SURV-IV, 1997,

International Conference SURV IV,

Surveillance, Pilot and Rescue Craft
for the 21st Century, Paper No. 12,
RINA Small Craft Group,
13-14 May 1997, Gothenburg, Sweden.

Report 1083-P,
Delft University of Technology,
Ship Hydromechanics Laboratory,
Mekelweg 2, 2628 CD Delft,
The Netherlands.

Reprinted: 08-10-2000
Website: www.shipmotions.nl

Sea Trials with Regard to Design and Operational Limits

of Fast Pilot Vessel MS VOYAGER
Jakob Pinkster and J.M.J. Journe
Delft University of Technology

Abstract
The fast, 28 knots, seagoing 18 metres pilot vessel MS VOYAGER is a new type of tender,
which is propelled by two water jets, with a hull being constructed from aluminium and a
deckhouse constructed from specially developed composite materials.
In order to assess the actual quality of the design, the vessel was extensively tested in 1994
during normal operating conditions with regard to vibrational conduct, vessel motions and
manoeuvring characteristics, as well as actual hull mechanical stress levels. In the present
paper a description is given regarding these full scale tests and the resulting data and analysis
thereof.
It was found that mechanical vibration levels on board were well below acceptable levels.
Transient loading of the vessel's structure, during fast free sailing in adverse weather
conditions as well as unwanted collisions during pilotage, resulted in problems related to
fatigue of construction members.
Speed and manoeuvring characteristics and the vessel's motions in a seaway are assessed too.
Also, ship motions calculations were carried out and compared with the actual vessel motions
already monitored, in order to determine the range of reliability of the strip theory method for
high vessel speeds. Equipped with this motion feedback, special attention has been given to
the effect of lengthening of the vessel on slamming phenomena and acceleration levels.
A number of recommendations are finally made for the following generation of this type of
pilot vessel.

Authors Biographies

University of Technology. After his

graduation in 1979, he was R$\&$D
project engineer at Royal Boskalis
Westminster Group, R&D project engineer

Mr. Jakob Pinkster holds a Master's degree

in Naval Architecture from the Delft

at Damen Shipyards and lecturer at the

Polytechnic West Brabant, respectively.
Since 1991, he holds the position of
Assistant Professor in the Ship Design
Section of the Delft University of
Technology. He presents lectures for
students on ship design and supervises
them in their design and graduation work.
His particular areas of research interest are
ship design and advanced marine vehicles,
on which subjects he has published various
papers.
Mr. Johan M.J. Journe holds a Master's
degree in Naval Architecture from the
Delft University of Technology since
1975. In 1958, he started his career as
assistant metal worker at the Rotterdam
Dockyard Company. Since 1963, he has
been
employed
at
the
Ship
Hydromechanics Laboratory of the Delft
University of Technology and at present
he is an Associate Professor there. He
presents lectures for students on waves,
seakeeping, manoeuvring and offshore
hydromechanics. His particular areas of
research interest are loads on and motions
of ships and other floating structures in
waves, the behaviour of ships during and
after a sudden ingress of water due to a
collision and speed loss and fuel
consumption of ships in waves. On these
subjects he has published a variety of
papers.
1

tender, which is propelled by two water

jets with the hull being constructed from
aluminium and a deckhouse from specially
developed composite materials.
The general arrangement of MS
VOYAGER, with an overall length of
about 18 metres, is shown in Figure 1. The
vessel is equipped with two diesel engines
of 626 kW each and is capable of
developing a sustainable service speed of
28 knots, which is more than twice the
speed of the companys other pilot vessels.
The new vessel and her three man crew
have to be capable of simultaneously
transporting a total number of 12 pilots
anywhere along the treacherous North Sea
along the Dutch coast in sea conditions
with significant wave heights up to 2.75
metres. In order to meet the owners
demanding specifications, the vessel may
well be considered to have been designed
in that presently poorly defined zone
where defiance of acceptable safety levels
may well be a fact; not purposely so, but
solely
through
lack
of
present
understanding of how to design such
vessels. The owners, realising this,
requested the Department of Marine
Technology of the Delft University of
Technology therefore to thoroughly test
the vessel with regard to her design and
operational limits.
MS VOYAGER was consequently tested
with regard to vibrational conduct, vessel
motions and manoeuvring characteristics
as well as actual hull mechanical stress
levels. These extensive tests were carried
out using full-scale measurement on board
of the vessel during normal operating
conditions. These conditions included
sailing at sea at (full) speed, sometimes in
seas up to 2.5 metres significant wave
height. Another condition was the actual
transfer process of pilots to and from
ships, during which the pilot vessel also at
times incurred heavy transient loading due
to unwanted collisions between the
vessels.
The following describes these different sea
tests and results thereof for the hydro-

