Vous êtes sur la page 1sur 15

Journal of Marine Engineering & Technology

ISSN: 2046-4177 (Print) 2056-8487 (Online) Journal homepage: https://www.tandfonline.com/loi/tmar20

Computation of mooring chain with the


touchdown on an inclined seabed

Milan Batista & Marko Perkovič

To cite this article: Milan Batista & Marko Perkovič (2019): Computation of mooring chain with the
touchdown on an inclined seabed, Journal of Marine Engineering & Technology

To link to this article: https://doi.org/10.1080/20464177.2019.1572059

Published online: 28 Jan 2019.

Submit your article to this journal

View Crossmark data

Full Terms & Conditions of access and use can be found at


https://www.tandfonline.com/action/journalInformation?journalCode=tmar20
JOURNAL OF MARINE ENGINEERING & TECHNOLOGY
https://doi.org/10.1080/20464177.2019.1572059

Computation of mooring chain with the touchdown on an inclined seabed


Milan Batista and Marko Perkovič

Faculty of Maritime Studies and Transport, University of Ljubljana, Portorož, Slovenia

ABSTRACT ARTICLE HISTORY


In this article, we discuss the effects of touchdown on an inclined seabed of a mooring system con- Received 16 June 2018
sisting of one or two chains joined at the sinker that is subjected to horizontal tension or tension Accepted 14 January 2019
at the suspension point. We modeled the statics of the system using the classical non-extensional
catenary theory. For a system without a sinker, we provide an analytical solution for the governing
equations, whereas for a system with a sinker, we derive governing equations that can be solved
only numerically. Analytical formulae for the calculation of the restoring coefficient of the system
are also provided. Numerical examples and a real case of mooring buoy design illustrate the use of
the proposed model.

1. Introduction some commercial programmes (Ariane8 [Ariane 2015],


MIKE 21 [MIKE 2017]). However, the manuals of these
In this article, we consider the effects of an inclined
programmes do not describe a method for the calcula-
seabed on the statics of a mooring system consisting of
tions.
one or two chains joined by a sinker, where one chain is
The aim of this paper is to develop an analytical math-
attached to an anchor and the other is suspended from a
ematical model of the mooring system which interacts
buoy. Such a system can be found in harbours as a part of
with the slanted seabed. The model is to be used for pri-
a ship’s mooring system (Jones 1985; Tsinker 1995) or as
marily quasistatic analysis and should be simple enough
a part of a mooring system platform (Smith and MacFar-
for in-house implementation. Therefore we will utilise
lane 2001). In this article, we provide a straightforward
only the rigid body mechanics and the classical static
static analysis of such a system based on the classical
catenary theory. Consequently, we must assume the fol-
theory of non-extensional catenary (Appendix A). As is
lowing regarding a chain and its load:
well-known, such a static analysis can be used in the
initial design phase of a mooring system (Triantafyllou
• The chain is inextensible and homogeneous.
1982).
• The chain is completely submerged.
The statics of a catenary is one of the oldest problems
• The resistance of a chain in water is omitted.
in the mechanics of deformed bodies and has therefore
• The dynamic effects are omitted.
been a subject in a significantly large number of statics
textbooks (Bedford and Fowler 2008). Also, a signifi-
For the seabed, the assumptions are the following:
cant number of studies on the use of catenary theory for
designing a mooring system have been carried out. How- • The seabed is inclined and ideally flat, i.e. without
ever, notwithstanding the large volume of literature on irregularities.
the subject, it appears that a complete analytical model • The seabed is rigid such that a part of a chain lying on
of chains touching down on a flat rigid seabed is of a it cannot penetrate it.
more recent origin (Garza-Rios et al. 1997; Masciola et al. • The seabed is rough, that is, the friction force between
2013). To the authors’ knowledge, an inclined seabed has the seabed and chain and the seabed and sinker is not
been considered only by Chai et al. (2002); however, they limited.
used a semi-analytical approach on a complex model and
therefore did not provide a general formula. To that we Practically all the above assumptions about the seabed
add that a model of the inclined seabed is included in are more or less false. Therefore some comments are in

CONTACT Milan Batista milan.batista@fpp.uni-lj.si Faculty of Maritime Studies and Transport, University of Ljubljana, Portorož, Slovenia

© 2019 Institute of Marine Engineering, Science & Technology


2 M. BATISTA AND M. PERKOVIČ

order. The first assumption is geometrical and implies


that the part of the chain on the seabed is a straight
line. This assumption can be omitted. However, our goal
is to obtain explicit expressions for calculation of the
geometry of chain. Profiled seabed require an algorithmic
solution (Ariane 2015). The next two assumptions are
physical assumptions and in practical terms indicate that
the chain and sinker cannot move over the seabed, i.e.
they can only be lifted. We incorporated these assump-
tions because a discussion regarding the interaction of a
chain with the seabed and anchor mechanics are separate
subjects and are out of the scope of the present article.
For more details on these subjects, refer to Wang et al.
(2010), Aubeny and Chi (2010), Elosta et al. (2013), and
the references therein. In the following, we will, there- Figure 1. Initial (dashed line) and deformed (solid line) geome-
fore, consider only the suspended part of a chain, i.e. tries of chain, where PA is the anchoring point, PB is the initial
the catenary. We note however that we can obtain a sim- suspension point (buoy chain stopper), P0 is the chain apex point,
P1 is the chain touchdown point, and P2 is the current suspension
ple model of chain-seabed interaction by the assumption point.
of, for example, the validity of the Coulomb friction law.
But this again requires an algorithmic solution because of
possible stick-slip motion of the chain over seabed. that

 > hB and  > hA . (2)


