ITTC – Recommended 7.

5 – 02
07 - 02.1
Procedures Page 1 of 16
Testing and Extrapolation Methods
Effective Date Revision
Loads and Responses, Sea Keeping 2002 01
Sea Keeping Experiments

CONTENTS

1. PURPOSE OF PROCEDURE

2. SEAKEEPING EXPERIMENTS
2.1 Model Size
2.2 Model Completeness
2.3 Model Weight Distribution
2.4 Guidance System
2.5 Free Running Test
2.6 Measurement of Wave Loads
2.7 Measurement of Added Resistance
2.8 Measurement of Impact Load
2.9 Parameters to be Measured
2.10 Headings
2.11 Regular Waves
2.12 Transient Waves
2.13 Irregular Waves
2.14 Data Presentation
3. PARAMETERS
3.1 Parameters to be Taken into Account
3.2 Recommendations of ITTC For Parameters
4. VALIDATION
4.1 Uncertainty Analysis
4.4 Benchmark Tests
APPENDIX
A.1 EXAMPLE OF AN UNCERTAINTY ANALYSIS
A.1.1 Analysis of Elemental Errors
A.1.2 Response Amplitude Operator (RAO) and Overall Uncertainty
A.1.3 Results
A.1.4 Conclusions

Updated Approved

Load and Responses Committee of 23rd ITTC 23rd ITTC

Date 2002 Date

 ITTC – Recommended 7.5 – 02
07 - 02.1
Procedures Page 2 of 16
Testing and Extrapolation Methods
Effective Date Revision
Loads and Responses, Sea Keeping 2002 01
Sea Keeping Experiments

Seakeeping Experiments

1. PURPOSE OF PROCEDURE
To ensure that seakeeping tests to primarily 0.7

obtain linear response functions in waves are 0.6

performed according to the state of the art.
0.5

Fn . ω . √ ( Lm / g )
0.4

2. SEAKEEPING EXPERIMENTS
0.3

2.1 Model Size 0.2

0.1
The size of the model should be such that
tank wall interference is avoided for the range 0
0 0.5 1.0 1.5 2.0 2.5 3.0 3.5 4.0 4.5
of wave frequencies and model speeds to be BT / Lm

tested. Table 1, Fig.1 give, in dimensionless
form, a relationship among model length LM,
tank breadth BT, Froude number Fr and the
highest wave frequency ω at which interfer-
ence effects may occur in head waves.
Fig.1 Maximum frequency at which tank inter-
ference occurs in head waves
2.2 Model Completeness

