A

r
c
h
i
v
e

o
f

S
I
D

Journal of Faculty of Engineering, Vol. 42, No. 6. December 2008, PP. 809-821 809
* Corresponding author: Tel: 61112613 , Fax: 66403808 , Email: [email protected]


Numerical Modeling of Launching Offshore Jackets from
Transportation Barge & the Significance of Water Entry
Forces on Horizontal Jacket Members

Nikzad Nourpanah
1
and Moharram Dolatshahi Pirooz
*2

1
School of Civil Engineering, University College of Engineering, University of Tehran
2
Asisstant Professor, School of Civil Engineering, University College of Engineering,
University of Tehran
(Received 26.7.85, Revised Manuscript received 15.7.86, Accepted 29.2.87)
Abstract
Development of a numerical model which describes launching of offshore jackets from barge is
presented in this paper. In this model, in addition to capabilities of commercial softwares, water entry
forces on jacket members and an implicit Newmark solution technique are included. The results are in
general agreement with other numerical softwares available (SACS). Fluid forces acting on jacket and the
importance of each one is discussed. It is observed that water entry forces on horizontal jacket members are
very significant and may locally govern the design of these members. This force is more important for
horizontal slender members near the mud-line, which do not experience significant environmental loading
in operating conditions. Therefore the water entry impact force with large magnitude can cause over-stress
and/or ovalling of near mud-line members. It is also observed that taking water entry forces in account
modifies the jacket trajectory only in a little extent.

Keywords: Launch - Offshore Jacket - Hydrodynamic Force - Water Entry Force
Newmark - SACS

Introduction
In most relatively shallow waters,
fixed platforms are the optimum
economical choice. One of the important
aspects of offshore fixed platform design is
their installation procedure. Among the
several conditions during installation,
launching is probably the most critical.
This operation has to be modeled and
examined carefully in order to insure safe
separation of jacket from barge.
As the platforms get heavier,
launching technique for placing the steel
template jacket from transportation barge
onto its location is preferred, rather than
lifting it directly by means of crane barges.
This operation, besides the preparations,
takes only a few minutes, which shows the
operations dynamic and critical nature. A
complete time-history launch analysis is
generally required for three reasons: a)
checking the jacket will not hit mud-line,
b) checking that jacket members (launch
truss) can resist the reaction forces exerted
by rocker arm and c) checking that
members can resist the hydrodynamic
forces experienced during plunging in sea.
Vasicek and Lu [1] presented a
numerical model of jacket launching. They
used an iterative finite difference scheme
to solve the governing equations of
motions. Sphaier et al. [2] describe the
theoretical backgrounds of a launch
modeling software. In their work, the
system of governing differential-algebraic
equations is solved (algebraically) at the
beginning of time step with accelerations
and reaction forces as unknowns. These
values are assumed constant through the
time step and displacements at the end of
time step are calculated. Nelson et al. [3]
presents a numerical modeling of a similar
case; lifeboat launching. Similar to [2],
accelerations and reaction forces are
calculated at the beginning of each time
step. Reactions are considered constant
through the step and the resulting system
of differential equations is solved using the
Runge-Kutta method.
In this paper a mathematical
formulation of the problem accounting for
water entry forces using experimental slam
coefficients [4], on the launched body is
presented. A numerical solution using the
www.SID.ir
A
r
c
h
i
v
e

o
f

S
I
D

810 Journal of Faculty of Engineering, Vol. 42, No. 6. December 2008


Z
X
O
Y
rj
rb
rrb
rjr
CG j
CG b

Figure 1. Jacket rotating about rocker arm. Orientation of the global coordinate system and the relative
positions of jacket and barge CGs.

Newmark linear acceleration method is
used. Also, iteration is performed in each
step to obtain solutions consistent with the
constraint equations. The effect of each of
the fluid forces acting on jacket during
launching, including water entry are
investigated.

