Full scale measurement of a large container carrier on the far east – Europe route

FULL SCALE MEASUREMENT OF A LARGE CONTAINER

CARRIER ON THE FAR EAST

-

EUROPE ROUTE

H. C. Yu (AM), J. W. Choi (V), C. I. Park (V), S. Y. Han (V), S. C. Taj (V)

SMTC-074-2008

There is limited long term service experience with the modern generation of large container carriers and hence there is great interest in improving our understanding of the performance of these vessels. In an effort to assess the actual structural service performance of a large container carrier, a comprehensive full-scale measurement system was

developed to measure the wave environment, ship motions and structural response. The system was installed on an 8063 TEU container carrier built in 2006, and t he first year measurement campaign has successfully been completed

This paper presents a summary of noteworthy observations during the first year 's voyage records which includes ship motion, wind and wave conditions, and hull girder strains and derived hull girder bending and torsional moments.

The observed vibratory responses of the hull girder are also presented.

KEY WORDS: Full scale measurement, container carrier,

hull stress monitoring, onboard wave measurement

INTRODUCTION

The benefits of the full-scale measurement are once again

relevant as the recent growth in size of modern container

carriers has been very rapid. This has created challenges for the container shipping industry including the operators,

designers, builders and classification societies. The recent

trend in design and construction of large container carriers and

technical approaches taken by ABS has been presented in

recent publications [1] [2] [3].

The limitations of the

traditional classification rules criteria and linear seakeeping analysis were addressed, such as sag-hog wave moments, wave-induced shear force and torsional moment, and load

combination. These papers also presented the way in which the

nonlinear seakeeping program may be applied to overcome

these limitations.

Yu

The operators of modern large container carriers are also

concerned about the safety and performance of these vessels since they do not have a lot of experience with these ships.

The hull stress monitoring system is gaining popularity among

ship operators in order to provide assistance to the crew for better handling of difficult situations. For example, Orient Overseas Container Lines (00CL) has equipped all vessels of

Samsung Heavy Industries (SHI) built

8063 TEU class

container carriers with the hull stress monitoring system

(HSMS). The HSMS system has primarily been used by the ship's crew to monitor the hull girder bending moment and

bow acceleration to assure that the vessel operates within safe operational limits. ABS considers the monitoring of hull

condition important and has published the ABS Guide for Hull Condition Monitoring Systems (1995, 2003) [4].

The ship designers and builders are also interested in providing more value-added features to ships in order to enhance safety and operational efficiency to make their ships more attractive to potential owners. SHI has developed the HSMS system as a valued optional feature to be offered to interested owners.

1

Another desired feature to assist the ship's crew is the onboard wave measurement system. Although the ship's crew are well

trained to safely operate the ship in varying environmental

conditions, accurate wave information can assist them in

making correct decisions in potentially dangerous situations.

For example, fine hull form container carriers are susceptible

to very large roll angle due to parametric roll resonance at

certain speeds and wave conditions, such as relative heading to

the wave, wave height and period [5]. 1f such a condition

arises

in the dark, the assistance from the onboard wave

measurement system would be most valuable. SHI developed

the onboard wave measurement system, called WaveFinder,

based on the X-band navigational radar system.

00CL, SHI and ABS have been working together

in the

design, classification review, construction, and commissioning

of the 00CL 8063 lEU class vessels.

These three parties

recognized the opportunity for carrying out full-scale measurement in actual operating conditions. The objectives of

the full-scale measurement project are to validate the HSMS

and WaveFinder against analytical

results and

available

weather data and to investigate the long-term statistics of sea loads.

TEST VESSEL

MV 00CL Europe was one of a series of container carriers

built by SHl and delivered to 00CL on 29 July 2006. The first

of this series vessel was delivered in 2002, and principal

particulars are given in Table 1. The vessel was recognized as the largest container carrier at the time of delivery. The vessel

design was reviewed using full-ship finite element analysis

using ABS Dynamic Loading Approach. The profile and plan view of the vessel is shown in Figure 1.

Table I. Principal Particulars of the Test Vessel

Figure 1 Profile and plan view and of MV 00CL Europe 8063 TEU Container Carrier

HULL STRESS MONITORING SYSTEM

The full-scale measurement system consists of two main

component systems; the hull stress monitoring system (HSMS)

and the onboard wave monitoring system (WaveFinder). In

general, HSMS installed on commercial ships have been used

to monitor global hull girder stresses and related vertical

bending moment by using long base strain gauges (LBSG)

from a section near the midship. Existing guidelines for the

HSMS system require monitoring of the measured stress and moments to demonstrate the soundness of hull structures.

