Non-Linearly Restoring Performance and Its Hysteresis Behavior of

Proceedings of the ASME 2019 38th International
Conference on Ocean, Offshore and Arctic Engineering

OMAE2019
June 9-14, 2019, Glasgow, Scotland, UK
OMAE2019-95651
Downloaded from http://asmedigitalcollection.asme.org/OMAE/proceedings-pdf/OMAE2019/58806/V05AT04A016/6443285/v05at04a016-omae2019-95651.pdf by Shandong JianZhu Univ Library user on 02 May 2024
NON-LINEARLY RESTORING PERFORMANCE AND ITS HYSTERESIS BEHAVIOR OF
DYNAMIC CATENARY
Yilun Li Shuangxi Guo Yue Kong

School of Aeronautics Sciences and Institute of Mechanics, School of Aeronautics Sciences
Engineering, Beijing University of Chinese Academy of and Engineering, Beijing University
Aeronautics and Astronautics Sciences of Aeronautics and Astronautics
Beijing, China Beijing, China Beijing, China
Min Li Weimin Chen*

School of Aeronautics Sciences and Institute of Mechanics, Chinese Academy of Sciences
Engineering, Beijing University of School of Engineering Science, University of Chinese
Aeronautics and Astronautics Academy of Sciences
Beijing, China Beijing, China
ABSTRACT
Catenary is increasingly used as mooring-line and riser Keywords: dynamic response; catenary mooring-line;
system as the water depth gets larger due to its lower cost and dynamic behavior; hysteresis loop
easier installment. Its dynamic response and restoring
performance become more complicated, as the length of the NOMENCLATURE
mooring-line become larger, and the structural and fluid ωd angular frequency
dynamics the mooring-line become consequently more obvious. A0. amplitude of body motion
Compared to the quasi-static method where the static restoring T concentration force
force is mainly involved, the dynamic behaviors and its hysteresis q distributed force
of the catenary mooring-line are considered here so as to ρ density of catenary
comprehensively examine the non-linearly restoring A cross section area
performance of mooring-lines. Based on the 3d dynamic vector r position vector
equations along with the modified FEM simulations, the EI bending stiffness
hysteresis character of the restoring stiffness and the influences λ effective tension
of the catenary dynamics on its restoring performance are ε strain
presented and discussed. f hydrodynamic force
It is found that, principally owing to the damping and CD drag coefficient
inertial effect coming from the fluid and structural dynamics, the CA added mass coefficient
restoring force of the mooring-line depends on both the D structural diameter
structural displacement and velocity. Moreover, the dynamic u structure displacement
stiffness behaves as a hysteresis loop, instead of a curve. Our V fluid velocity
numerical results show that the energy consumption during one M mass matrix
period rises nonlinearly with the increase of the body frequency M added mass matrix
ωd and amplitude A0. And, the influence of nonlinear restoring U displacement vector
stiffness on the structural response along with the slack-taut C damping matrix
phenomenon caused by structural /hydrodynamic inertia and K stiffness matrix
damping is discussed.
*
Contact author: [email protected]
1 Copyright © 2019 ASME