Description of Problem
A calm sea/weather is required for
launch. Jacket is transported on launch
barge to installation site. Barge
compartments are ballasted; trimming the
vessel a large angle (2
o
-4
o
) and sea-
fastenings are cut. This is the start point in
a launch analysis. Jacket slides toward
rocker arm located at barge aft. After
passing the rocker arm it rotates and
plunges in sea. After separation of barge
and jacket they both oscillate a few times
and come to rest [5], [6]. Depending on
the relative motions of jacket and barge,
five phases of motion are possible [7]:
Phase 1: Jacket is sliding on barge deck
due to hydraulic jack pushing or winch
pulling.
Phase 2: Jacket is sliding on barge deck
under action of its own weight.
Phase 3: Jacket is only rotating about
rocker arm pin.
Phase 4: Jacket is rotating about rocker
arm pin and sliding on tilting beam (rocker
arm) simultaneously.
Phase 5: Jacket and barge have separated.
If initial trim angle of barge is greater
than dynamic friction coefficient angle,
after cutting sea-fastenings, phase 2 will
occur, however due to large friction at
beginning, an initial pull/push by
winch/jack is required. If not, phase 1 will
occur and the jacket will slide on barge
deck with the constant velocity of winch.
There are two conditions necessary for the
jacket to start rotating on rocker arm: 1)
jacket CG (center of gravity) passes rocker
arm pin and 2) reaction moment is
negative. The necessity of condition 2)
implies that jacket CG might slide a few
meters past the pin and then start rotating.
If the two mentioned conditions occur
while jacket is still in phase 1, phase 3 will
occur. Jacket stops here and starts to only
rotate about the pin. Rotation continues
until angle of jacket exceeds static friction
coefficient angle. From this point on,
jacket rotates and slides simultaneously
(phase 4). Therefore, Phase 3 can only
occur after phase 1. Phase 4 can only occur
after phase 2 or 3. Geometry of barge and
jacket during phases 3 or 4 is shown in Fig.
1. A right-hand-sided coordinate system
with its origin located at S.W.L. and above
CG of barge is used to describe the
problem. Positive Z axis is upward and X
axis lies in water-plane with positive
direction pointing toward barge aft (Fig.
1).
Equations governing the motion are
Newton-Euler equations of motion. These
www.SID.ir
A
r
c
h
i
v
e

o
f

S
I
D

Numerical Modeling .. 811


equations are the same for all phases. In
addition to equations of motion, constraints
are needed to describe the motion. These
constraints vary in each phase. By
assembling the motion and constraint
equations, one gets a system of
differential-algebraic equations, with
accelerations and reaction forces as
unknowns.

Equations of Motion
The Newton-Euler equations of motion
state that rate of change of system
momentum equals the forces acting on
system. We consider jacket and barge as
two separate systems, with the common
reaction forces acting on both. In practice
jacket is placed on barge deck such that
their CGs lay in a vertical plane above
each other, eliminating any yawing
moments. Therefore launch is essentially a
two dimensional problem. In this regards
and by assuming jacket and barge as rigid
bodies, position of each body can be
represented by three components, namely
X, Z and coordinates of their CG.
Equations of motion for these six degrees
of freedom are as follows:
X
j F
X
w
X
c j j
F P P X m
,
+ + =
& &

j
Z
j , F
Z
w
Z
c j j
W F P P Z m + + =
& &

) ( ) (
) ( ) (
,
j c
Z
w
Z
c
j c
X
w
X
c
Y
j F
Y
c j j
X X P P
Z Z P P M M I
+
+ + + =
& &
X
b , F
X
w
X
c b b
F P P X m + =
& &

b
Z
b , F
Z
w
Z
c b b
W F P P Z m + =
& &

) ( ) (
) ( ) (
,
b c
Z
w
Z
c
b c
X
w
X
c
Y
b F
Y
c b b
X X P P
Z Z P P M M I
+ +
+ + =
& &
(1)
Where:
m: Mass
I: Moment of inertia about CG of body
X, Z, : Components of position vector.
P, F: Forces. M: Moment. W: Weight.
Subscripts j, b denote jacket and barge
respectively
Subscripts c, w, F denote contact, winch
and fluid (hydrodynamic and hydrostatic)
forces respectively
Double dot denotes second derivative with
respect to time.
Eleven unknowns appear in these six
motion equations, namely six
accelerations, three contact forces and two
winch forces. Therefore one needs
constraint equations to complete the
system of equations.

