The International Conference on Marine Safety and Environment, 12-13 November 2013, Johor Bahru, Malaysia

23
Semi-Submersibles Response Prediction by Diffraction
Potential Method
C. L. Siow
1
, Abby Hassan
1,
and Jaswar
1,*

1. Department of Aeronautic, Automotive and Ocean Engineering, UniversitiTeknologi Malaysia, Skudai, Johor Baharu, Malaysia

Abstract:This paper propose of three dimensional diffraction potential methods to predict motion response of multiple hulls
semi-submersible structure. Besides, three dimensional green functions also applied here to estimate wave velocity potential at each
panel on semi-submersible surface. In previous stage, this method was showed it capability to apply on single hull structure.
However, some of modification is required to make into the original method to able it applied for multiple hull structure. Upon this
stage, the modification only made into the meshing system in purposes to test the tendency of this numerical result. This paper also
presented the comparison between the numerical result and experiment result for the selected semi-submersible model. In the
comparison, it is obtained that the tendency between the numerical result and experimental result is agreed between each other.
However, the numerical result is slightly under predicted compared to experimental result in most cases.

Keywords:Multiple Hulls,Semi-submersible, Motion Response,Diffraction Potential, Green Function.

1. Introduction
In recent development, multiple hull structures are
popular structure often used in deep water oil and gas
exploration activity.

As an example, semi-submersible
structure and tensional leg platform are frequently used
for deep water oil and gas exploration. This is because
both the structures have good response characteristic to
incident wave.
However, the numerical analysis method for
multiple hull structures is slightly different compared
to single hull structure. Existing of hydrodynamic
interaction effect in multiple hull structure can causes
the response amplitude for the structures increased or
decreased depended on the structures characteristic and
environment condition such as incident wave
frequency and incident wave angle.In previous study,
hydrodynamic behavior of multiple hull structures can
be studied by using experimental method and
numerical method. There arealso much available

* Correspondingauthor:J aswar,Dr,research fields: Marine
Hydrodynamic. E-mail:jaswar@fkm.utm.my.
commercial software able to analyses the behavior of
this type of structure.
On the other hand, available discussions on the
numerical method use for analyze multiple hull
structures still not sufficient. Hence, this study is
proposed to expend the numerical method which
previously use for single hull structures to able it
apply for multiple hull structures especially for
offshore structure such as semi-submersible. However,
in this paper, the discussion is only focused on the
meshing system which required modifying for
multiple hull structures. After that, this paper also will
discuss the comparison between then tendencies of the
numerical result compared to the experimental result.
The comparison obtained that few improvement is
needed for the suggested method in order to improve
the accuracy of the executed result.
2. Literature Reviews
Oscillating of floating structure caused by wind,
wave and current affects loading and offloading
operation systems. Hess and Smith, Van Oortmerssen
andLoken studied on non-lifting potential flow
The International Conference on Marine Safety and Environment, 12-13 November 2013, Johor Bahru, Malaysia
Pusat Teknologi Marin, UTM