It is desirable that the model is complete up

Table 1 – Maximum frequency at which tank to the uppermost weather deck, including fore-
wall interference occurs in head waves. castle and bulwarks. A more complete model-
ling of deck fittings, deck houses and freeing
BT / LM Fr ⋅ ω LM / g ports may be necessary if parameters such as
0.50 0.635 deck wetness are to be measured.
0.75 0.458
1.00 0.378 All appendages should be fitted and the re-
1.25 0.335 port should state which appendages were fitted
1.50 0.309 during the experiments
1.75 0.292
2.00 0.280
2.25 0.271 2.3 Model Weight Distribution
2.50 0.265
2.75 0.260 If the bending moments, shears, and torsion
3.00 0.255 experienced by the model in waves are to be
3.25 0.252 measured, the longitudinal and transverse dis-
3.50 0.249 tributions of mass must be reproduced as cor-
3.75 0.247 rectly as possible, and must be correctly re-
4.00 0.245 ported. In other cases, only the radii of gyration
 ITTC – Recommended
Procedures
Testing and Extrapolation Methods
Loads and Responses, Sea Keeping
Sea Keeping Experiments
need to be simulated. For tests in head or fol-
lowing waves with a model restrained in roll, it
is not necessary to simulate transverse weight
distribution.
If the longitudinal radii of gyration for pitch
or yaw are unknown, a value of 0.25 L pp
should be used. If the transverse radius of gyra-
tion is unknown, a value between 0.35B and
0.40B, depending on the ship type, should be
used. (These values are those without including
the effect of added mass.)
For experiments in which roll is not re-
strained, the metacentric height should be
simulated. If the vertical position of the centre
of gravity is unknown, it should be determined
and recorded. As an alternative to ballasting the
model to a specified transverse radius of gyra-
tion, the natural period of roll may be simu-
lated.
2.4 Guidance System
The guidance system should be such as to
impose the minimum restraint on the motions
of the model. It is desirable that in head or fol-
lowing waves the model should have the free-
dom to roll, that is, to rotate about the longitu-
dinal axis through the centre of gravity. In
oblique waves, care must be taken on the re-
straint of sway and yaw motions.
2.5 Free Running Test
None
2.6 Measurement of Wave Loads
None
7.5 – 02
07 - 02.1
Page 3 of 16
Effective Date Revision
2002 01
2.7 Measurement of Added Resistance
None
2.8 Measurement of Impact Load
None
2.9 Parameters to be Measured
The amplitudes and phases of hull motions
in desired degrees of freedom (as defined in
Reference 6) should be measured as appropri-
ate to the purpose of the test.
Wave height measurements should be made
with a probe mounted close to the model, but
not causing interference. The probe should
preferably be fixed to the carriage, but meas-
urements may be made at a fixed point in the
tank. In the latter case, the measuring point
should be selected in the position where waves
are fully formed without being affected by the
waves reflected at the wave maker and the tank
walls & beaches.
Establishments should develop the capabil-
ity to measure the following additional parame-
ters:
- Accelerations. In order to provide corrobo-
rating data for computation of accelera-
tions from measured motions.
- Relative motion. Preferably measure-
ments of the relative motion between the
model and the water surface at points that
allow correlation with wave and other mo-
tion data.
 ITTC – Recommended
Procedures
Testing and Extrapolation Methods
Loads and Responses, Sea Keeping
Sea Keeping Experiments
- Rudder angle. In cases where active rudder
control is employed, the rudder angle
should be continuously monitored.
- Impact pressures on the hull or on deck at
selected locations.
- Still water resistance and added resistance
in waves (if not freely running).
- Water on deck.
- Propeller revolutions. Whenever a self-
propelled model is used, the shaft revolu-
tions should be recorded.
- Visual records. Tests should be recorded
visually, either by film or video, prefera-
bly in a way allowing scaling of time.
The following parameters may be measured
depending on the test requirements.
- It is recommended, if possible, that propel-
ler torque and thrust are also continuously
recorded.
- Encounter angle. The angle between the
mean model heading and the wave direc-
tion.
- Leeway (or drift) angle. The angle be-
tween the mean model course and the
mean track of CG.
2.10 Headings
When performing tests in oblique seas, the
range of encounter angles between zero and
180 degrees should be selected appropriately in
accordance with the stated test objectives.
7.5 – 02
07 - 02.1
Page 4 of 16
Effective Date Revision
2002 01
2.11 Regular Waves
For conventional ship forms, a sufficient
number of tests should be carried out at each