Constraint Equations/Relations
Constraint equations are
geometrical/force relationships that relate
the motion of jacket and barge. These
constraints are required in all phases
except phase 5. In phase 5 the two bodies
have separated and reaction forces are all
zero, therefore only the six accelerations
are unknown and the six equations of
motion Eq. (1)-(6) suffice.
In phases 1 to 4 where the two bodies
are connected, positions of their CGs are
related. Differentiating this relation with
respect to time yields the velocity relation
and another differentiation results in the
acceleration constraint of the system:
jr rb b j
r r r r
r r r r
+ + =
jr jr j rb b b j
r r r V V
r
&
r
r
&
r
r
&
r r
+ + + =
jr jr j jr j j
jr j rb b b rb b b j
r r r
r r r a a
r
& &
r
&
r
&
r
r
&
r
&
r
r
& &
r
r
&
r
&
r
r
& &
r r
+ + +
+ + + =


2 ) (
) (
(2)
Where:
r : Position vector
V : Velocity a : Acceleration
Dot denotes first derivative with respect to
time and subscript r denotes rocker arm
pin.
The vector form of acceleration
constraint results in two constraints in
scalar form:
) ( ) ( 2
) ( ) (
) ( ) (
2
2
j t j j t
j r j j r j
b b r b b r b j
Cos V Sin V
X X Z Z
X X Z Z X X



& &
& & &
& & & & & & &
+
+
+ =

) ( ) ( 2
) ( ) (
) ( ) (
2
2
j t j j t
j r j j r j
b b r b b r b j
Sin V Cos V
Z Z X X
Z Z X X Z Z



& &
& & &
& & & & & & &


=
(3)
Where:
www.SID.ir
A
r
c
h
i
v
e

o
f

S
I
D

812 Journal of Faculty of Engineering, Vol. 42, No. 6. December 2008


t
V : Jacket relative sliding velocity on
launch skids/rocker arm.
t
V
&
: Jacket relative sliding acceleration on
launch skids/rocker arm.
Equation (3) is valid throughout
phases 1 to 4. As can be seen in Eq. (3)
another unknown, Jacket relative sliding
acceleration, is introduced in the constraint
equations. Therefore in phases 1 to 4 there
are a total of 12 unknowns. In addition to
Eq. (3), four more equations are needed to
construct a system of equations for the 12
unknowns. These additional equations vary
for each phase.

Constraint Equations during Phase 1
During phase 1 jacket is sliding on
barge deck due to winch pulling or
hydraulic jack pushing, therefore relative
velocity of jacket sliding on launch skids is
constant and equal to velocity of winch.
This implies that relative sliding
acceleration is zero, resulting in the
following relation:
0 V
t
=
&
(4)
The following equations result from
these facts respectively: winch force is
parallel to barge deck, contact forces result
from normal reaction and friction force
which are related by the dynamic friction
coefficient, and finally, barge and jacket
rotate together.
0 ) ( Cos P ) ( Sin P
j
Z
w j
X
w
= + (5)
0 )) ( ) ( (
)) ( ) ( (
=
+
j d j
Z
c
j d j
X
c
Cos Sin P
Sin Cos P


(6)
j b

& & & &
= (7)
Where:
d
: Dynamic friction coefficient between
jacket and launch skids.

