S~zlhan~, Vol. 20, Parts 2-4, April-August 1995, pp. 513-527. © Printed in India.

Stochastic response of guyed tower under second-order

wave force

T K DATTA, R S BISHT and A K JAIN

Civil Engineering Department, Indian Institute of Technology, Hauz Khas,

New Delhi 110 016, India

Abstract. Guyed towers have low natural frequencies compared with

generally encountered wave frequencies. 'They are therefore highly suscep-
tible to low frequency second-order wave forces resulting from the non-
linear phenomenon in the integrated fluid pressure on the submerged
structure. A simplified approach is presented here to obtain the second-
order wave forces encountered in random sea state for offshore guyed
towers using a simulation procedure. Nonlinear response of a guyed tower
to second-order and first-order wave forces, and a combination of the two
are obtained in time domain, The time histories of responses and the wave
forces are analysed to investigate their stochastic characteristics.

Keywords. Offshore guyed tower; second-order wave force; drift force;

nonlinear vibrations; sto¢hastic characteristics.

1. Introduction

The basic feature of a guyed tower platform is that it is designed to move with the waves, in
at least some of its degre~ of freedom. The system opposes wave forces predominantly by
inertial effects.This is in contrast to the fixed platform which opposes wave forces primarily
by the stiffness effect in the structure-foundation system. By their very nature, guyed
towers have low fundamental frequency and, therefore, are susceptible to excitation at
much lower frequencies than those of waves. Analysis of these structures for first-order
wave forces would then invariably predict very low responses. The effect of second-
order wave forces, which have generally significant energy content at low frequencies,
on the response of such platforms must be included in their overall analysis.
Investigation of the second-order drift forces, exciting low frequency responses, has been
extemively conducted for floating structures like buoys, semisubmcrsibles etc. Potential
theory was used to derive the second-order drift forces by Pinkster (1979), Pinkster &
Huijsmam (1982), Standing et al (1981), Chakrabarti & Cotter (1984), Eatock Taylor (1986)
and Karppinen (1979), where the wave drift forces were defined as the second-order term in
the integrated fluid pressure on the submerged body and was found to be proportional
to the.square of the wave amplitude. Sal~esen et al (1982), Wahab (1974), Pijfers
& Brink (1977) and Ferretti & Berta (1980) studied the slowly varying motion caused by
513
514 T K Datta et al

the second-order viscous drift force due to interaction between current, wind and wave
force. The drift force was defined as the mean value of the total force over a wave period
and was derived by using Morison's equation. The time histories of wave drift force for
random waves were generated with the help of the square of the slowly varying wave
envelope (hence, second order) and the drift force operator (i.e. drift force coefficient
plotted against frequency of encounter).
Peak mooring forces were analysed by Hsu & Blenkarn (1970) which were caused by
the vessel's slow drift oscillation in a random sea. They calculated the wave drift force
based on conservation of momentum principles and the concept of "radiation stress"
introduced by Longuet-Higgins & Steward (1964). Remery & Hermans (1971) also
investigated the slow drift oscillations of a moored object in random seas. The drift
force was calculated as a function of the square of the reflected and scattered wave
elevation. The reflection coefficient was determined from experimental tests on a rec-
tangular barge in head waves.
Kim & Yue (1989) have presented a new method for predicting slowly varying wave
drift excitations in multidirectional seas. The method retains the assumption of narrow
bandedness in frequency but treats the wave directional spreading exactly. Naess (1989)
has emphasised the importance of estimating the extreme values of the combined first order
and slow drift response of an offshore structure, subjected to a stationary irregular sea
wave. The study indicates a strong coupling between the first order and the slow drift
response. Kato et al (1990) presented an approximate solution for calculating the power
spectral density function of total responses including first-order as well as second-order
motions. Also, they showed that there exists a coupling between the first- and second-order
responses. Grue (1988) has examined wave drift damping and low frequency oscillations of
a moored elliptic cylinder. For long incoming waves he has obtained positive damping.
But for short incoming waves damping forces were found to be negative. Langley (1987)
has derived probability density function for low frequency second-order forces and
motions using a method which represents the random sea state as a sum of regular
waves. Grue & Palm (1986) have analysed the influence of a uniform current on slowly
varying forces and displacements and the results show that current may have a great
impact on the slowly varying motion of a moored body.
Jain & Datta (1991) have carried out investigations on articulated towers for the
response due to viscous drift forces. They have calculated the slowly varying drift force
by assuming that the drift force is proportional to the product of the square of the wave
envelope and the drift force coefficient in regular waves.
Herein, a simplified method for obtaining the second-order wave force, encountered
in the nonlinear oscillation of guyed tower in random sea states, is determined by
a procedure similar to that proposed by Pijfer & Brink (1977). The guyed tower is
analysed in time domain for the second-order wave force, first-order wave force and
a combination of the two. The time histories of the nonlinear response as well as the
wave force are analysed for a particular sea state in order to highlight the stochastic
behaviour of the guyed tower in low frequency oscillation.

