You are on page 1of 8

Collision and Grounding of Ships and Offshore Structures – Amdahl, Ehlers & Leira (Eds)

© 2013 Taylor & Francis Group, London, ISBN 978-1-138-00059-9

Ship-ice collision analysis to define ice model according to


the IACS Polar Rule

M.J. Kwak, J.H. Choi, O.J. Hwang & Y.T. Oh


Daewoo Shipbuilding & Marine Engineering Co., LTD, Korea

ABSTRACT: Nowadays, the exploration and development of Arctic region are getting more active than ever
before due to the increasing demands for oil and gas supplies. There is a need for large crude oil tankers to
transport the oil that will be produced from these far northern locations. For the safety of hull structure, the
interaction between ship and ice is a critical issue to be settled for the safe sailing in Arctic region. However
the behaviors of a sea ice are very complicated to define reliable mechanical model since there are various
characteristics corresponding to its age and temperature. Recently, unified the IACS Polar Rule was released
to provide standard design guidance for Arctic Vessel. Despite that, many research activities are performed to
define ice mechanical properties for reliable ship-ice collision analysis. For reliable ship-ice collision analysis,
the definition of mechanical properties of ice is very important. In this study, ship-ice collision analyses were
performed to find the proper ice model which can describe ice failure feature such as crushing and flexural
phenomenon and resultant contact force complying with design ice force of IACS Polar Rule.

1 INTRODUCTION

Vast reservoirs of oil and gas are expected to be


exploited in the Russian Arctic region including the
Barents, the Pechora and the Kara Sea. There is a need
for large crude oil tankers to transport the oil that will
be produced from these far northern locations. There-
fore, the interests for arctic vessel have been increased
recently. However, sailing in the Arctic area is more
difficult than in the mild climate area since there are
many things to overcome, such as ice load and cold
temperatures.
Many engineering activities had been carried out
to predict reliable ice load based on clear defini-
tion of the ice features for Arctic vessels. But they Figure 1. Analysis methods for determination of ice model.
reached slightly different conclusion on design method
because they do not have a standard rule. Finally, uni-
fied the IACS Polar Rule was released to provide for structural assessment of 107k DSME STD Arctic
standard design guidance. In the IACS Polar Rule, the Tanker considering the ice collision was suggested by
ice load that links with collision scenario of glancing the ultimate limit sate and in addition, by some numer-
ice is defined. ical simulation in time domain. However, it was found
Many ship owners are worrying about hull safety that there still remains some difference from IACS
due to ship-ice collision as accidental event and requir- Polar Rule.
ing reliable assessment of ship-ice collision analysis In this paper, works achieved form previous stud-
considering dynamic effect although Arctic Rules pro- ies were more highlighted. In performing the collision
vide the design methods. For reliable ship-ice collision analysis to find the ice mechanical properties based on
analysis, the definition of ice mechanical properties is IACS Polar Rule, two simulations were conducted as
very important. In IACS Polar Rule, the crushing and shown in Figure 1.
bending modes of ice failure are considered to simplify In 1st simulation, rigid hull structure & ice colli-
the failure phenomenon of sea ice. sion analysis was performed to find ice model which
DSME conducted some engineer activities to can describe ice failure feature such as crushing and
understand the IACS Polar Rule in the view point of flexural phenomenon and resultant contact force com-
hull structure (Kwak et al. 2009) in which the method plying with design ice force of IACS Polar Rule since

205
Figure 2. Design scenario of IACS Polar Rule.

Figure 3. Design load cases.


hull deflection was ignored to calculate ice load in
IACS Polar Rule. After determined ice model, flexible
hull structure & ice collision analysis was addition- angles, ice strength and thickness determine the force
ally performed to evaluate response of ice and hull limit due to bending as shown in Eq. 2.
structure.
Before ship-ice collision simulation, non-linear
analysis was conducted to verify accuracy of hull
structure against ice load considering plastic limit
state. where hice = ice thickness m; σ f = ice flexural strength
MPa.
In this paper, ice model was defined to include
2 THE DEFINITION OF ICE MODEL above mentions.

