INTERACTION

by Dr. Ian W. Dand, B.Sc., Ph D, F.Eng, FRINA BMT SeaTech Ltd.

1. INTRODUCTION

When ships come close to each other, one is apt to 'feel' the presence of the
other to a greater or lesser degree. This may be manifest in a number of
ways, ranging from involuntary speed changes to catastrophic (and again
involuntary) course changes which may lead to collision, or grounding.

Of course, it is wise to avoid these sudden, unexpected occurrences by leaving

plenty of sea room between one ship and another. But this is not always
possible : ships in a long narrow approach channel have to pass each other
and some (tugs and pilot launches for example), by the very nature of their
work, must go in close to the ships they are attending.

The hydrodynamic phenomenon which causes ships in close proximity to

exhibit this behaviour is known as interaction and its causes and effects are the
subject of this paper.

First we look at the causes, then give some practical aspects which incline
toward operations in shallow or confined waters, and finally consider some
famous examples of interaction which themselves have spurred research into
this important and fascinating topic.

2. WHAT CAUSES INTERACTION?

Interaction at sea has a number of resonances with similar effects on land.

The sway that is felt when a car passes (or is passed by) a large lorry on a
motorway, the suction created by a high speed train passing through a station
(giving rise to the yellow warning lines painted on platforms) and ground
effect experienced by aircraft when taking off or landing are all examples.
Indeed they all arise from the same basic fluid dynamic cause, although with
ships the phenomenon is more insidious and can strike unexpectedly.

In the marine world, interaction is caused when the pressure fields around
ships interact. All ships are in a state of balance in the water, held in position
at rest by hydrostatic pressures to which are added dynamic pressures when
the ship begins to move. Dynamic pressures are those which cause (and
indeed are represented by) the familiar diverging and transverse wave systems
generated by a ship moving calm water.
 If the pressure system acting over the underwater hull of the ship is changed
in any way, its state of balance will be affected. It may go faster or slower,
move off course, sink deeper into the water or rise on to its surface (squat),
or possibly trim differently. If a ship moves from deep into shallow water the
presence of the seabed will have an effect on hull pressures as the water gets
more shallow. Waves also become steeper in shallow water and so do the
ship's own waves. The upshot of this is that the ship expends more energy
making its own wave system, so its resistance increases and, unless it changes
its engine setting, it will slow down. It will also exhibit an increased tendency
to squat. Finally it will cause, and feel, enhanced interaction.

In shallow water, ship-ship interaction becomes more severe from these purely
hydrodynamic causes. Figure 1, from reference 1, attempts to demonstrate
that, as ships pass close by each other, they may be sucked together or pushed
apart, turned toward or away from each other, all due to interaction.

Figure 2 shows some measurements of the surge and sway forces, together
with the yaw moments experienced by one model ship when passed by another
on a reciprocal course. The effect of reducing water depth can be seen
readily and it is of interest to note that transient changes in mean sinkage and
running trim (squat) also occur.

As well as ship-ship interaction, there is another form of interaction which

occurs when a ship is near a canal or fairway 'bank'. As shown in Figure
3 the bank acts in many ways as a mirror and a ship close to a bank behaves
as if it was close to its mirror image. The result is that the ship will be
generally sucked toward the bank. But the wave system of the ship will also
be affected and the bow wave close to the bank will increase in size and form
a pressure 'cushion'. This is enhanced if the bank is sloping, when the wave
may locally 'go critical' and get even steeper. This 'cushion' will tend to
push the bow away from the bank and, if the speed is high enough, the 'push'
from the cushion can overcome most of the suction pulling the ship toward the
bank so that it tends to be pushed bodily away.

3. SHIP-SHIP INTERACTION

Having looked briefly at the causes of interaction, what about its effects and
how can these be overcome?

Model tests have shown that, in general terms, ship-ship interaction varies.

• as the square of the speed.

• inversely with distance off.

• roughly as the inverse square root of the underkeel clearance to

draught ratio.
 These mean that:

• the faster the ship moves, the worse interaction becomes.

• the greater the lateral separation between ships, the better.

• the smaller the underkeel clearance, the bigger the effect.

So the lesson to be learnt is that the correct speed and distance off are vital
if interaction is to be avoided, or at least its effects minimised. Of course, in
some situations it is not possible to reduce speed or increase distance off in
which case an awareness of the possibility of interaction and its effects is
important. In such cases "fore-warned is fore-armed".