speed to provide adequate data for a minimum
range of wavelengths from 0.5 Lpp to 2.0 Lpp,
and also to ensure a good definition of the
resonance condition. Either the ratio of the
wave height to Lpp or the ratio of wave height
to wavelength should be maintained constant.
(The advised value of the ratio of wave height
to wavelength is around 1/50.)
In determining the motions, it is recom-
mended that the average amplitude and period
of at least 10 cycles be obtained, if possible.
Alternatively, a spectral analysis following the
procedures outlined below for irregular waves
could be followed to obtain the average ampli-
tude and period of waves and responses.
2.12 Transient Waves
The transient wave technique is an experi-
mental technique in which a wave train that
contains wave components of all the relevant
frequencies is produced in such a way that the
component waves reach a certain place in the
wave tank at the same time so that a single
large wave packet is formed. If a model struc-
ture is posted at the place where the single
large wave packet is produced, response char-
acteristics to regular waves of all the frequen-
cies contained in the wave packet is obtained in
a single experiment (provided the linear super-
position assumption holds). This technique
proves to be very efficient as a standard tool
for evaluating RAO’s of stationary offshore
structures or towed/self propelled ships.
 ITTC – Recommended
Procedures
Testing and Extrapolation Methods
Loads and Responses, Sea Keeping
Sea Keeping Experiments
2.13 Irregular Waves
Tests should be carried out in waves appro-
priate to the sea conditions in which the vessel
may be required to operate. In the absence of
specific wave spectrum data, an ITTC spec-
trum should be used. In generating irregular
waves in a tank, the input signal to the wave
maker should be produced in such a way that
the generated waves are non-repeatable.
Data should preferably be digitised before
analysis, using sample rates appropriate for the
avoidance of aliasing with the individual meas-
ured parameters. Care must be taken for the
duration of the data acquisition so that enough
data are recorded for the objective of the test.
The sample rate in the data acquisition needs to
be fast enough in order that a sufficient resolu-
tion is achieved.
Energy spectra of waves and relevant re-
sponses should be produced through spectral
analysis using either the indirect method of
Fourier transformation of the autocorrelation
function, or the direct method of splitting the
record into suitable blocks and subjecting these
to Fast Fourier Transform.
In addition to the spectral analysis, statisti-
cal analysis should be performed to produce
the mean, maximum, minimum, and the mean
of 1/3 highest values. Techniques utilised to
smooth spectral shapes, such as block overlap-
ping, should be documented in the presentation
of results. When reporting statistics, the num-
ber of events and number of encounters should
be reported independently, as well as the over-
all statistics.
7.5 – 02
07 - 02.1
Page 5 of 16
Effective Date Revision
2002 01
2.14 Data Presentation
The coordinate system in which data are
presented should be defined. Motion compo-
nents should also be defined. Linear transla-
tions and rotations may be presented in non-
dimensional forms as being divided by wave
elevation and wave slope respectively. Rudder
angles may be presented in other appropriate
non-dimensional forms. Dimensional presenta-
tions can be more appropriate depending on the
objectives of the experiment. Phase angles
should be given in degrees and increases in
resistance and propulsion parameters should be
presented in nondimensional forms defined in
Reference 6. Accelerations should be made
non-dimensional by LPP/(g ς a ). It is recom-
mended that the results are plotted to a base of
ω (LPP/g)1/2 or ω e (LPP/g)1/2, although, depend-
ing on the objectives of the experiment, other
bases such as wavelength-ship length ratio or
wavelength may be appropriate. The limit of
tank wall interference effects should be indi-
cated on the plot.
For tests in irregular waves, the correspond-
ing wave-energy spectrum should be defined.
When appropriate, performance in irregular
waves should be presented in a nondimensional
form involving a characteristic wave period or
frequency and a characteristic wave height.
The results of statistical analyses should be
presented in histograms to depict probability
density, and as cumulative probability distribu-
tion plots for selected responses.
Tabular presentation of all results should be
made in addition to plots
 ITTC – Recommended 7.5 – 02
07 - 02.1
Procedures Page 6 of 16
Testing and Extrapolation Methods
Effective Date Revision
Loads and Responses, Sea Keeping 2002 01
Sea Keeping Experiments