2. Chain without sinker
These assumptions exclude a vertical chain from the dis-
2.1. Initial geometry
cussion. To discuss only a system with a seabed, we also
We assume that in the initial state, the chain is unloaded, assume that
and thus, its shape is composed of two line segments (see
Figure 1): one lying on the seabed and one vertical from − 900 < α < 900 . (3)
the seabed to the buoy. The initial geometric parameters
are as follows: The geometrical parameters listed above exhibit certain
interdependencies amongst themselves, and they permit
• Chain specific weight, q several possibilities of the initial data.
• Chain length, 
• Length of the part of chain on seabed, 0 1. When hA , hB , and LB are specified, then, with the
• Angle between seabed and horizontal, α (we assume help of Figure 1, we can establish the following for-
the clockwise direction of α to be positive) mulae for calculating p, , and 0 :
• Water depth above anchor, hA
• Horizontal distance from anchor to buoy, LB hB − hA
p ≡ tan α = , (4)
• Water depth under buoy, hB LB

Using Archimedes’ principle, we can calculate the specific 


weight of a chain in the water as = L2B + (hB − hA )2 + hB , (5)
 
ρw
qw = 1 − q. (1) 0 =  − hB . (6)
ρc
Using popular trigonometric identities, we can express
If we consider the density of steel, i.e. ρc = 7800 kg/m3 , sin α and cos α, which are required in future calculations,
and the density of seawater, i.e. ρw = 1027 kg/m3 , then using tan α as follows:
qw ≈ 0.868 q.
To avoid an extensive discussion of special circum- p 1
stances, we assume that the values of all the geometrical sin α =  , cos α =  (7)
1 + p2 1 + p2
parameters above are positive. Moreover, we shall assume
JOURNAL OF MARINE ENGINEERING & TECHNOLOGY 3

2. If hB , LB , and p are specified, then, using Eqs. (4) and Figure 1):
(5),
X1 = XA + 0 cos α, Y1 = YA − 0 sin α, (15)
hB = hA + pLB , (8)
  Now, if we assume that H is specified, then, λ is expressed
as
 = hA + LB 1+p +p .
2 (9)
H
λ= , (16)
3. If p, , and hA are specified, then from Eq. (9) qw
  and therefore, the fundamental unknowns of the prob-
LB = ( − hA ) 1+p −p ,
2 (10) lem are arc-length coordinates s1 , s2 , of catenary start and
endpoint and length 0 of the chain on seabed. These
and hB is expressed by Eq. (8). unknowns are connected by the following two geomet-
4. Finally, if hA , hB , and  are specified, then, from rical conditions:
Eqs. (4) and (5),
• The total length of the chain is
hB − hA
p=  , (11)
( − hB )2 − (hA − hB )2  = 0 + s2 − s1 . (17)

and LB can be calculated from Eq. (10). • The vertical distance between the catenary end-points
We note that the chain cannot be straightened, and is (see Figure 1)
thus, chain span L is constrained to
y2 − y1 = hA + 0 sin α. (18)

LB ≤ L < 2 − h2A . (12)
From Eq. (17), we have
With zero horizontal force, the vertical force at the s2 =  − 0 + s1 . (19)
suspension point is composed of only the weight of
the vertical part of the chain. We thus obtain the Using Eqs. (17) and Eq. (88) for y, we can express condi-
following values of the vertical load at the chain tion (18) as
touchdown point (index 1) and chain end (index 2):  
p
λ2 + ( − 0 + s1 )2 − λ2 + s21 = hA + 0  .
V1 = 0 V2 = qw hB . (13) 1 + p2
(20)
This expression contains three parameters: s1 , 0 , and λ.
2.2. Calculation when the horizontal force is
When two of these are specified, then, the third can be
specified
calculated from Eq. (20). We distinguish the following
With the application of horizontal tension force H to three cases:
the suspension point, the chain gradually rises from the
seabed; thus, its shape consists of a line of length 0 on the Case 1: The chain is only to be raised from the seabed
seabed and the catenary. It is evident from the equations
in Appendix A that the shape of the catenary and its load In this case, 0 = 0, and at the catenary touchdown point
is completely determined if we know the chain param- (tangency point), we have α = −θ1 ; thus, tan θ1 = −p.
eter λ as well as the chain arc-length coordinates s1 of Therefore, from Eq. (89), we have s1 = −λp. Substituting
its start point and the arc-length coordinates s2 of its end this into Eq. (20), we obtain the equation for λ. By replac-
point. Further, the position of the chain in the plane coor- ing λ with Hmin /qw , where Hmin is the minimal force
dinate system OXY is completely determined if we know required to raise the chain from the seabed, we obtain
the plane coordinates of its start point (see Appendix B).
qw 2 − h2A
These coordinates are arbitrary. For definitiveness, in this Hmin =  . (21)
study, we define the coordinate origin to be on the water 2 hA 1 + p2 + p
surface immediately above the anchor: When α = 0, i.e. p = 0, it reduces to a popular formula
XA = 0, YA = −hA , (14) for describing a chain rising from a horizontal seabed
(Garza-Rios et al. 1997):
where XA , YA are the coordinates of the anchor in the
qw (2 − h2A )
plane coordinate system. The coordinates of the con- Hmin = . (22)
tact point between the catenary and seabed are then (see 2 hA
4 M. BATISTA AND M. PERKOVIČ