Constraint Equations during Phase 2
In phase 2, jacket is sliding on barge
deck under action of its self weight.
Physically this means that angle of barge
has exceeded angle of dynamic friction.
This implies that relative sliding velocity
exceeds winch velocity; so equation (4) is
not valid. Also winch force is zero. Rests
of the equations are the same as phase 1.
0 P
X
w
=
0 P
Z
w
= (8)

Constraint Equations during Phase 3
In phase 3, jacket is rotating about
rocker arm pin without sliding. Therefore
Eq. (4) is valid. Equation (8) is also valid.
Note that because there is no sliding,
friction force relation, Eq. (6), is not valid.
In addition, rocker arm pin does not resist
moments, therefore:
0 M
Y
c
= (9)

Constraint Equations during Phase 4
Additional constraints/relations in this
phase are the same as phase 3, except that
due to sliding, Eq. (6) is valid and equation
(4) is not.
The 12 equations required in each of the
first four phases to set up the system of
differential-algebraic equations are
summarized in Table (1).

Table 1. Summary of equations used in each
phase.

Phase
Motion
Equations
Constraints
1 (1) (3), (4)-(7)
2 (1) (3), (5)-(8)
3 (1) (3), (4), (7)-(9)
4 (1) (3), (6)-(9)
5 (1) -

Fluid Forces Acting on Jacket
Since the jacket is made of tubular
elements, fluid force acting on jacket is
sum of fluid forces acting on individual
tubular elements that are submerged.
Therefore we derive the fluid forces acting
on a typical submerged tubular element.

Buoyancy
Every element is buoyed up by the
surrounding fluid with a force equal to
weight of displaced fluid. This force passes
through center of submerged volume. The
calculation is straight forward, except for
members partially submerged (near SWL).
www.SID.ir
A
r
c
h
i
v
e

o
f

S
I
D

Numerical Modeling .. 813


These members are divided into strips for
calculation of submerged volume.
w w
Z
B
V g F =
) X X ( F M
j m
Z
B
Y
B
= (10)
Where:
m
X : X coordinate of element axis center
w
V : Volume of displaced water by element
w
: Density of sea water
g : Gravitational acceleration

Added Mass
Acceleration of body disturbs the fluid
and causes a pressure field. Resultant of
these pressures on body is a force
proportional to acceleration of body,
known as added mass. Normal acceleration
of a point p (
n p
a
,
r
), on axis of a submerged
tubular element, is derived as:
k Z i X a
j j j CG
r
& &
r
& &
r
+ =
,

) (
, pj j j pj j j CG p
r j j r j a a
r
r
&
r
&
r
r
& &
r r
+ + =
) (
,
u a u a
p n p
r r r r
= (11)
u
r
: Unit vector along element axis
pj
r
r
: Position vector from jacket CG to
point p
Added mass force and moment of a
submerged element due to this acceleration
are as follows:
dl a k F
X
n p
L
X
j A ,
0
1 ,

=
dl a k F
Z
n p
L
Z
j A ,
0
1 ,

=
dl a X X k
dl a Z Z k M
Z
n p
L
j p
X
n p
L
j p
Y
j A
,
0
1
,
0
1 ,
) (
) (

=
(12)
L : Length of submerged element
OD: Outside diameter of element
a
C : Added mass coefficient of elements
section which theoretically equals 1
L
OD
C k
a w
4
2
1
=
Calculation of added mass forces and
moments can be simplified by use of added
mass matrix [8].



Drag
There are two sources for drag force:
friction of fluid on body surface, known as
friction or viscous drag and a force due to
the unbalanced pressures on body, known
as form drag. Drag is proportional to
square of velocity. Drag forces
experienced by tubular members of jackets
during launching is mainly form drag, due
to the high Reynolds number of the
member. Therefore velocity of element
perpendicular to its axis is considered.
Because the motion of jacket during
launching is not harmonic, drag force
linearization schemes cannot be used.
Velocity of point p perpendicular to
element axis (
n p
V
,
) is derived as:
k Z i X V
j j j , CG
r
&
r
&
r
+ =
jp j j , CG p
r j V V
r
r
&
r r
+ =
) u V ( u V
p n , p
r
r
r
r
= (13)
The three components of drag force
are calculated as:
dl V V k F
X
n , p
L
0
n , p 2
X
j , D

=
r

dl V V k F
Z
n , p
L
0
n , p 2
Z
j , D

=
r

dl V V X X k
dl V V Z Z k M
Z
n p
L
n p j p
X
n p
L
n p j p
Y
j D
,
0
, 2
,
0
, 2 ,
) (
) (

=
r
r
(14)
D
C : Drag Coefficient
) OD ( C
2
1
k
D w 2
=
The integrals are evaluated
numerically using three point gauss
quadrature.