There are numerous studies

indicating

that the

torsional moment, in addition to the vertical and horizontal bending

moments, is one of the most important design parameters for a container ship having a channel-type cargo hold structure. The

detailed method of deriving external moments from the

measured strains, therefore, is to be developed by the HSMS designers.

Kenichiro et al [6] has proposed an approach to derive external

moments based on strain signals from a midship section and

correlation parameters that are

derived by applying FE

analysis. This is quite a useful suggestion but the limitation is

that the torsional component is derived from one transverse

section of the ship rather than from two adjacent sections. lt is

believed that the derivation of the torsional moment from the

strain measurements at two adjacent sections may yield more accurate results.

Choi and Kang [7] have proposed a method based on two

sections which conform to the theoretical formulation.

The derived bimoments using the strain decomposition method from two sections, which have four long base strain gauges at

each section, are used in the formulation. This systematic

approach is selected in this full-scale measurement project. Figure 2 illustrates the arrangement of long base strain gauges

for the hull girder stresses at three transverse sections: at two

sections in the midship region (Fr. 183, Fr. 207) and at the fore part (Fr. 320). The sensors in the midship region are four long base strain gauges at each section, and two LBSG are installed

at the forward part of the upper deck. All sensors have been

calibrated by the strain values using calculated levels in a calm sea condition. An accelerometer is installed at the forward bosun store to monitor the bow acceleration as well as the

effects of impact loads due to slamming.

Lengtho.A. 3230m

Length B.P. 3080m

Breadth (MId.) 428m

Depth (MId.) 246m

Draught (Scant.) 145m

Speed (Design) 26.0 knots

Container Capacity 8063 TEU

Cb 0.68

1,2.3

Figure 3 Sample Replay Screen of the Hull Stress Monitoring

System

ONBOARD WAVE MONITORING SYSTEM

The ship's crew typically receives weather information through

a weather services company during ocean voyages. The

forecast is typically updated every 12 hours. The crew of the ship also use visually observed wave height and period in

making operational decisions, such as

slowing down or

changing course. It is not practical, however, to observe the

weather conditions in the dark or in severe weather conditions.

Wave measurement using marine X-band radar has been

introduced to overcome this drawback and has been installed

on some offshore structures. The radar images by near range

setting have suitable wave information of direction, period, and height [8]. The wave parameters can be estimated by applying Fourier analysis with a unique filtering procedure to

accommodate radar dynamics and to reduce certain undesirable

signals such as those from ships. The feasibility study for the

shipboard application has also been carried out by considering

the speed effects of the ship. However, the validation was

difficult due to the lack of true information on waves.

Recently, Park et al [9] have completed the development of the

system, using marine X-band radar, named WaveFinder, and

have performed several sea trial tests comparing this data with

buoy data. Their results illustrate the system's validity.

Hence, the system

has been included in this full-scale measurement such that the measured hull stresses may be

correlated with respect to the

related wave and wind

conditions.

The marine X-band radar at the fore mast is selected as the

scanner, and a scan converter is added to obtain radar images from the existing radar. Interfaced radar images are connected to the WaveFinder system. The system can separate the swell

and wind-generated wave components, which is quite useful information for estimating ship motion and sea loads in the waves.

DERIVATION OF HULL GIRDER LOADS

FROM STRA[N MEASUREMENT

A ship in a seaway is subject to irregular wave excitation that would produce the vertical, horizontal and torsional moments

on the hull girder. The strains measured from the full-scale

measurement are the combined effects of these external

moments. Therefore, it is necessary to develop a formulation such that allows the measured strains to be directly related to

the bending and torsional moments. Choi and Kang have

proposed a procedure [7]. The decomposition method can

derive the independent strain components due to the individual

moments from the combined measured strains. Figures 4

through 8 are sketches which define the arrangement of strain

gauges and the derived strain components due to different

external moments [12]. BBC') 4C'ABEGAI) 6 )IOGBB) 6 c,c6) 4 L2.3.5 S (ON 0100K) S 106R 6 .C'106 5 (ON 0100<) 6 (oN OBI j 6 (101010K) 6 LOSCO) 6 (co 010.0< 6(ooroo( 5(00CG) SMTC-074-2008 Yu 3

Figure 2 Arrangement of long base strain gauges for Hull

Stress Monitoring System installed on the test vessel A motion sensor is also installed at the accommodation area to

record the

roll and pitch motions.

The motion data

is

accessible to HSMS by serial connection for this ship.

