E Book Hydronav2021

1
Resistance and Propulsion
Hydrodynamic state of the art review: rotor-stator marine propulsor system design. Przemysław Król....................................4
PLASTEX Sprint Boats – Research Investigations 1998-2020. Tomasz Bugalski........................................................................................20
Short review on evolution of the actual Ice class requirements for ship propulsion
and its implementation by PRS. Marek Narewski....................................................................................................................................................... 28
Ship manoeuvrability
Ilawa Ship Handling Research and Training Centre: the establishment of the centre,
and its present day. Jacek Nowicki....................................................................................................................................................................................... 34
Towards the Maritime Autonomous Surface Ship (MASS). Anna Miller, A. Rak, M. Rybczak.............................................................37
Numerical assessment of hull hydrodynamic derivatives in ship maneuvering prediction.
Radosław Kołodziej, Paweł Hoffman.................................................................................................................................................................................. 38
Influence of propulsion-steering system on the position of pivot point on a large LNG carrier during
standard manoeuvres. Maciej Reichel............................................................................................................................................................................... 39
Investigation of heading and manoeuvring properties of stern trawler B-280 with Hubble type rudder blade.
Paweł Dereszewski, Włodzimierz Hebel, Paweł Dymarski..................................................................................................................................... 44
Advanced computational and experimental modeling
Measurement Stand for Experimental Model Tests of the Dynamically Positioned Vessels.
Marek Kraskowski, Piotr Grymajło........................................................................................................................................................................................ 46
Digital twins concept in the activities of Classification Societies. Janusz Narewski.............................................................................. 55
Quasi-real verification of autonomous navigation. Zbigniew Pietrzykowski. Piotr Wołejsza ........................................................ 65

Mesh dependence study for numerical assessment of hydrodynamic characteristics of windsurfing fin.
Hanna Pruszko..................................................................................................................................................................................................................................66
Ship seakeeping
Simulation of fishing vessel motion for seakeeping analysis.

Monika Warmowska–Gadzińska, Andrzej Laskowski, Henryk Purowski, Maja Malinga.........................................................................73
Prediction of ship motions in irregular wave based on response amplitude operator evaluated experimentally
in filtered noise waves. Sebastian Bielicki.........................................................................................................................................................................81
Review of selected methods for prediction of added resistance in following waves.
Michał Kalbarczyk, Sebastian Bielicki................................................................................................................................................................................. 93
Analysis of lifeboat emergency launching from a passenger ro-ro ferry during its motions in stochastic waves.
Andrzej Jankowski........................................................................................................................................................................................................................104
Offshore structures. Dynamics of the marine environment. Hydroacoustics.
Impact of steep waves underneath a horizontal deck. Dawid Majewski, Wojciech Sulisz................................................................113
Numerical modeling results of harbor wave penetration versus in-situ measurements.
Tomasz Marcinkowski, Piotr Szmytkiewicz.................................................................................................................................................................... 114
Modeling the dynamics of the marine environment for the design of offshore structures. Magdalena Kempa................ 115
Modeling the dynamics of a floating wind turbine in the southern Baltic Sea area with a depth of 84 m.
Preliminary analyzes. Paweł Dymarski, Ewelina Ciba.............................................................................................................................................. 116
Analysis of the recorded underwater hydroacoustic signal of a cargo ship. Katarzyna Warnke-Olewniczak....................... 117
2
Resistance
and Propulsion
3
22nd International Symposium on Hydrodynamics in Ship Design, Safety,
Manoeuvring and Operation
22nd International Symposium on Hydrodynamics

HYDRONAV 2020 in Ship Design, Safety,
Manoeuvring
Sopot and Operation
- Iława, 14-16 October 2020
HYDRODYNAMIC STATE OF ART REVIEW: ROTOR – STATOR

MARINE PROPULSOR SYSTEMS DESIGN
PRZEMYSŁAW KRÓL1
HYDRONAV 2020
Sopot
1 Maritime Advanced Research Centre - Iława,
(CTO 14-16
S.A.) Gdańsk, Poland,October 2020
[email protected]
HYDRODYNAMIC STATE OF ART REVIEW: ROTOR – STATOR

The paper covers importantMARINE
topic of rotorPROPULSOR
– stator propulsor systemSYSTEMS DESIGN
design and operation. For stand – alone marine screw
propeller both design criteria for loading distribution as well as theoretical efficiency limits are well described in basic literature.
It is in contrast to combined propulsor system like a propeller cooperating with a pre swirl device. The paper describes
nowadays state of art, summarising results obtained by various Researchers by installing Energy Saving Devices on particular
vessels. Utilised design methods
PRZEMYSŁAW KRÓLare 1 briefly outlined, with underlined main characteristics. Rough analysis of gathered data
confirms expected trend that higher efficiency gain due to ESD installation if possible fir higher propeller loading.