With this in mind, the following general effects of interaction for ships
passing on parallel or reciprocal courses are given. They are illustrated
diagrammatically in Figures 4 and 5.

3.1 Head-on Passing

1. Interaction begins to be felt with the bows of both ships being pushed
away from each other accompanied by a slight increase in speed.

2. At the same time, the vessels feel a slight bodily repulsion.

3. As the ships pass, the 'bow-out' moment turns to 'bow-in' and the
repulsion reduces.

4. The 'bow-out' moment then returns as passing continues, but is now

stronger. Indeed it may cause both ships to sheer away from each
other once they have passed. A slight reduction in speed may also be
felt.

5. Finally a weak 'bow-in' moment accompanied by a repulsion may be

felt.

Comment

Passing on reciprocal courses has the merit (from an interaction point of view)
of being quick so that often the ship does not have time to react to the various
interaction forces and moments it feels. Usually the dominant effects are the
'bow-out' turning moments as the ships begin to pass and the stronger bow-
out moments once passing is almost over. The former is beneficial and is
usually small enough to control, while the latter is much stronger and, if not
anticipated, could cause one or other of the vessels to sheer toward the bank
of a narrow channel.
3.2 Overtaking

1. As the overtaking vessel overhauls the overtaken vessel, two things

happen:

• a small bow-in moment is experienced by both ships.

• the overtaking ship speeds up and the overtaken ship slows

down.

2. As the relative velocity when overtaking may be low, interaction has

time to take effect and at this juncture, the overtaken ship may be
caused to turn across the bows of the overtaking ship which may
perversely turn toward her. As a result both ships may collide (see
Figure 5).

3. If a collision does not occur (perhaps because the vessels are on

slightly converging courses) then the overtaking vessel will move past
the other and both will feel powerful bow-out moments together with
a mutual attraction. This may cause both ships to 'fly apart' and their
sterns to collide as shown in Figure 5.

4. Usually an overtaking manoeuvre, affected by interaction, does not get

as far as the final stages w ithout collision or a violent change of
course. If it did, it would find the overtaking ship experiencing an
increase in resistance which slows it down. At the same time, the
overtaken ship feels its resistance reduce, so it speeds up. The result
is that the overtaking ship finds it more difficult to complete its passing
manoeuvre and may, in extreme cases, get 'trapped'.

Comment

Overtaking manoeuvres should always be treated with caution. Relative

velocities are low so the ships are in proximity long enough for interaction to
have an effect. C ollisions may result, or the vessels may get
hydrodynamically trapped together; the former can be avoided by allowing
sufficient distance off, (or not overtaking at all), the latter by one or other of
the vessels slowing down.

Interaction when overtaking depends on the relative velocity; the lower it is

the more likely it is that problems will occur. If it is zero then the ships are
moving along together, as in a Replenishment at Sea (RAS) operation carried
out by warships. In such cases the effects of interaction (albeit in deep water
and therefore more controllable) must be known so that the most benign
position alongside can be found.

Other vessels must move in concert as part of their daily routine, the most
obvious being the harbour tug.
3.3 Ship and Moored Ship Interaction

If one of the ships in a passing manoeuvre is stationary (moored alongside a

jetty for example) it can still be affected by interaction. Just the same
sequence of forces and moments takes place but, because the ship's mooring
system is perhaps least stiff in surge, the moored ship may move ahead and
astern on her berth. This, coupled with sideways and rotary motions may
give rise to snatch loads in any slack or poorly-tended moorings which could
break. Once one line has broken, others may soon follow.

Comment

Speeds past moored ships should be kept as low as practicable and should be
at their lowest when underkeel clearances are small. Distance off should also
be kept as large as is practicable.

3.4 Tug-Ship Interaction

The tug is generally much smaller than the ship it is attending and while a
given depth of water may be deep for the tug, it may well be shallow for the
ship. This means that, whereas the ship will have a big interactive effect on
the tug, the tug will, naturally, have virtually no effect on the ship.

M odern tractor or reverse-tractor tugs have enough pow er and

manoeuvrability to be in less danger from the effects of interaction than their
conventionally propelled counterparts. This is not to say that they are
unaffected. Figure 6 shows measurements of the interaction sway force and
yaw moment felt by a tractor tug model keeping pace with a large ship; it is
seen that large forces develop. However, the fast response and enhanced
manoeuvrability of such tugs means that they are much more able to
manoeuvre out of difficulty.