F external load vector practical catenary with less assumptions, and few results of the
θ rotation angle extreme situations such as slack-taut based on the current
approaches are seen.
INTRODUCTION
In this study, the non-linearly restoring performance of
In recent years, more and more ocean energy, oil/gas and
catenary mooring-lines, under consideration of its dynamic
wind industries have been developing towards deeper ocean
behaviors, is comprehensively examine based on our 3d dynamic
area. The supporting systems of the floating bodies, such as oil
vector equations along with the modified FEM simulations
rig/production platform and wind turbine, includes catenary
model. Particularly, the hysteresis character of the restoring
mooring-lines, tendon lines and vertical tension legs[1]. Among
stiffness and the influences of the catenary dynamics on the
the three ones, the first one, i.e. catenary mooring-line, is
structure’s performance are presented and discussed. It is found
increasingly used due to its lower economic cost and easier
that the restoring force of the mooring-line depends on both the
installment implementation in deeper water. And, as the floating
structural displacement and velocity due to the fluid and
platform are developed toward deeper water depth, the length of
structural dynamics, and the top-end response can be
the mooring-line become larger and consequently the dynamics
significantly decreased.
such as the structural inertia and hydrodynamic damping of the
mooring-line become more obvious. Moreover, the translating 1 The Governing Equations and FEM Simulation of a
and rotational motions of the floating body should be no less the Dynamic Catenary
threshold value, and the economic cost of mooring-lines is
The static restoring force and tension distribution of the
supposed to be as low as possible under the condition of required
mooring-line can be calculated according to the classical
structure strength. Therefore, more reasonable analysis and
catenary theory[2], where only the static force of the mooring
results of mooring-lines’ restoring performance, along with its
system is considered. Here, to consider the catenary dynamics
dynamic behaviors is crucial to safety design of mooring lines
along with nonlinear geometry/structural and the fluid dynamics,
along with floating bodies.
the dynamic equations based on 3d curved flexible beam
There have been fruitful researches of the mooring-line,
approach is employed. The governing equations of a 3d catenary
which are mostly on restoring force (top tension) and floating
(see Fig.1) in terms of vectors[9] can be written as:
body stability[2-4]. Quasi-static method is one of the most
popular approaches to calculate the static restoring force of ( EIr )  ( r )  q   Ar  (1)
catenary. For examples, Qiao[2] presented the static restoring where λ is a scalar variable represents the effective tension, r is
force of a catenary, made up of three wire ropes and chains, with the position vector of the catenary, q is the distribute force, ρ and
different buoyancy unit weights and structural properties based A are structural mass density and area respectively, EI is the
on classic static catenary theory and piecewise extrapolation. bending stiffness. And the deformation equation is:
Van Den Boom[3] found that the nonlinearities coming from the r   r   (1   )2 (2)
geometry, elastic deformation and acting loads can significantly where ε is the strain of the catenary. If the value of the bending
enlarge top tension. Fan[4] studied the mooring line damping for moment in Eq.(1) is zero, we will have the dynamic equation of
the design of a truncated mooring system using an improved a catenary of which the external loads include the gravity,
quasi-static method. buoyancy and hydrodynamic forces.
However, as water depth and structural length increase, the
dynamic characteristics of mooring-line become more
T
significant. Chen[5] calculated the dynamic response of a system q M
including a spar and its mooring-lines based on a linear coupling
approach. He pointed out that if the inertial and damping forces y
s
of the mooring-line are involved during dynamic response, the r(s,t)
top tension would get larger. By now, there have been two kinds
of methods, i.e. lumped-mass method and flexible-bar
method[6,7], were used to involve the dynamic behavior of x
catenary. In the lumped-mass method, the mooring-line is z
modeled by a series of concentrated-mass points connected by
linear springs, the inertial force and the added mass coming from FIGURE 1: CATENARY MOORING-LINE MODELS
the fluid is calculated by the motion of the concentrated-mass Generally, the hydrodynamic force acted on per unit
points. For the flexible-bar model, the nonlinear dynamic structure length can be expressed by the Morison formula[10] as:
equations of curved flexible bar are build and to be numerically 1  D2  D2
f  CD  D V  u (V  u )  C A  (V  u )  V (3)
solved so as to model nonlinear structural stiffness and to get 2 4 4
large displacement of catenary. Given an ideal assumption of the where D and u are the structural diameter and displacement
two ends of catenary being at same level, Zhang[8] studied the respectively. V is the fluid velocity. Combing Eqs.(1), (2) and (3),
nonlinear dynamic response of a catenary. Still, the dynamic we have a group of nonlinear equations of the dynamic catenary,
behaviors of mooring line need further research to model a