Water Entry/Exit
When jacket members enter water,
they experience forces similar to wave
slam. These forces have very large
magnitudes but affect the member for a
very short interval. It is observed that
considering them, modifies the jacket
trajectory to a little extent. On the other
hand they are important in local member
design-checks. Mathematically this force is
equal to rate of change of added mass
www.SID.ir
A
r
c
h
i
v
e

o
f

S
I
D

814 Journal of Faculty of Engineering, Vol. 42, No. 6. December 2008


momentum, i.e., during water entry added
mass of member is zero in air, an instant
later when it is submerged, it has a large
added mass. As shown in Ref. [9],
excluding buoyancy and drag, the total
force in Z direction acting on a tubular
section during water entry is:
( )
dt
dV
m V OD C
dt
dV
m V
dh
m d
dt
dV
m V
dt
dh
dh
m d
V m
dt
d
F
a s
w
a
a
a
a
a E
+ = +
= + = =
2 2
) (
2
) (
) (

(15)
a
m : Added mass of section
h : Distance between bottom of tubular and
SWL
s
C : Water entry (slam) coefficient
Theoretical value of C
s
is . But
experimental results, [4], show that C
s
is
5.15 at the instant of entry and after that it
decays (Eq. 16.). Also, m
a
is the time
varying added mass during water entry,
which reaches its asymptotic value m
w

(mass of fluid displaced by tubular
element), after complete submergence. It is
assumed that water entry/exit forces act on
any section while 0< h/OD < 1.
OD
h
OD
h
C
s
55 . 0
17 1
15 . 5
+
+
= (16)
The velocity considered in Eq. (15) is
that component of V
p,n
lying in the u-Z
plane. Three components of water
entry/exit forces are:
dl V V k F
X
L
X
j E
r r

=
0
3 ,

dl V V k F
Z
L
Z
j E
r r

=
0
3 ,

dl V V X X k
dl V V Z Z k M
Z
L
j p
X
L
j p
Y
j E

=
0
3
0
3 ,
) (
) (
r
r
(17)
) (OD C
2
1
k
s w 3
=
s
C is a function of h , therefore
considering water entry/exit force as
mentioned, needs high computation times
and has prevented commercial softwares
from using it.
Finally, the three components of fluid
force acting on jacket are calculated by
summing up the mentioned forces on all
elements of jacket:

) (
, , , ,
X
j E
X
j A
elements
X
j D
X
j F
F F F F + + =


) (
, , , , ,
Z
j E
Z
j A
elements
Z
j D
Z
j B
Z
j F
F F F F F + + + =


) (
, , , , ,
Y
j E
Y
j A
elements
Y
j D
Y
j B
Y
j F
M M M M M + + + =

(18)
Fluid Forces Acting On Barge
Fluid forces acting on barge are the
same as jacket, with the exemption of
water entry/exit. Similar to jacket, barge is
made up of a number of plates. The drag
force formulation for each plate is identical
to that of jacket element; considering
velocities normal to plate and drag
coefficient of plate. In lieu of a rigorous
evaluation of barge added mass (for
example numerical methods), one can use
approximate values. Values given in CEM
[10] are used: added mass for vertical
motion is approximately equal to mass of
displaced fluid; added mass for horizontal
motion is approximately fifteen percent of
mass of displaced fluid. Therefore barge
added mass force is:
)
`

=
)
`

b
b
w w
w w
Z
b A
X
b A
Z
X
V
V
F
F
& &
& &

0
0 15 . 0
,
,
(19)
Accordingly, the total fluid force on
barge is sum of buoyancy, drag and added
mass forces.