2. Analysis
2.1 Guyed tower model

The Exxon guyed offshore tower model has been considered to carry out the present
 Stochastic response of guyed tower 515

Nelipocl

/
Drilling deck f See level Markerbuoy- ~

~ Fair lea~'

,.~/, ~ S , u . c,,, An,ho~ . I . . . .

Figure 1. Schematic Exxon
prototype guyed tower.

investigations. The tower is shown in figure 1. The bottom of the tower is provided with
a restricted rotational restraint. For the present study the tower has been idealized as
an equivalent uniform shear beam with a rotational spring at the base of the tower as
shown in figure 2. Varying degrees of rotational movement of the tower may be
achieved by varying the rotational spring stiffness. The guylines have been idealized by
a nonlinear spring. The contribution of buoyancy to the restoring force in these
structures is generally small when compared with that from the guyline stiffness, and is,
therefore neglected in this study (Dutta 1984). Further, it is assumed that the dynamics
of the guylines does not significantly influence the global tower motion. Thus, the
stiffness of the nonlinear spring is obtained from a separate static analysis of the
guylines under its own weight and current induced forces. The typical force excursion
relationship of the guylines, for the example problem is shown in figure 3.
The offshore tower is discretized into a number of plane 2D beam elements. At each
node, the dynamic degrees of freedom consist of sway translations and in plane
rotations. The equation of motion in structural co-ordinates takes the form
[M] {~} + [C] {~} + [K] {x} = F(t). (1)
In the mass matrix [M], added mass of the element is considered up to still water
level. The effect of variable added mass due to the instantaneous position of the tower is
assumed to be negligible and is ignored in the formulation oftbe problem. The damping
matrix [C] of the system is assumed to be a linear combination of (constant) mass and
stiffness matrices; {~}, ~ } and {x} are, respectively the vectors of structural acceler-
ation, velocity and displacement. The stiffness matrix [K] is made time invariant by
considering in it only the initial tangent stiffness of the nonlinear spring representing
the guy system. As a result of this, correction to the resistive force at any instant of time
is required and it is incorporated in the right-hand side load vector F(t). The equations
of motion are finally uncoupled using normal mode theory and an iterative technique.
As a result, the damping matrix [C] need not be explicitly known. Modal damping
ratios are only required for the solution of equation of motion. F(t) depends upon the
type of analysis being performed. For the first order wave force analysis, F(t) is due to
 O~

i I n~-I
(~a t i ,Nodes number ll to n 4-1)
0 = E~ment number ( 1 to n I

~.~ I '÷, .,.,

• i ÷lI X2i + 2
iI' k'ljtlj

2t x2i_, o,~,. ~..<, ",

Nodal displacements Force distribution Equivalent
nodal loads

• | IJl~"
k/
Gulmd )over
Equivalent beam

Figure2. Finiteelementmodelof guyedtower.

 Stochastic response of guyed tower 517

2001~0 .