The Polar Rule bases the ice loads on a specific design


scenario. The design scenario is a glancing collision on 3 STRENGTH VERIFICATION
the shoulders of the bow as shown in Figure 2. The ice
load is derived from the solution of an energy based on Before ship-ice collision analysis, accuracy of hull
collision model, where kinetic energy equation is equal structure was verified through the non-linear analysis.
to ice crushing energy. This means that hull structure is In the IACS Polar Rules, the design methods are
assumed as rigid structure. This theory was supported based on plastic capacity of the plate and stiffeners
by various studies, including numerical method, model along with a conservative design loads. For example,
test, ship trial and experiment data, Daley (2000). there should be two load cases considered for trans-
And the parameters, such as ice thickness, ship verse frame in the ice belt, which are symmetric load
speed, hull form, are considered based on class level, case and asymmetric load case to design the ship struc-
and these are described in the IACS Polar Rule. ture against ice load, Kendrick & Daley (2000), IACS
The rule scenario is strictly valid only for the bow (2007).
region, and for the stern of double-acting ships. In Symmetric loading states that ice load is applied at
order to produce a balanced structural design, loads mid-span and the second load case is reflecting that
on other hull areas are set as a proportion of the bow the load is concentrated at the end of frames. Figure 3
area by using empirical hull area factors (AF). The ice shows the cases of symmetric and asymmetric load
load is derived for an oblique collision on the bow. case.
The ice load model assumes a ‘Popov’ type of colli- The limit state in the IACS Polar Class Rule rep-
sion. The force is found by equating the normal kinetic resents a capacity comparable to that labeled “mecha-
energy (left side in Eq. (1)) with ice crushing energy nism 1” that called design limit state in Figure 4. Prior
(right side in Eq. (1)) as shown below: to mechanism 1, the load-deflection curve is essen-
tially linear and follows the slope of the original elastic
trend. Yielding occurs well before mechanism 1. This
is followed by the expansion of the plastic zone, during
which stress redistribution takes place. Further along
where δ = normal ice penetration; Fn = normal force; this curve, additional mechanisms can occur, including
Me = effective mass; and Vn = normal velocity. buckling and fracture as shown at mechanism 2.
The maximum force can be found by equating the This mechanism was validated through the compar-
normal kinetic energy with the energy used to crush ison study between experiment results and nonlinear
the ice. The ice crushing force cannot exceed the force analysis, Kendrick & Daley (2000), Daley et al. (2007),
required to fail the ice in bending. The combination of Kwak et al. (2009) .

206
Figure 4. Structural behavior mechanism.

Figure 6. Load & deflection curve for nonlinear analysis.

considering corrosion margin. The details of structure


members are given below:
Shell plate = 37 mm
Transverse frame = 500 × 35 mm
Longi.-stiffener = 1100 × 31 + 275 × 31 mm

For boundary condition, aft end of the FE model


Figure 5. FE model and boundary condition for nonlinear was constrained longitudinally and symmetric condi-
analysis. tion was applied to centerline elevation due to the half
model. Also, the connections of deck and main frame,
Table 1. Ice load Information. deck and shell plate were constrained vertically.
The nonlinear FE analysis was performed using
Patch Value Unit load patch in Table 1.
As a result of the analysis, load – deflection curve
Width of patch 4.328 m was obtained as shown in Figure 6. The ice patch load
Height of patch 0.675 m in Table 1 is found at around mechanism 1 described
Area 2.863 m2 in Figure 3. It means that the bow structure has enough
Force 19.200 MN strength under ice load defined in the IACS Polar Class
Patch pressure 6.720 Mpa
Rule.

4 SHIP-ICE COLLISION ANALYSIS

3.1 Nonlinear analysis Actually, the ship-ice collision analyses done before
DSME STD Arctic Tanker (107k) was selected as were conducted in DSME, Han et al. (2008) to evaluate
target vessel. The hull structure of target vessel was the hull safety and CCS of LNGc against the ice load
initially designed based on CSR rule for the open sea but these analyses were not based on IACS Polar Rule.
operation and the IACS Polar Rule (PC4) for the Arctic In this paper, two ship-ice collision analyses with
operation. In this paper, strength verification against level ice were performed using MSC.Dytran based on
ice load was only mentioned. IACS Polar Rule after strength verification of target
The ship’s main dimensions are shown below: vessel.
At first simulation, collision force obtained from
Length = 250 m the numerical simulation was compared with design
Breadth = 44 m ice load and properties of ice were iteratively adjusted
Draft = 14 m in order to derive proper mechanical property of ice
Block coefficient = 0.8 which can describe the ice failure feature and proper
ice load. Also, responses of hull structure and ice were
To carry out the strength analysis for bow struc- evaluated at second ship-ice collision analysis.
ture (Figure 5), the ice patch load was calculated
based on the IACS Polar Rule (PC4) and the ice load
4.1 Assumption
information is as shown in Table 1.
Transverse frames were added to reinforce the bow For these analyses, some concepts and assumption
area against ice load in initial design stage. The thick- were reflected to save effort and conduct the analysis
nesses of structures in FE model are net-scantling effectively.