Introduction

Any ship that arrives at or departs from a

harbour generally makes use of the
services of a pilot who knows the local
navigational hazards and has experience
with the manoeuvring of vessels. A pilot
tender is a vessel that brings these pilots to
and from
such ships. This exchange of pilots takes
place at sea, often under severe weather
conditions. The fast pilot vessel MS
VOYAGER, owned by the Rotterdam
based Dutch company Facilitair Bedrijf
Loodswezen B.V., was commissioned in
early 1994. This is a new type of pilot

mechanic and hydro-elastic behaviour of

the vessel.
2

of between 5 and 10 m. The remaining

position error is a slow drift phenomenon
around the correct position. However, as
only relative accuracy was important and
several of the tests were short (about 1
minute), the actual accuracy for these tests
was estimated to be between 1 and 3
metres. For the longer lasting speed tests,
the influence of the absolute error due to
the DGPS system decreases rapidly with
distance. During the manoeuvring tests a
Sperry C14 course gyroscope was used for
the instantaneous course registration.
Furthermore, a potentiometer was used to
measure the water jet angle ("rudder
angle").
The waves were measured by the
WAVDEL wave buoy of the Delft
University of Technology, which was
anchored near the measuring area. The
acceleration signals were transmitted to the
ship, sampled and stored for later
processing. This wave buoy measures the
total energy supplied by the waves to the
buoy, so information on a directional
spreading of the waves will not be
obtained. The mean wave direction has to
be determined visually.
During the ship motion tests a Schaevitz
LMP-05 accelerometer was placed next to
the fore peak bulkhead to measure the
vertical accelerations at the bow in shipbound co-ordinates. Three identical
accelerometers were placed in the
accommodation just aft of the helmsman's
chair, to measure the accelerations along
the x , y and z axes of the ship.
Roll and pitch angles were measured using
a Sperry VG14 vertical gyroscope. Each
signal was fed through a second order
filter with a cut-off frequency of 2 Hz.
Because this bandwidth was too low to
measure the often short peak accelerations,
the signal from the bow accelerometer was
also fed through two peak detectors that
caught and stored the positive and negative
peak values of the bow acceleration in
each sample interval.
All signals from filters and peak detectors
were then send to a PC with an analogue-

Hydromechanic Behaviour of the

Vessel

During several days a large number of fullscale tests, related to the hydromechanic
behaviour of the vessel, have been carried
out. Speed trials in calm water have been
performed out to determine the relation
between the forward ship speed and the
number of revolutions of the water jet
engines. Among others, the results of these
tests are needed for the determination of
the magnitude of the speed loss of the ship
in a seaway.
Acceleration and stopping tests have been
performed to get an impression of the
safety of the vessel during pilotage
operations. In this connection, also the
manoeuvring characteristics of the vessel
have been determined. For this, turning
circle tests, spiral tests and Kempf zig-zag
manoeuvres have been carried out.
Finally, during two days seakeeping tests
have been performed to determine the
motion and acceleration behaviour of the
vessel in a seaway. Special attention has
been paid to the vertical acceleration peaks
in the passengers room and forward due
to slamming.
A full description of these experiments and
the results is given by Ooms and Journe
(1994) in a technical report to the
contractor.
2.1

Experimental Set-Up

For the calm water tests and the ship

motion tests, different test equipment was
required. For the speed and trajectory
measurements use was made of a
Magnavox MX200 Differential GPS
receiver with an MX50R correction
receiver. The correction signals came from
the Hook of Holland reference station at a
distance of between 25 and 50 km.
Using DGPS rather than standard GPS
resulted in an absolute position accuracy

to-digital converter, where they were

sampled 8 times per second and
subsequently stored for later processing.
2.2

lowering the buckets with causes a

reversal of
the direction of the thrust. During the
natural stopping test given in figure 3, the
ship slowed down to a speed equal to 10%
of the initial speed of about 17.7 knots
after a distance of 73 metres (4 ship
lengths) in 45 seconds. This means an
average acceleration of -0.18 m/s2 with an
extreme value of about -0.60 m/s2 .
During the forced stopping test given in
figure 3, the ship slowed down from the
initial speed of about 17.7 knots to zero
speed after a distance of 7 metres (less
than half a ship length) in 8.3 seconds.
This means an average acceleration of 0.87 m/s2 with an extreme value of about 2.05 m/s2 .

Speed Trials

The speed trials in calm water have been

carried out in the North Sea Channel,
under weak transverse wind conditions.
Figure 2 shows the measured forward ship
speed V as a function of the number of
revolutions N of the water-jet engines. At
about 15 knots the ship speed starts to
increase strongly with the number of
revolutions, because of a decrease of the
wetted surface of the hull due to the
planing behaviour of the vessel.
Figure 2 also shows the same data in
irregular waves defined by Beaufort 4 and
6 at various wind and wave directions. The
data shown in this figure have been
obtained during sea trials when measuring
the motions of the vessel, as will be
described further on. In head wind and
waves a small speed loss was observed,
while in following wind and waves there
was a slight increase in speed due to the
(wind)surfing behaviour of the vessel.
2.3