All sensor signals of the HSMS are digitized at 20 samples per second. The sensor signals are processed to generate statistics

for each 5-minute time history, and then are recorded on the

hard disk of the processing computer. The statistics include the maximum, mean, minimum, peak-to-peak, standard deviation,

root mean square (RMS) and mean crossing period.

In addition to the statistics, the raw time histories of the 1-ISMS

sensors are also recorded for 5 minutes on every hour during the first round trip voyage between the Far East and Europe.

The system software was modified to record all measurements

continuously then write a data file on every hour from the

second voyage.

The HSMS provides an interactive user interface to display the

bending moments, torsional moment, statistics, bottom slam occurrence, cumulative fatigue cycle count and the real-time sensor signal display. The HSMS also provides an option to

replay the recorded data.

Figure 3 illustrates a sample replay screen of the HSMS, which

shows the option to display the measured strain, the derived

bending and torsional moments, and the navigation route from the recorded data. Hence, hull structural conditions are easily

measured and analyzed using this hull

stress monitoring system. S

1

6) I LGOGEN 0601*0 10<01 2 SIGNAL C14IOI000SINO *01 7 4 1 ZENOI 004010k 004 N JLOACflGN *0* 6 LONG 00000 01440< GAUGE 00001.04000000

Long hase strain gauge

Figure 4 Arrangement of long base strain gauges for

two-section measurement

Cl

y

Yr

y

Figure 5 Arrangement of gauges at the

midship of a

container vessel

Figure 6 Graphical illustration

horizontal bending moment

y

YB.

z4

-_-,<__ y,

rs

-1_---í.>-Figure 7 Graphical illustration of strains

in the case of

vertical bending moment

Strain due to horizontal moment: =

of strains

in

the case of

Strain due to vertical moment: e. = e2

Warping strain due to torsional moment: e, = St. Venant strain due to torsional moment: e. =

V

Figure 8 Graphical

illustration of warping strains due to

torsional moment

The strain from the long base strain gauge measures the

combined quantity of vertical, horizontal, warping and St.

Venant strains. Therefore, one can write the strain components at one midship section as follows.

. 6,,

Here, the measured strain components are counter-clock-wise and the symbol stands for the strain vector consisting of strain components of horizontal, vertical, warping, and pure torsion

at the Ist sensor, and stands for the warping function that can be obtained by the finite element modelling: Fujitani(1990)

[10]

for complex geometry, Kim et

al.

[Il]

for simple geometry.

The decomposed strain component vector, therefore, is defined as the following, with respect to the strain at the Ist sensor.

Similar results can be obtained at an adjacent plane with

subscript a and the distance d of in-between sections, and

then the subtraction of the two equations can be expressed as

=

Here,

Mm =

[,0

E3 S3 e4

M=

_-

e e

-

e

-

aT

-The transformation matrix A will be the same because the

two planes will be located at the plane having the same cross-sectional properties. Finally, the torsional component due to

warping can easily be obtained by the following:

= dz2 6aw -U1d

--i

I i I

-1

A=

YB/

ZB/

U3/

(r3/

/Yr

/ZT

/U

yB/

z/

U3/

(r3,.

/

YT

/Z

/U1/

SMTC-074-2008 Yu 4 rn = A where, em= [1 e2 e3 64fr

-Fiere, f' is the torsional constant and multiplication of Young's modulus and the second moment of cross section at shear center. w is the slope of twisting angle. The warping strain ¿', at an adjacent cross section is required to obtain the

torsional moment.

ONBOARD

SYSTEM

TESTING

DURING

MAIDEN VOYAGE

The test ship MV 00CL Europe was commissioned to the trading route between the Far East and Europe, which would

take about 60 days for the round trip. lt was determined that

the installed hull stress monitoring system and the onboard

wave measurement system should be tested during the maiden voyage [12]. The onboard system was tested during the voyage from Busan to Port Suez, as shown in Figure 9.

HULL STRESS MONITORING SYSTEM TESTING:

Figure 10 is the selected records of the hull stress monitoring system taken during the voyage from Busan to Port Suez. Each data point in this figure represents a processed statistic, such as mean, maximum, etc., over a five- minute period.

Figure lO(a) shows the vessel

speed during the voyage

annotated with the names of the ports at which the ship had called.

Figure 9: The route of the onboard testing of the measurement

system

Figure 10(b) shows the ship's heading with key leeway points annotated. The annotated route segments are also shown in

Figure 12. The ship maintained 20 to 25 knots in the open ocean segment while reduced to lower speed in short coastal

segments. Figure 10(c) shows the mean wind speed measured by the onboard anemometer. The wind direction relative to the ship and ship's heading are also recorded. The true wind speed and direction may be determined for comparison to the weather forecast or hindcast data.

