Applied Ocean Research 73 (2018) 42–58

Contents lists available at ScienceDirect

Applied Ocean Research

journal homepage: www.elsevier.com/locate/apor

Damage assessment for submarine photoelectric composite cable

under anchor impact
Qiang Gao a , Menglan Duan a,b,∗ , Xiaoxia Liu c , Yi Wang b , Xu Jia c , Chen An b , Tao Zhang d
a
Department of Aeronautics and Astronautics, Fudan University, Shanghai, China
b
Institute for Ocean Engineering, China University of Petroleum, Beijing, China
c
Engineering Research Centre, CNOOC Research Institute, Beijing, China
d
Qingdao Hanhe Cable Co., Ltd, Qingdao, China

a r t i c l e i n f o a b s t r a c t

Article history: The impact of dropped anchor on submarine photoelectric composite cables may possibly cause electrical
Received 4 July 2017 faults, i.e. electricity and optical signal transmission failure. In order to study the impact capacity and
Received in revised form 10 January 2018 structural impact failure mechanism, a test setup is designed originally to examine the structural and
Accepted 15 January 2018
functional integrity. A detailed finite element model (FEM) is created, considering material nonlinearity
Available online 8 February 2018
and component interaction. A parametric analysis has been performed to predict the deformation of
components and impact forces, under different impact velocities and collision directions. Relationships
Keywords:
between the armor layer indentation rate and that of internal power and optical units are achieved.
Submarine photoelectric composite cable
Anchor impact
The impact deformation of internal entities can be evaluated intuitively by armor layer indentation. The
Damage assessment proposed experimental and numerical methods are well correlated, suitable to assess the impact capacity
Deformation rate of subsea power cables and assist the protection design of subsea power cables in engineering.
Finite element model © 2018 Elsevier Ltd. All rights reserved.
Parametric analysis

1. Introduction cables in the Atlantic Ocean, as presented in Fig. 1. In addition,

Submarine power cables can be divided into three categories
regarding to their functions, viz. power cables for electric transmis-
sion, optical fiber cables for signal transmission and photoelectric
composite cables with the above two functions integrated. Nowa-
days, photoelectric cables gain an extensive popularity in oil & gas
field and offshore wind farms for power supply and signal transmis-
sion [1]. Submarine cables are relatively fragile, lying on a seabed
that is shared with other users, such as the fishing industry and
ship anchoring. Vessel anchors are hostile intervention on sub-
marine cables. Especially, those submarine cables close to vessel
anchorages or shipping channels are at high risk of anchor dam-
age [2,3]. Anchor impact is one of the most critical accidental loads
encountered among loading conditions [4]. Attwood [5] mentioned
that the most frequent cause of cable failure was external damage
by anchors and heavy fishing tools. An example of two Skagerrak
cables fault caused by external mechanical impact from anchors
and beam trawlers was given. According to the statistics by Worzyk
[6], anchor caused 18% of the total faults of submarine telecom
∗ Corresponding author at: Institute for Ocean Engineering, China University of
Petroleum, Beijing, China.
E-mail address: mlduan@cup.edu.cn (M. Duan).
Yoon and Na [7] reported that contribution of anchoring activities
to cable faults increased to 48%, and became the largest cause of
submarine cable faults in recent years. Once the cable is impacted
to some extent, the electric transmission and/or signal transmis-
sion will be cut off, seriously affecting the crowd’s production and
life.
Drop of anchor will happen where ships manoeuvre, in har-
bour basins, ship lanes, or fishing grounds. When designing a new
cable route or designing protection measures for installed cable,
it’s necessary to assess the risk of cable fault caused by anchor col-
lision. Several researchers have conducted risk assessment. Allan
and Comrie [8] used a risk matrix to assess damage frequency and
consequence. The risk assessment consisted of frequency of occur-
rence of hazard events and evaluation of the consequence of the
event. Nakamura et al. [9] carried out reliability analysis of sub-
marine cables in Japan and built a mathematical model to calculate
the probable cable failure rate due to ship anchoring. Guo [10] used
probabilistic risk assessment method to study the possible failure
rate of submarine cables by anchor damage. For instance, the annual
failure frequency of a submarine cable installed in Qiongzhou Strait
by dropped anchor was 0.000726. It was relevant to the number of
vessels passing the routing area and the probability of loss of control
of anchor handling. Although it was quite low risk, the possibility
of being struck by a falling anchor could not be ruled out.