For any conventionally powered (and steered) tugs and other vessels similarly
equipped, Figure 7 shows diagrammatically the sort of interaction forces and
moments they will experience when they come alongside. Clearly there are
areas near the bow and stern that are best avoided because the control that the
rudder exerts adds to, rather than subtracts from, the effects of interaction.
Of particular interest is the tendency to turn under the bow of the larger vessel
brought about by interaction. This has caught a number of conventional tugs
unawares over the years with disastrous consequences. The sudden changes
in the interaction forces and moments acting on the vessel as it alters its fore
and aft position alongside the bigger ship are largely to blame; if they are not
anticipated by the helmsman, the smaller vessel will drive itself under the bow
of the bigger ship.
4. INTERACTION NEAR FIXED BOUNDARIES

It has already been mentioned that interaction can occur when a ship is near
a bank. In general, fixed boundaries to waterways, whether they be banks of
canals, rivers or fairways or whether they are the walls of enclosed docks, can
have effects on ships which may be sudden and unexpected. Some of these
are now discussed.

4.1 B an k Effects

It has already been shown that bank effects are manifest as a bow-out turning
moment together with a suction. This will be experienced whether the bank
is vertical (as in a waterway with piled sides), flooded (as in a fairway) or
sloping (as in a canal). It will also occur if the water shoals to one side of the
ship.

The practical outcome of this phenomenon is usually that the ship sheers away
from the bank. The 'bow cushion' dominates and turns the ship which then
moves away from the bank and, as it does so, experiences less and less
interaction as distance from the bank increases. This means that, to move
parallel to a bank, interaction is countered by steering toward the bank; if the
rudder is correctly set, a balance can, in principle, be found to cancel
interaction (see Figure 8).

Clearly, passage along the centreline of a waterway, midway between the

banks, should avoid bank effects as they will cancel. This will be true in a
waterway such as a canal with uniform banks, but in fairways and rivers
whose banks may be anything but uniform, it cannot be relied upon implicitly
It is often argued that in such circumstances, the ship will automatically 'find'
the centre of the river, the bank effects acting as a form of control device.
While it is true that bank effects will turn a ship toward the centre of a
waterway, their relationship to the vessel's mass, inertia and turning ability
is very unlikely to ensure that the ship does not simply over-shoot the
centreline and ground on the other bank.

A possible scenario in such a case is that the ship will sheer off one bank,
head across the centre of the waterway to approach the opposite bank at an
angle. If this angle is right, the vessel may turn, under the influence of the
growing bow cushion, to leave the bank, without touching, and head for the
other side of the waterway (Figure 8). Usually this process is divergent and
'reflection' does not occur a second time so that the ship runs aground.
4.2 Ships in Basins

Ships moving in enclosed basins in which other ships are moving or moored
can generate interaction-like effects. A few are now considered.

Swinging and M anoeuvring

A ship manoeuvring unaided in an enclosed basin may use a combination of

propellers and bow thruster. This may cause the water in the basin to move
and the resultant swinging o f the ship (which acts as a form of 'paddle') will
cause further movement and pressure changes. Ships moored in the vicinity
may feel these pressure changes and range or surge on their moorings.

Tug Pumping

The modern harbour tug is usually powered by one or more propellers or,
more generally nowadays, by two powerful thrusters. Not only are these
good propulsion devices, but in the confined space o f a basin they act as
effective pumps, setting water in motion. In a very confined space (especially
if the tugs are on short lines), they can cause the ships they are attending to
move in unexpected ways. Figure 9 (from reference 2) shows a situation
which was modelled physically; it shows how the flow induced by tug wash
causes local pressure changes which affect the ship. Notice how the ship
moves bodily toward the tug even though the direction in which the tug is
pulling does not suggest such behaviour. Sim ilar effects have been
experienced in lock-bell-mouths when tug action has inadvertently caused
ships to move in an unexpected direction. In extreme cases tug wash can
cause an effect which is directly contrary to that expected. Figure 10 (also
from reference 2) shows the turning moment measured on a ship model when
'towed' by a tug in the manner shown. Notice how the turning moment on
the ship actually changes sign (ie. acts in a direction opposite to the expected)
at the shallowest water depth. This is yet another example o f the powerful
effect of shallow water, and suggests that care should be taken when using
powerful tugs on short lines in enclosed basins.