Figure 10(d) shows the average still water bending moment measured in the midship region (Fr. 183, Fr. 203) and at the

forward location (Fr. 320). It is noteworthy that the mean

bending moments show very little change over the long open

ocean segments compared to those during the short segments. It was observed that the transfer of ballast water or fuel oil was the main reason for the changes. This ship is equipped with an

automatic heel control system that maintains the ship in an

SMTC-074-2008 Yu

upright position during the loading and unloading operation in the terminal. The system automatically transfers ballast water between port and starboard ballast tanks located in the midship region. The ship typically intakes ballast water for heel control before entering a port and discharges in open waters. The fuel oil is also transferred before entering port for bunkering. There

exist many other causes that may change the mean bending

moment, such as trim, speed, temperature, etc.

Figure lO(e) shows the measured wave bending moment and the standard deviation of vertical acceleration measured at bow. The vertical scales are determined to produce the two

records aligned on top of each other as much as possible. This can be considered as simple ways of checking correlation. It is observed that

the wave bending moment and the bow

acceleration

indicate consistent correlation except for the

period between 8/24/06 20:00 and 8/26/06 12:00, which is annotated as the segment between leeway points A and B in Figure 9.

Figure 10(f) is the mean, maximum and minimum of the hull

girder stress measured by LBSG located at Fr. 183 on the

starboard side of the upper deck. The mean stress time history follows very closely with those of the mean bending moment. ONBOARD WAVE MONITORING SYSTEM Testing:

The WaveFinder system was tested during the open ocean

voyage from Port Kelang to Jeddah. The measured significant

wave height

from WaveFinder was compared with

the

forecasted wave patterns from the weather service data in

Figure 11. The measured significant wave height is plotted

using a square symbol. The measurement results

of

WaveFinder are the average value of one-hour measurement. The discontinuity of the measurement is due to the rain or too

low level of radar echo signal.

Since the wave forecast

provides the anticipated wave conditions for the following five

days, the forecasted wave heights are overlapped in the plot. Hence, one can use the WaveFinder as a real-time onboard

wave measurement system.

4

Q4 Q5.3ß 082? 0878 0829 O83O O&31

ooc as.x 000o

oo

o000 oOE 00110 00:00

Measured 6y Was54rder1Hrsausrage) Førecs1ed on 24th

r)recasted on 270, Farecaslerl on -. Forecasted on 3116 Hdcast

Fgure 11 Wave heights during the voyage from Port Kelang

3 .E +01 2.E01 Q) 1.E+01 (J.) 8.E+05 - E+05 4.E+05 j 2.E+05 0E+0D 8.EO1 6.E+U1 4EU1 2. E+ 01 ° 0.E+00 () Ship Speed n (h) Headiii ngI

aE.0i (c) Wind Speed

Shanghai

0.E±00 '

'--'

'I

_'

'i'''

''''

8/13 8/14 8/15 8/16 8117

0113 0/14 8115 9/16 0/17

(d) Vertical Still Water Bending moment

2.E+05I(e) Wave Vertical Bending moment

f) Strain Mid-Starlioard-Upper SMTC-074-2008 Yu Forward(Er.320) 8/17 8118 8/19 8120 8/21 8)22 8/23 8,24 8125 8,2 8.27 8128 8/29 8/30 8131 9/1 9/2 9/3 0/13 8/14 8/15 8/16 8/17 8/18 8/19 8i20 8/21 /22 8123 0/24 825 8/26 8/27 8/20 828 8130 8131 9/1 912 9/3 Bow acceleration

Wave Bending moment

O E+OCI 000

8113 0114 0/15 0/16 8/17 8118 0/19 0/20 0121 0i22 0123 0/24 025 0.26 0127 8/28 0129 9130 0131 9/1 9/2 9/3

B,13 8/14 8/15 8/16 8/17 8/18 8/19 8.20 8/21 8122 8123 8124 8/25 8/26 0/27 0128 8/29 8/30 8/31 9/1 92 9/3

Figure 10 Selected records of measurements taken during the onboard test period

0.04

8/18 8/19 8/20 8/21 8/22 0123 8124 8/25 0/26 8/27 8i28 8/29 8130 8131 9/1 9/2 9/3

;:

I I

_____________________

0/10 8/19 0/20 9121 9/22 0123 0i24 8/25 8i26 0,2/ 0120 9i29 0/30 0/31 9/1 9i2 9/3

