
based on linear hydrodynamic theory
and have proved their worth on many
occasions.
Nonlinear, mean wave drift forces on
semisubmersible type structures can
be computed based on the application
of linear, diffraction
theory computational methods combined
with either a farfield method or a
nearfield method for the evaluation
of the second order wave loads on the
structure. In case a farfield method
is applied, generally only the mean
second order horizontal drift forces
can be calculated. See reference [2]
and reference [3]. If a nearfield or
pressure integration method is
applied the mean and lowfrequency
components of the drift forces can be
computed for 6 degrees of freedom.
See reference [4]. This type of
computational method assumes the flow
to be inviscid thus excluding any
effects which might arise from
separated flow around the structure.
In the past efforts have been made to
verify the computational methods for
the mean and lowfrequency or slowly
varying wave drift forces on semi
submersible type structures. See
reference [5]. It has been surmised
that the drift forces on semi
submersible type structures, which
consist of relatively slender
surfacepiercing columns and sub
merged floaters, are in some cases
significantly affected by viscous
effects in the flow around the
structural elements. Very little ex
perimental data is available which
can give insight in such effects. See
reference [10].
In order to increase insight in these
effects MARIN, in cooperation with a
number of offshore operators, design
ers and manufactures, carried out
extensive model test programs and
computations among others of the mean
and slowly varying wave forces on
slender and full semisubmersibles.
In this paper a number of aspects of
this research including the model
test programs and the correlation
between model test results and
602
results of computations are
discussed.
The findings of these studies have
confirmed that significant viscous
effects can be present in the low
frequency wave forces on such
structures. As a result, a research
program has been initiated by the
Delft University of Technology into
determining such effects on structu
ral elements of semisubmersibles
such as the columns and the pon
toons. The purpose of this research
is to determine for which of these
elements the viscous effects play an
important role and, if possible, to
develop a rational computational
procedure for taking such effects
into account when determining the
mean and slowly varying drift forces
on the complete structure. In this
paper some results of recent model
tests carried out on a fixed verti
cal cylinder in waves are presented.
SECOND ORDER WAVE DRIFT FORCES
ON A SEMISUBMERIBLE
The subj ects of this investigation
were a slender 8column SemiSubmer
sible I with circular columns and a
displacement of 23,270 tonnes and a
full 6column SemiSubmersible II
with square columns and a displace
ment of 56,300 tonnes. The body plans
of the semisubmersibles are given in
Figure 1 and Figure 2. In the follow
ing all results of measurements and
computations will be given for the
full scale structures.
The aims of the study were as
follows:
To increase insight in the mean and
lowfrequency horizontal wave excit
ing forces and motion responses of
large semisubmersibles
To check the validity of computa
tional methods for the prediction
of first order wave frequency mo
tions and lowfrequency wave drift
forces based on 3dimensional
potential theory.
The total scope of the research does
not allow all aspects to be treated
here. In this paper the results of
the following investigations are
presented:
Results of model tests in regular
waves to determine the mean hori
zontal wave drift force response.
Results of tests in irregular waves
to determine the mean and lowfre
quency wave drift force records.
The model tests were carried out at a
scale of 1:40 in the Seakeeping Basin
of MARIN. This basin measures 100 m x
24 m x 2.5 m.
MODEL TEST SETUP IN THE BASIN
Measurements of the mean horizontal
wave drift forces on a model in re
gular waves can be carried out using
a softspring restraining or mooring
system which consists of horizontal
wires incorporating soft linear
springs which are connected to force
transducers mounted on the model. The
mooring wires are connected at deck
level. The setup for tests in
regular waves is shown in Figure 3.
In order to measure the mean and
slowly varying horizontal wave drift
forces in irregular waves, ideally
the model should be moored in such a
way that all low frequency motion
response is suppressed while leaving
the model completely free to carry
out the motions at wave frequencies.
The first requirement ensures that
the measured force is not affected by
dynamic magnification effects. The
second requirement can be deduced
from theoretical analysis of the
second order wave drift forces which
show that part of the total second
order excitation forces are directly
dependent on the structural motions
at wave frequencies. See reference
[4].
As a consequence, the model re
straining system must possess the
characteristics of an ideal Dynamic
Positioning system. For the model
.....::...=..
= = 
tests a system consisting of hori
zontal restraining wires connected to