Keywords: Energy Saving Device, Guide Vane, Pre-Swirl Stator, Propulsor Design, Efficiency
1Maritime Advanced Research Centre (CTO S.A.) Gdańsk, Poland, [email protected]
The1.paper
Introduction
covers important topic of rotor – stator propulsor system design and operation. For stand – alone marine screw
propeller both design criteria for loading distribution as well as theoretical efficiency limits are well described in basic literature.
It isVast literature
in contrast exists that
to combined coverssystem
propulsor topiclike
of marine propulsor
a propeller design
cooperating with and
a preits optimization.
swirl device. The Due
paper to rising
describes
ecological
nowadays staterequirements
of art, summarisingmore attention
results obtained byisvarious
being paid nowadays
Researchers by installing to theSaving
Energy propulsor
Devices efficiency.
on particular
vessels. Utilised design
Consequently Energy methods
Saving are briefly
Devicesoutlined,
(ESDs)with underlined
are gaining mainmore
characteristics.
and more Rough analysisof
interest of gathered
subsequent data
confirms expected trend that higher efficiency gain due to ESD installation if possible fir higher propeller loading.
Researchers and Marine Engineers. Important types of these are pre- and post swirl devices, meant to
recover rotational energy losses occurring during the screw propeller operation. This paper addresses
Keywords: Energy Saving Device, Guide Vane, Pre-Swirl Stator, Propulsor Design, Efficiency
these ESDs, of various forms like pre swirl stators (PSS), pre swirl ducts (PSD) and post swirl stators
(PostSS).
The paper is divided into three parts: in the first one most important features of design methods met
in1.literature
Introduction
are sketched. Employed fluid flow models are named and ESD design criteria are
mentioned. It is however to be said that vast majority of modern studies adopt CFD computations
Vast literature
combined exists thatoptimization
with variational covers topicofofparametrically
marine propulsor designESD.
described and its optimization. Due to rising
ecological
The second part of the paper is the largest one and summarises mosttoimportant
requirements more attention is being paid nowadays the propulsor efficiency.
results reported in
Consequently
each article. InEnergy
the mostSaving
of studiesDevices (ESDs) areand
both numerical gaining more and
experimental datamore
were interest ofInsubsequent
available. such case
Researchers
numerical andwere
data Marine Engineers.
mostly omitted Important typesdue
in this review, of to
these
spacearelimitation.
pre- and post swirl devices, meant to
recover rotational
The last, energypart
concluding losses occurring
of this reviewduring
presents theresults
screw reported
propellerfromoperation. This
selected paperinaddresses
papers common
these ESDs, of various forms like pre swirl stators (PSS), pre swirl ducts (PSD)
base – declared efficiency increase along with propeller loading. In fact ‘selected papers’ means and post swirl stators
here
(PostSS).
‘all for which it was possible’. The latter is due to huge variety of data presented in the literature. In
The paper is divided into three parts: in the first one most important features of design methods met
in literature are sketched. Employed fluid flow models are named and ESD design criteria are
1
mentioned. It is however to be said that vast majority of modern studies adopt CFD computations
combined with variational optimization of parametrically described ESD.
The second part of the paper is the largest one and summarises most important results reported in
each article. In the most of studies both numerical and experimental data were available. In such case
numerical data were mostly omitted in this review, due to space limitation.
The last, concluding part of this review presents results reported from selected papers in common
base – declared efficiency increase along with propeller loading. In fact ‘selected papers’ means here
‘all for which it was possible’. The latter is due to huge variety of data presented in the literature. In
1
4
22nd International Symposium on Hydrodynamics in Ship Design, Safety, Manoeuvring and Operation HYDRONAV 2020
some papers detailed results of model tests, along with vessel and propeller particulars and full scale
prediction are given. In sharp opposition there are also publications where only given value is percentage
of delivered power reduction – with hardly any dimensional particulars provided.
2. Design methods for stator type Energy Saving Devices
Stand – alone screw propeller has wide literature and huge variety of well – recognised methods
already exists in the literature. This is somewhat contradictory to the propeller – stator propulsor system.
Due to increased complexity of flow phenomena occurring in such system there are much more methods
existing in the literature for design of such system. None of them however can be stated canonical or as
reliable as classical approaches for propeller operating alone. Reported methods can be distinguished
mainly by employed flow representation; it can be either inviscid vortex flow model or a viscous one
(mainly RANSE). Second factor is stator’s particular that undergoes a design process: it can be either
bound circulation distribution (in such case stator foils geometry results directly from this) or geometry
optimization. In the latter case information on bound circulation is most often omitted in the papers and
it is in fact the only stator design concept available for RANSE – based approach.