(ω ) = 155 (ζ )
3. PARAMETERS ~ 2
 − 944 
exp  4 4 (3.3)
W 1/ 3 Y
SJ
3.1 Parameters to be Taken into Account T ω
1
4 5
 T1 ω 
(m²/s)
The amplitudes and phases of hull motions in where
six degrees of freedom   0.191ωT1 − 1  2 
Wave height Y = exp  −   
Accelerations.   2σ  
Relative motion
Rudder angle. and
Encounter angle σ = 0.07 for ω < 5.24 / T1
Leeway (or drift) angle σ = 0.09 for ω > 5.24 / T1
Impact pressures on the hull or on deck
Added resistance in waves This formulation can be used with other char-
Water on deck. acteristic periods by use of
Number of propeller revolutions, torque and
thrust. T1 = 0.924 T-1 = 0.834 T0 = 1.073 T2
Sea spectra
Where
T-1 is energy average period 2πm −1 / m 0
3.2 Recommendations of ITTC For Parame-
ters T0 is the peak period 2π / ω 0
where ω 0 is the circular frequency of the spec-
1975 Performance in irregular waves should trum peak.
be presented in non-dimensional form involv-
T1 is the average period 2πm 0 / m1
ing wave characteristics period and characteris-
tic wave height. T2 is the average zero crossing period,
2π m 0 / m 2
1978 Recommendation for open ocean spectral
formulation:
4. VALIDATION
A −B / ω 4
S( ω ) = e (m²/s)
ω5 4.1 Uncertainty Analysis
where
The detailed procedure of an uncertainty analy-
~ sis is shown in the appendix using the sample
A = 173(ζ W )1 / 3 / T14
2
anlysis of the S-175 ship.
B = 691 / T14 , T1 = 2πm0 / m1