2.4

Manoeuvring Tests

Turning circle and spiral tests have been

performed in coastal sea areas SW of
IJmuiden, at N = 1200, 1800 and 2200
rpm with nominal water jet angles of 6,
12 and 18 degrees to port side as well as to
starboard. The observed wave height was
about 0.75 metres.
An example of the path of the ship,
corrected for current, during one of these
turning circles is given in figure 4. Also,
the relations between the rate of turn and
the water jet angles are given in this figure.
From these experiments, it appeared that
the turning ability at high speeds of this
ship is about 0.7 degrees of turning per
second per degree of water jet angle. The
figure shows an offset in the water jet
angle of about +2.5 degrees. Probably, this
offset can be explained by the equal
rotational directions of the two water jet
engines.
At about 1800 rpm and a water jet angle of
18 degrees, the diameter of the turning
circle was less than 50 metres, while the
forward ship speed slowed down from
17.0 to 10.8 knots.
Additionally, a range of Kempf zig-zag
tests have been carried out which
confirmed the excellent manoeuvring

Acceleration and Stopping Tests

At about 1200 and 1800 rpm of the water

jet engines, a range of acceleration and
stopping tests have been performed in the
North Sea Channel at moderate wind
conditions. An example of some of the
results of these experiments is given in
figure 3 for about 1800 rpm.
During the acceleration test given in figure
3, the ship reached a speed equal to 95%
of the obtainable speed of about 15.3 knots
after a distance of 90 metres (5 ship
lengths) in 18.0 seconds. This means an
average acceleration of 0.43 m/s2 with a
maximum value of about 0.75 m/s2 .
Two types of stopping tests have been
performed; natural stopping by simply
reducing the number of revolutions of the
water jet engines and forced stopping by

characteristics of the vessel, found during

the turning circle and spiral tests.
2.5

the ship and 600, 1200, 1800 and 2200

rpm of the water jet engines.
It should be mentioned that a visual
observation of the mean wave direction on
board of small vessels is rather difficult,
because of the low altitude of the observer
above the waves.
For the two highest ship speeds, the
measured RMS -values of the vertical
accelerations amidships and forward
during these two days, are given in figure
5. These experimental data have been
compared with theoretical data, obtained
with the linear frequency domain strip
theory ship motions computer code
SEAWAY, of which a users manual is
given by Journe (1992) and the
underlying theory is described by Journee
(1996).
The time-averaged significant wave
heights and mean wave periods, during
each of the two measuring days, are input
for the computations. A constant wave
direction has been used and the directional
spreading of the waves has not been taken
into account; uni-directional long crested
waves are supposed here. This choice was
more or less justified by visual
observations of the waves during the
experiments.
At about 1800 rpm, the vessel will start
planing. At this ship speed and at lower
speeds, a fairly good agreement between
experiments and predictions has been
found; see figure 5. When judging the
predictions, it should be kept in mind that
constant wave parameters have been used
and that the effect of the S to SSW wind
waves during the measuring days has not
been accounted for.
At higher speeds the ship is planing and
computations, while assuming that the
vessel is a displacement ship, would have
less sense. But, even then, figure 5 shows
that the maximum values of the vertical
accelerations were fairly well predicted,
especially at Beaufort 4.
Also, the peak values of the vertical
accelerations amidships and forward have
been measured. Figure 6 shows the

Seakeeping Tests

During the seakeeping experiments in

coastal sea areas SW of IJmuiden, the
motions of the centre of gravity of the
vessel in 6 degrees of freedom, the vertical
motions forward at the fore peak bulkhead,
the number of revolutions of the water jet
engines, the forward ship speed and the
heading have been measured. The water
depth varied between 10 and 15 metres.
The waves have been measured by the
WAVDEL wave buoy, anchored near the
sailing area. During two days, these
experiments were carried out for a range of
headings and number of revolutions of the
water jet engines.
During the first measuring day, after a
severe storm in the days before, the sea
conditions were characterised by Beaufort
6. The significant wave height H1 / 3
decreased from 2.1 to 1.5 metres at the end
of the day, while the zero up-crossing
wave period T2 varied between 5.1 and 5.8
seconds. The heading of the waves,
influenced by a S to SSW wind with a
speed of about 5 m/s, measured ashore,
and the refraction of the waves in shallow
waters, was estimated to be about 080
degrees. The experiments have been
carried out at a wide range of headings of
the ship and 600, 1200, 1800 and 2050
rpm of the water jet engines.
During the second day, the stable sea
conditions were characterised by Beaufort
4. The significant wave height H1 / 3 varied
between 1.0 and 1.1 metres, while the zero
up-crossing wave period T2 was about 4.5
seconds. The heading of the waves,
influenced by a S to SSW wind with a
speed of about 3 m/s, measured ashore,
and the refraction of the waves in shallow
waters, was estimated to be about 100
degrees. The experiments have been
carried out at a wide range of headings of

measured peak values forward at the two

highest ship speeds in Beaufort 4 and 6.
The figure shows extreme negative peak
values of about -1 g and extreme positive
peak values of about +4 g . It is obvious
that such data can not be obtained with a
linear frequency domain computer
program.
Figure 7 shows the measured positive and
negative peak values of the vertical
accelerations during all the experiments as
a function of the corresponding RMS values. The derivatives of these curves, i.e.
the values of the tangents to the curves for
small RMS -values, are equal to plus or
minus 5 RMS . This coincides with
acceleration thresholds for the linear
harmonic motions with a probability of
exceedence of 3.7 10 6 .
Also, figure 7 shows that in rough weather,
a small decrease of the RMS -value may
result in a considerable reduction in the
peak values of the bow accelerations. For
instance, a decrease of 10% of the RMS value may result in a reduction of 20% in
the peak acceleration values.
3