Table 1 Design approach in selected papers
Paper Designed particular Design criteria Flow model
Aryawan et al. 1999 Circulation distribution Rotation cancellation Vortex
Celik, Guner 2006 Circulation distribution Rotation cancellation Vortex
Chen et al. 2016 Geometry Optimization Viscous
Coney 1989 Circulation distribution Optimization Vortex
Hooijmans et al. 2010 Circulation distribution Optimization Vortex
Hou et al. 2015 Circulation distribution Rotation cancellation Vortex
Hyun-Joon et al. 2013 Geometry Optimization Viscous
Kerwin et al. 1988 Circulation distribution Optimization Vortex
Joon-Hyoung et al. 2017 Geometry Optimization Viscous
Kim et al. 2004 Circulation distribution Optimization Vortex
Kim et al. 2015 Geometry Optimization Viscous
Krasilnikov et al. 2019 Geometry Optimization Viscous
Król 2019 Circulation Optimization Vortex
Kwi-Joo et al. 2014 Geometry Optimization Viscous
Nielsen et al. 2019 Geometry Optimization Viscous
Simonsen et al. 2012 Geometry Streamline adjustment Viscous
Xing-Keading et al. 2015 Geometry Optimization Vortex
Yong-Jin et al. 2015 Circulation distribution n/d Viscous
Among the vortex models the most basic tool is a Lifting Line Model (LLM), replacing stator foils
with straight line bound vortices. It is sufficient simplification for determination of induced tangential
velocity coming from the stator, which is in fact the vital part from cooperating propeller point of view.
In such approach stator vortex wake deformation can be taken into account, what allows more accurate
determination of stator – induced axial velocity component (Król 2019). Stators, with foils of low aspect
ratio, may also be successfully designed with LLM. Such approach seems however to be highly
outmoded nowadays as already in 1988 Kerwin used Lifting Surface Model (LLS) for determination of
stator foils geometry for given circulation distribution.
Decision on stator bound circulation magnitude is a vital one for the efficiency of the propulsor
system. Due to the lack of theoretical criterion – variational optimization is done in most cases.
Most advanced vortex flow model, Boundary Element Method (BEM) can not be directly utilised in
determination of foil geometry for prescribed loading. Thus it is rather met in design via variational
geometry optimization and / or analysis of given propulsor system geometry.
Vortex methods have several advantages which encourage many designers to use them, especially at
initial stages of work. Easiness of individual code implementation (comparing to RANSE
computations), relatively short calculation time and possibility of reverse problem solution stand main
advantages of these methods. They have however serious limitations also. Vortex wake deformation
that could affect designed foils geometry theoretically can be computed directly – in reality it requires
much care and may pose a difficulty in case of heavily loaded stators, especially if the propeller –
9
5
induced velocity field is to be taken into account during the simulation. Moreover there are certain
phenomena, like flow separation and vorticity dissipation, which can not be included directly within the
vortex model and have to be handled with semi – empirical formulas or other way. Flow separation is
likely to occur on highly loaded profiles. Within frame of vortex model it is possible to design a section
with arbitrary lift force coefficient (in fact one can receive a convergent profile geometry solution even
for NACA a=1.0 circulation distribution) but exceeding reasonable values will result in fail to realize
assumed operation parameters in real fluid. Finding such limit may be done upon systematic blade
section profiles data, like collected in (Abbot 1959). The basic factor allowing the stator to make it’s
influence on the propeller rotor is it’s vortex wake. The intensity of shedding vortices is however
decreasing with the distance from the stator, due to dissipation. According to this, vortex flow models
predictions on propeller – stator cooperation are limited to rather short distances between them.
Once RANSE (or other viscous flow model) is adopted, most of mentioned difficulties are driven
away; vortex wake shape, separation and vorticity dissipation are included directly in the simulation.
Meanwhile however all advantages favouring vortex models are gone alike. Viscous calculations require
notable computer resources and – what is most disadvantageous – reverse problem solution is no longer
possible. One has to employ rather parametric optimization of both stator and propeller geometries for
prescribed operating conditions. It may notable increase required number of analyzed cases.
Due to issues outlined above – most reasonable approach seems to begin the design process with
vortex model to determine initial geometry of the system. Once it is given, the optimization process may
be finalized by viscous computations, conducted for narrowed number of cases.