controllable tension winches was
selected. See Figure 4. The winches
were operated based on an active
control system with a feedback loop
supplemented by a feedforward
control loop. See Figure 5.
The feedback loop 'acted on the hori
zontal position error and the time
derivative of the error (Proportio
nalDifferential control). The feed
forward control loop was based on the
realtime measurement of the re
lative wave elevation on the upwave
columns of the semisubmersible. It
has been shown that a major part of
the mean and slowly varying second
order wave drift forces as predicted
by potential theory, is due to terms
related to the square of the instan
taneous relative wave elevation
around the waterline of a floating
structure. This has been demon
strated, among others, from model
tests on a tanker. See reference [4].
Application of feedback and feed
forward control still does not result
in full suppression of lowfrequency
motions however. This due to the fact
that the feedforward loop is supply
ing an imperfect estimate for the
instantaneous low frequency horizon
tal force. As a result, the total
restraining force is not equal and
oppos i te to the low frequency wave
exciting force thus resulting in
residual low frequency motions. See
Figure 6. In order to obtain a best
estimate of the total low frequency
wave force on the model, the measured
restraining force is corrected for
the residual horizontal motions of
the vessel. This is carried out off
line after a test has been carried
out. The bas ic assump tion behind this
process is that the instantaneous
discrepancy between the true wave
force and the measured restraining
force results in horizontal motion
accelerations which are described by
the following relationship:
m x(t) = Fd(t)  Fm(t)
603
~ ~ ~ ~  ~ ~ ~ = =
; ; ; ; = ~ ~
in which m represents the virtual
mass of the vessel and x(t) the mo
tion acceleration. Assuming that the
virtual mass is constant, the accele
ration force can be determined in the
time domain by passing a double
differentiating filter over the time
record of the low frequency horizon
tal motions. The best estimate of the
time record of the horizontal drift
force then follows from:
Fd(t)
= Fro(t) + mx(t)
An example of time records of
measured restraining force, residual
surge motion, correction force and
total drift force are shown in Figure
7. The results apply to SemiSubmer
sible II.
In order to verify the accuracy of
this procedure model tests were
repeated using different settings of
the dynamic restraining system.
Anexample of the results found for
the spectral density of the slowly
varying wave drift force on the Semi
Submersible 11 in irregular head seas
is shown in Figure 8.
The results apply to the system of
restraint being adjusted to repre
senting a spring system (Proportio
nal control), a spring and damper
system (ProportionalDifferential
control) and a PD control including
Feedforward based on the relative
wave elevation measurements. The
results shown in the figure indicate
that the spectral density of the
drift force obtained from tests with
significantly different characteris
tics of the restraining system are
reasonably consistent.
TESTS IN REGULAR WAVES
Tests in regular waves were carried
out for both semisubmersibles for a
range of wave frequencies, wave
amplitudes and wave directions. For
the slender SemiSubmersible I tests
in head seas were carried out with
out and with bracings. The results
are given in Figure 9 through Figure
13 for both structures in head waves
and in beam waves as mean drift force
transfer functions,
In the figures the theoretical values
found on the basis of 3dimensional
potential theory computations ex
cluding the contribution from the
bracings are also given.
Comparison between the results of mo
del tests and computations show that
in the lower wave frequency range the
mean drift forces tend to be consi
stently underestimated by the compu
tations. The effect of the bracings
on the mean drift forces on the
slender SemiSubmersible I in head
seas is to increase slightly the mean
drift forces as can be seen from the
comparison between the measured
results shown in Figure 9 and in
Figure 11.
TESTS IN IRREGULAR WAVES
Tests were again carried out for both
semisubmersibles. The slender Semi
Submersible I was tested without
bracings.
Results of tests in irregular waves
are given in the form of time traces
of the measured low frequency drift
forces compared with time traces of
the corresponding predicted low fre
quency force based on 3dimensional
potential theory. In some cases the
spectral density of the computed and
measured forces are compared. For the
time domain predictions, use was made
of second order impuls response
functions combined with the measured
time trace of the undisturbed ir
regular wave record in the basin. See
reference [5]. The time domain second
order impuls response functions for
the drift forces are obtained from
the complete second order quadratic
transfer functions computed in the
frequency domain. The quadratic
transfer functions were computed
based on the pressure integration
method, See reference [4].
604
.=
_ _= _ = _ 
._._
.=. ____ ... ___ ., s. .< .=
.. .=..=, _.s L = _ s =:. _=__=_ . .
.
..
=.
..
__=~ ~._ _=.= ~_.__ .. _
..:
.