and a numerical simulation based on FEM is used to solve the 2.1 Top Tension and Displacement along Catenary
dynamic equations.
The typical dynamic response of the catenary for case of a
The catenary is uniformly divided into N two-node Euler
regular motion, i.e. 0.05Hz frequency and 5m surge amplitude,
beam elements. For representativeness and simplicity, only the
of the floating wind turbine (FWT) is shown in Fig. 2a. In Fig.2a,
translation displacement in x-y plane [ui,vi] and one rotation
the top tension of the quasi-static method is also plotted as a
around z axis θi, of per node, are considered. Then the governing
comparison. It is seen that either the value of the peaks or the
equation of the catenary with many degrees of freedom can be
trough get extremer than, the maximum tension increases by
written as follow： 30% and the gap value between the peak and trough is about 3
(M + M)U + CU + KU = F (4) times of the quasi-static ones owing to the inertial and damping
where M and M are respectively the structure mass matrix and effects, while the period of dynamic tension is consistent with
the added mass matrix. C is the structure damping matrix. K is that of the top-end surge.
the stiffness matrix. F is the hydrodynamic force and gravity As the motion of the top-end gets larger, the top tension
acted on the catenary. U is the displacement vector. In order to response under the condition of the top movement with 0.1Hz
model simultaneously its original catenary shape and the large frequency and 5m amplitude is give in Fig.2b. There is an abrupt
rotation/translation flexibilities the rotation motion between two increase of the dynamic tension while the minimum tension
neighboring beam elements is released in our model, it means almost reaches zero value, which means slack-taut happens. In
that the rotation angles of the two beam elements are no longer this case, the maximum tension is around three times of the static
consistent with each other at same grid. Subsequently, the system method and even the peak-trough value is about ten times of the
rotational degrees of the freedom θ would double as θ,θ ′ static method.
because of the additional rotation angle. The displacement vector
of beam element changes from the original form:
Ui  [ui , vi , wi ,i , ui 1 , vi 1 , wi 1 ,i 1 ]T i  1, , N (5)
as
Ui  [ui , vi , wi ,i , ui 1, vi 1, wi 1,i 1 ]T i  1, N
(6)
Ui  [ui , vi , wi ,i ,i, ui 1, vi 1, wi 1,i 1,i1 ]T i  2, , N  1
Then, given the statically indeterminate characteristics
along with stronger stiffness singularity of the system, here the
original shape and top tension based on traditional static catenary
theory is used as the definite conditions so as to eliminate the (a)
singularity of the stiffness matrix. To run the dynamic response
analysis, the Newmark method is employed here so as to adjust
the distribution of the structural acceleration and the nonlinearity
of the catenary during the integration range by properly changing
the integration parameters.
2 Dynamic Response and Restoring Performance of
the Catenary Mooring-lines
The structural and geometrical parameters of the catenary
are listed in Table 1. The dynamic response of the catenary under
surge motion of the top-end body is analyzed based on our (b)
numerical simulations, i.e. the displacement, velocity and FIGURE 2: TOP TENSION RESPONSE OF THE CATENARY (a)
tension of the catenary along structural span are calculated. Then TOP-END SURGE AMPLITUDE=4m, PERIOD=20s (b) TOP-END
the restoring performance and its hysteresis are discussed. SURGE AMPLITUDE=5.5m, PERIOD=10s
TABLE 1: THE GEOMETRICAL AND MATERIAL The displacements and velocity change significantly as the
PARAMETERS OF THE CATENARY motion of the top-end gets larger, see Fig.3 where the phase track
Geometrical Value of the catenary middle point. It is noted that the center of the blue
Length 800m curve, corresponding to smaller motion of the top-end (T=20s,
Initial horizontal projection 706m A=5.0m) is close to the zero position, that indicates the dynamic
Initial vertical projection 350m balance position deviates from the static one. While the center of
Diameter 0.19m the red curve, i.e. the top-end motion of T=10s, A=5.5m, deviate
Young’s modulus 210GPa from the zero position, and the shape of the track phase looks no
Density 2513kg/m3 longer like elliptical one. Moreover, in Fig.3b, the vertical
Poisson’s ratio 0.3 velocity almost keeps being a constant value, i.e. the minimum
velocity, as the displacement change from its positive peak to