https://doi.org/10.1016/j.apor.2018.01.006
0141-1187/© 2018 Elsevier Ltd. All rights reserved.
 Q. Gao et al. / Applied Ocean Research 73 (2018) 42–58
Fig. 1. Cause of damages to telecom cables in the Atlantic.
In addition, some specifications emphasize the risk of anchor
collision. DNV RP J301 [11] points out that the accidental limit
state, e.g. objects dropped onto cables and anchor impact should be
considered in the design of subsea power cables. In the protection
method of rock installation, the impact of falling rock onto the cable
should be confirmed to not compromise the cable integrity. The
International Telecommunications Union (ITU) [12] specifies the
test methods and conditions to examine the impact resistance of
optical fibre submarine cable systems. API SPEC 17E [13] indicates
that the armor layer is used to sustain tensile loads in the umbilical
and also has the additional function of providing impact protection.
It shall be documented that structural integrity is not endangered
due to impact loads. Extreme impact load shall be evaluated. Impact
forces due to rock dumping shall be analysed. Impact load and
energy relationship with consequence shall be determined. Two
important common specifications for submarine cables of the Inter-
national Council on Large Electric Systems (CIGRE), ELECTRA 171
[14] and ELECTA 189 [15], recommend the methods of mechani-
cal tests, like coiling test, tensile test, tensile bending test et al. but
without clues about mechanical impact.
However, the aforementioned international specifications do
not give out the test or simulation methods to determine the
impact resistance of submarine photoelectric composite cables
against anchor collision. A detailed understanding of impact fail-
ure mechanism of such cables is quite useful in order to achieve
the required performance. But only few researchers have investi-
gated the impact capacity of submarine photoelectric composite
cables. Jia et al. [16] performed impact tests of composite cables
on the seacoast. It was found that the cable would produce global
flexural deflection and local indentation. The optical unit was more
vulnerable to be crushed than power units during impact.
Static lateral crushing of umbilicals, flexible pipes and cables
provide some reference for impact analysis of cable. Guttner et al.
[17] presented and compared two different numerical models for
prediction of stress and strain fields in the steel tube umbilical com-
ponents under crushing. The elastic-plastic behavior was adopted
for the steel tubes. Alsos et al. [18] performed impact analyses for an
8-inch (bore diameter) flexible production riser. The cross section
deformation, impact forces and indentation were achieved. Tayama
et al. [19] conducted static lateral compression tests to confirm the
ability of 6.6 kV XLPE submarine cable with optical fiber sensors
for detecting anchor damage and defacement of wire armor. The
relationship between compression load and transmission loss in
optical fiber sensor and cable deformation rate was achieved. Nishi-
moto et al. [20] investigated the ability and sensitivity of 66 kV XLPE
submarine cable with a damage-detecting optical fiber sensor. The
submarine cable was placed on a steel test bench and compressed
gradually by a metal weight having a 60◦ angle with edge radius
43
Fig. 2. Impact test setup.
of 10 mm. The relationship between the deformation rate and the
increase loss in optical fiber was also attained. The cross section of
deformed cable at 30% deformation rate was given.
Although field tests can be expected to yield accurate data,
in general, field impact tests are expensive and quite inconve-
nient, limited by test locations, weather conditions, instrument and
human support. And it is difficult to accurately describe using a the-
oretical formula. Therefore, factory or laboratory test methods and
numerical procedures need to be developed.
In this present work, experimental and numerical methods
are proposed for damage assessment of submarine photoelec-
tric composite cables during anchor impact. Firstly, a factory test
arrangement is established to examine the impact properties of
composite cable, including structural integrity and electrical per-
formance. Secondly, a nonlinear finite element model (FEM) is
proposed and then verified by test results. The complex interaction
within composite cable, such as contact and friction between cable
strands is taken into consideration. Finally, a parametric analysis
is performed, considering different collision velocities and impact
directions. Damage assessments are conducted from the perspec-
tive of structural and functional integrity. The impact responses
of cable are discussed in detail, including sectional deformation
and impact force. The proposed experimental and numerical meth-
ods show a good potential for ascertaining the impact capacity for
submarine cables and for cable protection design, such as deter-
mining the proper impact energy of rock placement. The methods
can also provide some references for determination of allowable
impact limits for subsea power cables or umbilicals.
2. Experimental investigation
2.1. Test setup and procedure
The test setup is designed to study the structural deformation
and resulting performance degradation of photoelectric compos-
ite cable impacted by dropped anchors. As shown in Fig. 2, the
test setup mainly consists of a portal carne, electric release device,
admiralty anchor, test specimen cable. Other auxiliary tools are
44 Q. Gao et al. / Applied Ocean Research 73 (2018) 42–58
Fig. 3. Section of composite cable.
measuring tape, cable cutting saw and vernier caliper, etc. Inno-
vatively, the electric release device is altered from linear actuator,
connected through wire to a reversible switch. It’s very simple and
convenient to operate and release the anchor in the air during
impact tests.
Though stock anchors are tedious to handle, and abandoned
by large ships, they are still used due to their remarkable holding
power in seabed [21]. In China coastal waters where cable route
passes through, small crafts and fishing vessels are dominant in
ships quantity, posing the greatest threat. Taking Hangzhou Bay for
instance, 50% of the small-type vessels are equipped with anchors
less than 100 kg, most of which are admiralty anchors [22]. Thus
stock anchor of 100 kg is considered in this investigation. It is a
standard product [23], with bottom edge radius of 100 mm. The
anchor weight is sufficient to examine the cable impact capacity
and its structural impact failure mechanism by adjusting the falling
height. The experimental and numerical methods proposed in this
article can be extended for the impact investigation of other anchor
types on submarine cables.
The test procedures are as following. Firstly, the cable is pulled
out from the reel, and put on the steel plate. The anchor is lifted
to a certain height by portal crane through the electric release
device, which is slung by the hoist hook. The impact points are
calibrated by vertical line. The bottom of anchor will impact the
top longitudinal axis of the cable. After impact, the indentation of
armor layer will be measured with vernier caliper and ruler. The
insulation resistance of each power cable is measured by megohm-
meter. Selectively, before the cable is cut into segments, the cable
would again be wound on the reel and be sent to the high voltage
laboratory for alternating current (AC) voltage withstanding test
and partial discharge test. Since insulation defects often give rise
to partial discharge (PD), PD measurement is a powerful tool for
identifying and evaluating insulation defects [24].
After that, the cable is cut and disassembled. The visual inspec-
tion is undertaken on various layers of the cable. The deformations
of internal cable entities, i.e. optical unit and power units are
measured with vernier caliper. The minimum and maximum thick-
nesses of XLPE insulation are measured with digital measuring
projector. The purpose of electrical tests is to confirm the functional
integrity of power units. The transmission loss in optical fiber is not
measured and will be studied for later work.
2.2. Specification of cable
In this work, the submarine photoelectric composite cable pos-
sesses three power cores and an optical fiber core. The voltage of
cable is 12/20 kV, with three power cores and an optical fiber unit.
It’s a medium voltage alternate current cable. The construction pro-
file of the cable is shown in Fig. 3. A similar real photo of cable cross
section is given for instance. The interstices of multi-core subma-
rine power cables contain polypropylene filler materials. The armor
is covered by a sheath of polypropylene threads with bitumen.
2.3. Test results and analysis
The consequence evaluations of anchor impact include struc-
tural and functional integrity. The former contains armor dent
depth (indentation) ı, the insulation minimum and maximum
thickness tmin , tmax and insulation eccentricity eP , and deforma-
tion rate of optical unit eO . The latter is electrical performance
 Q. Gao et al. / Applied Ocean Research 73 (2018) 42–58 45