_=. :__:=_ .,. ._
=.

. ..
. .
 .=... _+ _
.
. ... .. =s>s===. .  =_.===_.  = : 
.
.

..=. = .==__ _
_ ___~
~ ,__ _=_=_. .._._: _ : .:=~ ~:_=_,._, ._
.
.. ._ =s=z_
s _= .=. . __ ..
=.=+=_+,=::. : ~ __:===.V=: s_ .= ._._ .__.~z ~~_+= . . . _: =
.. 

E!si__ ._=D ==== .O==.. ~_.>=__  ~.:_c= ,.e. =. < ==== ~.. ____ : .
. .
= .. __a _ =..=.< =
.
determined
equation:
Fvd(t) = %/J
in which:
v(t) 
D
C(t) =
from the following
r(t)
cd~ v(L).lv(t)l.D dz
o
relative horizontal
velocity between the
fluid and the column
column diameter
relative wave elevation
Cd = drag coefficient
This contribution to the drift force
could be evaluated for each column in
the timedomain based on the un
disturbed wave elevation record, the
frequency domain motion characteris
tics of the semisubmersible and an
assumption regarding the drag coef
ficient in the above equation. The
summation of the drag force on each
column results in the estimated vis
cous drag force contribution to the
drift force. The total drift force is
found by adding the viscous and
potential contributions.
The results of thes computations are
shown in Figure 22 for the slender
SemiSubmersible I in irregular head
seas. In this figure the wave eleva
tion record, the potential part of
the drift force and the viscous part
of the drift force are shown in the
top three traces. The lower trace
shows the sum of the viscous force
and the potential force compared with
the total measured force.
It is clear that the result of adding
the viscous contribution is a clearly
improved correlation with the
measured force.
overall effect
contribution to
bution spectra
In order to show the
of adding a viscous
the potential contri
of the lowfrequency
surge force in irregular head seas
and sway force in beam seas on Semi
Submersible I are given in Figure 23
for three different sea conditions.
Each figure shows the drag coeffi
cient Cd used for the computations of
the viscous force contribution. The
Cd values used for the computations
of the viscous contribution in some
cases had to be adjusted in order to
achieve a reasonable fit with the
measured data. This clearly is an
unsatisfactory aspect of the simpli
fied model for the viscous effect
which will need to be refined in the
future. An important effect not
accounted for is for instance, the
shielding effects due to the
I
proximit~ of the columns. However,
the above results tend to confirm
that there is a significant viscous
effect in the drift forces on semi
submersible type structures which, in
irregular waves without current seems
to be concentrated in the splash zone
of the columns. The analysis has been
based on a rather simple model for
the viscous contribution which has
not been verified to any great
extent. In the next section some
results of ongoing detailed research
carried out at the Delft University
of Technology into such effects is
described.
VISCOUS EFFECTS IN DRIFT FORCES
ON A FIXED VERTICAL CYLINDER
In the previous section it was indi
cated that the most significant vis
cous contribution to the horizontal
drift force on a semisubmersible
seems to originate from the splash
zone of the columns. In order to gain
more insight in such effects, model
tests have been carried out to deter
mine the distribution along the ver
tical of the mean horizontal drift
force on a single vertical cylinder
in regular waves. The work is part of
an ongoing Ph.D. project.
See
The
No.
reference [11].
model tests were
2 towing tank of
I
carried out in
the Ship Hydro
606
I
mechanics Department. This facility given for the model scale.
measures 80 m x 2.75 m x 1,25 m and The model tests were carried out for
is equiped with a single flap a range of wave frequencies corres
hydraulically operated wavemaker pending to the longer waves for a
capable of generating regular and semisubmersible. At scale 1:100 the
irregular waves. The basin is fitted wave frequencies tested in the model
out with a towing carriage with a correspond to 0.3 r/s to 0.8 r/s at
special low speed carriage control full scale. This is a range of
for the simulationof current effects frequencies relevant for extreme sea
by towing. conditions.
The model cylinder which had a dia The results shown in Figure 25 and
meter of 0.075 m is shown in Figure Figure 26 confirm that the greatest
24. Afiscale 1:100 this could be re discrepancies between the potential
preventative of a column with a 7.5 m computations and the measurements of
diameter. The splash zone and the the mean forces are found for the
subsurface part are independently splash zone of the cylinder. The
attached to a central core through measured mean forces are consistently
force transducers measuring the significantly larger than the com