1984 (pp 533-34) Recommendation fng crested 4.4 Benchmark Tests

limited fetch sea spectral formulation:
 ITTC – Recommended
Procedures
Testing and Extrapolation Methods
Loads and Responses, Sea Keeping
Sea Keeping Experiments
1) Seagoing Quality of Ships
(7th 1955 pp.247-293)
A Model of the Todd-Forest Series 60,
Cb=0.60:
7 tanks used 5ft. models, 2 tanks used 10 ft.
models, and 1 tank used 16 ft. model
Froude Numbers 0,0.18,0.21,0.24,0.27 and
0.30
The Ratio wave height to the Length of the
Model: 1/36 1/48 1/60 1/72
for Wave Length 0.75L 1.0L 1.25L 1.5L
2) Comparative Tests in Waves at Three Ex-
perimental Establishments Using the Same
Model
(11th 1966 pp.332-342)
British Towing Tank Panel: A 10 ft. Fibre-
Glass Model of the S.S. Cairndhu
A Series of Experiments on a Ship Model in
Regular Waves Using Different Test Tech-
niques
Data Obtained in Irregular and Transient
Waves and Some Result Predicted by the
Theory (Based on Korvin Kroukovsky's
Work and Employing the Added Mass and
Damping Coefficients Calculated by Grim)
3) Full Scale Destroyer Motion Measurements
(11th l966 pp.342-350)
Full Scale Destroyer Motion Tests in Head
Sea
Comparison among Motion Response Ob-
tained from Full Scale Tests, Model Ex-
periments and Computer Calculations
The Destroyer H.M. "Groningen” of the
Royal Netherlands Navy
A Scale Ration 40 to 1
4) Comparison of the Computer Calculations
of Ship Motions (11th l966 pp.350-355)
ShipResponseFunctionsfortheSeries60
Cb=0.70 Parent Form
7.5 – 02
07 - 02.1
Page 7 of 16
Effective Date Revision
2002 01
5) Computer Program Results for Ship Behav-
iour in Regular Oblique Waves
(11th l966 pp.408~-411)
Series 60, Cb=0.60 and 0.70 Parent Form
DTMB Model 421OW and 4212W
6) Experiments in Head Seas
6-1) Comparative Tests of a Series 60 Ship
Model in Regular Waves
(11th 1966 pp.411-415)
Series 60 Cb=0.60
6-2) Experiments on Heaving and Pitching
Motions of a Ship Model in Regular Longi-
tudinal Waves (11th 1966 pp.415-418)
Series 60 Cb=0.60
6-3) Experiments on the Series 60, Cb=0.60
and 0.70 Ship Models in Regular Head
Waves
(11th 1966 pp.418-420)
Series 60, ~Cb=0.60 and 0.70
6-4) Comparison of Measured Ship Motions
and Thrust Increase of Series 60 Ship Mod-
els in Regular Head Waves (11th 1966 pp.
420-426)
Series 60, ~Cb=0.60 and 0. 70
6-5) Estimation of Ship Behaviour at Sea from
Limited Observation (11th 1966 ~pp.426-
428)
7) Computer Results, Head Seas
7-1) Theoretical Calculations of Ship Motions
and Vertical Wave Bending Moments in
Regular Head Seas (11th 1966 pp. 428-430)
Series 60, Cb=0.70
 ITTC – Recommended
Procedures
Testing and Extrapolation Methods
Loads and Responses, Sea Keeping
Sea Keeping Experiments
7-2) Comparison of Computer Program Results
and Experiments for Ship Behaviour in
Regular Head Seas (11th l966 pp.430-432)
Series 60, Cb=0.60 and 0.70
7-3) Computer Program Results for Ship Be-
haviour in Regular Head Waves
(11th 1966 pp.433-436) Series 60,
~Cb=0.60 and 0.70 Parent Form DTMB
Model 421OW and 4212W
7-4) Comparison of Calculated and Measured
Heaving and Pitching Motions of a Series
60, Cb=0.70 Ship Model in Regular Longi-
tudinal Waves (11th l966 pp.436-442)
Series 60, Cb=0.70
7-5) Computer Calculations of Ship Motions
(11th 1966 pp.442)
7-6) Comparison of the Computer Calculations
of Ship Motions and Vertical Wave Bending
Moment (11th 1966 pp. 442-445)
Series 60, Cb=0.60 and 0.70
8) Comparison of the Computer Calculations
for Ship Motions and Seakeeping Qualities
by Strip Theory (14th 1975 Vol.4pp.341-
350)
A Large-Sized Ore Carrier
9) Comparison on Results Obtained with
Computer Programs to Predict Ship Mo-
tions in Six Degrees of Freedom
(15th 1978 pp. 79-90)
S-175,Cb = 0. 572
10) Comparison of Results Obtained with
Compute Programs to Predict Ship Mo-
7.5 – 02
07 - 02.1
Page 8 of 16
Effective Date Revision
2002 01
tions in Six-Degrees-of-Freedom and
Associated Responses(16th 1981 pp.217-
224)
To Identify the Differences in the Vari-
ous Strip Theories and Computation
Procedures utilised by the Various Com-
puter Programs and Provide Guidance
for Improvement if Necessary
S-175 Container Ship for Fn = 0.275
11) Analysis of the S-175 Comparative Study
(17th 1984 pp.503-511)
12) S-175 Comparative Model Experiments
(18th 1987 pp.415-427)
13) Rare Events
(19th 1990 pp.434-442, Seakeeping)
14) Validation, Standards of Reporting and
Uncertainty Analysis Strip Theory Predic-
tions ( 19th 1990 pp.460-464)
15) ITTC Database of Seakeeping Experi-
ments (20th 1993 pp.449-451)
Two Dimensional Model, Wigley Hull
Form, S-175
16) Validation of Seakeeping Calculations
(21st 1996 pp.41-43)
Basic Theoretical Limitations
Numerical Software Engineering Aspects
17) ITTC Database of Seakeeping Experi-
ments (21st 1996 pp.43)
S-175, High Speed Marine Vehicle
 ITTC – Recommended 7.5 – 02
07 - 02.1
Procedures Page 9 of 16
Testing and Extrapolation Methods
Effective Date Revision
Loads and Responses, Sea Keeping 2002 01
Sea Keeping Experiments