Case 2: Part of the chain is on the seabed Because H ≥ 0, the lower value of TB is specified based
on the condition H = 0. Using Eq. (13), we obtain
When H < Hmin , it implies that part of the chain is on the
seabed. At the catenary touchdown point, we have s1 = TB ≥ qw hB or λB ≥ hB , (30)
−λp; thus, Eq. (20) becomes an equation with 0 as the where equality implies that H = 0. Equations (20) and
unknown. A solution to this equation is (27) have three parameters: s1 , 0 , and λ. Thus, one of
  these must be specified, and the remaining two can be cal-
2
0 = (1 + p ) + 1 + p hA p
2 culated. As described in section 2.2, we now distinguish
  three cases:

− h2A + 2hA (λ + p) 1 + p2 + 2λp+ (2 + h2A )p2 . Case 1: The chain is only to be raised from the seabed
(23)
In this case, we have 0 = 0, H = Hmin , and s1 =
When α = 0, this equation reduces to the formula for −Hmin p/qw , where Hmin is expressed by Eq. (21). Substi-
describing a chain on a horizontal seabed (Garza-Rios tuting these into Eq. (27) for H, we obtain the minimum
et al. 1997): tension force required for raising the chain from the
 seabed:

0 =  − 2hA λ + h2A . (24)
TB,min = Hmin 2 + (q  − H
w
2
min p) . (31)
Case 3: The chain is completely raised from the seabed
Case 2: Part of the chain is on the seabed
If H > Hmin , then 0 = 0; thus, Eq. (20) becomes an
If qw hB < TB < Tmin , then a part of the chain 0 is on the
equation with s1 as the unknown. Its solution is
seabed. At the catenary touchdown point, we have s1 =

−λp; thus, in this case, for λ, Eq. (28) yields
1 2 − h2A + 4λ2
s1 = hA − . (25) 
2 2 − h2A 0 =  − λp − λ2B − λ2 . (32)

When s1 < 0, the catenary part of the chain contains its Note that we assume  − 0 + s1 > 0 because the ver-
apex. When s1 = 0, the apex is at the anchor point. This tical force V2 > 0 at the suspension point. Substituting
occurs when the horizontal force is this into Eq. (20), we obtain an equation with λ as the
unknown, which has the following solution:
2 − h2A 
H = qw . (26)  
2hA
λB − r 1 + p + p 2λB 1 + p2 − r r(1 + p2 )
2

When H exceeds this value, s1 becomes positive, and we λ=  ,


have only the monotone part of the catenary. (1 + p2 ) 1 + p2
(33)
where

2.3. Calculation when the tension force at the buoy
r ≡ p + hA 1 + p2 . (34)
is specified
For p = 0, this equation reduces to
When the tension force TB at the suspension point is
specified, then, by means of Eqs. (19) and (92), H can be λ = λB − hA . (35)
expressed as
 Case 3: The chain is completely raised from the seabed
H = TB2 − q2w ( − 0 + s1 )2 , (27)
If TB > Tmin , then 0 = 0. From Eq. (27), we then have
or alternatively as 
 s1 =  − λ2B − λ2 . (36)
λ= λ2B − ( − 0 + s1 )2 , (28)
Substituting this into Eq. (20), we obtain an equation with
where λ as the unknown. Its solution is

TB 1
λB = . (29) λ= (2 − h2A )[(2λB − hA )2 − 2 ]. (37)
qw 2
JOURNAL OF MARINE ENGINEERING & TECHNOLOGY 5

2.4. Chain-restoring coefficient and when H ≤ Hc , i.e. when the chain is lifted from the
seabed, then
The chain-restoring coefficient, which is used for lineari-
sation of H in, for example, mooring lines analysis (Jones
1985), is defined as d0 ds1 λh2A
=0 and = 2 . (45)
dλ dλ ( − h2A )(s1 + /2)
dH
K≡ , (38)
dδ We observe that the chain-restoring coefficient can be
where δ is the horizontal displacement of the buoy, which calculated in a purely analytical manner once the shape
can be expressed as the following using Eq. (88) for x: of the chain is known.
s   s 
δ ≡ L − LB = λ sinh−1 − sinh−1
2 1
λ λ 3. Chain with sinker
0
+ − LB . (39) When the chain has a sinker, the sinker divides the chain
1 + p2
into two parts: the part from the anchor to the sinker
This coefficient is used for linearisation of H. (referred to as the anchor chain and denoted by super-
Now, δ = δ̂(λ, s1 , s2 , 0 ). As observed from Eqs. (25) script -) and the part from the sinker to the buoy (referred
and (23), s1 and 0 rely only on λ; meanwhile, based to as the buoy chain and denoted by superscript +). In
on Eq. (19), s2 = ŝ2 (0 , s1 ). The total differential of δ is addition to the parameters from section 2.1, we now have
therefore the following parameters:

∂ δ̂ ∂ δ̂ ∂ δ̂ ∂s2 d0 • Sinker-chain specific weight, q− , and length, −
dδ = + +
∂λ ∂0 ∂s2 ∂0 dλ • Buoy-chain specific weight, q+ , and length, +
 • Initial water depth above the sinker, hS
∂ δ̂ ∂ δ̂ ∂s2 ds1
+ + dλ. (40) • Initial horizontal distance from the anchor to the
∂s1 ∂s2 ∂s1 dλ sinker, LS
∂s2
From this as well as ∂ 0
= −1, ∂s 2 dH
∂s1 = 1, and dλ = q , the The total chain length, , is now expressed as
chain-restoring coefficient is expressed as follows:

K=  
q
  , (41)  = − + + . (46)
∂ δ̂ ∂ δ̂ ∂ δ̂ d0 ∂ δ̂ ∂ δ̂ ds1
∂λ + ∂0 − ∂s2 dλ + ∂s1 + ∂s2 dλ
3.1. Initial geometry
where
∂ δ̂ s  s  s1
When there is no horizontal force on the buoy, then,
= sinh−1 − sinh−1
2 1
+ similar to a chain without a sinker, the chain shape is
∂λ λ λ λ2 + s1 2 composed of two straight lines (Figure 2).
s2 We now distinguish two cases:
− , (42)
λ + s2 2
2
1. The sinker is initially on the seabed
∂ δ̂ 1 ∂ δ̂ λ
= , = − , In this case, + ≥ hB . If LS is specified, then the water
∂0 1 + p2 ∂s1 λ + s1 2
2
depth above the sinker, hS , and the length of anchor chain
∂ δ̂ λ are expressed by
= . (43)
∂s2 λ + s2 2
2
hS = hA + LS p, (47)
Now, when H ≤ Hc , i.e. part of the chain is on the seabed,
then
   
d0 p + hA 1 + p2 (1 + p2 ) − = LS 1 + p2 . (48)
=−    and
dλ  −  + p p + h 1 + p2
0 A Evidently, these relations are valid if LS < LB . The initial
ds1 load of the buoy chain is identical to that without a sinker
= −p, (44) (see Eq. (13)).

6 M. BATISTA AND M. PERKOVIČ

Figure 2. Initial state of a chain with sinker on the seabed (left) and that of a chain with sinker lifted (right).

If − is specified, then, based on Eq. (48), the distance (Qw − V1+ )p + H (λS − s+
1 )p + λ
+

between the anchor and sinker is F=  = q+


w  , (57)
1 + p2 1 + p2
−
LS =  . (49) where
1 + p2
Qw
λS ≡ . (58)
The end forces on the buoy chain are q+
w

V1+ = 0, V2+ = q+ Here, H will lift the sinker when N = 0. According to


w hB . (50)
Eq. (56), this occurs when
2. The sinker is initially lifted from the seabed s+ +
1 = λS − pλ . (59)

In this case, + < hB . Thus, we have Evidently, N = 0 can be achieved only when + 0 = 0, i.e.
when the buoy part of the chain is completely raised from
LS = LB , (51) the seabed. This can also be demonstrated analytically. If
hS = hB −  . +
(52) +
0 > 0, then, the contact point of the chain and seabed is
the catenary tangency point. Thus V1+ = pH, and there-
When the sinker is lifted, we have the following vertical fore, N = √Qw 2 > 0 for any Qw > 0. Thus, for H ≤ HS ,
forces at the chain’s characteristic points: 1+p
where HS is the minimum force that raises the sinker,
V1− = 0, V2− = q−
w (hB − hS ), (53) the sinker assumes the role of the anchor; moreover, the
V1+ = V2− + Qw , V2+ = V1+ + q+ chain is considered through a replacement of all the for-
w hS . (54)
mulae based on the methods described in the previous
section, i.e.
3.2. Equilibrium of sinker
hA ← hS and  ← + (60)
If Q is the weight of the sinker in air, then, its weight Qw
in water is
  To determine HS , we equate Eqs. (25) and (59) for s+
1 . The
ρw solution to the resulting equation is
Qw = 1 − Q. (55)
ρs

(+ − h2S )[(+ + 2λS )2


2
where ρs is the density of the sinker. For concrete of den- hS
−h2S + (+ − h2S )p2 ]
2
sity 2400 kg/m3 , we have Qw ≈ 0.572 Q, whereas for steel
q+ −p(+ − h2S )(+ + 2λS )
2
of density 7800 kg/m3 , we have Qw ≈ 0.868 Q.
When the sinker is in equilibrium on the seabed, then, HS = w (61)
2 h2S − (+ 2 − h2S )p2
the normal component N of the reaction force and the
tangential component F of the reaction force are (see When the sinker loses contact with the seabed, its equi-
Figure 3) librium requires the following (see Figure 3):
Qw − V1+ − pH λS − s+ − pλ+ Qw = V1+ − V2− = q+ + − −
N=  = q+ 1 ≥ 0, (56) w s1 − qw s2 . (62)
w
1 + p2 1 + p2
JOURNAL OF MARINE ENGINEERING & TECHNOLOGY 7

Figure 4. Chain with sinker.

Figure 3. Equilibrium of (a) sinker on seabed and (b) lifted sinker.


• the length of the anchor chain

3.3. Calculation when the horizontal force is − = − − −


0 + s2 − s1 , (67)
specified
• the length of the buoy chain
When the horizontal force H lifts the sinker, the chain
shape becomes composed of three parts: the catenary cor- + = s+ +
2 − s1 , (68)
responding to the buoy chain (specified by the parame-
• and the vertical distance between the chain suspension
ters s+ +
1 and s2 ), the catenary corresponding to the anchor points (see Figure 4)
chain (specified by the parameters s− −
1 and s2 ), and the
part of the anchor chain on the seabed (length − 0 ). The y2− − y1− + y2+ − y1+ = hA + −
0 sin α. (69)
anchor chain and buoy chain parameters are as follows:
Using Eq. (88) for y, we can express this last equation as
H H
λ− = and λ+ = . (63)   
q− q+
(λ ) + (s2 ) − (λ ) + (s1 ) + (λ+ )2 + (s+
− 2 − 2
+ 2 − 2 2
2)