Table 1
Test results of impact direction A.

Damage description
Falling height h Impact velocity v Impact energy E Armor dent depth
(m) (m/s) (kJ) ı(mm) Deformation of power unit Deformation rate of
optical unit eO (%)
tmin (mm) Eccentricity eP (%)

1 4.427 0.98 10.12 6.43

1.5 5.422 1.47 12.78 12.50
2 6.621 1.96 14.82 4.13 18.2 17.96
2.5 7.0 2.45 17.36 3.43 31.4a 22.51
3 7.668 2.94 18.68 25.61
3.5 8.283 3.43 21.82 33.15
4 8.854 3.92 23.44 35.45

Note: the specified minimum thickness of insulation is 4.85 mm; the specified insulation eccentricity is not beyond 15%.
a
Partial discharge is 30 pC.

Test-eO
Test point

0.62
y = 12.0 x
2
R = 0.99136

Fig. 4. Armor indentation rate vs. impact energy. Fig. 5. Deformation rate of optical unit vs. armor indentation rate.

Insulation eccentricity brings electrical field distortion. The elec-

assessment, including insulation resistance, voltage withstanding
capacity and partial discharge; see Appendix B. Test results are
listed in Table 1.
Impact energy is defined as following: E = mgh, where E, m, h
are impact energy (kJ), the mass of anchor (kg) and falling height
(m). The anchor falls in air for convenient change of impact energy.
The air friction on anchor is neglected. According
 to the formula of
free falling body, the impact velocity is v = 2gh. As was defined
by Nishimoto et al. [20], the deformation rate was presented as
the amount of deformation/outside diameter of original cable as
percent. Fig. 4 shows the relationship of armor indentation rate eA
and impact energy in tests. The optical unit deformation rate of
optical unit is related to the armor indentation rate, as shown in
Fig. 5.
IEC 60502-2 [25] requires that, for each power core, the smallest
value of XLPE insulation thickness measured shall not fall below
90% of the nominal value by more than 0.1, i.e. tmin ≥ 0.9 tn − 0.1.
The insulation eccentricity is (tmax − tmin )/tmax ≤ 0.15, where tn is
the nominal thickness. In the target cable, the nominal thickness of
XLPE insulation is 5.5 mm, so the minimum thickness is calculated
to be 4.85 mm. Please see Appendix A for the deformation of various
components.
The insulation resistances between conductor and lead sheath
are measured to be 5000 megohms. The megohmmeter demon-
strates “+∞”, which means open circuit. The impact tests didn’t
give rise to insulation cracking. But the insulation eccentricity and
minimum thickness can’t meet the requirement of the specification
IEC 60502-2 [25].
tric field intensity will be higher in thin insulation point. It will
affect the safe operation and service life of cable. More seriously,
local electric field intensity concentrates and leads to insulation
breakdown, resulting in power delivery stop. Therefore, anchor
impact will expose the power cable to a larger risk of breakdown.
3. Numerical analysis
Experiments have advantages for attaining some results
directly, such as permanent indentation of armor and cable sec-
tional deformation. However, the maximum deformation and
impact force are not easy to be measured in experiments. Finite
element analysis is capable of achieving more impact consequence
details and dealing with different impact scenarios. Firstly, the
finite element model developed using common finite element soft-
ware ABAQUS/CAE (version 6.13) is simplified for computation
efficiency. The ABAQUS/Explicit solver is used for solving the non-
linear dynamic problem. And then, the FEM is verified by test in the
case of impact direction A. Last, a parametric analysis is performed,
considering different impact directions and impact velocities.
3.1. Simplification of model
The conductor shield, insulation shield, semi-conductive
waterblocking layer and anti-corrosion layer are neglected for their
quite small thicknesses and slim strength. The contribution of opti-
cal fibres to the cable strength is negligible. The copper conductor
strands are treated as a complete one. The optical unit is reduced
to a three layer structure, i.e. the central polymeric core, thin steel
46 Q. Gao et al. / Applied Ocean Research 73 (2018) 42–58

Table 2
Dimensions of 20 kV submarine composite cable.