Nirigho Singapore Port Kelaroq Jeddah

Q) a) a ci] 3.E.01 i.E 01 2.E.01 I.E .01 5.E.00 0.E.00 8/13 8114 9/15 flh1 Bow Acceleration E 1.EU5 5EÛ4 (13 4.E+02 3.E+02 2 E+02 i E02

Figure 12 Trading Route of the Test Vessel

VESSEL TRADING ROUTE AND ONSHORE

DATA PROCESSING

After the onboard system testing, the test vessel continued

service between the Far East and Europe. Figure 12 is the

trading route of the vessel. The round trip voyage takes about

60 days. The vessel completed three round trip voyages on this

route, and then was placed on a slightly different route. This

paper presents the findings from the first three voyages.

The project team engineer visited the vessel in Busan after completion of the round trip voyages where he recovered the recorded data, conducted system check and cleared hard disk

space ready for the following voyage. Onshore data processing

was carried out with the recovered data. There are three types

of data available for onshore processing. The first group of data consists of strains from LBSG, bow acceleration, roll and pitch

motion.

The second group consists of navigational data

including the Global Positioning System (GPS) position, ship's heading, engine RPM and speed. The third group consists of

derived hull girder loads at midship - vertical bending moment

(VBM), horizontal bending moment (HBM), and torsional

moment (TM). Statistics (maximum, minimum, mean, and

standard deviation, etc.) of the first group of data are processed

onboard and recorded every 5 minutes. Statistics of derived

channels are calculated during onshore data processing using the time histories of LBSG recording.

Selected statistics of all measured and derived channels were

plotted to evaluate the quality of recording and to screen out any erroneous data. The recorded data was found to be of high

quality without electronic noises or instrument drift, etc.

Occasional gaps were found in the recording, which were

caused by the crew turning off the computer system. The

measurement system is designed to start the software by itself

after the power to the computer has been restored, hence the loss of data recording can be minimized. Each record of data, i.e., S minute statistics, is further tagged with a flag indicating whether

the vessel was in port or at sea, since the allowable VBM ¡s

different for these two conditions.

After reviewing all three voyage records, it was found that the vessel encountered most severe weather conditions during the

westbound voyage number 3 in winter months.

lt was

determined to examine the recorded data in detail for the voyage segments in the Mediterranean Sea, along the coast of Spain in the Atlantic Ocean up to the entrance to the English Channel.

2

Figure 13 Selected Recordings, Westbound Voyage Number 3 Figure 13 shows measured mean wind speed, mm/max/mean of hull girder loads - VBM, HBM and TM - from 12January 2007

to 29January2007. It is shown that the vessel encountered two

storms, one in the Mediterranean Sea and another between

Gibraltar and Rotterdam.

STORM AND VESSEL RESPONSES IN THE

MEDITERRANEAN SEA

The vessel encountered a short and intense storm soon after

entering the Mediterranean Sea from Port Suez. Figure 14

shows the location of the vessel, vessel speed, wind direction, wave direction and heights along the vessel path during the

storm. lt is shown that the vessel passed through the storm area

within24hours. The wave record shows a very rapid growth of

wave height up to 5 meters in the first six hours then a gradual decrease in the following 18 hours as the vessel passes through the storm area. The true wind direction is fairly constant during the storm. Relative wave heading changes from bow seas to

stern quartering seas during the same time. It is also observed that the vessel reduced speed from23 knots to20knots near the peak of the storm.

L

UIIU$ UIMIUU:

_IIi

UhIflhI' IUEIII

U0111

rtt

R9.W Iodng thnge

from mati sea. toS*ern quartering

Figure 14 Vessel Location and Weather Conditions During the

Storm in the Mediterranean Sea

SMTC-074-2008 Yu 7

zm i2iuuuuur4.

g

11111 Hunai..um 1111 PUIIIIIUIIIUIIISIIIDÍI

Figure 15 shows the engine RPM, min/maxlmean of bow

acceleration, VBM and HBM. It is

also shown that the

voluntary reduction of the ship speed occurred when the

maximum bow acceleration became greater than 02g while the

vessel's course was kept unchanged. The reduction of the

maximum VBM is observed when the speed was reduced.

These records show typical practice of the crew in heavy

weather to keep the vessel in safe condition. The ship's crew does not have the option to change the ship heading to avoid

large storms in the Mediterranean Sea because leeway points for

navigation cannot be changed in such a restricted area. The

speed reduction is the most frequently used method to reduce

large ship motions and hull girder loads. If the vessel were on a

route through the ocean then the crew would have a choice of

change of heading and/or reduction of speed. These practices

will affect the long term extreme values of ship motion or hull

girder loads that the vessel may encounter in its lifetime.