horizontal force on each of the two puted values, For the subsurface
sections. part of the cylinder, differences
Model tests were carried out in regu also occur between measurements and
lar waves with and without current. computations. In a relative sense
For each test the vertical position they appear to be of the same order
of the cylinder was adjusted so that as for the splash zone part. However,
the through of the wave passing the the absolute value of the forces is
cylinder passed just above the sepa considerable lower and the differen
ration between the splash zone part ces between measurements and compu
of the cylinder and the subsurface tations are less consistent.
part of the cylinder. This ensured It can be concluded that these model
that the subsurface part of the tests poit to the splash zone contri
cylinder was fully submerged at all bution to the viscous part of the
times. Results of measurements in mean drift force as being the most
regular waves without current of the important one.
mean horizontal drift force on the
splash zone and the subsurface zone
FINAL REMARKS
are compared with results of calcula
tions of the relevant contributions In this paper we have shown some
to the drift forces based on 3dimen results of an extensive series of
sional potential theory and the ap model tests on two semisubmersibles
plication of the pressure integration which confirm differences between
or nearfield method in Figure 25 and computed and measured mean and low
Figure 26 respectively. frequency horizontal wave drift
According to the nearfield theory forces in regular and irregular
for drift forces, the splash zone waves.
contribution is dependent on the Application of a simple model for the
square of the relative wave elevation viscous contribution to the drift
around the cylinder while the drift forces indicated that irregular waves
force on the subsurface element is without current the major source of
due to the nonlinear pressure con the viscous contribution was to be
tribution in the Bernoulli pressure found at the splash zone part of the
equation. For this reason the results columns of a semisubmersible.
of mean force measurements have been Model test in regular waves with a
divided by the square of the undis fixed vertical cylinder representing
turbed wave amplitude. Results are a single column of a semisubmersible
607
or a TLF confirm that the largest
discrepancies between computed and
measured drift forces are indeed to
be found in the splash zone.
Further experimental investigations
are required in order to be able to
formulate a more detailed model for
the viscous effects which can also
take into account such aspects as the
interaction effects due to the proxi
mity of the columns of a semisubmer
sible.
REFERENCES
[1]
[2]
[3]
[4
[5]
[6]
Haoft, J.P.: Hydrodynamic
Aspects of SemiSubmersible
Platforms , Publication No.
400, Netherlands Ship Model
Basin, 1972
Newman, J.N,: The Drift Force
and Moment on Ships in Waves ,
Journal of Ship Research, 1966
Faltinsen, O.M. and Michelsen,
F,C,: Motions of Large Struc
tures in Waves at Zero Froude
Number , Symposium on Marine
Vehicles, London, 1974
Pinkster, J.A.: LowFrequency
Second Order Wave Exciting
Forces on Floating Structu
res , Publication No, 650,
Netherlands Ship Model Basin,
Wageningen, 1980
Pinkster, .J.A. and Huijsmans,
R.H.M, : The Low Frequency Mo
tions of a SemiSubmersible in
Waves , Boss82, Boston, 1982
Pijfers, J.G.L. and Brink,
[7]
[8]
[9]
[10]
[11]
Huse, E.: Wave induced Mean
Force on Platforms in Direc
tion Opposite to Wave Propa
gation, Norwegian Maritime
Research, VO1.5, No.1, 1977
Standing, R.G., Brendling,
W.J. and Jackson, G.E.: Full
scale Measured and Predicted
LowFrequency Motions of the
SemiSubmersible Support Ves
sel Uncle John, First In
ternational Offshore and Polar
Engineering Conference ,
Edinburgh, 1991
Ferretti, C. and Berta, M.:
Viscous Effect Contribution
to the Drift Forces on Float
ing Structures , International
Symposium on Ocean Engineering
Ship Handling, Gothenburg, 80
Chakrabarti, S.K.: Steady
Drift Force on Vertical Cylin
der  Viscous vs. Potential,
Applied Ocean Research, VO1.6,
No.2, 1984
Dev, A.K.: Experimental In
vestigations of Viscous Mean
Drift Forces on a Fixed Verti
cal Circular Cylinder in Waves
and Currents Part I , Report
No. 928M, Ship Hydrodynamics
Department, Delft University
of Technology, 1992
A.W. : Calculated Drift Forces
of Two SemiSubmersible Plat
form Types in Regular and Ir
regular Waves , Paper No. OTC
2977, Offshore Technology
Conference, Houston, 1977
608
9.14 20.57
a I
! 1 I I
i I I I
5,49 5.49 3.14
I 1 !
I
I
3.051 19.i41 I 3.05
59.44
Dimensionsin
22.86 t 22.86 ! 22.86
1.52 68.58
I
Fig. 1General arrangement of Semisubmersible 1.
Dimensionsaregiveninmetres