24

calculation about arbitrary 3D objects [1, 2, 3]. They
utilized a source density distribution on the surface of
the structure and solved for distribution necessary to
lake the normal component of fluid velocity zero on the
boundary. Plane quadrilateral source elements were
used to approximate the structure surface, and the
integral equation for the source density is replaced by a
set of linear algebraic equations for the values of the
source density on the quadrilateral elements. By
solving this set of equations, the flow velocity both on
and off the surface was calculated. Besides, Wu et al.
also studied on the motion of a moored
semi-submersible in regular waves and wave induced
internal forces numerically and experimentally [4]. In
their mathematical formulation, the moored
semi-submersible was modeled as an externally
constrained floating body in waves, and derived the
linearized equation of motion.
Yilmaz and Incecik analyzed the excessive motion
of moored semi-submersible [5]. They developed and
employed two different time domain techniques due to
mooring stiffness, viscous drag forces and damping. In
the first technique, first-order wave forces acting on
structure which considered as a solitary excitation
forces and evaluated according Morison equation. In
second technique, they used mean drift forces to
calculate slowly varying wave forces and simulate for
slow varying and steady motions. Sylemez developed
a technique to predict damaged semi-submersible
motion under wind, current and wave [6]. He used
Newtons second law for approaching equation of
motion and developed numerical technique of
nonlinear equations for intact and damaged condition
in time domain.
Clauss et al. analyzed the sea-keeping behavior of a
semi-submersible in rough waves in the North
Seanumerically and experimentally[7]. They used
panel method TiMIT (Time-domain investigations,
developed at the Massachusetts Institute of
Technology) for wave/structure interactions in time
domain. The theory behind TiMIT is strictly linear and
thus applicable for moderate sea condition only.
An important requirement for a unit with drilling
capabilities is the low level of motions in the vertical
plane motions induced by heave, roll and pitch. Matos
et al.were investigated second-order resonant of a
deep-draft semi-submersible heave, roll and pitch
motionsnumerically and experimentally [8]. One of the
manners to improve the hydrodynamic behavior of a
semi-submersible is to increase the draft. The low
frequency forces computation has been performed in
the frequency domain by WAMIT a commercial
Boundary Element Method (BEM) code. They
generated different number of mesh on the structure
and calculated pitch forces.
Wackers et al. was reviewed the surface
descretisation methods for CFD application with
different code [9]. Besides, simulation of fluid flow
Characteristic Around Rounded-Shape FPSO was also
conducted by A. Efi et al. using RANs Method [10].
J aswar et al. were also developed an integrated CFD
simulation software to analyze hull performance of
VLCC tanker. The integrated CFD simulation tool was
developed based on potential theory and able to
simulate wave profile, wave resistance and pressure
distribution around ship hull [11].
In addition, few experiment tests were carried out to
obtain the motion response of structure. A model test
related to interaction between semi-submersible and
TLP was carried out by Hassan Abyn et al. [12]. In
continue Hassan Abyn et al. also tried to simulate the
motion of semi-submersible by using HydroSTAR and
then analyze the effect of meshing number to the
accuracy of execution result and execution time [13].
Besides, C. L. Siow et al. also make a comparison on
the motion of semi-submersible when it alone to
interaction condition by using previous experimental
result [14]. Besides that, K.U.Tiau (2012) was
simulating the motion of mobile floating harbor which
The International Conference on Marine Safety and Environment, 12-13 November 2013, Johor Bahru, Malaysia
Pusat Teknologi Marin, UTM

25

have similar hull form as semi-submersible by using
Morison Equation [15]
3. Basic Mathematic Model
3.1 Diffraction Potential
In this study, diffraction potential method was used
to obtain the wave force act on the semi-submersible
structure. The regular wave acting on floating bodies
can be described by velocity potential. The velocity
potential normally written in respective to the flow
direction and time as below:
(x, y, z) = Rc|(x, y, z)c
wt
] (1)
(x, y, z) =
g
c
w
{
0
(x, y, z) +
7
(x, y, z)]

+ iwX
]

]
(x, y, z)
6
]=1
(2)

Where,
g : Gravity acceleration

u
: Incident wave amplitude
X
]
: Motions amplitude

0
: Incident wave potential

7
: Scattering wave potential

]
: Radiation wave potential due to motions
] : Direction of motion
From the above equation, it is shown that total
wave potential in the system is contributed by
potential of the incident wave, scattering wave and
radiation wave. In addition, the phase and amplitude
for both the incident wave and scattering wave is
assumed to be the same. However, radiation wave
potentials are affected by each type of motion of each
single floating body inside system, where the total
potential for radiation wave for the single body is the
summation of the radiation wave generates by each
type of body motion such as roll, pitch, yaw, surge,
sway and heave.
Also, the diffraction wave potential
(d)
must be
satisfied with boundary conditions as below:
v
2

(d)
= u or u z b (S)
o
(d)
oz
+k
(d)
ot z = u (k =
w
2
g
) (4)
o
(d)
oz
= u ot z = b (S)

(d)
~
1
r
c
-k
0

sboulJ bc u i r (6)
o
(d)
on
= -
o
0
on
on tbc boJy bounJory (7)
3.2Green Function and Wave Potential
By considering the wave potential only affected by
structure surface, S
H
, the wave potential at any point
can be presented by following equation:
(P) = _
o()
on