The coordinates of the contact point between the sinker −
0p
− (λ+ )2 + (q+ s+
2
catenary and seabed are (Figure 2) 1 ) = hA +  . (70)
1 + p2
X1− = XA + −
0 cos α, Y1− = YA − −
0 sin α. (64) We thus have five unknowns, − − − + +
0 , s1 , s2 , s1 , and s2 , and
four equations for disposal. From Eqs. (67), (68), (62),
Then, the sinker coordinates are the coordinates of the
and (59) we have
endpoint of the anchor chain
s− − − −
2 =  − 0 + s1 , (71)
XS = X2− = X1−
 
s− s− q− −
+λ −
sinh −1 2
− sinh −1 1
, s+
1 = s + λS , (72)
λ− λ− q+ 2

YS = Y2− = Y1− s+ + +
2 = s1 +  . (73)
⎛ ⎞
 2  2 Introducing these expressions into Eq. (70), we obtain
⎜  s−  s−

+ λ− ⎝1 + 2− − 1 + 1− ⎠ , (65) the relation among − −
0 , s1 , and H. Two of the parameters
λ λ must be specified; then, the third can be calculated. Sim-
ilar to a chain without a sinker, we consider the following
and the coordinates of the sinker are the coordinates of three cases:
the start point of the buoy chain, i.e.
Case 1: The sinker is only to be raised from the seabed
X1+ = XS , Y1+ = YS . (66)
In this case, − − − −
0 = 0 and s1 = −λ p = −Hmin p/q . In

Five parameters of the chain, − − − +


0 , s1 , s2 , s1 ,
and are s+
2,

this manner, Eq. (70) becomes the equation for Hmin , that
connected by the sinker equilibrium equation (62) and is, the horizontal force that completely lifts the anchor
the following four geometrical equations: chain from the seabed.
8 M. BATISTA AND M. PERKOVIČ

Case 2: Part of the sinker of the chain is on the seabed To obtain the tension force TB,min that lifts the anchor
chain from the seabed, we first calculate Hmin ; moreover,

In this case, H < Hmin , and at the anchor chain touch- from the above formula, we express the following:
down point, we have s− −
1 = −λ p. Thus, the remaining
− 
unknown in Eq. (70) is 0 . 2 + (q+ + + q− − − H 2
TB,min = Hmin w w min p + Qw ) .
(76)
Case 3: The chain is completely lifted from the seabed

− Case 2: Part of the anchor chain is on the seabed


In this case, H > Hmin , and therefore −0 = 0. From

Eq. (70), we can calculate the unknown s1 .
In this case, TB < TB,min . From Eq. (75), we have
All the cases above result in equations that do not have
an analytical solution and must be solved numerically. 
Reasonable initial estimates of the numerical solutions TB2 − H 2 + Hp − Qw − q+
w
+

of these three equations are Hmin = Hc , − − − −


0 = −
0 =  /2, and
. (77)
− − q−
w
s1 = −λ p, respectively.
Alternatively, we can transform the equations for the
Substituting this into (74), we obtain an equation with H
first two cases into a polynomial sixth-order equation, the
as the unknown.
second case into an eighth-order equation, and the third
case into a twelfth-order equation (see Appendix C). No
Case 3: The anchor chain is completely lifted from the
initial estimates are required in these cases; however, as a
seabed
result, we obtain multiple roots. In general, from among
such roots for each of these cases, we must select one
In this case TB > TB,min . From (75), we have
that is real and that satisfies the original equation (70).
In particular, in the first case, we must select a root that is 
positive. In the second case, we must select one that yields TB2 − H 2 − Qw − q+
w
+

the value 0 ≤ − − s−
1 = − − (78)
0 ≤  . This is the smallest positive root. q−
w
In the third case, we must select the root that is proximate
to zero. Substituting this into (74), we obtain an equation with H
as the unknown.
3.4. Case when the tension force at the buoy is The resultant expressions are long and cumbersome;
specified therefore, we omitted them herein. However, the pro-
gramming of these equations with the aid of a computer
To address this case, first, we use the definitions λ− = algebra programme is straightforward.
H/q− and λ+ = H/q+ to re-express Eq. (70) in the fol-
lowing form:

 
  3.5. Restoring coefficient of chain with sinker
H 2 − 2 H 2 In this case, the horizontal displacement of the buoy, δ, is
+ (s−
2
+ (s ) − 1)
q− 2
q−

 
   
+ s+ −1
+
−1 s1
H 2 + 2 H 2 δ = λ sinh 2
− sinh
+ (q+ s+
2
+ + (s ) − 1) λ+ λ+
q+ 2
q+
 
+
− p s s+
= hA +  0 . (74) + λ− sinh−1 2− − sinh−1 1−
1 + p2 λ λ
−
In addition, in this case, when applying Eqs. (71)–(73), + 0 − LB . (79)
Eq. (27) for H becomes 1 + p2

H = TB2 − [q+ + + q− (− − − − 2 To calculate K = dH dδ , we have two equations: one has
0 + s1 ) + Qw ] .
(75) seven parameters, i.e. δ = δ̂(− − − − + + +
0 , λ , s1 , s2 , λ , s1 , s2 );
Again, we consider three cases: the other is Eq. (70), which has the form ĥ(− − − −
0 , λ , s1 , s2 ,
+ +
λ+ , s1 , s2 ) = 0. If we add λ− = H/q− and λ+ = H/q+ ,
Case 1: The chain is only to be raised from the seabed we have sufficient equations to compute K. A tedious
JOURNAL OF MARINE ENGINEERING & TECHNOLOGY 9

albeit straightforward application of partial differenti- When H > Hc , i.e. when −


0 = 0 (the chain is completely
ation results in the following equations for the chain- raised),
restoring coefficient:
When H ≤ Hc , i.e. when s− −
1 = −λ p, −1 1 ∂ δ̂ 1 ∂ δ̂ 1 ∂ ĥ 1 ∂ ĥ
K = + ++ − −− + − −
q ∂λ q ∂λ q+ ∂λ+ q ∂λ