3.1

Experimental Set-Up

For the registration of the SB and/or PS

water jet revolutions, 2 Honeywell
922AA3XM-ASN-L magnetic pick-ups
have been utilised. For the measurement of
the
vibration
accelerations
42
accelerometers were installed, of which 41
single axial accelerometers (makes
Sundstrand, Endevco and Bruel & Kjaer)
distributed along the ship and 1 Bruel &
Kjaer 4322 tri-axial seat-accelerometer for
use on helmsman's chair. The measured
signals were conditioned, analogue filtered
and digitised. In order to protect
overshooting of these signals, equal time
delay filters (linear phase response) were
used.
The same equipment was, more or less,
utilised for the mechanical stress
measurements. For the measuring of these
mechanical stresses in the vessel, 40
Hottinger
Baldwin
Mess\-technik
3/12OLY13 strain gauges were installed.
Furthermore,
3
Endevco
2262-25
accelerometers were used for measuring
the accelerations in the fore peak and on
the engine foundation.
Vibration levels were monitored at 43
different locations in different directions.
Hull stress levels were measured by strain
gauges, mounted at 40 different locations
in different directions, and monitored.
Also, the signals of accelerometers for
measuring the ship motions fore and aft
were monitored, together with the water jet
engine revolutions, measured with the aid
of an infrared sensor and markings. A 50channel recorder was used to store these
monitored signals on a time basis.
The location of all monitoring devices and
the designation of the signal channels are
shown in tables 1 and 2. Figure 1 also
shows some of the stress measurement
positions used on frame 10.

Hydroelastic Behaviour of the

Vessel

Also, the vessel was consequently tested

with regard to vibrational conduct as well
as to actual hull mechanical stress levels.
These extensive tests were carried out,
using full scale measurements on board of
the vessel during normal operating
conditions, i.e. during sailing at sea at
(full) speed, at times in seas up to 2.5
metres significant wave height, as well as
during the actual transferal process of
pilots to and from ships. During the latter,
the pilot vessel also at times incurred
heavy transient loading due to unwanted
collisions with the (to be) piloted vessel.
A full description of these experiments and
the results are given by Pinkster and
Hylarides (1994a and 1994b) in two
technical reports for the contractor.

3.2

Vibration Measurements

In order to assess the comfort and well

being of the individuals on the vessel,

extensive vibration measurements were

carried out on board under different
service conditions. In essence, the
vibrational signature of the vessel is to be
found resulting from the presence of
excitation forces, created by the water jets
and/or the main engines as well as the
presence of excitation forces due to
slamming. This latter force can be
regarded as being a transient (impact)
excitation force, i.e. zero frequency.
Vibration levels were registered on the
water jet, on the engine foundations and
the engine itself, on the hull, on the
mounting of the superstructure and in the
superstructure, eg. the helmsman's chair.
Within the choice of measurement
locations the vessel's symmetry was taken
into consideration.

sailing in severe seaway in the vicinity of

the entrance to the outer port of IJmuiden.
Three such vibration measurement runs
were made. Due to severe slamming in the
given sea conditions the vessel's speed was
reduced to a maximum of about 18 knots.
Each run lasted about 5 minutes and the
signals were recorded on a hard disc
discretely at a rate of 2000 samples per
second. Thereby, vibrations could be
clearly registered up to a frequency of 500
Hz. After each run the vessel returned to
IJmuiden where the registered data results
were investigated on the computer screen.
If the results were considered to be clear
enough, then a tape backup was made of
the data and then the next run was made.
In total 45 Gbytes of signal data for all
three runs has been backed up on tape.
Using the aforementioned vibration data,
table 1 shows the maximum vibration
velocity peak values of each of the
accelerometers, encountered during any
given measurement run for a discrete
number (6) of frequency values. These
frequency values are repectively 18, 36,
54, 72, 90 and 108 Hz and coincide
directly with 0.5th , 1st , 1.5th , 2nd, 2.5th and
3rd engine frequency order. These
frequency values also coincide with 1 st ,
2nd, 3rd, 4th , 5th and 6th order of water jet
shaft-frequency values.
For the sake of clarity, the gearbox
reduction ratio is 2:1, the main engines are
of the V-type with a total of 12 cylinders
per engine each producing 626 kW at 2200
rpm and each water jet is fitted with a six
bladed impeller.
It may be expected that if any resonance
occurs at all, then these should be found at
the same frequencies as the excitation
frequencies coming from the engines and
the water jets. Therefore, the maximum
vibration velocity peak values of each of
the accelerometers, encountered during
any given measurement run for these 6
frequency values, have been filtered from
the measured data.