negative trough. That means owing to the additional dynamic The maximum top tension and tension gap at different
behavior of the catenary, the structural initial/damping forces and amplitudes and frequencies of the top-end motions are given in
fluid drag force can balance the structural gravity which is Fig.4. It shows that the values of the maximum top tension (and
supposed to principally cause the structural tension, and the tension gap) gets larger, e.g. up to 3.5 times of the static value
consequently, slack would happen if no tension is caused. particularly for case of snap, as the amplitude and/or frequency
of the top end get larger.
2.2 The Non-Linear Restoring Stiffness and Its
Hysteretic Behaviors
As presented above, the mooring-line dynamics could
produce an increase of top tension along with tension amplitude
difference. In fact, because of involvement of catenary
dynamics, i.e. the inertial and damping effects, the restoring
stiffness may change too. The horizontal restoring stiffness will
(a) be examined here, which are calculated under conditions of
different top-end amplitudes, i.e. A=7, 5, 4 and 3m, and different
periods, i.e. T=5, 10, 20 and 40s. The selected results are
presented in Fig.5 and 6.
1.2E3
Dynamic
1.0E3 Static
Restoring Force (kN)

8.0E2
6.0E2
(b)
FIGURE 3: PHASE TRACK OF THE VELOCITY VERSUS
4.0E2
DISPLACEMENT OF THE MIDDLE POINT (a) HORIZONTAL
MOTION (b) VERTICAL MOTION
4000
2.0E2
-6 -5 -4 -3 -2 -1 0 1 2 3 4 5 6
Quasi-Static Displacement (m)
3500 T20s A2m
T20s A3m (a)
3000
Top tension amplitude (kN)
T10s A2m
1.0E3
T10s A3m
2500 T10s A6m
2000
Tension Amplitude (kN)
1500 1.0E2
1000
500
0 1.0E1
1 2 3 4 5 6
Load cases
(a)
4000
1.0E0
Quasi-Static 0.0 0.2 0.4 0.6 0.8 1.0 1.2 1.4
3500
T20s A2m Frequency (Hz)
3000 T20s A3m
(b)
Top tension gap (kN)
T10s A2m
2500 T10s A3m FIGURE 5: RESTORING CHARACTERISTICS OF SINGLE
T10s A6m
2000
MOORING-LINE (a) RESTORING STIFFNESS OF THE MOORING
LINE (b) SPECTRUM OF THE RESTORING FORCE
1500
1000
Fig.5a shows the horizontal restoring stiffness curve of the
catenary at 20s period and 5m amplitude, and the static stiffness
500
is also plotted as a comparison. Interestingly, the dynamic
0 stiffness curve shows that the top tension is no longer linearly
1 2 3 4 5 6
Load cases related to only the top-end displacement as it does for case of
(b) quasi-static scenario, but, notably, it depends on both the top-end
FIGURE 4: TOP TENSION RESPONSES OF THE CATENARY displacement and velocity approximately in a way of
MOORING-LINE UNDER DIFFERENT CONDITIONS (a) approximately ellipse loop, which is called hysteresis loop. And
TENSION AMPLITUDE OF THE CATENARY (b) TENSION GAP differently from the static stiffness, the dynamic restoring force
OF THE CATENARY
does not get its maximum value at the maximum displacement