APPENDIX the photo coupler during the time period t. The

A.1 EXAMPLE OF AN UNCERTAINTY
ANALYSIS
An example of an uncertainty analysis based
on the S-175 ship is shown.
A.1.1 Analysis of Elemental Errors
Accuracy of Model Geometry and Weight Dis-
tribution. The error sources in model geometry
are the length between perpendicular (LPP),
breadth (B) and draft (d) as shown in Table 3.
The error limits for the first two items are esti-
mated from model manufacturing errors. The
error in draft comes from the error in model
displacement. For the execution of motion test,
the weight should be distributed properly inside
model ship to satisfy predetermined value of
vertical centre of gravity, KG, and longitudinal
radius of gyration, kyy. Swinging table type
device is used for the determination of KG and
kyy. Tables 3 and 4 show the error limits of
above mentioned error sources and error limits
for KG and kyy. The detailed procedure of ob-
taining these values can be found in Yum et al.
(1993).
Model Speed. The error limit in model
speed is estimated following the procedure
suggested by Fogash (1992). Under the as-
sumption that model speed, V., through the
water is equal to the speed of towing carriage,
the model speed is determined from
Vm =
(n / 5000)πD = fπD (1)
t 5000
where D(m) is the diameter of carriage wheel
and n is the number of light pulses sensed by
measured quantities and error sources for the
estimation of model speed and error limit are
the diameter of carriage wheel and the pulse
frequency f(= n/t). The results are summarised
in Table 3.
Measuring Device and Calibration Errors.
The measurement items in regular wave motion
test are the wave encounter frequency, ( ω e ,
wave amplitude, ζ a , heave amplitude, H, and
phase, α H , and pitch amplitude, θ , and
phase, α θ . The servo needle type wave probe is
used for the measurement of wave and 4-
component motion measuring device is used
for heave and pitch motions. The equation used
for the analysis of wave amplitude is
ζ p −ζt
.ζ a = (2)
2
where subscript p and t represent wave peak
and wave trough respectively. Using Eq. (2),
the bias limit of wave amplitude, Bζ a , becomes
 ∂ζ
2

2
 ∂ζ 
( )
Bζ a =  a Bζ p  +  a Bζ t 
2
 ∂ζ  ∂ ζ
(3)
 p   t 
∂ζ a ∂ζ a
+ 2 ρ pt Bζ p Bζ t
∂ζ p ∂ζ t
If we consider the fact that the peak and
trough of wave are continuously measured us-
ing one servo needle type wave probe, the bias
limit of peak value Bζ p , and the bias limit of
trough value, Bζ t , are perfectly correlated ( ρ pt
= 1.0) and equal in their magnitudes. This
means that the value of Eq. (3) becomes zero
and furthermore the bias error of the wave
probe does not influence the bias limit in
measured wave amplitude at all. The same
 ITTC – Recommended 7.5 – 02
07 - 02.1
Procedures Page 10 of 16
Testing and Extrapolation Methods
Effective Date Revision
Loads and Responses, Sea Keeping 2002 01
Sea Keeping Experiments

conclusion can be made for heave and pitch

measuring transducers. Signal analysis There are many different
methods for the analysis of measured regular
Other error sources in measuring devices signals in motion test. Two most frequently
are calibration errors. These errors are repre- used methods among them are the method of
sented by Fourier Transform and the method of peak-to-
peak counting. For the present analysis peak-
1 M to-peak method is used. In this method, the
SEE 2 = ∑ ( yk − y LS ,k )2 . (4) peak and trough values are searched and ana-
M − C k =1
lysed to get the amplitude and frequency of a
measured signal. Using peak-to-peak method,
M: number of data used for calibration
the wave amplitude, heave and pitch ampli-
C: number of variable for fitting function
tudes and wave encounter frequency and period
Yk: data
are obtained by following equations.
YLS,k: fitted value

SEE (standard error of estimate) is considered

ζ p −ζt
as precision limit because this value is obtained ζa = (5)
by a statistical analysis of repeated independent 2
measurements.

All the errors in measuring devices consid- H p − Ht

H= (6)
ered so far are errors in the outputs for static 2
inputs. But, for the case of dynamic inputs like θ p − θt
motion test, dynamic response errors need to be θ= (7)
considered. The traditional way to investigate 2
2π
the dynamic response of an instrument is to ωe = (8)
consider the differential equation that describes T
T = ∑ (T p ,n +1 − T p ,n )
output. And measuring instruments are classi- 1 N
(9)
fied by the order of the governing differential N n =1
equation. The potentiometers used for the
measurement of heave and pitch amplitudes are The bias errors which occur during the pro-
zero-order instruments in which there are no cess of amplitude analysis of wave, heave and
errors in the output due to the dynamic re- pitch are obtained by using regular sinusoidal
sponse. The characteristic (order) of the servo waves which have same amplitudes and fre-
needle type wave probe is not known clearly. quencies as the measured signals. The errors
In the present study 0.5 mm is used for the dy- related to the half of data sampling interval are
namic error of the servo needle type wave used to obtain these bias errors. These errors
probe (Hirayama et al, 1988). This error is con- are known to be asymmetric.
sidered to be asymmetric error because the
measured values of wave amplitudes are al-
ways smaller than the true values. Table 3 lists
all the error limits of measuring devices.
 ITTC – Recommended 7.5 – 02
07 - 02.1
Procedures Page 11 of 16
Testing and Extrapolation Methods
Effective Date Revision
Loads and Responses, Sea Keeping 2002 01
Sea Keeping Experiments