∂ δ̂ ∂ δ̂ ∂ δ̂ ∂ δ̂
−1 1 ∂ δ̂ 1 ∂ δ̂ ∂s−
+ ∂s−
+ ∂s+
+ ∂s+
K = + ++ − × 1 2 1 2
. (81)
q ∂λ q ∂λ− ∂ ĥ ∂ ĥ ∂ ĥ ∂ ĥ
 ∂s−
+ ∂s−
+ ∂s+
+ ∂s+
1 2 1 2
∂ δ̂ ∂ δ̂ ∂ δ̂ ∂ δ̂
− + −+ ++ + p
∂s− 1 ∂s2 ∂s1 ∂s2 The partial derivatives in the above equations are as
 follows:
1 ∂ ĥ ∂ ĥ ∂ ĥ ⎫
− + − + − ∂ δ̂ − − ⎪
q ∂λ + ∂s1 −
∂s2 −1 s2 −1 s1 ⎪

= sinh − sinh ⎪

 ∂λ− λ− λ− ⎪



∂ ĥ ∂ ĥ 1 ∂ ĥ − −
s1 /λ − −
s2 /λ ⎪

+ ++ + p + − − + − ⎪

∂s1 ∂s2 q ∂λ ⎪

− − 2
1 + (s1 /λ ) − − 2
1 + (s2 /λ ) ⎪



∂ δ̂ ∂ δ̂ ∂ δ̂ ∂ δ̂ + + ⎪

− − − ∂ δ̂ s s ⎪

∂− ∂s− ∂s+ ∂s+ = −1 2
− −1 1 ⎪
× 0 2 1 2
. (80) sinh sinh ⎪

∂ ĥ ∂ ĥ ∂ ĥ ∂ ĥ ∂λ + λ + λ + ⎪

− − − ⎪

∂− ∂s− ∂s+ ∂s+ + +
s1 /λ + +
s2 /λ ⎪

0 2 1 2
+ − ⎬
+ + 2 + + 2 ,
1 + (s1 /λ ) 1 + (s2 /λ ) ⎪



Table 1. Calculated values for q− + ∂ δ̂ 1 ∂ δ̂ 1 ⎪

w = 100, qw = 20, hA = 1000, ⎪
+ = 500, p = 0, − = 896.6, Qw = −10 × 103 , and TB = − =  − = − ⎪


∂0 1+p 2 ∂s1 1 + (s1 /λ ) ⎪
− − 2
100 × 103 . The unit of length is m, and the unit of force is 9.81 N. ⎪



∂ δ̂ 1 ∂ δ̂ 1 ⎪

Present (Smith and MacFarlane 2001) =  = −  ⎪

− + ⎪
∂s2 ∂s + + ⎪
− −
1 + (s1 /λ ) ⎪
2 2
H 44290.134 44.29 × 103 1 + (s2 /λ ) 1 ⎪

XS ⎪

644.210 644.2
∂ δ̂ 1 ⎪

YS −442.901 442.9
=  ⎪

X2+ − XS + ⎪

231.925 231.9
∂s2 + +
1 + (s2 /λ )
2 ⎭
θ1− 0 0 (data)
θ2+ 63.711 63.7
(82)

Table 2. Calculated values for q− + −


w = 2, qw = 5, hA = 3,  = 20, p = 100 %,  = 8, and
+ −
Qw = 50 and various values of H. For these data, Hmin = 7.122, HS = 36.368 , Hmin =
47.835, LB = 7.042, hB = 10.042, LS = 5.657, and hS = 8.657. The units are unspecified
and are assumed to be dimensionally consistent. The angles are expressed in degrees (°).
Case 1 2 3 4
H 5 20 45 70
L 9.983 12.925 14.359 16.773
Anchor chain
−
0 8 8 1.639 0
Point 1 2 1 2 1 2 1 2
s− – – – – −22.5 −16.139 −22.706 −14.706
θ− – – – – −45 −35.651 −32.973 −22.790
V− – – – – −45 −32.278 −45.412 −29.412
T− – – – – 63.640 55.379 83.440 75.928
Buoy chain
+
0 0.572 0 0 0
Point 1 2 1 2 1 2 1 2
s+ −1.000 10.429 0.008 12.008 3.544 15.544 4.118 16.118
θ+ −45 84.522 0.117 71.577 21.496 59.930 16.390 49.022
V+ −5.000 52.139 0.041 60.041 17.722 77.722 20.588 80.588
T+ 7.071 52.398 20.000 63.284 48.364 89.809 72.965 106.745
10 M. BATISTA AND M. PERKOVIČ

  ⎫
∂ ĥ ⎪

= 1 + (s− /λ− )2 − 1 + (s− /λ− )2 ⎪


∂λ− 2 1


s− /λ− s− /λ− ⎪

+ 1 − 2 ⎪



1 + (s− − 2 1 + (s− − 2 ⎪

1 /λ ) 2 /λ ) ⎪

  ⎪

∂ ĥ ⎪

= + (s + + 2
/λ ) − 1 + (s+ + 2 ⎪
1 1 /λ ) ⎪

∂λ+ 2


s+ /λ+ s+ /λ+ ⎪

+ 1 − 2 ⎪



1 + (s+ + 2 1 + (s+ + 2 ⎪

1 /λ ) 2 /λ ) ⎪


∂ ĥ p ∂ ĥ s−
1 /λ

.
= −  = −  ⎪
∂− 0 1 + p 2 ∂s −
1 1 + (s1 /λ− ) ⎪
− 2 ⎪



− − ⎪

∂ ĥ s2 /λ ⎪

=  ⎪

∂s2−
− − 2 ⎪

1 + (s2 /λ ) ⎪



∂ ĥ + +
s1 /λ ⎪



+ = −  ⎪

∂s1 + + 2 ⎪

1 + (s1 /λ ) ⎪



∂ ĥ + +
s2 /λ ⎪

=  ⎪

+ ⎪

∂s2 + +
1 + (s2 /λ )
2 ⎭
Figure 6. Force–deflection diagram for the chain described in
(83)
Table 1.

