensional Pump Characteristics

BASELINE DESIGN POINT JET VELOCITY RATIO IS 0 . 6 2 3

0.5 0.6 0.7

operating point on the non-dimensional head-flow curve, or

290 Marine Waterjet Propulsion

Table 2. Sample Page From Jet Efficiency Tables

WAKE FACTOR= 1.00000

JET PROPULSIVE EFFICIENCY WITH

MAX. JET EFF = 0.70806 AT VCVJ = 0.70806

Marine Waterjet Propulsion 291

o4 , , ',. ',, ~ ',

57 58 5~ 60 61 -~-- S-L3Z ----"--~--

66L 54t- ' 60 L- I I _ _ [ . . . . L_ L---3--