150oo
v
o

Figure 3. Force excursion re-

. . . . . . . . i.b. . . . . . . . s.b. . . . . . . . .1a.0'. . . . . . . . . . . . . 1. .e.o
.. "~ lationship for prototype guyed
Excursion at 9uy node (rn) tower.

the hydrodynamic loading calculated based on inertia and drag effects produced by the
water particle kinematics. For the second order wave force analysis, the time history of
F(t) is simulated by considering low frequency wave forces resulting from nonlinear
phenomenon in the integrated fluid pressure acting on the structure. For complete
analysis, F(t) is a combination of first-order and second-order wave forces. It is to be
noted that the first-order wave force F(t) depends on both structural velocity (due to
nonlinear drag) and structural displacement (due ~o nonlinear force excursion relation-
ship of guylines) apart from the water particle kinematics. The second-order wave force
depends upon the nonlinear response of the guyed tower in regular waves of different
periods as explained later. Thus (1) becomes highly non-linear especially where first-
and second-order wave forces are considered together.

2.2 Simulation of second.order wave forces

The time histories of the drift forces are obtained using a method proposed by
Pijfers & Brink (1977). According to this method, a mean hydrodynamic force at
various nodes is calculated, which is the time average of all hydrodynamic forces acting
on the structure, obtained through the Morison's concept. The wave drift force is then
defined as the mean force in consequence of the waves and can be calculated by
subtracting the true current force from the hydrodynamic force. The method is based
on the assumption that the wave spectrum is sufficiently narrow. A random, long
crested sea may then be written as an amplitude modulated signal:
fi(t) = Re ll(t) = Re JR(t) {exp i r (t)} ], (2)
where, ~i(t) = wave amplitude, R(t) = slowly varying wave envelope, ~(t) = a complex
variable whose real part provides the wave amplitude F(t) the phase angle, equal to
coot + e(t), where toc is the central frequency of the narrow band spectrum and E(t) is
a slowly varying phase angle.
The above formulation leads to the expression for a momentary frequency of the
spectrum as:
[7(0 = o9~+ ~(t). (3)
518 T K Datta et al

The slowly varying drift force is then calculated assuming that the drift force is
proportional to the product of the square of the wave envelope R(t) and the drift force
coefficient in regular waves at the momentary frequency I'(t). The drift force Di(t ) at any
node i of the tower is then given by
Di(t)=0"5 pwgC,(F) R2(t). (4)
Evaluation of the value of Di(t), (5), requires the time histories of R2(t), r(t) and drift
force coefficient C i at each momentary frequency I'(t)

2.3 Determination of time histories of R2(O and ['(t)

The wave envelope R is the modulus of a complex variable i.e.

R 2 = (Re f~)2 + (Im fl) 2. (5)
Also,
F = t a n - 1 [Im 12/Re ~ ] , (6)
and
= [(Re f~) * (|m f~) - (Im f~)* (Re D)]/[ (Re D) 2 + (Im D) 2]
= (1/g2)[(Re D)*(lm fl) - (Im D)*(l~e fl)]. (7)
The irregular wave amplitude 3(t) is defined as Re D(t) and is given by the sum of a large
number of regular waves, i.e.

b(O = Refl(t) = ~ A. cos(rn.t + O.), (S)

n

in which A. is obtained according to the well-known formula;

A. = [2 S6(o~.)Aco.J1/2. (9)
S6(o.) is the spectral density ordinate of the wave spectrum at frequency oJ.; O. is the
random phase angle between 0 and 2u radians; ~o. is the randomly varying frequency
and Ao~.is the frequency interval. Since 6(0 is considered as Re [~(t), the imaginary part
of fl(t) is found to be

Im[l(t) = ~ A.sin(o~.t + O.). (lO)

n

The time derivatives of Re f~ and Im ~ are therefore,

l~ef~ = - ~ A.oJ.sin(co.t + 0.), (11)

n

im l'~ = ~ A.~o.cos(rn.t + O,). (12)

R

Using (8)--(12), the time histories of R2(t) and F(t) can be determined from (5) and (7).