1
~, +_
1
I I I
%: )3
1s.5
1~
J 
. 
~! ,,
II II
,1 II
II II
,1 {1
73.5 ,. . _.__._[.l
il II
,1
,1
,1
,1
,1 If
al

:.r
14.0
w
I \ !  * :
L
! I
# I
25.0
I
I !5.0
55.3 I
Fig. ZGeneral arrangement of Semisubmersible Il.
609
Fig.3Test setup for tests in regular waves.
. .
.
. .
1
wave s
Drift forces
1
Feed.f cmwasd ~Q1ative
control
Wzlve dev.
system
Ffg. 4Tast srrtup for tests in Irregular waves.
20==)
Fig. 5Block diagram of control system for tests in irregular waves.
I
FB (Forcefrom control system)
Fig,6Block diagram of forces acting on the structure.
~r BSSM..I Surge motions .
m ~L., . .
>, ~
.=
Fig. SSpectra of drift forces obtained for different restralnhrg system characteristics for the
Fig. 7Exemple of (a) measured restraining form, (b) correction force for motions,
scme sea conditions,
and (8+ b) total drift force record.
610
I I
.20 
calculation
A 0 a Rqularwaves,
b
a
A
lo
o
A
o ,
A I
07
5
t
5000
0 0.25 0.50
Fig. 18Spectral density on surge drift force of Fig. 1S on
Semisubmersible II.
1
Measu]
i
#
...
calculi
II
,1
20000 t
,1
,1
,1
1:
,1
I
II
~:
lsooo +;
, I II
I 1};]
! It, , ,
I q, t
; t,l ~
I
~1
t
1 II
t
10000
t
9
!
I
t
1
1
1
I I
:
,
:
!
1
5000 +
(
I
0. I
0 0.25
Fig. 20Spectral deneity of surge drlff force of Fig. 17 on
Semlsubmereible Il.
*
!
I
Measured
.. ..
Cslculskd
l(W)OO
m
5
t
&
II
~
~b?i
,
I
o .
o 0.2.s 0.50
@ ~ radla
Fig, 19Spectral density of surge drift force of Fig. 16 on
Semisubmersible W.
Fig. 21Wave eet.down in irregular wave8.
CD = 0.80
25.02
Calculated (potential part) a
tf o
25,CQ
J
2S.W
Measured
1
Force Calculated (viscouspart) b
tfo,
.~
..
%
t~:k=~.<~:=clat 
25,1YI
rfwT
0 50 100
tin,
Fig. 22Lowfrequency surge drift force on Semlsubmersibla I in irregular head seas,
. .
.
.
._. _
=.. .:=.
. .= 
..
..

.
___ =__~= ._.. ==.,== .=a _. , __= 
.=
. . . .
 .=... _+
s==.=. _ . . .: .= . > ~_ .,

.
.
_ ___ ~ ~ ? ,  =.=__._ ._:_ : .:= : ~ =
.
. _ & ..s.=a.=.
_._== ._=__ .=. __
= .=, .==, ____:. ___ :_.. :_
_____. __ ~. .. .. = .= .... .=. .
.
.==e.= . _ ,= =. .:
.
. . _=. = __= _ =...=< v. . __= _ __: ___ ____=  :.== .. =._ _ . .
_.
.
Calcufakd %,  7.1 *
dir  180.
mrm
0 50 100
tir. ,
Fig. 15Low.frequency surge driff force on Semisubmersible II in
irregular head aeaaHa = 3.1 m, T, 7.1 G.
5.03 I
1
Wave
mO
. Mtaturcd
C4kuhted
Wo 5.5 m
*1  11.3 s
diz  130
m
0 50 100
tic!,
Fig. 16Low.frequency surge drlff force on Samiaubmeraible II In
irregular head aeaaHs x 5.5 m, T, =11,3 a.
614
S.@.l ~ Wave 4W  3.09; $=  7.12 s
m o .,!3. .  !ti..  ;,,?.;:.. ..:;. . :...; ..!j:.. ,,,.... . ;
5.@I J
Measured
.. . .
Cahdared
10.03,
1
Force
tf o
 ,,=.~ 
Io,lx J ~
~...

/=
.._ ,,
4m 585 =; $1 = 11.30 =
10.027
Forc e
! 4. 13..24 M; %1 = 14.25 s
10.C4
Forc e
to* . .......

=./ks ... ... {+ ;7%;,:,:
10.W
7,
1 L
\ ;
\,
u
1, I , t
I
I
1
0 50 100
t in seconds
Fig. 14Low.frequency eurge driff force on Samiaubmemible I in
Irragular head aeaa.
4=0  Io.3 =
Mu.rurtd
%1  14.5 G
. .
Catculatrd
dir  180.
.
Fmmm
o 50 100
tins
Fig. 17Low.frequency surge driff force on Semisubmaralble II In
Irregular head seasHa = 10.3 m, T, x 14.5 s.
Mult mai mult decât documente.
Descoperiți tot ce are Scribd de oferit, inclusiv cărți și cărți audio de la editori majori.
Anulați oricând.