207
Since time step size is usually controlled by the
smallest element size over all elements in the ship-
ice collision simulation using explicit code, such as
MSC.Dytran, the modeling of ship and level ice is
very important for stable computation. In these analy-
ses, the microscopic failure for ice was ignored since it
needs the very small size element that can require too
small time step. Actually, mesh size of ice will affect
the ice failure mode under ship-ice collision. However,
the simulation of microscopic failure is not critical
for the calculation of collision force and therefore
macroscopic ice failure mode was only applied.
Based on MSC handbook, MSC (2006), 10−6 time
step is recommended for explicit simulation. In these
analyses, the mesh size was set such that time step is
more than 10−6 level in this analysis.
Also, the partial FE model for bow structure and
ice was used to avoid too much analysis time due to Figure 7. The simulation model for bow structure (rigid) &
the unnecessary element. Besides, to reflect the real ice.
situation, the ice model and bow structure was given
Table 2. The values for calculation of ice load.
symmetric boundary condition.
The mass of ship included added mass factor (1.1) Content Symbol Value
to take into account surrounding fluid effect in bow
collision. Ice pressure (at 1 m2 ) [MPa] Po 2.45
And the level ice thickness was 3.5 m and collision Ship mass [kton] ship 126.5
velocity was 2.5 m/s according to the background of Ship velocity [m/s] Vship 2.5
class factor in the IACS Polar Rule (PC4). Shape coefficient (minimum) fa1 0.29
fa2 0.19
fa3 0.6
4.2 Collision analyses
Sea ice has various characteristics by age and temper-
ature and its behavior is very complicated to describe (2010). However, the values of elastic modulus, fail-
mathematically. At present the failure mechanism of ure stress and yield stress were adjusted to meet the
ice is not fully understood and not clearly defined. The collision force defined the IACS Polar Rule formula
crushing and bending mode of ice failure are consid- in the collision analysis between rigid bow structure
ered to simplify the sea ice failure mode in the IACS and ice. These re-defined mechanical properties of ice
Polar Rule. The ice characteristics were considered were used to evaluate response of hull structure and
according to the IACS Polar Rule for the ship-ice colli- ice in collision analysis between flexible bow structure
sion in this study. In the IACS Polar Rule, flexibility of and ice.
hull structure is not considered to derive the ice load.
Therefore, in this numerical simulation, there were two 4.2.1 Rigid bow structure & ice
types of collision cases as below. As described above, the rigid bow structure & ice
collision analysis was performed to find the proper
– 1st Simulation: Rigid bow structure and ice
mechanical properties of ice which can make similar
(Without flexibility of hull structure)
collision force to that calculated by the IACS Polar
– 2nd Simulation: Flexible bow structure and ice
Rule.
collision analysis (Responses of hull structure
The rigid bow structure model was used, as the
and ice).
absorbed energy of hull is ignored in the calculation
The rigid bow structure and ice was used to identify of ice load in the IACS Polar Rule. Figure 7 shows the
the appropriate mechanical properties of ice reflecting rigid bow structure and ice FE model.
failure phenomenon, which was validated by compar- Ice load defined in the IACS Polar Rule was cal-
ison with collision force defined by the ICAS Polar culated to consider the crushing and flexural failure
Rule. of ice and the rule formula of ice force was as shown
The elastic behavior of ice should be characterized below:
by moderate anisotropy and its mechanical property
varies widely depending on age, salinity and tempera-
ture. However, ice material was assumed to be isotropic
for this numerical simulation. Elastic modulus and In Equation 3, the values to calculate ice force are
Poisson’s ratio are 6.25 GPa and 0.33 respectively. as shown in Table 2.
Also, as a failure criterion of ice, the compressive The crushing force is 28.9 MN when fa1 is used for
stress of 8 MPa was used based on ISO 19906, ISO calculation of ice load in Rule. The fa2 is flexural term

208
Table 3. Analysis cases for calculating crushing force.

Elastic modulus Failure stress


Cases (MPa) (MPa)

A 6250 8
B 625 8
C 625 45
D 6250 400

Figure 9. Results of crushing forces for analysis cases.