01

-3.!

Maximum Acceleration

A B C D F

Figure 15: RPM. Bow Acceleration, VBM and HBM in the

Mediterranean Sea

VBM during the storm was further analyzed to investigate the

frequency component which reveals some interesting features. First the time histories of VBM were created by combinìng the

recorded time series of four strain gauges.

Then spectral

analysis was carried out on the generated time histories of

VBM.

Spectrums of VBM are produced by Fast Fourier

Transform (FFT) method analyzing one hour time histories.

One hour time series were divided to six 10 minute records, and

the spectra from each time series were averaged to generate a representative spectrum over one hour. Figure 16 shows the hourly spectra of VBM plotted along the time axis where the

vessel location is marked as A to F in Figure 14.

S1,ortìncoutit« waves (6-7 s&

frombow producedlarge2-node vibratory responses

Encounterwvepanodbecomes ong (15 sec.) towrd stern quarteringseas

':

-.

-.. SMTC-074-2008 Yu 8 A I C O A C O p

Figure

16 Changes of Spectra of VBM in Time in

the

Mediterranean Sea

First, the peaks of VBM spectra shift from high frequency

(around 0.15 Hz) to low frequency (around 0.05 Hz) from

location B to C. This is mainly due to the change of relative

wave heading from bow seas to stern quartering seas that result

in

longer encounter wave period towards the

location B.

Secondly, the 2-node hull girder vibration component is visible

around 0.5 Hz which is the natural frequency of the 2-node

vibration of hull girder. The peak of the 2-node vibration

component diminishes as the wave heading changes to stern

quartering seas as well as the wave height becomes small. The

third observation is the spectral peaks around 0.15 Hz that

persist throughout the storm from location B to C. This appears

to be the VBM responses to the wind-generated waves in

developing seas.

Figure 17 shows the spectra of VBM and corresponding time histories at two instances, one at 14:00 1-IR and 18:00 1-IR. Relative wave directions are bow quartering seas and stern

quartering seas from starboard. The spectrum of VBM at 14:00 HR shows two distinctive peaks, one around0.1 Hzrepresenting the VBM due wave frequency component and the other around 0.5 Hz representing the 2-node hull girder vibration component.

The spectrum of VBM at 18:00 HR shows three distinctive

peaks, one around 0.05 Hz, the second around 0.15 Hz and the

third around 0.5 Hz.

The peak around 0.5 Hz is a 2-node

vibration component of VBM of which the magnitude

is

reduced compared to the 14:00 HR spectrum. The reduction is

due to the change in wave direction and wave height. Bow flare

impact is the main cause of 2-node vibration of hull girder for

fine hull form vessels. The vessel in stern quartering seas would

produce less bow flare impacts than bow quartering seas that would contribute to the reduction of 2-node vibration of hull girder. The other cause of reduction is due to decreased wave

height. Two spectral peaks, around 0.5 Hz and 0.15 Hz, can be explained that the sea conditions in the storm area may consist

of swell and locally generated wind waves. The encounter

frequency of swell becomes lower in stem quartering seas

compared to those in bow quartering seas while those of locally

generated wind waves would remain the same over the storm

area. lt is believed that the encounter frequencies of swell and

locally generated wind waves are close at 14:00 1-IR showing

one spectral peak, while these are separated

at 18:00 HR showing two spectral peaks when the relative wave heading

changes to astern.

VSM Spectrum 1-140714:00

Figure 17 Spectra of VBM and Corresponding Time Histories

of VBM in the Mediterranean Sea

STORM AND VESSEL RESPONSES IN THE

ATLANTIC OCEAN

The next storm event occurred where the vessel sailed along the

coast of Spain towards the English Channel. Figure 10 shows the position of the vessel and wave direction along the route

with selected channels. lt is shown that the wave environment changes abruptly from calm to heavy when the vesse! enters into

the Atlantic Ocean passing Gibraltar. It

is noted that the

predominant waves in the ocean are from the west while relative wave heading changed from head seas to following seas.

i

F aoe

1-14.0? 14:00

VBM Spectrum 1-14-0? 18:00

peaks of spectra of VMB show the change of encounter

frequency from high to low as the wave heading changes from head seas to following seas. It is noteworthy to compare spectra

of VBM in the Mediterranean Sea to those in the Atlantic

Ocean. The VBM spectra in the Atlantic Ocean show very

small 2-node hull girder vibratory responses compared to those in the Mediterranean Sea. The plausible explanation is that the Atlantic waves are dominantly swells that have been propagated from far away hence there is very little wind wave component.