4. Numerical example
demonstrate that the various calculated values are highly
We implement the discussed model in the Maple and consistent.
the Matlab programme. For verifying the model, we use The second example provided in Table 2 and the cor-
data from Smith and MacFarlane (Smith and MacFar- responding Figures 5–7 are for demonstration purposes
lane 2001), where instead of θ1− = 00 , we use the sinker- and can be used to verify the results when one imple-
chain length calculated by them. The results in Table 1 ments their own programme.

Figure 5. Chain shape for cases presented in Table 2.


JOURNAL OF MARINE ENGINEERING & TECHNOLOGY 11

mooring ropes to the bollards at the loading ramps, while


at the bow the ship is anchored and fastened to a moor-
ing buoy or buoys, their sides exposed to winds. Recently,
during stormy weather, the sinker of a buoy was pulled
out, the anchors did not hold the ship, and the ship col-
lided with the crane at the coil terminal. A pier was also
destroyed. With the developed methodology, the design
of the new mooring layout was analysed. The buoy is
secured with anchor and a mooring chain, which is lay-
ing on a slope, while vertically loaded with the sinker. The
mooring buoy design is shown in Figure 8, applied load,
catenary and geometry of the chain is shown in Figure 9,
and analysis case study results are shown in Table 3.

5. Conclusions
In this article, we provided a complete analytical solution
for a system consisting of an anchor, chain(s), a sinker,
and a buoy. The assumptions listed in the introduction
Figure 7. Restoring-coefficient diagram for the chain described
in Table 1. permitted us to fabricate a relatively straightforward ana-
lytical description of the system, which can be conve-
niently implemented using a computer programme.
The developed methodology was used in a real case in When the presented model is used for a buoy and/or
the Port of Koper, where large ro-ros are berthed in the an uphill seabed, a few additional examinations of the
Mediterranean fashion, the ships’ sterns made fast with input data are required because an arbitrary selection of

Figure 8. Mooring buoy design.


12 M. BATISTA AND M. PERKOVIČ

Figure 9. Screenshot of the Matlab programme used for calculations. Example is load case 1 from Table 2.

Table 3. Tension, displacement and catenary geometry at differ- Disclosure statement


ent loads for the case from Figure 8.
No potential conflict of interest was reported by the authors.
F/SWL 1/10 1/2 3/4 1
Applied horizontal force kN 150 750 112.5 1500 ORCID
Buoy displacement m 11.63 15.27 15.39 15.44
Sinker displacement m 0.15 1.27 1.28 1.28 Milan Batista http://orcid.org/0000-0002-4004-8098
Sinker lift m 0.58 7.95 8.74 9.15 Marko Perkovič http://orcid.org/0000-0002-7493-6587
Chain angle at sinker deg 27.7 7.0 5.5 4.8
Vertical force at sinker kN −59.8 −92.6 −108.6 −125.9
Max. chain tension kN 182.1 762.3 1135.4 1509.5 References
Ariane. 2015. Ariane8 user guide. Paris: Bureau Veritas.
Aubeny CP, Chi C. 2010. Mechanics of drag embedment
anchors in a soft seabed. J Geotech Geoenviron Eng.
data could result in illegitimate mathematical operations. 136(1):57–68.
Bedford A, Fowler WL. 2008. Engineering mechanics: statics.
That is, not all the input data have a physical realisa- 5th ed. Upper Saddle River (NJ): Pearson Prentice Hall.
tion for the cases considered. We also emphasise that Chai YT, Varyani KS, Barltrop NDP. 2002. Semi-analytical
for an uphill seabed, we must assess the interaction of quasi-static formulation for three-dimensional partially
the chain with the seabed more carefully. It is notewor- grounded mooring system problems. Ocean Eng. 29(6):627–
thy that the model can also be used for a case in which 649.
Elosta H, Huang S, Incecik A. 2013. Dynamic response of
the chain is suspended directly from a ship. However, in
steel catenary riser using a seabed interaction under random
this case, part of the chain may not be submerged; thus, loads. Ocean Eng. 69:34–43.
the results of the calculation include a model error that Garza-Rios LO, Bernitsas MM, Nishimoto L. 1997. Catenary
must be assessed separately. In general, any application mooring lines with nonlinear drag and touchdown. Depart-
of the present model requires a critical consideration of ment of Naval Architecture and Marine Engineering, Uni-
its fundamental assumptions. For example, one of the versity of Michigen.
Jones JP. 1985. Fleet moorings. Alexandria (VA): Naval Facili-
limitations of the model is its profile of the seabed. A ties Engineering Command.
complete treatment of a chain rising from an arbitrarily Masciola M, Jonkman J, Robertson A. 2013. Implementation of
profiled seabed requires an algorithmic approach. Thus, a multi segmented, quasi-static cable model. 23rd Interna-
the formulae derived above can be of significant aid. tional Ocean and Polar Engineering Conference Anchorage,
Finally, we should mention that the governing equa- Alaska.
MIKE. 2017. MIKE 21 mooring analysis user guide. DHI.
tions (Eqs. (20) and (84)) of the model can be solved for
Smith RJ, MacFarlane CJ. 2001. Statics of a three component
other combinations of data. For example, one may spec- mooring line. Ocean Eng. 28(7):899–914.
ify the chain direction at the suspension points and then Triantafyllou MS. 1982. Preliminary design of mooring sys-
calculate the required chain length(s). tems. J Ship Res. 26(1):25–35.
JOURNAL OF MARINE ENGINEERING & TECHNOLOGY 13