TNO-CMC at Delft was sub-contracted to

install and carry out the vibration
measurements
according
to
the
specifications of the Delft University of
Technology. The installation of the
vibration measurement equipment took
two days, followed by one day of
measuring at sea.
The weather conditions during the
measurements, Beaufort 8 with a
significant wave height of about 2.5
metres, were considered to be very good as
far as expected input levels of external
impact forces (slamming) on the ship's
structures were concerned. The idea is that
a fitting input value should lead to a firm
output value and thereby to good (strong)
vibrational response signals.
Two types of measurement runs were
made.
Firstly, so-called sweeps were carried out,
whereby vibrations were monitored while
the revolutions of both main engines were
slowly increased from minimum number
of revolutions to the maximum number.
This run was carried out within the outer
harbour of IJmuiden.
Secondly, vibrations were monitored for a
longer time, while the engines were
producing constant (maximum) power and

Along with this, the maximum of the

maximum vibration velocity peak values
per measurement location has been found
and is highlighted in table 1 by means of
italic bold printing. If these maximum
values are found to be within the
acceptable values, which are given in the
ISO 6954 vibration norm diagram, then no
adverse comments are probable with
regard to the comfort and well being of
people and materials on board of the vessel
in the given service scenario.
No extreme values were found for the peak
values of the vibration velocities as
described above. The highest value
obtained was 2.05 mm/s for the topside of
the deckhouse aft on starboard side at a
frequency of 18 Hz. This coincides with
the 1st order of water jet shaft-frequency.
According to the ISO 6954 vibration norm
diagram, no adverse comments are
probable with regard to the comfort and
well being values when the velocity level
lies below 4 mm/s for this frequency
value. Adverse comments are probable
when velocity levels are above 9 mm/s and
there is a grey zone for velocities between
4 and 9 mm/s.
In conclusion, one may say that this vessel
should receive no adverse comments at all
with regard to vibration levels. Indeed to
gain such good low vibration levels are the
dream of many naval architects.
3.3

measure the stresses in the hull under

different operational conditions. Analyses
of these results were expected to lead to
the cause of the cracking phenomenon and
suggestions should then be made for
measures to be taken to solve the problem.
Stress levels were registered in web frames
number 10, 11, 13, 14, 15 and 17.
Particular attention was taken in the
placing of strain gauges on the second
spray rail at web frame number 10, since it
was difficult to find uncracked material at
this location as the cracks had initiated at
this exact point. A solution was found by
placing 3 of these strain gauges on port
side and 2 on starboard side. In order to
investigate the stress levels along the
frames themselves, web frames 10 and 14
were well fitted with 20 and 12 strain
gauges, respectively. Stress levels in the
frames along the length of the vessel were
investigated with the aid of 2 strain gauges
fitted on web frames 11, 13, 15 and 17
each. Within the choice of measurement
locations, the vessel's symmetry was taken
into consideration. The location of all
monitoring devices is shown in table 2, as
well as the designation of all 44 signal
channels.
Again, TNO-CMC at Delft was subcontracted to install the necessary
equipment and to carry out the hull stress
measurements
according
to
the
specifications of the Delft University of
Technology. The installation of the
vibration measurement equipment took 4
days. The actual stress measurements were
carried during 2 days; one day for static
tests and one day for dynamic tests. The
weather conditions during the dynamic
tests, Beaufort 8 with a signnificant wave
height of about 2.5 metres, were
considered to be very good as far as
expected input levels of external
(slamming) impact forces on the ship's
structures were concerned. The idea is that
a fitting input value should lead to a firm
output value and thereby to good (strong)
hull stress level response signals.

Hull Mechanical Stress Levels

Measurements

About six months after receiving the

vessel from the builders, the owners
noticed that cracking of certain parts of the
aluminium hull was occuring. The cracks
were appearing at the second spray rail
which forms an integral discontinuity
within the main frame structure. This part
of the main frame structure and the cracks
are described in figure 8.
The Delft University of Technology was
asked to investigate this phenomenon and
make suggestions to solve this problem. In
order to realise this, it was decided to

Two types of measurement runs were

made.
Firstly, so-called pilotage measurements
were performed, whereby hull stresses
were monitored while the vessel bashed
off the quay wall in an attempt to simulate
the actual transferring of pilots to and from
seagoing vessels. The revolutions of both
main engines were somewhat slowed
down and the vessel had a reasonable
speed during this measurement run, which
was carried out within the inner harbour of
IJmuiden.
Secondly, so-called sailing measurements
were performed, whereby hull stresses
were monitored during about 5 minutes,
while the engines were producing constant
(maximum) power and sailing in severe
seaway in the vicinity of the entrance to
the outer port of IJmuiden.
Three such sailing measurement runs were
made. Due to severe slamming in the
given sea conditions the vessel's speed was
reduced to a maximum of about 18 knots.
Each run lasted about 5 minutes and the
signals were recorded on a hard disc
discretely at a rate of 2000 samples per
second. Thereby, hull stresses could be
clearly registered up to a frequency of 200
Hz. After each run the vessel returned to
IJmuiden where the registered data results
were investigated on the computer screen.
If the results were considered to be clear
enough, then a tape backup was made of
the data and then the next run was made.
In total 160 Mbytes of signal data for all
three runs has been backed up on tape.
Using the hull stresses as measured during
the pilotage simulation and sailing runs,
table 1 gives a quantitative description of
the extreme value of the hull stress signal
that each of the strain gauges showed.
This quantitative description of the stress
signal is built up using the following
stresses and stress ratios:
TABEL