Components Simplified model

Wall thickness Outer diameter

(mm) (mm)
Copper conductor – 9.9
XLPE insulation 5.5 20.9
Alloy lead sheath 1.5 23.9
PE sheath 2 27.9
Polypropylene filler and tape – 60.8
Bedding layer 3 66.8
Galvanized steel wire armor 5.0 76.8
Outer sheath 3 82.8
Optical cable filler 8.6
Optical unit Steel–plastic tape 0.2 9
Lead sheath and PE sheath 2.3 13.6

in direction A for example, corresponding to 4 m falling height, the

Fig. 6. Three-dimensional model.
tube and the PE sheath. Fig. 6 shows the three-dimensional cable
model. The anchor has much larger rigidity than impacted cable
zone. Thus, the deformation of anchor is neglected. It is idealized
as rigid body which has the identical bottom shape with real anchor.
This measure aims to simplify the model and reduce the element
quantity and the computational cost, as other researchers did in
lateral crushing of umbilicals by Guttner et al. [17] and transverse
impact on pipelines by Zheng et al. [26]. The stock part of anchor
is not modelled because they do not affect the collision simulation.
However, the anchor weight is retained, giving the same kinetic
energies [21].
As we know, there are possibly various seabed soil types on a
cable route, such as mud, soft or firm clay, loose or dense sand,
gravel and rock and so on. If the cable is lying on a rigid seabed like
rock, the anchor collision will induce the most severe damage to
cable. Thus, the steel plate support for cable in tests is idealized as
a rigid plate in ABAQUS to improve computational efficiency. More-
over, if the test is performed on stiff supports, it will reflect the true
capacity of cable, as elaborated in DNVGL RP F107 [27]. Moreover,
the models can be modified to simulate other support conditions
if the subsoil mechanical properties are known. For example, Zein-
oddini et al. [28] studied the response of submarine pipelines to
impacts from dropped objects, considering bed flexibility effects.
The dimensions of 20 kV AC submarine composite cable are
listed in Table 2.
Considering the symmetry of the model, there are four typical
impact directions, A, B, C and D, as shown in Fig. 7. In order to
optimize the finite element model for computation efficiency and
accuracy, four finite element cable models of different lengths and
mesh sizes are analysed, as illustrated in Fig. 8. When the falling
height reaches 3 m, the armor indentation of four different direc-
tions are nearly identical. Then take collision velocity of 8.854 m/s
permanent indentation of armor is 26.6 mm. The deviation from
test is 13.6%. Moreover, the longitudinal length of impact inden-
tation is about 5.5 times the anchor bottom width, as shown in
Fig. 9. All the four models provide consistent deformation results.
The shortest model is the most efficient in computation. Therefore,
the 400 mm length model is finally adopted for parametric analysis.
3.2. Materials, element types and mesh
There are various metallic and polymeric materials. Compo-
nents are here considered with a linear elastic or elastic-plastic
behavior [29]. The polypropylene outer sheath and filler are lin-
ear elastic. The material nonlinearity due to plastic deformation is
considered. The alloy lead sheath, PE sheath and XLPE insulation
are treated as ideal elastic-plastic model, as done by Opgård [30].
The Ramberg-Osgood description of material deformation charac-
teristics is employed for the copper conductors and steel armor
wire, as given in Fig. 10. For the two materials, isotropic plastic
hardening and von Mises yield criterion are adopted. Because the
armor deforms in quite short time by the order of millisecond under
impact, the Cowper-Symonds relation [31] is adopted for the steel
armor to model the strain rate effect, which is given as follows:
0d
 ε̇ 1/n
=1+
0 D
Where 0d is the dynamic flow stress;  0 is the static flow stress;
D (= 40) and n (= 5) are constants in the strain rate hardening law.
The constants D and n were determined on the basis of test data
by Cowper and Symonds [32]. They are used to calculate the yield
strength of armor steel under different strain rates.
As for the meshing, outer serving, inner bedding and HDPE
sheath have only one element through the thickness. For XLPE insu-
lation, optical fiber core steel tube and armour wires, which are
components of greater interest, a finer mesh is defined to enable

Fig. 7. Four different impact directions.

 Q. Gao et al. / Applied Ocean Research 73 (2018) 42–58
Fig. 8. Model optimisation.
Fig. 9. Armor denting.
Steel
Copper
Fig. 10. Material properties of armor and conductor.
a more detailed examination of the results. A mesh refinement
analysis is carried out in order to ensure the convergence of the
results according to element size. There are 187,176 elements of
type C3D8R, and 286,336 nodes. The element type C3D8R denotes
three- dimensional continuum elements with eight nodes, linear
interpolation functions and reduced integration. This element type
can avoid shear locking problem.
3.3. Contact and boundary conditions
Interactions between all layers are taken into consideration,
including lateral contacts between adjacent steel wires. Abaqus
general contact algorithm is employed. This algorithm automati-
cally detects all possible contact pairs and calculates the contact
47
stiffness of its penalty-method, greatly simplifying the develop-
ment of the model [33]. The normal behavior of contact pairs is
hard contact, not allowing the mutual penetration of contact sur-
faces. The tangential behavior of contact pairs is represented by a
penalty friction formulation assuming a coefficient of friction of 0.2,
as obtained by Olsen et al. [34] and Nasution et al. [35] from friction
tests. An exception is made that surface-to-surface contact is used
between lead sheath and XLPE with penalty contact method. The
friction coefficient is set as 0.7, as what has been done in Foti and
Martinelli [36]. It is considered that the lead sheath is extruded on
the rough semi-conductive waterblocking layer, which wraps XLPE
insulation tightly.
In the ABAQUS Interaction module, two points are created as
reference points (RP) to apply rigid body constraints to anchor and
steel plate respectively. Motion or constraints that we apply to the
reference point are then applied to the entire rigid part, namely
anchor and steel plate. In the ABAQUS Load module, the predefined
velocity field is applied at the anchor reference point RP-1, imply-
ing that the whole anchor has the same velocity. The fully fixed
boundary conditions are applied to the steel plate reference point
RP-2, while the anchor reference point RP-1 is free, consistent with
actual tests. The acceleration of gravity is applied on the whole
anchor, see yellow arrows in Fig. 11. Herein, the effect of gravity is
taken into account, as what has been done by Woo et al. [4,21]. In
tests, the cable ends far from the impact point are fixed by U shape
clamp. Therefore, in FEM, both ends of cable are pinned, leaving
three rotational freedoms, as depicted in Fig. 11.
4. Comparison and discussions
The maximum and permanent indentations of armor layer and
maximum impact force, at different impact scenarios, as are listed
in Table 3. The minimum and maximum thicknesses of XLPE insu-
48 Q. Gao et al. / Applied Ocean Research 73 (2018) 42–58