Tsinker GP. 1995. Marine structures engineering specialized for the plane coordinate system OXY, the catenary geometry is
applications. New York: Chapman & Hall. expressed as
Wang LZ, Guo Z, Yuan F. 2010. Three-dimensional interac- s
tion between anchor chain and seabed. Appl Ocean Res. X = X0 + λsinh−1
λ
32(4):404–413. 
 s 2
Y = Y0 + λ 1+ −1 (s1 ≤ s ≤ s2 ), (A9)
λ
Appendices
where X0 and Y0 are the coordinates of the catenary apex, and
Appendix A. Catenary s1 , s2 are the natural coordinates of the catenary start and end
points, respectively. The coordinates of the apex can be calcu-
A catenary equation that has the apex at its coordinate origin
lated once we know a point that lies on the catenary. In our
(Bedford and Fowler 2008) is (see Figure A1)
case, this point is the contact point between the catenary and
 x  the seabed, i.e. the catenary start point with coordinate s1 . If
y = λ cosh − 1 , (A1) the coordinates of the contact point are X1 and Y1 , then, using
λ
Eq. (A9), the coordinates of the catenary apex are
where λ is the catenary parameter (characteristic length) s 
X0 = X1 − λsinh−1
1
defined by horizontal force H and catenary specific weight q ,
as λ

 s 2
H 1
λ≡ . (A2) Y 0 = Y1 − λ 1+ −1 . (A10)
q λ

The length of catenary s (arc-length coordinate) from its apex Thus, we can alternatively obtain the catenary through the
to a point with coordinate x is following formulae:
s  s 
X = X1 + λ sinh−1 − sinh−1
1
x ,
s = λ sinh . (A3) λ λ
λ  
 s 2  s 2
1
From Eqs. (A1) and (A3), we can express x and y as functions Y = Y1 + λ 1+ − 1+ (s1 ≤ s ≤ s2 ).
of s using the trigonometric identity cosh2 x − sinh2 x = 1, i.e. λ λ
 (A11)
   s 2
−1 s
x = λsinh , y=λ 1+ −1 . (A4)
λ λ Appendix C.
In this appendix, we consider an alternative method to compute
The inclination of the catenary at the point with coordinate s is
an unknown quantity for a chain with a sinker for the three
s cases discussed in section 4.3.
tan θ = , (A5) The governing expression for the first two cases has the
λ
following form:
where θ is the angle between the x-axis and the tangent to the √ √ 
catenary. The vertical component of the tension force at the X1 − X2 = X3 − A. (A12)
point with coordinate s is expressed as By eliminating the square roots, we obtain
s 64A2 X1 X2 X3 − [4(AX3 + X1 X2 )
V = qs = H . (A6)
λ
− (A2 + X3 − X1 − X2 )2 ]2 = 0. (A13)
In particular, at the catenary end-points, we have
Now, in the first case, H is unknown. Thus, if we substitute the
V1 = −qs1 and V2 = qs2 , (A7) following into Eq. (A13), we obtain a polynomial equation of
the sixth order for unknown H:
where the minus sign at the start point results from Newton’s  2  − − 2
H q  − Hp + Q
law of reciprocal actions. The total tension force T at the point X1 ≡ +
with coordinate s is q+ q+
  2  2
  s 2 H + q− − − Hp + Q
X2 ≡ +  +
T = H2 + V 2 = H 1 + . (A8) q+ q+
λ  2  2 
H Hp H
Note that V, T, and θ are expressed as functions of s and are X3 ≡ −
+ − − + A ≡ − 1 + p2 + hA .
therefore independent of the position of the catenary in the q q q
plane. (A14)
For the second case, when −
0 is unknown, we substitute the
following into Eq. (A13):
Appendix B. Catenary in the plane
 2 − − 2
In Appendix A, the geometry of the catenary is expressed in H q ( − −0 ) − Hp + Q
X1 ≡ +
the coordinate system located at the catenary apex. In general, q+ q+
14 M. BATISTA AND M. PERKOVIČ

In the third case, the basic equation has the following form:
√ √  √
X1 − X2 = X3 − X4 − a. (A17)
To eliminate
√ the square roots from this equation, we substitute
A = X4 + a into Eq. (A13). After a few manipulations, we
obtain an expression of the following form:
X4 f12 (X1 , X2 , X3 , X4 , a) − f22 (X1 , X2 , X3 , X4 , a) = 0, (A18)
where f1 and f2 are rather long and are thus omitted herein. If
we substitute the following into the above expression,
 2 − − 2
H q ( + s− 1)+Q
X1 ≡ +
q+ q+
 2 2
H + q− (− + s−1)+Q
X2 ≡ +  +
q+ q+
 2
H
+ (− + s−
2
Figure A1. Catenary. X3 ≡ 1)
q−
 2
 2 2 H
+ (s−
2
q− (− − − X4 ≡ 1) a ≡ hA , (A19)
H + 0 ) − Hp + Q q−
X2 ≡ +  + (A15)
q+ q+
we obtain a twelfth-order polynomial equation with s−
1 as the
 2  
H − − Hp 2 unknown.
X3 ≡ +  −  − The described procedure is evidently manageable using a
q− 0
q+
computer algebra programme.
 − p
H
A ≡ − 1 + p2 + hA +  0 . (A16)
q 1 + p2
After a rearrangement, we obtain an eighth-order polynomial
equation with −
0 as the unknown.