the hull stresses, occuring at the second

spray rail during slamming, is shown in
figure 9.
As may be seen from figure 8, the hull
stress levels are well below the 125 N/mm2
allowable 0.2% yield stress level of the
aluminium (AlMg4.5Mn, 5083) material
of which the hull is constructed. The
maximum stress level encountered was 90
N/mm2 and this incorporates a safety
factor of 1.4.
In general, it may be expected that, if
stress values within the hull sufficiently
exceed the 125 N/mm2 value, then cracks
or breaking of the material may occur. If
however such excessively high stress
values are not encountered during
measurements under such extreme
working conditions of the vessel then one
can only conclude that the cracking of the
material results from metal fatigue. An
investigation into the Whler fatigue curve
of aluminium in welded condition shows
that even at such low stress levels as those
encountered during the hull stress
measurements presented in this paper
fatigue cracking may occur. This is true
even though the vessel, after six months
service, has had to bear such a short
amount of fatigue loading cycles. The
more so when one considers the
discontinuity of the structure at the
position of the second spray rail, where
initially the cracks occured. As is well
known, in such discontinuities the actual
stresses are even higher due to the fact that
the stress concentration factor is greater
than 1, generally 2-3. This higher stress
causes a significant reduction in fatigue
life. Also, it should be well noted that no
less than three welds have been placed on
that small area of material. With regard to
this last point, one should also keep in
mind that welding aluminium requires
much more heat than when welding steel,
because of the high thermal conductivity
of aluminium. This leads to a large heat
affected zone (HAZ) which in turn yields
very poor material quality of the
aluminium in the vacinity of the welds.

Figure 8 shows the maximum stress levels

measured in the frames during pilotage
and sailing conditions. A typical record of

The latter results in even poorer fatigue

resistance of the material when compared
to similar virgin aluminium material.
Another problem related to the welding of
aluminium is the residual stresses in the
weld itself, which may be as high as the
yield stress of the material itself. Preening
of the weld to reduce such residual stresses
is recommended.
With regard to the actual amount of stress
cycles encountered by the vessel, these are
far larger than the number of slams
incurred by the vessel alone as can be seen
in figure 9 from the measured stresses on a
time basis. This is so as the following
occurs directly after transient loading of
the hull structure; firstly a peak
disturbance, i.e stress, in the structure then
a damped free oscillation of the frame in
the plane of the frame itself. This leads to
a larger number of stress cycles than the
number of slams alone. Due to these in
plane oscillations a rotation point, about
which the frame locally oscillates, may be
deduced and, as result from this, also a
hard spot where the cracking indeed
occurs, see figure 10 (problem and
mechanism respectively).
Regarding all these unfavourable points
related to the material, construction and
production method of the 2nd spray rail, it
was recommended that the fatigue
sensitivity of this construction detail be
positively reduced by undertaking the
following steps, see also figure 10
(solution): The frame be detached from the
2nd spray rail locally and the web thickness
was also increased locally by adding extra
plating on both sides (approx. 2 mm
thickness). The extra plating was glued to
the existing web-plating of the frame. Also
the flange on the frames were fully welded
instead of intermittently.
4

forward of about 4 g . Since the ship has to

operate in sea conditions with wave
heights of 2.75 metres, peak accelerations
forward up to 6 or 7 times the acceleration
of gravity g can be expected, which is too
high from an operational point of view.
During vibration and stress tests, peak
accelerations forward of 6 g were
regularly measured in sea conditions with
wave heights of 2.5 metres. Lengthening
of the ship will reduce the RMS -values of
the vertical accelerations. Then, as can be
seen in figure 7, the peak accelerations
will be reduced much more.
In a study, the ship has been lengthened by
increasing the two intervals between the
three aftmost ordinates in the lines-plan
from 1.51 to 2.71 metres, which means
that the ship length at the waterline will be
increased from 15.10 to 17.50 metres. The
relative position of the fore peak bulkhead,
at which the (peak) accelerations have
been investigated, has been maintained.
Calculations with the strip theory
computer program SEAWAY for head
waves and a sea state defined by Beaufort
6, showed a reduction of about 9% of the
vertical accelerations forward at 1800 rpm
of the water jet engines. Slamming
pressures however, are mainly determined
by the square of the vertical relative
velocity between the ship and the waves.
From the computations it appeared that
this velocity will be reduced by about 25%
in Beaufort 6. Because of the quadratic
relation between pressures and velocities,
a reduction of the slamming pressures can
be expected of about 45%.
This trend for the pressures will be
confirmed in figure 7 for the vertical peak
accelerations caused by the slamming
pressures. The figure shows a very nonlinear relation between these peak-values
of the accelerations and the RMS -values.