The Following Wake

When a ship slows down too abruptly, the water moving with it may not be
so obliging. The ship's wake takes time to slow down and, in shallow,
confined alters it should be remembered that the body of water which moves
with the ship takes time to slow down and in so doing, will overtake the ship.
This may often affect the vessel and can move it ahead or, in extreme cases,
turn it in an uncontrolled manner (see Figure 11). The lesson is clearly to
reduce speed, or a swinging manoeuvre gradually.
5. EXAMPLES OF INTERACTION

There are a number of marine accidents where interaction has played the
main, or at least an important, role. Some are listed here and their
importance lies not so much in their details as in the fact that most of them
played an important part in developing our understanding of the phenomenon
known as interaction.

'Olympic/Hawke' Collision

The 'Olympic', the sister ship of the Titanic' was in collision with the
cruiser HMS 'Hawke' in the Solent in 1911. Both ships were on similar, but
converging, courses and the 'Hawke' suddenly and unexpectedly sheered to
port into the 'Olympic'. There was suggestion of both ship-ship interaction
and bank effect on the 'Hawke', as well as a demonstration of the effects of
speed and converging courses. This case really began the modem study of
interaction, for a number of investigations were done for the subsequent
litigation and beyond. The 'Olympic' was taken out of service for repair and
her presence at Harland and W olff delayed completion of the Titanic.

Titanic '/'New York'

While leaving Southampton on her maiden (and only) voyage, the Titanic',
sister ship of the 'Olympic' was almost in an interaction-induced collision.
While passing the Dock Head she passed close to two smaller passenger liners
moored abreast. The outer one, 'New York', broke free and drifted toward
the Titanic'. Only the quick actions of a tug prevented a collision, although
the Titanic' was delayed. It is of interest to note how interaction played a
part in the short life of the Titanic'; her completion was delayed by the
'Olympic'/'Hawke' incident and her voyage was delayed by her incident with
the "New York'. Had these incidents not occurred, the story of the Titanic'
might have been very different.

'Queen Mary7'Curacao ’

During the Second World War the 'Queen Mary', in use as a troop transport,
cut the light cruiser 'Curacao' in two. Both vessels were moving at speed in
deep water and the 'Queen Mary' was carrying out a zig-zag manoeuvre.
Model tests carried out after the event showed that interaction from the larger
ship caused the warship to move toward her from a significant distance off.
The investigation (reference 3) provided some illumination of the powerful
effect of speed and ship size on interaction.

'HMS N elson' Grounding

Just before the Second World War, HMS Nelson, on leaving Portsmouth
Harbour, took a sudden sheer to starboard and ran hard aground on the
shallows off Haslar Wall. Subsequent model tests (reference 4) showed
evidence of strong bank effect and gave the first published evidence of the
interaction caused by a nearby shelving beach.
 'Royston Grange9/Tien Chee'

In 1972 the reefer 'Royston Grange' collided with the tanker Tien Chee' in
the River Plate. The 'Tien Chee' was heavily constrained by her draught and
was moving more or less on the channel centreline. This caused the 'Royston
Grange', who was making good speed, to move well to starboard. This
caused her to sheer, from bank effect, into the Tien Chee'. Both ships
caught fire and there was heavy loss of life. The subsequent investigation
showed the importance of speed in any interaction incident as well as the
magnitude of bank effect. It also initiated studies of interaction aimed at
explaining the effect and informing the maritime community. This paper is
the latest example of that continuing task.

6. REFERENCES

1. DAND, I.W.: The Physical Causes o f Interaction and its

Effects'. Nautical Institute Conference on Ship
Handling, Plymouth, 1977, p. 34-73.

2. DAND, I.W.: Tug Wash Effects in Confined Waters' Seventh

International Tug Convention, Thomas Reed
Publications Ltd., London, 1982.

3. ROBB, AM .: 'Interaction Between Ships : A Record of Some

Experiments and Evidence of Wave Effect'
Trans. Royal Institution of Naval Architects,
vol. 91. 1949, p. 324-339.