Table 3
Indentation and impact force.

Impact points
Drop height h (m)
A B C D

1 20.79 21.18 22.74 22.22

Maximum Indentation
2 29.49 29.55 29.43 29.03
(mm)
3 33.85 34.05 33.74 33.61

1 11.91 9.11 8.29 8.74

Permanent Indentation
2 18.81 18.34 16.24 16.71
(mm)
3 22.84 23.73 21.14 21.96

1 109.79 103.42 104.10 101.50

Maximum impact force
2 187.32 166.90 174.54 173.04
(kN)
3 221.90 204.82 226.15 223.27

Table 4
Power units and optical unit deformation.

Impact points
h (m)
A B C D

tmin tmax eP tmin tmax eP tmin tmax eP tmin tmax eP

Power units deformation 1 5.46 5.56 1.8 5.41 5.59 3.2 5.48 5.7 3.9 5.37 5.58 3.8
2 4.60 5.67 18.9 4.5 5.61 19.5 4.38 5.58 21.5 4.48 5.57 19.6
3 3.30 5.81 43.2 3.21 6.18 48.1 2.7 55.4 66.3 3.41 5.71 40.3

1 9.7 1.5 2.2 1.5

Optical unit deformation
2 26.4 11.8 9.6 12.5
rate (%)
3 34.5 23.5 18.4 17.6

4.1. Comparison between test and FEM results

As shown in Fig. 12, the cable is indented by falling anchor,

Fig. 11. Loads and boundary conditions.
lation, eccentricity of power core insulation and deformation rate
of optical unit are listed in Table 4. The above data are noteworthy
and of interest for engineers in practical engineering. The defor-
mation of structural components will cause some loss of functional
integrity, e.g. power and signal transmission.
at collision velocity of 7.668 m/s (falling height of 3 m). The steel
wire armor with black bitumen-impregated is curved. To evalu-
ate the armor layer deformation, the indentation ␦ is defined as
the difference between deformed and undeformed diameters, as
depicted in Fig. 13. The indentation of armor layer is also called
dent depth or the crushing distance. The cable is squashed verti-
cally, and expanded horizontally. The permanent indentations of
3 m falling height in test and FEM are 18.68 mm and 22.84 mm,
respectively, differing little.
Fig. 14 presents the XLPE insulation shape before and after defor-
mation in test and simulation, at falling height of 2 m and 2.5 m,
respectively. It is obvious that the insulation becomes approxi-
mately elliptical, because of crush induced by impact.
As shown in Fig. 15, the test and FEM results in general represent
almost the same growth trend with falling heights. FEM result are
larger than tests, because the model simplification and neglect of
steel plate’s vibration. Factually, the steel plate which the cable is
lying on should vibrate under impact. This part of energy is instead
applied on the cable, because the steel plate is treated as rigid body
in FEM. The permanent indentation of armor layer increases with
falling height, obviously with increase tendency slowing down. It

Fig. 12. Armor indentation at collision velocity of 7.668 m/s.

 Q. Gao et al. / Applied Ocean Research 73 (2018) 42–58 49

Fig. 13. Armor indentation ␦ (mm).

Fig. 14. XLPE insulation deformation.

Test-eO
Test--A
FEM--A FEM-eO

0.57
y = 12.17 x
2
R = 0.99237

0.62
y = 9.82 x
2
R = 0.99136

Fig. 15. Armor indentation of tests and FEM. Fig. 16. Deformation rate of optical unit.

is explained in Fig. 18 for the slowing down increase tendency.

From the initial collision contact, the indentation grew with time
to its maximum, until the anchor stopped. The anchor began to
bounce back under the action of contact reaction force. And the
armor layer also partially restored and vibrated with decreasing
magnitude, because of energy dissipation by interface friction, leav-
ing some permanent indentation. As falling heights increase, the
differences between maximum indentations diminished gradually.
And the cable sooner ceased vibrating, especially at higher collision
speed. Fig. 16 presents the comparison of optical unit deformation
rates of tests and simulations, the latter being slightly higher.
50 Q. Gao et al. / Applied Ocean Research 73 (2018) 42–58
Fig. 17. Cross sectional deformation at four stages.
4.2. Section deformation and impact force
Similar with the impact deformation of flexible pipe [18], a
sequential illustration describing the cable cross sectional behavior
is presented in Fig. 17, the drop height being 2 m. At the moment
t = 0.8 ms (milliseconds), the anchor had already crushed the armor
wires, the top wires contacting the anchor are flattened more
obviously than those at the bottom. At t = 7.8 ms, the indentation
reached its maximum. The section profile of top armor layer coin-
cides with the arc anchor bottom. The cross sectional deformations
of structural components are similar with that by Nishimoto et al.
[20] in static compression tests.
As shown in Fig. 19, the impact force increases quickly as
the falling heights grow. The impact duration is about 10–15 ms,
A-1m
A-1.5m
A-2m
A-2.5m
A-3m
A-3.5m
A-4m
Fig. 18. Armor indentation time history.
becoming shorter as the collision velocity that’s corresponding to
falling height increases. The peak of impact force can reach 274 kN
at 4 m falling height. The relationship of impact force and indenta-
tion is collaborated in Fig. 20. The loading and unloading curves
of impact force share similar variation tendency with different
heights. The cable indentation consists of elastic and permanent
parts. DNVGL RP F111 [37] illustrates that the kinetic energy of
anchor may partly be dissipated during the impact by elastic defor-
mation and possible plastic denting. The elastic deformation will
reverse along with the anchor bouncing back. The areas under the
loading and unloading curves are the work done by the impact force
on the cable.
A-1m
A-2m
A-3m
A-4m
Fig. 19. Impact force time history.
 Q. Gao et al. / Applied Ocean Research 73 (2018) 42–58 51