Lengthening of the Ship

During the hydromechanic experiments,

the ship has been tested in sea conditions
with wave heights up to 2.0 metres, which
results in maximum peak accelerations

This very non-linear behaviour of the peak

accelerations and the resulting peak impact

10

pressures, was one of the reasons for the

owner's decision to increase the length of
his newly ordered ships by about 2.50
metres.
5

Conclusions

With regard to improved design of the

vessel, a number of conclusions are finally
made for the following generation of this
type of pilot vessel:
In calm water, the vessel can maintain
a speed of about 27 knots at 2200 rpm
of the water jet engines. In the higher
speed range ( N = 1800 rpm) and in
head wind and waves with a height up
to 2.0 metres, the speed loss appeared
to be less than about 2 knots. In
following wind and waves, an increase
of the speed with this same amount has
been found, due to the wind and
surfing of the ship on the waves.
The acceleration and natural stopping
tests confirm a good behaviour of the
ship in this respect. Unexpected forced
stopping by reversing the direction of
the thrust of the water jets causes large
longitudinal accelerations, which can
be
dangerous
for
non-seated
passengers.
The ship has excellent manoeuvring
characteristics. To maintain a certain
heading, a small water jet angle of
about 2.5 degrees is required.
Probably, this is mainly caused by the
similar rotational directions of the two
water jet shafts.
In the high speed range ( N = 2000
rpm)
and
maximum
allowable
operational sea states defined by a
significant wave height of 2.75 metres,
very high peak accelerations up to 6 g
or 7 g can be expected.
A lengthening of the ship by 2.40
metres will considerably reduce these
peak accelerations.
It was found that mechanical vibration
levels on board were well below
acceptable
levels.
The
highest

measured vibration velocity was 3.95

mm/s, but, in the majority of cases, this
value was well below 2 mm/s.
The transient loading of the vessel
structure, during fast free sailing in
adverse weather conditions and during
unwanted collisions during pilotage,
although yielding measured stress
levels that were significantly below
normally well acceptable values, still
resulted in problems related the fatigue
of some construction members. This
was also underlined by the advent of
fatigue cracking, discovered around the
time of the measurements.
With regard to an improved structural
design of the vessel, a number of
recommendations are made for the
following generation of this type of
pilot vessel:
The spray rails should not be an
integrated discontinuity in the hull
structure; an externally fitted spray
rail is therefore recommended.
Avoid too many welds being made
in a small area.
Increase the web frame thickness
from 5 to 7 mm, locally in the area
of the spray rails.
If integrated spray rails are utilised,
free the web of the frame locally in
the vicinity of the horizontal
plating of the spray rail.
Pay particular attention to welding
in aluminium with regard to the
fitting tolerances and the residual
welding stresses.
Up to a speed of about 16 knots ( N =
1800 rpm), where the ship is not actual
planing,
linear
strip
theory
computations with the computer
program SEAWAY give fairly reliable
predictions of the RMS -values of the
ship motions and the accelerations in
waves.

As a result of these tests related to design

and operational limits, the owner has since
ordered and received another 3 pilot
vessels, which incorporate many of the
11

recommendations as given in this paper.

At the end of 1996 the owner again placed
an order for yet another 3 similar modified
pilot vessels.
6

References

Journe (1992)
J.M.J. Journe, SEAWAY-Delft, User
Manual of Release 4.00, Report 910,
March 1992, Ship Hydromechanics Laboratory, Delft University of Technology,
The Netherlands.

Acknowledgements

Although the results and views expressed

in this paper are entirely those of the
authors, accurate design prototype studies
of this type are not possible without "realtime design information" from the field
itself and the opportunity to make all of
the measurements discussed within this
paper.
Therefore, the authors are very grateful to
Ing. A.C.M. Baaten of the ship owner
"Facilitair Bedrijf Loodswezen B.V." at
Rotterdam and the crew of MS
VOYAGER for their co-operation during
this project and to the ship owner for their
permission to publish the results of this
research.
Also, thanks to Prof.dr.ir. S. Hylarides
(former professor of the Delft University
of Technology) for his advisory work
concerning the vibration and mechanical
stresses measurements as described in this
paper and the analysis thereof.
A special word of thanks is indebted to Ir.
J. Ooms of the Delft University of
Technology for his contribution to the
preparation, the performance and (last but
not least) to the analysis of the
hydromechanic experiments.

Journe (1996)
J.M.J. Journe, The Behaviour of Ships in
a Seaway, Report 1049, May 1996, Ship
Hydromechanics Laboratory, Delft University of Technology, The Netherlands.
Ooms and Journe (1994)
J. Ooms and J.M.J. Journe, Pilot Vessel
Voyager, Part IIa: Hydromechanic
Behaviour (in Dutch), Report 1004-O,
October 1994, Ship Hydromechanics
Laboratory,
Delft
University
of
Technology, The Netherlands.
Pinkster and Hylarides (1994a)
Jakob Pinkster and S. Hylarides, Pilot
Vessel Voyager, Part III: Vibration
Measurements (in Dutch), Report OEMO
94-20-O, October 1994, Ship Design
Section, Delft University of Technology,
The Netherlands.
Pinkster and Hylarides (1994b)
Jakob Pinkster and S. Hylarides, Pilot
Vessel Voyager, Part V: Hull Stress
Measurements (in Dutch), Report OEMO
94-21-O, October 1994, Ship Design
Section, Delft University of Technology,
The Netherlands.