The precision errors of wave amplitude, variables are obtained analytically using ship
heave and pitch amplitudes and wave encoun- motion program based on the strip method
ter frequency are obtained by (Park and Kang, 1989).

1 The heave RAO, bias limit, precision limit

 ∂ζ 
2
 ∂ζ 
2
2 and overall uncertainty are
Sζ a =  a S ζ p  + a S 
 ∂ζ ζ t 
 (10)
 ∂ζ p 
  t  
  H
1 H'= (14)
 ∂H 2 2
 2
ζa
  
S H =  SH p  +  ∂H S H   (11)
 ∂H p   ∂H 
 
t

  t
  ∂H ' ± 
2 2
 ∂H ' ± 
1 B ±
=  Bζ a  +  BH  (15)
 ∂ζ a
H'
 ∂θ   ∂H 
 
2 2 2
 
S ζ a =  S  +  ∂θ S   (12)
 ∂θ p θ p   ∂θ θ t  
   t   2 2
 ∂H '   ∂H ' 
1 PH ' =  Pζ a  +  PH  (16)
 ∂ω e  
2 2  ∂ζ a   ∂H 
S ω e =  ST   (13)
 ∂T  
where S ζ p and S ζ t are the precision index for U H± ' = (B ) + (P )
± 2
H' H
2
(17)
the mean value of wave peaks and troughs re-
spectively. The similar notations are used for
The pitch RAO, bias limit, precision limit
heave and pitch motions. S T is the precision
and overall uncertainty are
index for the mean value of peak-to-peak peri-
ods. θ
θ'= (18)
 360ζ a 
 
A.1.2 Response Amplitude Operator (RAO)  λ 
and Overall Uncertainty 2 2 2
 ∂θ ' '   ∂θ ' ±   ∂θ ' ± 
B = 
±
θ '' Bλ  +  Bθ  +  B ζ a 
This section describes the procedure of  ∂λa   ∂θ   ∂ζ a 
combining all the errors obtained in A.1.1 and (19)
data reduction equations (DRE's) to get heave 2 2 2
and pitch RAO's and their overall uncertainties.  ∂θ '   ∂θ '   ∂θ ' 
Pθ ''' =  Pλ  +  Pθ  +  Pζ a 
 ∂λ   ∂θ   ∂ζ a 
For the variables such as LBP, B, d, KG, kyy
and model speed (Vm), the data reduction equa-
tions which describe the functional dependency U θ±' = (B ) + (P )
± 2
θ' θ'
2
(21)
of the heave and pitch RAO's to these variables
cannot be found as functional forms. Therefore,
the error propagation coefficients for these
 ITTC – Recommended 7.5 – 02
07 - 02.1
Procedures Page 12 of 16
Testing and Extrapolation Methods
Effective Date Revision
Loads and Responses, Sea Keeping 2002 01
Sea Keeping Experiments