Figure 10. Contact geometry during collision with an ice


edge.

Figure 8. The description of calculation for normal force.


Table 4. The values for contact geometry parameters.
and fa3 is the limiting case of the crushing equation.
In the IACS Polar Rule, the minimum value (0.19) is Symbol Value Unit
used to calculate the final ice force (18.9 MN), which
W 5.77 m
means that ice force depends on the combination of H 3.32 m
crushing and flexural term and calculated ice load will δ 1.45 m2
be limited by flexural failure. β 60.32 deg
Firstly, the collision analysis between rigid bow φ 90 deg
structure and ice was conducted to investigate crushing
term only. To find the appropriate mechanical property
of ice only for crushing force, some analysis cases
were carried out with adjusting the elastic modulus
and failure stress of ice as shown in Table 3.
Collision force should be transformed into normal
force in order to compare with IACS Polar Rule since where W = Width of the nominal contact area;
the collision force from FE result was calculated in the H = Height of the nominal contact area; δ = Normal
direction as shown on Figure 8. penetration depth; β = normal frame angle; and
As shown on Figure 9, the ice force is increasing φ = Ice edge angle.
with time since contact area is widened gradually. The It was assumed that the flexural failure occurred in
ice load for Case C and Case D were above the ice force approximately 3 seconds considering the initial gap
defined in rule formula. According to the IACS Polar between ship and ice and reduced velocity due to col-
Rule, normal penetration depth (δ) can be calculated lision. The properties corresponding to Case C were
using Eq. (4) and Eq. (5) below which is about 1.5 m, selected to consider crushing failure of ice.
and the normal speed of ship is 0.74 m/s according As a second step of the analysis, the very small truss
to rule formula. Therefore, the flexural phenomenon elements between solid elements of ice were used to
will occur within 2∼3 second. Figure 10 shows the consider the flexural failure. The solid element which
contact geometry and parameters which were applied had the mechanical property of Case C was locally
in Eq. (4) and Eq. (5). Also, Table 4 shows values of crushable during the ship-ice collision, and global flex-
contact geometry parameters. ural phenomenon of ice which breaks the connection
of solid elements could be simulated by truss element
simultaneously due to collision force on the contacted

209
Figure 13. The FE model for Hull structure (Flexible
material) & Ice.

Table 5. FE model data.


Figure 11. Simulated ice failure mode (Crushing & Flexural
Failure). Type Element NO. Node NO.

Ship Shell 102590 100880


Beam 5938
Ice Solid 34270 62964
Truss 26133
Total 142798 163844

Rupture strain depending on the steel grade for the


bow structure was determined as below, NTS (2004):
Mild 0.2 (20%)
HT 32 0.167 (16.7%)
HT 36 0.15 (15%)
Also, Cowper-Symonds rate enhancement formula
was used to consider strain rate as follows:

where D and P = Constant in the Cowper-Symonds


enhancement formula.
The applied values of D and P are 40.4 and 5 for
mild steel, and 3200 and 5 respectively for HT steel.
Figure 12. Comparative ice forces of two ice models.
The ice model from the rigid bow structure & ice
was used to perform this simulation as well.
FE model for target vessel was constructed by
3-node or 4-node shell element and beam elements.
area. Figure 11 shows the broken ice due to crushing As described above, only half of the bow structure was
and flexural phenomenon. modeled but mass of ship was fully applied in order to
Some parameters such as failure strain or yield realize the kinematic energy of the whole ship.
stress for truss element were adjusted, but elastic mod- The FE model and data for collision analysis are
ulus of truss element was same as the solid element. shown in Figure 13 and Table 5.
The analysis results are shown in Figure 12. In this collision analysis, ice force is a little
different from the rigid bow structure-ice collision
4.2.2 Flexible bow structure & ice analysis. It may be inferred that different structure
To investigate the hull structure response (bow struc- response between rigid and flexible structure occurred
ture) during the ship-ice collision, flexible bow struc- as the flexible structure can absorb the kinematic
ture that includes the strain rate and plastic property energy. Therefore, the collision force from the flexible
was used. bow structure-ice collision was lower than rigid bow

210
Figure 14. Comparison rigid bow structure & flexible bow Figure 16. Relationship between pressure and contact area.
structure (contact force).