12

Figures

Figure 1 General Arrangement MS VOYAGER and Frame Section Number 10

13

Speed in waves

Speed in calm water

30

30
Mean values
in calm water
Experiments
in calm water

25

25

20
Speed (kn)

Speed (kn)

20

15

15

10

10

Calm water curve

Experiments
in Beaufort 6
Experiments
in Beaufort 4

ms Voyager
0

500

1000 1500 2000 2500

ms Voyager
0

500

N (rpm)

1000 1500 2000 2500

N (rpm)

Figure 2 Speed Trials in Calm Water and in Waves

Stopping test

Acceleration test
20

20

Experiments

Experiments
with N = 0 rpm
Experiments with
reversed thrust

15

15

Speed (m/s)

Speed (kn)

N = 1750 rpm

10

10
N=
875 rpm
5

N = 0 rpm

N = 0 rpm
ms Voyager

drift
ms Voyager

0
0

10 20 30 40 50 60 70 80
Time (s)

-20 -10 0

10 20 30 40 50 60
Time (s)

Figure 3 Acceleration and Stopping Tests in Calm Water

14

Rate of turn

Turning circle

15

10

rate of turn (deg/s)

Y (m)

150

N = 2200 rpm
N = 1800 rpm
N = 1200 rpm

100

-5

50

-10

N = 1200 rpm
V = 8.9 kn
= 6.3 deg
0

ms Voyager

50

-15
-20 -15 -10 -5

100

X (m)

ms Voyager
0

10 15 20

waterjet angle (deg)

RMS vertical acc. (m s)

Figure 4 Manoeuvring tests in calm water

N
E
S
Beaufort 4
5 N = 1800 rpm
V = 15.5 kn
4
6

N
E
S
Beaufort 4
5 N = 2200 rpm
V = 25.4 kn
4

RMS vertical acc. (m s)

90

180

N
E
S
Beaufort 6
5 N = 1800 rpm
V = 16.0 kn
4
6

270

360

450

1
90

180

270

Heading (deg)

360

180

N
E
S
Beaufort 6
5 N = 2050 rpm
V = 21.2 kn
4
3

90

450

90

Head waves
SEAWAY
Acc. amidships
Acc. forward

180

270

360

450

270

360

450

Heading (deg)

Figure 5 Measured and computed RMS values of vertical accelerations

15

50

E
S
W
N
E
Head waves
N = 2200 rpm, V = 25.4 kn
N = 1800 rpm, V = 15.5 kn

Peak accelerations (m s)

40

S
W
N
E
Head waves
N = 2050 rpm, V = 21.2 kn
N = 1800 rpm, V = 16.0 kn

30
20

10

10

-10

-10
90

40

20

Beaufort 6

Beaufort 4

30

50

180

270

360

450

90

180

Heading (deg)

270

360

450

Heading (deg)

Figure 6 Measured peak values of vertical accelerations forward}

50

Derivative
Mean values
Experiments

Peak of accelerations (m/s )

40

5 RMS +0.0922 RMS

30
20

5 RMS

10
0
-5 RMS +0.552 RMS
-10

ms Voyager
0

-5 RMS
1

4
2

RMS of accelerations (m/s )

Figure 7 Peak versus RMS values of vertical accelerations forward}

16

C
3

D
4

E
5

F
6

Measurement location along frame

90
Frame 10
80 nd
70 2 spray rail
60
50
40
30
20
10
0
-10
1
2
3

Maximum stress (N/mm )

B
2

A
B
90
80 Frame 14
70
60
50
40
30
20
10
0
-10
A
B
1
2

A
1

Smax outer (sailing)

Smax inner (sailing)
Smax outer (pilotage)
Smax inner (pilotage)

Maximum stress (N/mm )

Measurement location along frame

90
Location A
80
70
60
50
40
30
20
10
0
-10
10 11 12 13 14 15 16 17

Maximum stress (N/mm )

90
80
70
60
50
40
30
20
10
0
-10

Maximum stress (N/mm )

Smax flange (sailing)

Smax outer (sailing)
Smax inner (sailing)
Smax flange (pilotage)
Smax outer (pilotage)
Smax inner (pilotage)
B
C
D
E
F
Frame 10

C
3

D
4

E
5

F
6

nd

Frame number
Smax outer (sailing)
Smax inner (sailing)
Smax outer (pilotage)
Smax inner (pilotage)

Measurement location along 2 spray rail

Smax PS (saling)
Smax SB (sailing)
Smax PS (pilotage)
Smax SB (pilotage)

Figure 8 Hull stress levels during pilotage and sailing conditions

17

Figure 9 Hull Stress Level Measurement Results for Frame 10 at 2nd Spray Rail PS, in
Sailing Condition BF 8, Significant Wave Height 2.5 Metres

Figure 10 Structural Improvement 2nd Spray Rail

18

19

20