0(P; )
S
H
- ()
o0(P; )
on

_ JS() (8)
Where P =(x, y, z) represents fluid flow pointed at
any coordinate and = (, p, ) represent any
coordinate, (x, y, z) on structure surface, S
H
. The green
function can be applied here to estimate the strength of
the wave flow potential. The green function in eq. (8)
can be summarized as follow:
0(P; ) = -
1
4n(x - )
2
+ (y - p)
2
+(z -)
2
+ E(x - , y - p, z + ) (9)
Where E(x - , y -p, z +) in eq. (9) represent
the effect of free surface and can be solved by second
kind of Bessel function.
3.3Wave force, added mass and damping
The wave force or moment act on the structure to
cause the motions of structure can be obtained by
integral the diffraction wave potential along the
structure surface.
E

= -

(x, y, z)n

JS
S
H
(11)
Where,

is diffraction potential,

=
o
+
7

Potential for diffraction wave
Potential for Radiation wave
The International Conference on Marine Safety and Environment, 12-13 November 2013, Johor Bahru, Malaysia
Pusat Teknologi Marin, UTM

26

Also, the added mass, A
ij
and damping, B
ij
for each
motion can be obtained by integral the radiation wave
due to each motion along the structure surface.
A
]
= -p Rc|
]
(x, y, z)]n

JS (12)
S
H

B
]
= -pw Im|
]
(x, y, z)]n

JS (1S)
S
H

n

in eq. (11) to eq. (13) is the normal vector for each

direction of motion, i =1~6 represent the direction of
motion and j =1~6 represent the six type of motions.
4. Multiple hull numerical method
4.1Rule for meshing data and offset data
In this study, the expansion of mono hull numerical
method to multiple hull numerical method is limited to
meshing system only. Interaction between hulls by
radiation wave was ignored in this study. The purpose
of this study is only limited to discuss the possibility to
produce the multiple hull numerical method bases on
the mono hull method.
In this numerical method, the right hand rule is
applied. The panel coordinate must arranged follow
this rule to ensure the execution of normal vector and it
direction is in correct manner. The panel for this
method is shown as fig. 1 below:

Fig. 1 Right hand rule for panel use in the numerical method
In addition, offset data covered the half breadth of
the one side of hull is also requiredto execute the wave
force on the structure. The offset data should be made
separately for both the column and pontoon to avoid
the programming code wrong reading the data and then
generate wrong meshing on the structure.
The selected semi-submersible model in this study is
constructed based on GVA4000 type. Total panels used
in the execution are 272 where 25 panels on each
column and 222 panels on pontoon surface. The
example meshing constructed by this numerical
method for the semi-submersible model is shown in
fig.2.

Fig.2meshing for semi-submersible model
4.2Programming Flow chart
In the general, this numerical method for multiple
hull structure is still largely similar with the mono hull
method. The numerical method will only execute the
wave force acting on one side of hull and then multiply
the magnitude of force according to the number of hull
for the selected structure.
As similar with other CFD method, this numerical
method start with meshing generation and then execute
the normal vector, center point of each panel and area
for each panel. After that, the program will
constructmatrix element for distribution of sources and
normal dipoles over the panel.
Next, wave force on each panel will executed by
using green function and Bessel function. At this
moment, radiation force and diffraction will be
considered. After that, the total wave force acting on
the structure to cause the motion can be obtained by
sum up the total diffraction force on each panel. At the
The International Conference on Marine Safety and Environment, 12-13 November 2013, Johor Bahru, Malaysia
Pusat Teknologi Marin, UTM

27

same time, added mass and damping for the structure at
same wave condition can be obtained by summing up
the real part of radiation potential and imaginary part of
radiation potential. Lastly, the structure motion and it
response to wave can be obtained by coupled motion
equation. The flow chart for this numerical method is
shown in fig. 3.

Fig.3Flow Chart for Diffraction Potential Numerical
Method
By the explained process, wave force on the body
and the motion response of the structure can be
executed. However, until this step, the interaction
effect between the hulls for the same structures is not
executed. Hence, it is predicted that the motion
response obtained from this numerical method will be
smaller compared to the exact one.
5. Model Parameter and Test Data
The Semi-submersible model was constructed based
on GVA 4000. The model has four circular columns
connected to two pontoons and two braces. Two pieces
of plywood are fastened to the top of the
Semi-submersible to act as two decks to mount the test
instruments. The model was constructed from wood
following the scale of 1:70 (Table 1).
Upon the model complete constructed, few tests
were carried out to obtain the model particulars.
Inclining test, swing frame test, oscillating test, decay
test and bifilar test were carried out to identify the
hydrostatic particular for the semi-submersible. The
dimension and measured data for the model was
summarized as in table 1.
Table 1 Principal particular of the Structures
Length
0.954 m
Width
0.835 m
Draft
0.239 m
Displacement
0.043501 m
3