2.4 Determination of drift force coefficient

The drift force coefficient Ci at any given momentary frequency [~ is obtained from the
plot of Ci vs ~e, where ae is the frequency of encounter. For a given regular wave of
 Stochastic response of 9uyed tower 519

amplitude 6. and frequency tr, the drift force coefficient C~ is given by

C, = 0,/(0-5flo,96~), (13)
where/)/is the mean drift force and is found by subtracting the true current force from
the mean hydrodynamic force i.e.
/), = F, -- Fc, (14)
where Fc is current force due to constant current profile across the depth,/~ is mean
hydrodynamic force and i is the node number.
The frequency of encounter is given by:
oe = o(1 + o 0el0). (15)
The mean hydrodynamic force F~ at node i is obtained by determining the mean
hydrodynamic forces over one period of encounter T i.e.
F, = (1/ T) fro f ,(t)dt, (16)
fi(t) = hydrodynamic force at node i.
The hydrodynamic forcef~(t) is calculated from Morison's equation viz.
f ,(t) = 0"5flwCdDalO~l(O,) + 0"25C~flwnD2 fj,
-- 0-25rcflw(Cm-- 1)D2.~/. (17}
The relative velocity U comprises mass transport velocity, Vml,water particle velocity,
0/, the current velocity Uc and structural velocity X i and is given by
Oi = Vmi "~ ~Ji "~- [Jc - Xi, (18)
In which Vm~,in deep water, according to Stokes theory, is given by
Vmi = 62 K o e x p ( - 2Ky), (19)
where K is the wave number and y is a function of the water depth.
Since the computation offi(t) as given by (17) requires Xi and Xi, the response of the
tower is to be obtained for different combinations of the regular wave of amplitude 6,
and frequency a. The response analysis of the guyed tower for regular waves of
amplitude 6, and frequency o is performed by (1) in time domain using the Newmark
fl method. Because of the dependence of F(t) on structural velocity and displacement,
the equation of motion is solved by numerical integration using iteration at each time
step to tackle all the nonlinearities. The efficiency of the analysis is augmented by the
use of the normal mode theory (Bathe & Wilson 1976).

2.5 Simulation of the first-order wave force

For a given random sea state the time history of first-order wave force at any node is
obtained by using (17). The time histories of water particle kinematics 0 i and [)i are
obtained from the PSDFs of water particle kinematics using the Monte Carlo
simulation technique (8).

2"6 Response analysis

The responses of the guyed tower for the first-order and second-order wave forces and
520 T K Datta et al

Table 1. Details of the idealised guyed tower.

Deck height 480 m

Mean sea level 457 m
Height at guy node 442 m
Guy diameter 9 cm
Number of guys 20
Guy system initial tension 1680 kN
EI 50000 G N m
Deck weight 69000 kN
Structural weight 3790 k N / m
Base restraint 20 GNm/rad
Effective drag diameter 35-05 m
Effective inertial diameter 5-79 m
Cd 0-90
Cm 2.00
Structural critical damping ratio 2%
Height at node 17 447 m
Structure's fundamental frequency 0.2689 rad/s
Structure's second frequency 1.3260 rad/s

for the combination of the two are obtained in time domain using N e w m a r k ' s fl method
with iteration required at each time step to tackle the nonlinearities (Datta & Jain
1988). The equations of motions are solved in normal coordinates thereby decoupling
the equations of motion. The details of the method are available in the reference cited
above.

3. Numerical results and discussion

The details of the idealised guyed tower are given in table 1. The tower is analysed for a
r a n d o m sea state with H1/3 = 15"0 m and Tz = 9.0 s. The sea state is characterised by the
one-sided Pierson Moskowitz (P-M) spectrum and has been simulated for a duration

200.aO

E Hs : 15.0 m
Tz = g.o s

~'100.00

! 50.00

J
0.00
Figure & Time history of wave
n~ (s ) envelope square.
 Stochastic response of guyed t o w e r 521

6,0
H8 = 15.0 rn
Tz = 9.0 s
4.0
"o
v2
2.0
c
10
g o.o
ta-

~-2.0

@
E -4.o.
o

-B.O :
50 I 1 2 2 o Figure 5. Time history of mo-
n.~ (e) mentary frequency.