but at a smaller displacement, that produces a smaller dynamic The load cases of the top-end motion are list in Table 2. It
stiffness. The hysteresis character of the dynamic stiffness is can be seen in Fig. 7b that, the value of energy consumption from
mainly due to the damping effect coming from the structure and Case1 to Case 6 gradually increases, except for Case 1 and Case
fluid of the mooring-line. The spectrum plot, shown in Fig.5b, 6, where the minimum top tension approaches to zero, or slack
indicates that the peak values at frequencies of odd times of the happens. Then we may say that, the energy consumed by
excitation frequency is much larger than others. hysteretic damping in one cycle may get smaller as the mooring-
It is found that the energy consumption during a period get line becomes slack.
larger as the amplitude (and/or the frequency) increases. In other
TABLE 2: THE LOAD CASES OF THE TOP-END MOTION
words, the hysteresis effect of the dynamic stiffness gets more
Load cases 1 2 3 4 5 6
obvious as the amplitude and/or frequency increase, as shown in
Period/s 40 20 20 10 10 5
Fig.6. The energy consumption and mooring-line top tension Amplitude/m 6 4 6 4 6 4
under different load cases are presented in Fig.7.
2.3 The Influence of Hysteresis Characteristics on the
Catenary Response
Fig.8 shows the trajectories of different positions along the
mooring-line while the motion frequency changes (the amplitude
of the top-end is 6m). It shows that the trajectory loop becomes
more obvious and the vertical/horizontal displacements become
smaller owing to the hysteresis characteristics of the mooring-
line, as the motion of the top-end gets faster (or the period gets
smaller). The trajectory loop becomes more obvious as the
motion of catenary points, e.g. point 2 and point 3, get larger,
because of the larger damping effect coming from larger catenary
motion.
FIGURE 6: RESTORING LOOPS AT DIFFERENT FREQUENCY
AND AMPLITUDE
(a) (a)
4.5E3
4.0E3
Maximum tension
3.5E3 Minmum tension
Top tension (kN)
3.0E3
2.5E3
2.0E3
1.5E3
1.0E3
5.0E2
0.0
1 2 3 4 5 6
Load cases
(b)
FIGURE 7: ENERGY CONSUMPTION AND TOP TENSION
RESPONSES UNDER DIFFERENT TOP-END AMPLITUDES AND
FREQUENCIES (a) ENERGY CONSUMPTION (b) TOP (b)
TENSION

Particularly, the hysteresis of point 5 appears to be the most
obvious, see Fig. 8e. And, Fig.9 shows the velocity spectrum of
Point 5. It can be seen that, as the motion period decreases, the
moving speed of Point 5 gradually increases. The peak values at
higher frequencies also gradually increases. If comparing the
peak values at double and triple excitation frequencies, we can
see that peak value at triple excitation frequency gets larger, or
even larger than that of double excitation frequency as the period
of top-end motion decreases. That also means that the hysteresis
characteristics of the mooring-line becomes more obvious.
(c)
(a)
(d)
(b)
(e)
(f)
FIGURE 8: TRAJECTORY OF THE CATENARY (a)
TRAJECTORY OF POINT 1 (b) TRAJECTORY OF POINT 2 (c)
(c)
TRAJECTORY OF POINT 3 (d) TRAJECTORY OF POINT 4 (e) FIGURE 9: VELOCITY SPECTRUM OF THE CATENARY POINT
TRAJECTORY OF POINT 5 (f) AXIAL LOCATION OF POINT 1-5 5 (a) TOP-END PERIOD 40s (b) TOP-END PERIOD 20s (c) TOP-
END PERIOD 10s