4.3. Influence of impact directions

The indentation versus time relations of different heights and

A-1m four impact directions are presented in Fig. 21. As shown in Fig. 22,
A-2m the variation tendency of indentation with time is quite similar.
A-3m At falling height of 1 m, the permanent indentations of direction B,
C and D are very close; the maximum difference is 0.37 mm. For
2.0 m height, the difference of maximum indentation is very small,
and the residual deformations came to closer. For height of 3.0 m,
the indentation time history curves almost coincide before the first
vibration magnitude. It can be concluded that the armor layer is the
most important structural component to withstand anchor impact.
The cushioning effect of internal cable entities, such as polymeric
filler material, power units and optical fiber unit is obvious under
lower collision velocities, for example, not beyond 4.427 m/s.
As shown in Fig. 23, the cable sectional deformations of direc-
tion A and D are nearly symmetric about the section symmetry axis
parallel to global Y-axis, while the cable sectional deformations of
direction B and C are not symmetric. The damage to optical unit
Fig. 20. Impact force vs. indentation.
of direction A is worst. The damage of optical unit of direction B is

Fig. 21. Indentation time history of directions A, B, C, D at different falling heights.

52 Q. Gao et al. / Applied Ocean Research 73 (2018) 42–58

Fig. 22. Comparison of indentation time history.

 Q. Gao et al. / Applied Ocean Research 73 (2018) 42–58
Fig. 23. Cable section deformation at collision velocity of 4.427 m/s (falling height of 1 m).
minor. Optical core deformations of C and D are quite slight com-
pared with A and B. At height of 1 m, anchor impact won’t cause
apparent change of XLPE insulation shape under four directions.
And that power units are not crushed heavily because of polymeric
fillers with smaller hardness than power unit materials.
The curves of impact force versus time are depicted in Fig. 24.
From Fig. 24(a)–(c), it’s interesting to note that the curves become
nearly coincident. The impact durations under different impact
directions become shorter as falling heights increase. Meanwhile,
the influences of impact directions on impact duration and peak
value of impact force decrease. The peak value of impact force is
largest at point A among the four impact directions.
4.4. Deformation of power and optical units
Fig. 25(a) shows that the minimum thickness of insulation
tmin lessens with falling height h increasing. It is obviously affected
by falling height, and is below the required minimum thickness of
4.85 mm at 1.6 m height. Here, the impact energy is 1.568 kJ. The
eccentricity of power units eP is also related with falling heighth,
as shown in Fig. 25(b). It is beyond the specification requirement of
15%, when the falling height is over 1.6 m. The optical unit defor-
mation rate eO is worst at impact point A, as shown in Fig. 25(c). It
means that anchor impact right above the optical unit leads to its
severe deformation. The relationships between deformation rate of
optical unit eO and armor indentation rate eA at different directions
(A, B, C and D) and heights are presented in Fig. 25(d). It pro-
vides important applicable significance for optical unit deformation
assessment from the armor indentation rate.
53
The test and finite element results mentioned above demon-
strate that the armor layer is the most important structural
component of cable to withstand anchor impact. The internal enti-
ties are manufactured with polymeric materials and lead and so
on, possessing much smaller elastic moduli than steel armor wire.
They are vulnerable to crush by anchor impact.
5. Conclusions
This study focused on the damage assessment methods of sub-
marine photoelectric composite cable under anchor impact. A test
arrangement is designed for structural and functional integrity
examination of cable under different impact velocities, corre-
sponding to certain falling heights. A finite element model is
created, considering material plastic hardening and strain rate
and interaction between components. The damage and perma-
nent deformation of various components are measured in tests and
calculated in FEM. The model is well refined and presented good
accordance with test results, implying its accuracy and reliability.
As it should be expected, the armor layer is the major com-
ponent to withstand anchor impact. But it can only bear smaller
impact energy. The eccentricity of power cables and deformation
rate of optical unit have positive correlation with armor indentation
rate. The damage of internal entities can be inferred from the armor
indentation rate when it is not convenient to disassemble the cable
for visual inspection. The eccentricity of power units is beyond the
specification requirement of 15%, when the falling height exceeds
1.6 m. The minimum thickness of insulation is less than the required
minimum thickness of 4.85 mm.
54 Q. Gao et al. / Applied Ocean Research 73 (2018) 42–58

A-1m
B-1m
C-1m
D-1m

(a) Falling height of 1 m

A-2m
B-2m
C-2m
D-2m

0.005 0.010 0.015 0.020 0.025

Time (s)

(b) Falling height of 2 m

A-3m
B-3m
C-3m
D-3m

Fig. 24. Comparison of impact force time history.