of about 17.5 minutes. Typical time histories of wave envelope square R2(t) and the
momentary frequency F(t) for the sea state are shown in figures 4 and 5.
Figure 6 shows the variation of the drift force coefficient with encounter frequency
computed at node 17 of the tower which is the node just above the fairlead point (very
near to the mean sea level). The nonlinearly varying drift force coefficients do not differ
significantly for different wave amplitudes up to an encounter frequency equal to 1.0
rad/s. Since the oscillation of the tower is predominantly guided by the low frequency
response, the range of instantaneous frequencies generally falls within 1.0 rad/s.
Therefore the values of drift force coefficient Ci(I') for any value of ['~(t) may be
obtained from any of the curves corresponding to wave amplitudes 6o = 1 m, 2 m
and 3 m.
Figure 7 shows the power spectral density function (PSDF) of second order wave
force at node 17. The PSDF has significant energy at lower frequencies, as expected. On
the other hand, the low frequency contents of the PSDF of the first-order wave force are
small (figure 8).

25.00

Wave Amplitude
vE 2 0 . 0 0
.... 3 m
........ 2 m
| m
"g~s.0o
Current - 1.5 r n / I
~ tO.O0

°'°°o.66 . . . . . . "1~66. . . . . . ~,[66 . . . . . . ~ ...... ~.:66 ...... ~l Figure 6. Variation of drift

Encounter Frequency ( r a d / s )
force coefficient (at node 17).
522 T K Datta et al

4.0E+009
Ha = 15.0 m
P l'z : 9.0= 1.5s m/s
"~o 3.0E+009
=

...~2.0E+009

4p
C~

1.0E+009

t~

O.OE+O00 l ......... = . . . . . . . . . i ......... i,

o 1,o 2.0 ~.o . . . . . . . ~,:o Figure 7. P S D F of 2nd or-
Frequency (rod/s )
der wave force (at node 17).

3.0E+010

Hs = 15.0 m
~'O2.5E+010
g. I•L TZ ffi 9.0 s m/s

m 2,0E+010
%
~ 1.5E+010 '

e :
f~ 1.0E+010 -
-6
U
X soe+~ i
U3

0.0£+000 :
0.0 ~.o 2.o ~.o 4.o Figcre & PSDF of 1st order
Frequency ( r o d / s )
wave force (at node 17).

18.0 ~"
~] He = 15.0 m
4 Tz = 9.0 s
Current = 1.5 m / *
~1,.o -1, t

i10.0

;iii T.~ (=)

Figure 9. Time history of tip
displacement due to 1st order
wave forces.
 Stochastic response of guyed tower 523

8.0
Ha= IS.Ore
TZ " 9.0 S

A 4.0
E
v
E
eo 0.0
o
o.
15
F~-4.0

-8.0
Figure 10. Time history of tip
: ....... i66. . . . . . . i66 ....... ~6 ....... &$6 ....... displacement due to 2nd order
~ (s ) wave forces.

The PSDFs of tip displacements and the corresponding time histories of response
due to first- and second-order wave forces are shown in figures 9-12. Low frequency
excitation of the structure at the fundamental frequency due to a second-order wave
force is evident from the plot of PSDF shown in figure 12. It is interesting to note that
the frequency contents of the response spectrum due to a second-order wave force are
practically insignificant in the vicinity of predominant wave frequency. On the other
hand, the response characteristics due to the first-order wave force are such that the
spectrum has large frequency content at the predominant wave frequency. Low
frequency contents of the spectrum are not large.
Figure 13 shows the PSDF of tip displacement due to the combined first and second
order wave forces. The spectrum is characterised by two peaks, one occurring near the
fundamental frequency of the structure (0-269 rad/s) and the other at the predominant
wave frequency (0-7 rad/s). The first peak shows the low frequency resonating effect due
to the second order wave force while the second peak is predominantly the effect of the
first-order wave force. Comparing figures 12 and 13, it is seen that the response peak
(the first peak in figure 13) due to the second-order wave force is reduced when first- and