Water Plan Area
0.108082 m
2

Number of Columns
4
Pontoon length
0.954 m
Pontoon depth
0.09 m
Pontoon width
0.19 m
Pontoons centerline separation
0.645 m
Columns longitudinal spacing
(centre)
0.651143 m
Column diameter
0.151286 m
GM
T

0.041 m
GM
L

0.058 m
K
XX

0.452 m
K
YY

0.385 m
K
ZZ

0.5 m
6. Result and Discussion
In this part, the response amplitude for GVA 4000
semi-submersible structure in head sea condition was
discussed. The result from the proposal numerical
result was also compared to the experiment result. The
input for the numerical program was also adjusted to
The International Conference on Marine Safety and Environment, 12-13 November 2013, Johor Bahru, Malaysia
Pusat Teknologi Marin, UTM

28

make the condition as close as the experiment
condition. However, the effect of soft mooring use in
the experiment test is not simulated in the numerical
method since the proposal method is unable to simulate
the mooring effect. Hence, from the logical of view, the
executed numerical result should be slightly higher
than the experiment result due to the missing of
mooring effect in numerical simulation.
As shown in fig. 4 to fig. 6, the tendency of surge,
pitch and heave motion response obtained by both the
numerical and experimental method is agreed between
each other. However, response amplitude for surge
motion and pitch motion calculated by numerical
method is slightly under predicted compared to the
response amplitude obtained from experimental
method. This observation was not agreed with the
previous predicted outcome because the absence of
mooring system in the numerical method should causes
the motion of the semi-submersible become higher
compared to the experimental result which completed
with soft spring to simulate the mooring force.

Fig.4Surge motion response for semi-submersible model

Fig.5Pitch motion response for semi-submersible model
From the theoretical point of view, one of the
reasons can be explained the under predicting
phenomena in numerical method is absence of
hydrodynamic interaction force in numerical method.
According to the previous study done by C. L. Siow et
al. (2013) to evaluate the hydrodynamic interaction
between two floating structure in tank, the interaction
force between two floating structures is around 39% for
surge motion and 22% for pitch motion at the selected
comparison case. Hence, it can be predicted that the
hydrodynamic interaction between the hulls in the
single semi-submersible structure also will give an
extra wave force to the structure and cause larger
motion to the structure.

Fig.6Heave motion response for semi-submersible model
In addition, the numerical execution result also
success to detect the resonant condition for the heave
0
0.2
0.4
0.6
0.8
1
1.2
0 2 4 6 8 10
R
e
s
p
o
n
s
e

A
m
p
l
i
t
u
d
e
,

m
m
/
m
m
WaveLength,m
ExperimentResult NumericalResult
0
0.01
0.02
0.03
0.04
0.05
0.06
0 2 4 6 8 10
R
e
s
p
o
n
s
e

A
m
p
l
i
t
u
d
e
,

m
m
/
R
a
d
WaveLength,m
ExperimentResult NumericalResult
0
0.5
1
1.5
2
2.5
0 2 4 6 8 10
R
e
s
p
o
n
s
e

A
m
p
l
i
t
u
d
e
,

m
m
/
m
m
WaveLength,m
ExperimentResult NumericalResult
The International Conference on Marine Safety and Environment, 12-13 November 2013, Johor Bahru, Malaysia
Pusat Teknologi Marin, UTM