The wave conditions in the Mediterranean Sea appear to be

developing seas which is a

mixture of swell and locally

generated wind waves.

-J u VM Spectrum 1.19-07 12:00 f VBM Spectrum 1-19-07 00:00, VBM Spectrum 1-18.07 00:00 VBM Spectrum 1.17-4)7 13:001 t'lì

Figure 19 Spectra of VBM in the Atlantic Ocean

SMTC-074-2008 Yu 9

Figure 18 Vessel Location, Wave Direction, Vessel Heading

and Hull Girder Loads during a Storm in the Atlantic

Ocean

Figures 19 and 20 show the spectra of VBM and corresponding

VBM Time Historyl.19.07 12:00 700000 100400 300000 24.1 000400 050000 040000 040000 7x034 $50007 004040 400 3 70000000x00 -504000 700400 100400 006407 000000 4040 '10 070 740 700 440 7. 1)40 %0 4040 VBM lime Histolyl -19.07 00:00 VBMTineHistotyl -18-07 12:00 430 4*4 400 VBM Tine Hoiy1 -18.07 00:00 7120 420 040 VBMThu,e Historyl.17-07 13:00

Figure 20 Time Histories of VBM in the Atlantic Ocean

DIURNAL THERMAL EFFECTS ON HULL GIRDER

LOADS

Another interesting observation is the effect of the sun on hull girder bending moments. Figure 21 is mm/max/mean of VBM

and HBM during the westbound voyage number 2. It shows

large daily

fluctuation of mean HBM in

the circled area

compared to those of VBM. The mean HBM fluctuation also shows a sharp decrease of HBM (increase of tension on port side) over a short duration of a day. Figure 22 is a similar plot

of VBM and HBM during the eastbound voyage number 3. It

shows similar hull girder responses that the fluctuation of mean

HBM is greater than those of VBM. What is interesting in this figure is that the pattern of daily fluctuation of mean HBM is

opposite of those observed on the westbound voyage. It shows a sharp increase of I-IBM (increase of tension on starboard side).

The circled portion of the voyages were both in the Indian

Ocean where the vessel maintained steady heading either to the

west or to the east. It can be explained that these fluctuations are due the temperature difference between port and starboard

side of hull structures, since HBM is

proportional to the difference of strains of port and starboard side of the hull. On

the other hand VBM is proportional to the difference of strains between deck and bottom structures.

It should be noted that the on-deck LBSG is housed in a

protective cover and positioned inboard side of the deck passage

towards the hatch coaming. Furthermore there is overhead

platform to store the most outboard stack of containers that

keeps the LBSG in shade all the time.

For other types of

vessels, such as oil tankers, the on-deck LBSG is more directly

exposed to direct sun which shows far greater fluctuation of

VBM than those of container carriers.

Figure 23 shows the fluctuation of strains at four parts of the

hull structure measured by LBSG. lt is clear that the deck strain on the starboard shows largest fluctuation due to position of the sun. The magnitude of daily fluctuation of the hull girder loads

is also affected by the atmospheric conditions, such as fog,

ovecast of cloud, etc.

>2 104.40 -.... 5x0.A4 47437040 - 040*44$1.---44*-44 600E_04 0U043* 8420,074 00.644.40 0.4660. .'7.),J,06 .8,44.740 0Oct.00

Figure 21 MinlmaxlMean of VBM and HBM during Westbound Voyage Number 2

----Vòyage-3 Eatb

Voyage -2 Westbound

44447.00 0.504440 !41E70 142101.20, 7*1.5.0.40

S.4..04 244,fl_7 7.4.6.20 42.0.807 42.044.47 72.141.-07 27.0 0.07

Figure 22 Mm/max/Mean of VBM and HBM during Eastbound Voyage Number 3 00*00... ¿... -. 01.400, ....t.&', -ee-05 DEbO .. ... 00.00 -I

.4...._.__.._...

...-SMTC-074-2008 Yu 10 0400 420 4440 7.0.1.07 12.0.1.07 '7.0-w 040.6.04 27.0.600

CONCLUSIONS

A comprehensive full-scale measurement system has been

successfully designed and installed on a large container carrier.

The system consists of the hull stress monitoring system, the

onboard wave monitoring system, and the voyage optimization

system. The HSMS and WaveFinder systems were tested

during the maiden voyage, and three round trip voyage data has been analyzed.