4. GAWN, R.W.L.: 'Steering and Propulsion of HMS 'Nelson' in

a Restricted Channel' Trans. Royal Institution of
Naval Architects, vol. 92, 1950, p. 82.
FIGURE 1
 FIGURE 2

SURGE FORCE

2 0
R E P U LS IO N
IO
SWAY FORCE 0 -2 06
--4 - 1

-1 -0
A T T R A C T IO N

■2 0
KEY
—— — 1 10
119
— 1 49
4 2 65
RESULTS MEASURED
ON THIS MODEL _ DEPTH/DRAUGHT

bow our
TURNING MOMENT

BOW IN

Cs

-2 2 ■18 - 14 -1 0 - 0 -6 -0 2 0 2 0 6
MEAN SINKAGE L _ L 1 1 1 L
i
- ■0 25
s i NKAGE
- 1 -O 5

Ct
1 BOW UP
10
-2 2 -18 -0 2 0 2 0 6 I
RUNNING TRIM L 1

I BOW DOWN
BOW TO BOW

HEAD-ON ENCOUNTER,
EFFECT OF DEPTH / DRAUGHT RATIO

11
 STRAIGHT ’ DIVIDING' MIRROR* VERTICAL BANK
STREAMLINE

STREAMLINES
-1 2

TWO IDENTICAL SHIPS SIDE-BY-SIDE ONE SHIP I ITS MIRROR IMAGE ONE SHIP PARALLEL TO
A VERTICAL BANK

BANK EFFECTS

FIGURES
COURSE OF SHIP PARALLEL TO LINE OF BANK
 FIGURE 4

WITHOUT

INTERACTION DURING RECIPROCAL PASSING

 FIGURES

COLLISION

INTERACTION WHEN OVERTAKING :

ALTERNATIVE SCENARIOS
14 -
 FIGURE 6

i-O - ATTRACTION SIOEFORCE

* o /L (
—I
1-2

TANKER MODEL

IO
BOW IN

TURNING MOMENT

BOW OUT
DIAGRAM OF MODEL D

INTERACTION FORCES & MOMENTS INDUCED ON

-r-r^ a r * T r \ n t i iz >
CLJcOHS

•EI

FIguro 7 - lnterc)c6o entre navio e reboccldor.

SIM -Interclction
Curso Regular 2019
® Telepresenclol
 I < RU’D D E R -IN D U C E D FORCE I MOMENT

__ POWER REQUIRED TO M A i n TAI N STATION

'B' IS POINT OF SAFE APPROACH

FOR THE SMALLER VESSEL AS
ITS CONTROL ACTIONS WILL COUNTER
T U G - S H IP IN T E R A C T IO N INTERACTION

FIGURE?
 FIGURES

CONTROL UNCONTROLLED SHEERS

TRACKS DUE TO BANK EFFECT

(DIAGRAMMATIC ONLY)
 FASTER FLOW NEAR BOW OF SHIP FLOW INDUCED BY TUG W ASH- LOWER PRESSURE
SO LOWER PRESSURE THAN STERN IN WATER VOLUME TO STARBOARD OF SHIP

NET SWAY FORCE INDUCED ON SHIP

NET TURNING MOMENT INDUCED ON SHIP

L IT T L E OR NO FLOW INDUCED SO NO LOWERING
OF PRESSURE IN WATER VOLUME TO PORT OF SHIP

' .7 / ’ 7 7 7 7 7 7 7 / 7 7 / 7 7 7 7 7 7 7 7 7 7 7 7 ' 7 7 7 7 7 7 7 7 7 7 7 '7~7

INDIRECT EFFECT OF TUG W ASH IN C O N F IN E D WATER

FIGURE 9
 FIGURE 10
(ACTUAL RESULT)/(EXPECTED RESULT)%

LONGITUDINAL FORCE

1 L I .... I l 1 DEPTH/DRAUGHT
10 11 1 2 1 3 1 4 15

TUG W ASH EFFECTS IN A CONFINED SPACE

 FIGURE 11

FOLLOWING WAKE
TURNS, AND MOVES,
SHIP

SHIP STOPS OVER

GROUND, CANTED TO
STARBOARD

SHIP STOPPING RAPIDLY

SHALLOW CHANNEL

-STOPPING MANOEUVRE BEGINS

-*■ WAKE MOVING WITH SPEED OF SHIP

- 20 -

THE FOLLOWING WAKE