 Q. Gao et al. / Applied Ocean Research 73 (2018) 42–58
Fig. 25. Deformation of power cables and optical unit.
As a final conclusion, the submarine cable, as a kind of helically
wound composite construction, is vulnerable to lateral impact by
dropped objects. Even if the insulation resistance and high volt-
age test are qualified, partial discharge exceeds requirement. It’s
a very big potential danger. So some external protection measures
should be taken in practical engineering to avoid severe mechanical
impact. The test method and finite element model proposed in this
paper is also of use in industry to assess mechanical impact capac-
ity of umbilicals and subsea power cable. The methods provide
some references for the design of rock placement for cable protec-
tion, especially in determining the proper rock impact energy. It is
recommended that the signal transmission performance of optical
fiber unit be studied in future work and the influence of seawater
on functional integrity be considered.
Acknowledgments
The authors gratefully acknowledge the financial support pro-
vided by the China National Key Research and Development Plan
55
[Grant No. 2016YFC0303704], National Natural Science Foundation
of China (Grant No. 51509258) and the CNOOC Research Institute.
Appendix A. Electrical Test
Electrical tests are inclusive of partial discharge test and
alternating current (AC) voltage withstanding test, to check the
functional integrity. According to IEC 60502-2 [25], there shall be
no detectable discharge exceeding the declared sensitivity from the
test object at 1.73U0 (U0 = 20 kV). The partial discharge test shall
be carried out in accordance with IEC 60885-3 [38], the sensitivity
being 10 pC (picocoulombs) or better. The detectable discharge of
red power core is 30 pC, beyond the defined sensitivity 10 pC. The
other two power cores are below 10 pC. It means that the anchor
impacts degrade the insulation performance, leaving a failure haz-
ard (Fig. A1).
The power cable should conform to the specification require-
ment of IEC 60502-2 [25]. The insulation layers would not break
down under the test voltage of 3.5U0 within five minutes. No break-
down of the insulation shall occur.
56 Q. Gao et al. / Applied Ocean Research 73 (2018) 42–58

Fig. A1. Electrical test system.

Appendix B. Visual Inspection

After anchor impact, the indentation of armor layer is measured.

Then the impacted cable sample is cut off and disassembled. The
visual inspection is taken on the various components of the cable, to
check if there are any signs of damage or permanent deformations.
The indentation of optical unit is measured with vernier caliper.
A thin slice is cut from the severely deformed power cable. The
maximum and minimum thickness of XLPE insulation slice would
be measured with digital measuring projector. The two values are
used to calculate the eccentricity of insulation (Figs. B1–B9).

Fig. B4. Four cable cores.

Fig. B1. Cable sample after cutting.

Fig. B5. Dented power cables and flattened optical unit.

Fig. B2. Bedding with armor layer removed.

Fig. B6. Lead alloy sheath with dents and wrinkles, semi-conducting swelling tape
ruptured.

Fig. B3. Measure deformation of optical unit.
The anchor impact endangered the power cores and optical unit.
Some signs of damage and permanent deformations are observed.
Except ovalisation of cable internal cores, the steel armor wire left
dents on HDPE sheath of the power cable cores. The lead alloy
sheath is also wrinkled and dented. The semi-conducting swelling
 Q. Gao et al. / Applied Ocean Research 73 (2018) 42–58 57