6.0'
Hs ffi 15.0 m
Tz = 9.0 s
~°~.o, Current = 1.5 m / s

4.0'

.~_~3.0
tO
C
~2.0

" ~ 1.0

0.0 Figure 11. PSDF of tip dis-

Frequency (rod/s )
placement due to 1st order wave
force.
524 T K Datta et al

t 6.0
Hs = 15.0 rn
Tz = 9.0 s
Current = 1.5 rn/s

12.0
1/)

~E
~ 8.0

c-

~ 4.0

(.~
0.0 ,m i "
Figure 12. PSDF of tip dis-
o.o . . . . . . . . b ~ ........ "I.o. . . . . . . . 'i.~. . . . . . . . ~.b placement due to 2nd order
Frequency (rad/s ) wave force.

second-order wave forces are considered together. This reduction is presumably caused
by nonlinear hydrodynamic damping associated with the first-order wave forces.
Table 2 compares the mean square values of the tip displacements due to the two
wave forces and their combinations. The response due to the second-order wave force
alone is quite significant (about 26% of that due to the first-order one). If the effect of
the second-order wave force is not taken into consideration, the overall response is
underestimated by about 13%. The difference in the overall response, however,
depends upon the amount of hydrodynamic damping present in the first-order wave
force.
Figure 14 shows the cumulative distribution function (CDF) of the first order wave
force, the second-order wave force and the combination of the two at node 17. It is seen
that the CDFs of the forces deviate from the normal distribution, especially near the tail
ends. This clearly shows the effect of nonlinearities on the distribution of the hy-
drodynamic forces. The CDFs of responses corresponding to different types of wave
forces are shown in figure 15. Distributions deviate considerably from the Gaussian
distribution near the ends, because of considerable non-linearities present in the wave

5,00
Hs = 15.0 m
Tz =n 9.0= ~s m/s
~ 4.00

3.00

2 .O0

1 .oo
g
4,O

0.00
Figure 13. PSDF of tip dis-
,'~", = . . . . . . . . . i ......... p ....... •. . . . . . .
o: 5 ...... d.g ..... 1.0 1.5 2.0 '~.g ...... L~.b placement due to combined 1st
Frequency ( r a d / s ) & 2nd order wave forces.
 Stochastic response of guyed tower 525

Table Z Comparison of mean square values of tip dis-

placement response for different sea states.

First Second Combined

Sea order order wave force
states (m2) (m 2) (m 2)

H, = 15"0 m 6-492 1"705 7"476

Tz= 9"0s

16
1/,
t2
0 10
8
ID
6

0
4
E 2
0
-2
-4
0.1 5 10 50 95 99 99.99 Figure 14. Distribution of
Cumula'tive probability wave forces (in Gaussian plot).

16
l/,

1O
0
-:- 10
O
> 8
A
E

0
-2
I I I I I I Figute 15. Distribution of tip
0.1 5 10 50 95 99 99.99 displacement responses (in
Cumulative probability Gaussian plot).

20
18i
16
.o lt,
> 12
~0
""
0
8
E 6

Figure 16. Distribution of

510 50 95 99 100 positive peaks of tip displacement
Cumulative p r o b a b i l i t y response (in Rayleigh plot).
526 T K Datta et al

forces. However, the distribution of positive peaks of the responses (figure 16) tend to
(mildly) follow Rayleigh distribution.

4. Conclusions

A simplified approach for obtaining the second-order wave force in the r a n d o m sea
state is presented for offshore guyed towers. The responses due to first- and second-
order wave forces, and their combination are obtained in time domain considering
hydrodynamic nonlinearity and the nonlinearity present in the restoring force pro-
vided by the guy lines. The responses of the tower are analysed to investigate their
statistical characteristics. The results of the numerical study lead to the following
conclusions.
(a) The second order drift force has significant energy at the low frequency range.
(b) The response due to the second-order wave force only shows resonating effect at the
structures's fundamental frequency and could be quite significant.
(c) When the second-order wave force is combined with the first-order wave force, the
low frequency response caused by the effect of the second-order wave force is
reduced because of hydrodynamic damping associated with the first-order wave
force effect.
(d) Both first-order and second-order wave forces are non-Gaussian due to significant
nonlinearity present in the integrated fluid structure system.
(e) The distributions of the responses are non-Gaussian. However, the peak distribu-
tion tends to (mildly) follow Rayleigh distribution.