where the value of wavelength, λ , and its bias where ∂H ' / ∂ω ' and ∂θ ' / ∂ω ' are the slopes of
limit, B λ , and precision limit, Pλ can be ob- the heave and pitch RAO's respectively, as a
tained using the relation between the absolute function of ω .
frequency, ω , and the encounter frequency,
ω e , and wave dispersion relation in deep wa-
A.1.3 Results
ter.
Table 5 contains summary of the overall
For the comparison of RAO, nondimen- uncertainties of heave and pitch RAO's. The
sionalized absolute frequency, ω L / g , is ratios of overall uncertainties to heave RAO's
used as an independent variable. The value of range from ±2.0% to ±3.5% and for pitch
nondimensionalized absolute frequency, its RAO's they range from ±3.5% to ± 6.0%. At
bias and precision limits and overall uncer- higher encounter frequencies, as the magni-
tainty are tudes of the heave and pitch RAO's decrease,
the uncertainty limits tend to increase signifi-
2 2 cantly.
 ∂ω '   ∂ω ' 
Bω ' =  BV  +  BL  (22)
 ∂V   ∂L  Fig.1 and Fig.2 show the final combined
uncertainty limits of the heave and pitch
2 2 2 RAO's. Presented in these figures are the
 ∂ω '   ∂ω '   ∂ω ' 
Pω ' =  PV  +  PL  +  Pω e  analysis results of the RAO's using the strip
 ∂V   ∂L   ∂ω e  theory indicated by the triangles (Park and
(23) Kang, 1989). The computed values of the
heave and pitch RAO's by the strip method
were found to lie far outside of the uncertainty
Uω' = (Bω ' )2 + (Pω ' )2 (24) limits of 95% confidence level. These differ-
ences between the test results and analysis re-
sults have been noticed in the report of the 15th
Combining Eqs. (17), (21) and (24), the fi- and 16th ITTC Seakeeping Committee Com-
nal overall uncertainties for heave and pitch parative Study on Ship Motion Program (1978,
RAO's become 1981).

2
 ∂H '
~
U H± ' = (U )
± 2
+

Uω'  (25) A.1.4 Conclusions
 ∂ω '
H'

1) Uncertainty analysis has been successfully
 ∂θ '
2
applied to the motion test in regular waves.
~
U θ±' = (U )
± 2
θ' +

Uω'  (26)
 ∂ω '  2) 95% confidence intervals for heave and
pitch RAO's were approximately ± 3.0%
and ± 5.0% respectively.
 ITTC – Recommended 7.5 – 02
07 - 02.1
Procedures Page 13 of 16
Testing and Extrapolation Methods
Effective Date Revision
Loads and Responses, Sea Keeping 2002 01
Sea Keeping Experiments

3) The analysis results of motion RAO's were

found to lie outside of the uncertainty lim-
its of 95% confidence level of motion test.
Measurement
4) The extension of present uncertainty L, B, T (m) Model Dimension
analysis to a more complex situations such (Lpp, Beam, Draft)
as motion tests in irregular waves are ex- V (m/s) Carriage speed
pected KG (m) Distance from keel to
centre of gravity
Model Ship S-175 Kyy (m) Pitch radius of gyra-
LPP (m) 4.50 tion
B (m) 0.6531 ζa (m) Wave amplitude
D (m) 0.396
ωe (rad/s) Encounter frequency
T (m) 0.2443
Displacement Volume (m³) 0.4101 H (m) Heave amplitude
LCB (m) 0.0364 θ (deg) Pitch amplitude
CB 0.5716
GM 0.0257 Analysis
KM 0.2705 λ (m) Wave length
Kyy/LPP
KG
0.0364
0.2448
(
ω' = ω L / g ) Non-dimensional fre-
quency
Table A.1. Principal particulars of model ship H ' (= H / ζ a ) Heave RAO
Scale 38.889
θ ' (= (θ / ζ a )λ / 360) Pitch RAO
Table A.2. Objects of uncertainty analysis for
motion
 ITTC – Recommended 7.5 – 02
07 - 02.1
Procedures Page 14 of 16
Testing and Extrapolation Methods
Effective Date Revision
Loads and Responses, Sea Keeping 2002 01
Sea Keeping Experiments
 ITTC – Recommended 7.5 – 02
07 - 02.1
Procedures Page 15 of 16
Testing and Extrapolation Methods
Effective Date Revision
Loads and Responses, Sea Keeping 2002 01
Sea Keeping Experiments
 ITTC – Recommended 7.5 – 02
07 - 02.1
Procedures Page 16 of 16
Testing and Extrapolation Methods
Effective Date Revision
Loads and Responses, Sea Keeping 2002 01
Sea Keeping Experiments