[3] P.G. Allan, Selecting appropriate cable burial depths ?a methodology, in: A
Methodology IBC Conference on Submarine Communication, The Future of
Network Infrastructure, Cannes, France. November, 1998, pp. 1–12.
[4] J. Woo, D. Kim, W.B. Na, Damage assessment of a tunnel-type structure to
protect submarine power cables during anchor collisions, Mar. Struct. 44
(2015) 19–42.
[5] J.R. Attwood, Cable design for subsea power links, IEEE Power Eng. Rev. 20 (9)
(2000) 13–14.
[6] T. Worzyk, Submarine Power Cables: Design, Installation, Repair,
Environmental Aspects, Springer-Verlag Berlin Heidelberg, Berlin, 2009.
[7] H.S. Yoon, W.B. Na, Safety assessment of submarine power cable protectors by
anchor dragging field tests, Ocean Eng. 65 (2013) 1–9.
[8] P. Allan, R. Comrie, Risk assessment methodology and optimisation of cable
protection for existing and future projects, in: SubOptic 2004, Monaco, 2004,
March.
[9] M. Nakamura, N. Nanayakkara, H. Hatazaki, K. Tsuji, Reliability analysis of
Fig. B7. Dented XLPE insulation. submarine power cables and determination of external mechanical
protections, IEEE Trans. Power. Deliv. 7 (2) (1992) 895–902.
[10] T. Guo, Study on the risk of submarine cable anchor damage caused by ship,
in: Master Dissertation, Project Management, Tianjin University, China, 2013,
pp. 37 (in Chinese) June.
[11] DNV-RP-J301, Subsea Power Cables in Shallow Water Renewable Energy
Applications, 2014, February.
[12] ITU-T G.976, Test Methods Applicable to Optical Fibre Submarine Cable
Systems, 2000, October.
[13] API SPEC 17E, Specification for Subsea Umbilicals, 2010, October.
[14] Electra 171, Recommendations for Mechanical Tests on Submarine Cables,
1997, April.
[15] Electra 189, Recommendations for Testing of Submarine Cables, 2000, April.
[16] X. Jia, Q. Gao, L. Wang, M. Duan, T. Zhang, Field test research on failure of
photoelectric composite submarine cable under impaction by dropped
anchors, in: The 2015 World Congress on Advances in Structural Engineering
and Mechanics (ASEM 15), Incheon, Korea, August 25–29, 2015.
[17] W.C. Guttner, C.C. Santos, C.P. Pesce, A finite element method assessment of a
Steel Tube Umbilical (STU) cable subjected to crushing load: comparison
between two and three-dimensional approaches, Mar Struct 53 (2017)
52–67.
[18] H.S. Alsos, B. Laksáfoss, P.D. Rasmussen, C. Kristensen, G. Paulsen, Finite
Fig. B8. Insulation measurement on digital projector. element analysis of flexible riser impact, ASME 2010 29th International
Conference on Ocean, Offshore and Arctic Engineering. American Society of
Mechanical Engineers (2014), Paper No. OMAE 2010–20777.
[19] H. Tayama, O. Fukuda, Y. Inoue, Y. Koike, K. Yamamoto, 6.6 kV XLPE submarine
cable with optical fiber sensors to detect anchor damage and defacement of
wire armor, IEEE Trans. Power Deliv. 10 (4) (1995) 1718–1723.
[20] T. Nishimoto, T. Miyahara, H. Takehana, F. Tateno, Development of 66 kV XLPE
submarine cable using optical fiber as a
mechanical-damage-detection-sensor, IEEE Trans. Power Deliv. 10 (4) (1995)
1711–1717.
[21] J. Woo, D. Kim, W.B. Na, Safety analysis of rock berms that protect submarine
power cables in the event of an anchor collision, Ocean Eng. 107 (2015)
204–211.
[22] Y. Gu, Investigations on the safety of submarine cables, in: The 2003
Academic Seminar of the Salvage Committee of China Navigation Institute,
October 1, 2003. Shanghai, China, 2018, in Chinese.
[23] Chinese specification GB/T545-1996, Admiralty Anchor, 1996, November.
[24] A. Cavallini, G.C. Montanari, F. Puletti, A. Contin, A new methodology for the
identification of PD in electrical apparatus: properties and applications, IEEE
Trans Dielectr Electr Insul 12 (2) (2005) 203–215.
[25] IEC 60502-2-2014. Power cables with extruded insulation and their
accessories for rated voltages from 1 kV (Um = 1.2 kV) up to 30 kV (Um = 36
kV) − Part 2: Cables for rated voltages from 6 kV (Um = 7.2 kV) up to 30 kV
(Um = 36 kV). 2014, February.
[26] J. Zheng, A. Palmer, P. Brunning, C. Gan, Indentation and external pressure on
subsea single wall pipe and pipe-in-pipe, Ocean Eng 83 (2014) 125–132.
[27] DNVGL-RP-F107, Risk Assessment of Pipeline Protection, 2017, May.
[28] M. Zeinoddini, H. Arabzadeh, M. Ezzati, G.A.R. Parke, Response of submarine
pipelines to impacts from dropped objects: bed flexibility effects, Int. J.
Impact Eng. 62 (4) (2013) 129–141.
[29] G. Feld, D.G. Owen, R.L. Reuben, A.E. Crockett, Mechanical behaviour of the
metallic elements of submarine cables as a function of cable loading, Eng.
Struct. 17 (4) (1992) 240–253.
[30] M.F. Opgård, Torsion instability of dynamic cables during installation, in:
Fig. B9. Deformed optical unit with dents.
Master Dissertation, Marine Technology, Norwegian University of Science and
Technology, 2017, June, 54 pp.
[31] J. Travanca, H. Hao, Numerical analysis of steel tubular member response to
tape is ruptured. The XLPE insulation is dented. The optical unit also ship bow impacts, Int. J. Impact Eng. 64 (2014) 101–121.
had dents by armor wire under impact. [32] G.R. Cowper, P.S. Symonds, Strain-hardening and strain-rate effects in the
impact loading of cantilever beams, in: Technical Report No. 28 from Brown
University to the Office of Naval Research, 1957.
References [33] Abaqus Users’ Manual, version 6.13. Providence (RI): Dassault Systémes
Simulia Corp. 2013.
[1] R. Perveen, N. Kishor, S.R. Mohanty, Off-shore wind farm development: [34] E. Olsen, S. Karlsen, L. Jordal, K.A. Hansen-Zahl, Determination of friction
present status and challenges, Renew. Sustain. Energy Rev. 29 (7) (2014) within a riser umbilical, The ASME 2012 31 st International Conference on
780–792. Ocean, Offshore and Arctic Engineering. American Society of Mechanical
[2] J. Pyrah, Cable installation and burial: practical considerations, Mar. Technol. Engineers (2012), Paper No. OMAE2012–83089.
Soc. J. 44 (1) (2010) 52–56.
58 Q. Gao et al. / Applied Ocean Research 73 (2018) 42–58

[35] F.P. Nasution, S. Sævik, S. Berge, Experimental and finite element analysis of
fatigue strength for 300 mm2 copper power conductor, Mar. Struct. 39 (2014)
225–254.
[36] F. Foti, L. Martinelli, Mechanical modeling of metallic strands subjected to
tension, torsion and bending, Int. J. Solids Struct. 91 (2016) 1–17.
[37] DNVGL-RP-F111, Interference Between Trawl Gear and Pipelines 2017, 2017,
May.
[38] IEC 60885-3:2015, Electrical Test Methods for Electric Cables −Part 3: Test
Methods for Partial Discharge Measurements on Lengths of Extrudedpower
Cables, 2015, May.