In order to examine the influence of the mooring-line displacement significantly decreases due to the hysteresis of the
hysteresis on the motion the top-end, the dynamic response of mooring-line. The maximum amplitude decreases by around
the top-end is calculated and compared with the quasi-static 50%, and the amplitude decreases by 28% during the steady
method. The period of the excitation force is 20s, and the first- phase of the dynamic response.
order natural period of the system is 30s. The displacement of
the top-end is shown in Fig.10. It is noted that the displacement ACKNOWLEDGEMENTS
significantly decreases due to the hysteresis of the mooring-line. The authors of this paper would like to thank the financial
For examples, the maximum amplitude decreases from 9.4m to supports provided by the Strategic Priority Research Programme
4.0m, or by more than 50%, and the amplitude decreases from of CAS (Grant No. XDB100203 and XDA100201) and the
4.7m to 3.4m, or by approximately 28% during the steady phase National Natural Sciences Foundation (Grant No. 11372320).
of the dynamic response.
REFERENCES
[1] Musial, Walt, Sandy Butterfield, and Andrew Boone.
"Feasibility of floating platform systems for wind turbines." 23rd
ASME Wind Energy Symposium, Reno, NV. 2004.
[2] Qiao Dongsheng, Ou Jinping. Static analysis of a deepwater
catenary mooring system, Ship&Ocean Engineering, 2009,
Vol.38(2):120-124.
[3] Van Den Boom, H.J.J. Dynamic behaviour of mooring lines.
Proceedings of Behaviour of Offshore Structures, 1985:359-368.
[4] Fan T, Qiao D, Yan J, et al. An improved quasi-static model
for mooring-induced damping estimation using in the truncation
design of mooring system[J]. Ocean Engineering, 2017,
(a) 136:322-329.
[5] Chen X, Zhang J, Ma W. On dynamic coupling effects
between a spar and its mooring lines [J]. Ocean Engineering.
2001, 28(7): 863-887.
[6] Vassalos D, Huang S. Dynamics of small-sagged slack-taut
marine cables [J]. Computers & structures, 1996, 58(3): 557-
562.
[7] Huang S, Vassalos D. A numerical method for predicting
snap loading of marine cables [J]. Applied Ocean Research,
1993, 15(4): 235-242.
[8] Zhang S, Tang Y, Liu X. Analysis of Nonlinear Dynamic
Response of Mooring Lines in Deepwater[C]. ASME 2010 29th
International Conference on Ocean, Offshore and Arctic
(b)
FIGURE 10: COMPARISON OF THE TOP-END RESPONSES Engineering. American Society of Mechanical Engineers, 2010:
BETWEEN THE DYANAMIC AND STATIC RESULTS (a) TIME 489-494.
HISTORY OF TOP-END DISPLACEMENT (b) THE MAXIMUM [9] Guo S, Li Y, Chen W, et al. Non-Linearly Restoring
AND STEADY STATE AMPLITUDE Performance of Catenary Mooring-Line Under Consideration of
Its Dynamic Behaviors[C], ASME 2017, International
3 Conclusion Conference on Ocean, Offshore and Arctic Engineering.
The non-linearly restoring performance of catenary 2017:V07AT06A021.
mooring-lines under consideration of its dynamic effects is [10] Mavrakos S A, Chatjigeorgiou J. Dynamic behaviour of
comprehensively examined based on our modified flexible-beam deep water mooring lines with submerged buoys [J]. Computers
model combining with the FEM simulations. By our numerical & structures, 1997, 64(1): 819-835.
simulations, the hysteresis characteristics of the restoring
stiffness and the influences of the catenary dynamics on the
catenary and top-end responses are presented and discussed.
Based our numerical results we draw the following conclusion:
Owing to the damping and inertial effect coming from the
fluid and structural dynamics, the restoring force of the mooring-
line depends on both the structural displacement and velocity.
The dynamic stiffness behaves as a hysteresis loop, instead of a
static line. The energy consumption during one period get larger
as the amplitude (and/or the frequency) increases. The

Non-Linearly Restoring Performance and Its Hysteresis Behavior of

Uploaded by

Copyright:

Available Formats

Non-Linearly Restoring Performance and Its Hysteresis Behavior of

Uploaded by

Document Information

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Non-Linearly Restoring Performance and Its Hysteresis Behavior of

Uploaded by

Copyright:

Available Formats

Proceedings of the ASME 2019 38th International

Conference on Ocean, Offshore and Arctic Engineering

Yilun Li Shuangxi Guo Yue Kong

Min Li Weimin Chen*

1 Copyright © 2019 ASME

2 Copyright © 2019 ASME

3 Copyright © 2019 ASME

Restoring Force (kN)

4 Copyright © 2019 ASME

5 Copyright © 2019 ASME

6 Copyright © 2019 ASME

7 Copyright © 2019 ASME

You might also like