References

Bathe K J, Wilson E L 1976 Numerical methods in finite element analysis (Englewood Cliffs, N J:
Prentice-Hall)
Chakrabarti S K, Cotter D C 1984 Interaction of wave with a moored semi-submersible. Proc.
3rd Offshore Mech. and Arctic Engineering Syrup., New Orleans pp. 119-:127
Datta T K, Jain A K, 1988 Nonlinear surge response of tension leg platform to random waves.
Eng. Struct. 10:204-210
Dutta A 1984 Dynamic behaviour of offshore guyed towers. Ph D thesis, Dept of Civil Eng., Indian
Institute of Technology, Delhi
Eatock Taylor R 1986 Low frequency behaviour of compliant offshore systems. Keynote
Addresses in Proceedings. 3rd Indian Conf. on Ocean Engineering (Bombay: Indian Inst.
Technol.) pp. 17-55
Ferretti C, Berta N 1980 Viscous effect contribution to the drift forces on floating structures.
International Symposium on Ocean Engineering and Shiphandling (Gothenburg: University
Press)
Grue J 1988 Wave drift damping and low-frequency oscillations of an elliptic cylinder in irregular
waves: Appl. Ocean Res. 10:10-19
Grue J, Palm E 1986 The influence of a uniform current on slowly varying forces and
displacements. Appl. Ocean Res. 8:232-239
Hsu F H, Blenkarn K A 1970 Analysis of peak mooring forces caused by slow vessel drift
oscillation in random seas. paper no. OTC 1159, I135-I145
Jain A K, Datta T K 1991 Response of articulated towers to viscous drift forces. Int. Shipbuilding
Progr., 38:315-331
Karppinen T 1979 An approach to computing the second order steady forces on semi-submersible
structures. Report No. 16 Helsinki University of Technology, Ship Hydrodynamics Laboratory.
 Stochastic response of guyed tower 527

Kato S, Kinoshita T, Takase 1990 Statistical theory of total second order responses of moored
vessels in random seas. Appl. Ocean Res. 12:2-13
Kim M H, Yue D K P 1989 Slowly varying wave drift forces in short-crested irregular seas. Appl.
Ocean Res. 11:2-18
Langley R S, 1987 A statistical analysis of low frequency second order forces and motions. Appl.
Ocean Res. 9:163-170
Longuet-Higgings M S, Steward R W 1964 Radiation stresses in water waves; a physical
discussion. Deep Sea Res 11:529-536
Naess A 1989 Prediction of extremes of combined first-order and slow-drift motions of offshore
structures. Appl. Ocean Res. 11:100-110
Pijfers J G L, Brink A W 1977 Calculated drift forces of two semi-submersible platform types in
regular and irregular waves, paper no. OTC 2977 155-161
Pinkster J A 1979 Mean and low frequency wave drifting forces on floating structures. Ocean
Eng. 6:593-615
Pinkster J A, Huijsmans R H M 1982 The low frequency motions of a semi-submersible in waves,
Proc. 3rd International Boss Conference, MI T, pp. 447-466
Remery G F M, Hermans A J (1971) The slow drift oscillations of a moored object in random
seas. paper no OTC 1500 II 829-836
Salvesen N, Von Kerczek C H, Yue D K, Stern F 1982 Computation of nonlinear surge motions
of tension leg platforms, paper no. OTC 4394 199-215
Standing R G, DaCunha N M C, Matten R B 1981 Slowly varying second order wave forces:
theory and experiments. Report No. R138, National Maritime Inst.
Wahab R 1974 Wave induced motions and forces on a floating structure. Report No. 1865,
Netherland Ship Research Centre, TNO, Delft