It was found that the system was fully functional throughout

the test period and al! recorded data could be recovered in

good quality

The recording of operational data, such as ship's heading,

position, speed, etc., was found very usefu! in

understanding the recorded data and vesse! operational practices.

The onboard wave monitoring system can further provide important wave parameters that would assist in validating

the analytical results of ship motion and sea loads.

Hull girder responses of the vessel in the Mediterranean wave environment showed higher hull girder vibratory

responses than those observed in the Atlantic Ocean waves under similar significant wave heights.

Horizontal bending moment and torsional moment are more affected by the diurnal thermal effect than vertical bending

moment.

The recorded time histories of the sensors will also be

valuable in validating the phase relationship between the

ship motion and sea loads.

Long term distribution of the measured data has not been

studied due to short measurement duration. lt is planned to

carry out the comparison of measurements against analytical

methods in the future.

ACKNOWLEDGMENTS

The authors wish to express their appreciation to Captain K. H. Lee and all crews of MV 00CL Europe for their support during the onboard testing of the system. We would Ìke to give special

astbound

theindian Ocean

Figure 23 Fluctuation of LBSG Strains at four Location

thanks to Y. T. Kang of SHI who provided the data processing software for HSMS.

REFERENCES

I. American Bureau of Shipping, "Guidance Notes on

SafeHull-Dynamic Loading Approach for Container Carriers (2005)"

Bill Shi, Donald Liu, Christopher Wiernicki, "Dynamic

Loading Approach for Structural Evaluation of Ultra Large Container Carriers", SNAME Marine Technology

Conference & Expo, 2005

Sungeun (Peter) Kim, Yung-Sup Shin, Donald Liu, "Advance

'Dynamic Loading Approach' for Ultra Large Container

Carrier

Based on Nonlinear Time-Domain Seakeeping

Analysis", ISOPE Conference, San Francisco, May 2006

American Bureau of Shipping, "Guide for Hull Condition

Monitoring Systems (1995, 2003 Edition)"

American Bureau of Shipping, "Guide for the Assessment of

Parametric Roll Resonance in

the Design of Container

Carriers (2004)"

Kenichiro Miyahara, Ryuju Miyake, Norikazu Abe, Atsushi Kumano, Masanobu Toyota and Yoshiyuki Nakajima,

"Full-scale measurements on hull response of a large-container ship in service", Proceedings of OMAE2006, 25th

International Conference on Offshore Mechanics and Arctic Engineering, June 4-9, Hamburg, Germany, 2006

Choi J.W and Kang Y.T., "Two-plane hull girder stress

monitoring system for container ship, The Society of Naval Architects of Korea", Vol.8, No.4, pp.17-25, 2004

Park G.l et al, "The Application of Marine X-band Radar to

Measure Wave Condition during Sea Trial", Proceedings of 2nd PAAMES and AMEC2006, Jeju Island, KOREA, OCT.

17-20, 2006

Park G.l et al., "Introduction of Optimum Navigation Route Assessment System Based on Weather Forecasting and Seakeeping Prediction, Journal of Korean Navigation and Port Research", Vol.28, No.10. pp833-841, 2004. (in

Korean)

Fujitani, Y. 1990 Thin-walled Beam Structural Analysis,

Baihukan Book Company. (in Japanese)

Il. Kim, K.S., Kim, KS., Lee, J.W., Kim, LT. and Bang, H.S. "Mechanics of Ship Structures", lnha University Press, in

Korean Translation Version, Chapter 5. (in Korean).

(Original book: Yamamoto, Y., Ohira, H., Sumi, Y. and Hujino, M. 1997 Mechanics of Ship Structures, Seizando

Shoten, in Japanese), 2000.

12. Yu, H., M. K. Ha, J. W. Choi, S. C. Tai, "Design and

Implementation

0f

a Comprehensive Full-Scale

Measurement System For a Large Container Carrier", RINA Conference on Design & Operation of Container Ships 2006 conference, London, UK, November 22-23, 2006

MANUSCRIPT FOR REVIEW

E-mail your manuscript as a PDF file to

snamepapers@sname.org for review by the Papers Committee. Please carefully follow the instructions for properly making a PDF on our website at:

http://www.sname.org/authorinstructions.htm.

FINAL MANUSCRIPT SUBMISSION

E-mail a copy of the final manuscript to

snamepaperssname.org for

review and acceptance by the

Papers Committee.

Mail Copyright Transfer Agreement to: SNAMIE, ann: Susan Evans Grove, 601 Pavonia Avenue, Jersey City, NJ 07306, or fax to 201-798-4975.