Assembling System of Equations,
Partitioning and Solving
In phases 1-4, motions of the two
bodies are modeled by 12 equations and 12
unknowns. Due to the fact that some of
these unknowns are accelerations and
some reactions, the assembled system of
equations is a differential-algebraic one.
Therefore the system is partitioned with
accelerations as unknowns. The result is a
system 7 second order nonlinear
differential equations, which can be solved
using standard time integration techniques.
www.SID.ir
A
r
c
h
i
v
e

o
f

S
I
D

Numerical Modeling .. 815


In phase 5, there are no additional
constraints/relations and the equations of
motion alone, describe the motions
completely. Assembling Eq. (1) and the
relevant constraints/relations (Table. 1) we
get:
)
`

=
)
`

f
a
ff fa
af aa
F
F
R
q
M M
M M & &
(20)
Where:
| |
T
t j j j b b b
V Z X Z X q
& & & & & & & & & & & & &
& & =
| |
T
Z
w
X
w
Y
c
Z
c
X
c
P P M P P R=
By partitioning Eq. (20) in terms of
accelerations and reaction forces we
obtain:
a af aa
F R M q M = + & &
| |{ } { } f q m = & & (21)
Where:
aa
M m= ; R M F f
af a
=
Note that Eq. (20) is non-singular,
which physically means that the system is
not over-constrained [11]. We solve Eq.
(21) by the Newmark time integration
technique. By assuming linear acceleration
variation during a time step dt,
accelerations and velocities at time i+1 are
expressed as [8], [12]:
) (
3
2
2
1 1 i i i i i
q q
dt
q q
dt
q + =
+ +
& & & &
) (
6 6
2
1
2
1 i i i i i
q q
dt
q
dt
q q + =
+ +
& & & & & (22)
Upon writing Eq. (21) for two
successive time instances i, i+1 and
substituting the accelerations from Eq. (24)
in them we get:
| |{ } { } { } | |{ }
| |{ } | |{ } | |{ }
i i i i i i
i i i i i i
q m q m q m
dt
q m
dt
f f q m
dt
& & & & & + + +
+ =
+ +
+ + + +
1 1
1 2 1 1 1 2
2
6
6 6
(23)
Equation (23) is solved as a system of
algebraic equations for q
i+1
. m
i+1
is a
function of state (positions, velocities,
accelerations) of system at time i+1, and
therefore unknown. f
i+1
is calculated by
solving Eq. (20) as a system of algebraic
equations. Therefore, in order to calculate
m
i+1
and f
i+1
, a state must be assumed and
iteration is required. Using m
i+1
and f
i+1
,
state at time i+1 is obtained. If this new
state is reasonably close to the state
assumed for calculating m
i+1
and f
i+1
,
iteration stops, otherwise the new state is
used for calculating m
i+1
and f
i+1
in the
next iteration. This procedure converges
quite fast (less than 5 iterations).
A computer code has been developed
which determines the motion phase,
calculates fluid forces acting on barge and
jacket and time integrates the relevant
equations.

Verification and Discussion
A launch time history analysis has
been carried out using the developed code,
and the results are compared with that of a
commercial well known software, namely
SACS, and seen to be consistent. Barge
and jacket properties are described in
Table (2) Duration of analysis is 120
seconds with 0.02 s time steps.

Table 2. Summary of barge and jacket
properties used in example.
Barge
Mass 5561 ton
Height 6 m
Width 20 m
Bottom Length 60 m
Initial Trim Angle 2.39 Deg
Drag Coefficient for Tubulars 1
Jacket
Mass 537 ton
Total Buoyancy/Weight 120.2%
Height 50 m
Drag Coef. for Tubulars 0.65
Added Mass Coef. for Tubulars 1

d
0.05

Figure (2) shows the position, velocity
and acceleration time history of jacket CG.
In addition, results from the same launch
analysis carried out by SACS are presented
with the displacement plots. Results agree
quite well, which showcases the present
models reliability. It is seen that jackets
peak velocity and acceleration occur at
rotation phase.
www.SID.ir
A
r
c
h
i
v
e