5 CONCLUSION

In this paper, the mechanical properties of ice were


investigated to determine the ice model using the rigid
bow structure based on IACS Polar Rule.
By the determined ice model from the collision
analysis between ice & rigid bow structure based
on IACS Polar Rule, the ice failure modes such as
crushing and flexural failure could be described and,
resultant contact force, which is well complied with
design ice load of Rule could be obtained.
However, in the collision analysis for flexible bow
structure & ice, unacceptable failure strain computed
on bow structure, even though the hull structure had
been verified by non-linear FE analysis mentioned in
section 3. Besides, as shown in Figure 16, the loading
per area above the design load of Rule was computed.
Accordingly, to define ice model based on IACS
Figure 15. Results of plastic strain & deformed shape.
Polar Rule, it was one of key points that relationship
between pressure and contact area should come close
structure-ice collision analysis as shown in Figure 14. to IACS Polar Rule.
Also, flexural failure of ice does not occur. In the evaluation for structural & ice response analy-
As a result, the failure strain appeared at the shell sis based on FEA, the ice mechanical property should
plate as shown in Figure 15 since the patch area was be defined considering not only collision force but
smaller than rule patch. The patch size was about 2.1 m also penetration depth which had an effect on relation-
deep and 3.4 m wide in approximately 3 seconds after ship between pressure & area in performing ship-ice
the collision in this simulation and the size of rule collision simulation complying with the IACS Polar
patch defined the IACS Polar Rule is 3.31 m deep and Rule.
5.75 m wide as shown in Table 4. It means that the pres-
sure applied to the structure of ship in this simulation
was larger than Rule value in the light of similar colli- REFERENCES
sion force but smaller patch size. Figure 16 shows the
comparison results of relationship pressure & contact Daley, C. 2000. Background Notes to Design Ice loads:
area between simulations and Rule. Memorial University.
Daley, C.; Hermanski, G.; Pavic, M. & Hussein,A. 2007. Ulti-
As shown in this figure, loading per area in sim- mate Strength of Frames and Grillages Subject to Lateral
ulation of flexible bow structure is much higher than Loads – an Experimental Study: Faculty of Enginner-
that of the Rule value. Accordingly, to define ice model ing, Memorial University, Institute for Ocean Technology,
based on IACS Polar Rule, it was one of key points that NRC, BMT Fleet Technology Limited, 10th International
relationship between pressure and contact area should Symposium on Practical Design of Ships and Other
come close to IACS Polar Rule. Floating Structures Houston, Texas.

211
Han, S.; Lee J.Y.; Park Y.I. and Che J. 2008. Structural Kwak, M. J.; Choi, J. H.; Park, J. H. & Woo, J. H.
Risk Analysis of an NO96 membrane-type Liquified Strength Assessment for Bow Structure of Arctic Tanker
Natural Gas Carrier in Baltic Ice Operation. Journal of (107k) under Ship-Ice Interaction. Daewoo Shipbuild-
Engineering for the Maritime Environment 222 Part M: ing & Marine Engineering Co., LTD, RINA ICSOT
179–194. 2009.
International Associate of Classification Societies (IACS), Kwak, M. J.; Park J. H.; Choi J. H. and Woo J. H. 2009.
‘Requirements concerning POLAR CLASS’, IACS, Ultimate Strength Analysis of Stiffened Panel Subject to
2007. Ice Load, Daewoo Shipbuilding & Marine Engineering
International Organization for Standardiztion (ISO), ISO Co., LTD, The Society of Naval Architects of Korea.
19906, ISO, 2010. MSC, MSC.Dytran User’s Guied, MSC Software, 2006.
Kendrick, A. & Daley, C. 2000. Derivation and use of Norwegian Technology Center (NTC), NORSOK STAN-
formulations for framing design. AMARK Inc., Daley DARD N-004, NTS, 2004.
R&E.

212

You might also like