29

motion as shown in figure 6. From the same figure, it is
observed that both the numerical and experiment result
is agreed between each other and give an almost similar
result before wave length equal to six meters in model
scale. However, at the wave length approximate to nine
meters, the numerical result detected the resonant
condition for this structure and causes the executed
result overshooting compared to experiment result.
Unfortunately, this observation cannot be validated by
experiment result due to insufficient data at the
resonant condition. Besides, the reason for the
overshooting in numerical result is because the
damping force executed by three dimensional potential
methods is low at the resonant condition. Hence, to
improve the execution result, viscous damping may
take into consideration at resonant condition to reduce
overshooting of the response amplitude returned by the
execution.
7. Conclusion
In the conclusion, the three dimensional diffraction
potential methods for the single hull method is possible
to modify for analyze the wave force and motion
response for multiple hull structure such as
semi-submersible. From the comparison, it is observed
that the tendency executed result by numerical method
is agreed with experimental result with carried out at
head sea condition. However, some modification is
required to carry out to improve the execution result. In
future, hydrodynamic interaction between hull form in
the same structure will be completed with this proposed
method to able it estimate the interaction forec and then
improve the motion prediction.
Acknowledgments
The authors would like to gratefully acknowledge to
Marine Technology Center,
UniversitiTeknologiMaalysia with grant researchers
J 130000.2424.00G47.
References
[1] J . L.Hess, A. M. O. Smith, Calculation of Nonlifting
Potential Flow About Arbitrary 3D Bodies, J ournal of
Ship Research (1964).
[2] G. Van Oortmerssen, Hydrodynamic interaction between
two structures of floating in waves, Proc. of BOSS 79.
Second International Conference on Behavior of Offshore
Structures, London (1979).
[3] A. E. Loken, Hydrodynamic interaction between several
floating bodies of arbitrary form in Waves, Proc. of
International Symposium on Hydrodynamics in Ocean
Engineering, NIT, Trondheim (1981).
[4] S. Wu, J . J . Murray, G. S. Virk, 1997. The motions and
internal forces of a moored semi-submersible in regular
waves, Ocean Engineering, 24(7)(1997), 593-603.
[5] O. Yilmaz, A. Incecik, Extreme motion response analysis
of moored semi-submersibles, Ocean Engineering, 23(6)
(1996) 497-517.
[6] M. Sylemez, Motion tests of a twin-hulled
semi-submersible, Ocean Engineering, 22(6) (1995)
643-660.
[7] G. F. Clauss, C. Schmittner, K. Stutz, Time-Domain
Investigation of a Semi Submersible in Rogue Waves,
Proc. of the 21st International Conference on Offshore
Mechanics and Arctic Engineering (OMAE2002), Oslo,
Norway (2002).
[8] V. L. F. Matos, A. N. Simos, S. H. Sphaier, Second-order
resonant heave, roll and pitch motions of a deep-draft
semi-submersible: Theoretical and experimental results,
Ocean Engineering, 38(1718) (2011) 2227-2243.
[9] J . Wackers et al., Free-Surface Viscous Flow Solution
Methods for Ship Hydrodynamics, Archive of
Computational Methods in Engineering, Vol. 18 (2011)
141.
[10] E.Afrizal, F.M. Mufti, C.L.Siow, J aswar Study of Fluid
Flow Characteristic Around Rounded-Shape FPSO Using
RANS Method, The 8th Interna-tional Conference On
Numerical Analysis In Engineering, Pekanbaru, Indonesia
(2013).
[11] J aswar et al, An integrated CFD simulation tool in naval
architecture and offshore (NAO) engineering, The 4th
International Meeting of Advances in Thermofluids,
Melaka, Malaysia, AIP Conf. Proc. 1440 (2011)
1175-1181.
[12] H. Abyn, A. Maimun, J aswar, M. Rafiqul Islam, A.
Magee, B. Bodagi, M. Pauzi, Model Test of
Hydrodynamic Interactions of Floating Structures in
Regular Waves, Proc. of the 6th Asia-Pacific Workshop
The International Conference on Marine Safety and Environment, 12-13 November 2013, Johor Bahru, Malaysia
Pusat Teknologi Marin, UTM

30

on Marine Hydrodynamics (APHydro2012), UTM,
Malaysia (2012).
[13] H. Abyn, A. Maimun, J aswar, M. Rafiqul Islam, A.
Magee, B. Bodagi, M. Pauzi, Effect of Mesh Number
on Accuracy of Semi-Submersible Motion Prediction,
Proc. of the 6th Asia-Pacific Workshop on Marine
Hydrodynamics (APHydro2012), UTM, Malaysia (2012).
[14] C.L. Siow, J aswar, E. Afrizal, H. Abyn, A. Maimun, M.
Pauzi, Comparative of Hydrodynamic Effect between
Double Bodies to Single Body in Tank, The 8th
International Conference on Numerical Analysis in
Engineering, Pekanbaru, Indonesia (2013).
[15] K.U. Tiau, J aswar, Hassan Abyn and C.L. Siow, Study On
Mobile Floating Harbor Concept, Proc. of the 6th
Asia-Pacific Workshop on Marine Hydrodynamics
(APHydro2012), UTM, Malaysia (2012).