o
f

S
I
D

816 Journal of Faculty of Engineering, Vol. 42, No. 6. December 2008



Figure 2. Position, velocity and acceleration time histories of jacket CG (Jacket CG positions are
compared with SACS). (a) Position X (b) Position Z (c) Velocity X (d) Velocity Z (e) Acceleration X (f)
Acceleration Z (g) Pitch Angle (h) Pitch angular velocity (i) Pitch angular acceleration (j) Trajectory of
jacket CG .
www.SID.ir
A
r
c
h
i
v
e

o
f

S
I
D

Numerical Modeling .. 817




Figure 3. Time histories of fluid forces/moments acting on jacket (Sum of forces acting on all jacket
members). (a) Buoyancy force (b) Buoyancy moment (c) Drag Force X (d) Drag force Z (e) Added mass
force X (f) Added mass force Z (g) Drag moment Y (h) Added mass moment Y (i) Water entry/exit force Z
(j) Water entry/exit moment Y.
www.SID.ir
A
r
c
h
i
v
e

o
f

S
I
D

818 Journal of Faculty of Engineering, Vol. 42, No. 6. December 2008


As can be seen in the third plot of Fig. 2,
the maximum jacket rotation obtained
from the two analyses differs by 1.75:
o presented
j
75 . 31
max ,
= ,
o SACS
j
30
max ,
=
This difference is due to the water
entry moments on jacket which we have
considered. It should be noted that SACS
does not account for water entry forces.
Also because of this additional rotation,
the jacket plunges more deeply in our
model:
m Z
presented
j
41 . 10
min ,
= , m Z
SACS
j
6 . 9
min ,
=
In addition because of the differences
mentioned and some other minor
differences in fluid force formulation
(namely that our model considers
buoyancy of members as they enter
waterline), the jackets maximum
horizontal displacement differ by 0.86 m in
the two models:
m X
presented
j
56 . 58
max ,
= , m X
SACS
j
7 . 57
max ,
=
Figure 3 shows the time history of
fluid forces (buoyancy, drag, added mass,
water entry/exit) acting on jacket. In
general, buoyancy has the biggest
magnitude among fluid forces. Drag force
has a more pronounced effect than added
mass force. Drag force stops horizontal
motion of the jacket and its vertical
oscillations. By comparison of Fig. (2) and
Fig. (3) it is observed that drag and added
mass force have nearly the same trend as
velocity and acceleration respectively (plot
(c) has the same trend in both figures i.e.
drag force in X direction has the same
trend as velocity in X direction. The same
is true for plots (d), (e) and (f) from
Figures 2 and 3 respectively) although
each force component is a function of all
other motion components.
The time history of water entry/exit
has several peaks. These peaks correspond
to instances were a horizontal chord of the
jacket is entering water. For instance, state
of launch is shown graphically at time
92.52 s (when the largest peak occurs) in
Fig. 4, and the chord member entering
water is highlighted. An impact force of
161 kN is affecting the water entering
member.
The reaction forces between jacket and
barge are resisted by launch skids (
sk
N )
and rocker arm (
ra
N ), which can be
calculated from static equilibrium of
reaction forces. Figure (5) shows the
normal forces acting on rocker arm and
launch skids. It is seen that rocker arm
normal force reaches its maximum when
jacket starts rotating, at which instant the
launch skid force drops to zero. Lesser
normal reaction of rocker arm needs less

Figure 4. Graphical representation of jacket and barge at time 92.52 s and water entry of a jacket bottom
chord. The water entering jacket chord is highlighted.
www.SID.ir
A
r
c
h
i
v
e

o
f

S
I
D

Numerical Modeling .. 819


strengthening of launch truss. In this
regards it is economical to use deeper
drafts for barge so that buoyancy lessens
the rocker arm normal force.

0 40 80 120
0
2500
5000
Time (s)
N
R
A
,

N
S
K

(
k
N
)


N
RA
N
SK

Figure 5. Time history of rocker arm and launch
skid normal reaction force.

We note that if the acceleration
constraint equations Eq. (3) are not
satisfied accurately, displacements of
jacket and barge are erroneous, that is, the
jacket will either move away or interfere
with launch skids, which is impossible. A
descriptive output that indicates accuracy
of solution is the perpendicular distance of
jacket CG to rocker arm pin. This distance
should remain constant throughout phases
1~4. Figure (6) shows this distance and it
is seen that it is nearly constant until the
end of phase 4.
0 40 80 120
8.254
8.255
Time (s)
h

(
m
)

Figure 6. Variation of perpendicular
distance between jacket CG and rocker arm pin
(h) during launch.

Conclusion
Equations governing the operation of
launching a jacket from barge, including
motion and constraints, are derived from
principles of dynamics. Non-linear fluid
forces acting on both bodies are also
formulated, and the resulting system of
equations is solved. Time histories of
motions and forces are calculated. The
following conclusions are drawn:

- Water entry forces are important
regarding local member design,
especially slender horizontal jacket
chords. Jacket chords near mud-line
(Figure 4) are examples of slender
members which large water entry
forces act upon them.
- Considering water entry forces and
moments via Eq. 15 modifies the jacket
trajectory to a little extent.
- Dominance of drag force among the
hydrodynamic fluid forces, and its
effect on limiting and damping jackets
motion.
- Very good accuracy in satisfying the
acceleration constraint equations (Eq.
3) by using iteration.

www.SID.ir
A
r
c
h
i
v
e

o
f

S
I
D

820 Journal of Faculty of Engineering, Vol. 42, No. 6. December 2008


References
1 - Vasicek, Daniel and Lu, Cheng-Heng (1979). Launch and floatation analysis of offshore structures part 2
barge and jacket interaction on launch analysis. Petroleum Engineer International, Vol. 51, No. 6, PP. 10-16.
2 - Sphaier, S. H., Vasconcellos, J. M., Esperanca, P. T. T. and Ferreira, M. D. A. (1985). The study of jacket
installation using INPLA system. Proc. Of the 5
th
International Symposium on Offshore Engineering, Federal
University of Rio de Janeiro, Brazil, PP. 541-573.
3 - Nelson, J. K., Fallon, D. J. and Hirsch, T. J. (1991). Mathematical modeling of free-fall lifeboat launch
behavior. Proc. of International Conference of Offshore Mechanics and Arctic Engineering, PP. 695-702.
4 - Campbell, I. M. C. and Weynberg, P. A. (1980), Measurement of parameters affecting slamming final
report. Technology Reports Centre No. OT-R-8042, Rep. No. 440, Southampton University: Wolfson Unit for
Marine Technology.
5 - Gerwick, B. C., Construction of Offshore Structures (1986), John Wiley & Sons.
6 - Jo, C. H., Kim, K. S. and Lee, S. H. (2002). Parametric study on offshore jacket launching. Ocean
Engineering, No. 29, PP. 1959-1979.
7 - SACS 5.1, Users Manual, Release 5, Revision 4.
8 - Patel, Minoo H. and Witz, Joel A. (1991). Compliant Offshore Structures. Butterworth-Heinemann Ltd, 1
st

edition.
9 - Faltinsen, O. M. (1990). Sea Loads on Ships and Offshore Structures. Cambridge University press, 1
st

edition, 1990.
10 - US Army Corps of Engineers (2002). Coastal Engineering Manual. US Army Corps of Engineers.
11 - Shabana, A. A. (1989). Dynamics of Multibody Systems. John Wiley & Sons.
12 - Nachlinger, R. Ray, How MOSES Deals with Technical Issues: http://www.ultramarine.com.
13 - Nourpanah, N. (2006). Numerical Modeling of Launching Offshore Jackets. MSc Thesis, Faculty of Civil
Engineering, University of Tehran.














www.SID.ir
A
r
c
h
i
v
e

o
f

S
I
D

Numerical Modeling .. 821






1

2

1
- -
2
- -

.
-
.
.
.
SACS .
.
.

.

: - - - -
SACS


www.SID.ir