3. Energy saving effect
Considerable number of papers upon propeller – stator topic are published, most of them reporting
delivered power reduction, provided by ESD of Authors design. Energy saving effect is determined
either by means of numerical calculations or model tests – in some cases sea trials results are even
available. Due to wide differentiation of available data completeness (in some cases propeller diameter
or ship speed are not even provided) it was impossible to construct comprehensive table that would
summarize reported results. Accordingly subsequent papers are briefly outlined below.
In most of papers there were both numerical and experimental results available. In such case only
empirical data are repeated below, as main focus of this study is ESD operation, not CFD development.
Corresponding data may be found in original papers.
Due to requirements of further analyses there are some additional values calculated below. These
are mainly standard propeller loading coefficients like KT, KQ and CT, CN. After some considerations,
an effective advance speed for calculation of CT, CN was calculated with use of advance ratio / effective
wake coefficient determined via model tests with ESD installed (if such value was available). Such
approach was taken because in Author’s feeling it better reflects propeller loading conditions, even if
is not formally equal to mean axial velocity experienced by operating propeller. Unless otherwise stated
– water density was taken equal ρM = 998.6kg/m3 for model scale and ρS = 1025.9kg/m3 for full scale.
In the paper (Celik et al. 2007) the topic of Post Swirl Stator is described. Vortex representation of
the lifting blades is utilised. The propeller is designed individually and then the stator is designed as a
retrofit and the study: “therefore neglects the effect of the stator in the optimization process of the
propeller”. The design criteria for the stator demands cancellation of the fluid rotation passing through
it. This criteria gives direct relation between bound circulation on the stator, with respect to the one of
the propeller.
Two design examples are presented in the paper, for the “Trawler” and for the “Tanker”. The
particulars of the vessels are not provided; only operating parameters assumption are given:
Table 1 Assumed operation parameters
PDS VS nS D Z KT
[kW] [knots] [rpm] [m] [-] [-]
Trawler 648 10.0 380 2.12 4 0.1042
Tanker 1863 12.5 247 2.84 4a 0.1811
a – as stated in Table 1. of the original paper; in Figure 5. three-bladed propeller is visible
9
6
FVM simulation for the full scale was carried out to analyse designed propulsors (propeller + stator).
For two vessels following operation profile was determined at their design points:
Table 2 FVM simulation results
KT KQ η Ja CTa CNa
[-] [-] [-] [-] [-] [-]
Trawler 0.1042 0.0117 0.437 0.308 2.792 6.388
Tanker 0.1811 0.0232 0.557 0.448 2.294 4.119
a – calculated based on data provided in original paper
The paper (Chen et al. 2016) begins with presentation of data for validation case, which is Japanese
Bulk Carrier (JBC).
Table 3 Particulars of Japanese Bulk Carrier (JBC) and its propeller diameter
λ LPP LWL B T D
[-] [m] [m] [m] [m] [m]
Full scale 1 280.0 285.0 45.0 16.5 8.12
Model 40 7.000 7.125 1.125 0.413 0.203
For speed around VS = 7.5knots (Fr = 0.142, as given in the paper) the model test prognosis for
delivered power is PDS = 2863kW. Corresponding test after installing ESD (Accelerating Duct) brought
delivered power prognosis of PDS = 2691kW. Resulting power reduction of 6.03% was considered to
be in sufficient agreement with one determined via Finite Volume calculations (5.15%).
FVM simulation was applied to design via optimization the geometry of ESD for another vessel,
the Tanker:
Table 4 Particulars of Tanker and its propeller diameter
λ LOA B T D
[-] [m] [m] [m] [m]
Full scale 1 333.0 60.0 20.5 10.5
Model 50 6.660 1.200 0.410 0.210
For speed around 8.1knots (Fr = 0.141, as given in the paper) the model test prediction for delivered
power is PDS = 2813kW. Corresponding test after installing ESD (Pre Swirl Duct) brought delivered
power prediction of PDS = 2716kW. Resulting power reduction of 3.45% is closer to the one,
determined via FVM simulation (2.95%) than in case of JBC.
Table 5 Results of the model tests
VS PDS
[knots] [kW]
Trawler No ESD 2863
7.5
Trawler AD 2691
Tanker No ESD 2813
8.1
Tanker PSD 2716
The dissertation (Coney 1989) is not focused on Pre Swirl Stator issue, which is discussed along
with other propulsor systems. Circulation distribution between stator and the propeller is derived via
iterative solution of equations system, constructed for goal function optimization (minimum power for
prescribed thrust and speed). The design example is only briefly outlined, with bound circulation given
on propeller and stator. Design requirement is defined as to develop thrust loading coefficient CT =
0.69 during propulsor operation at advance ratio of J = 0.69. The paper presents obtained propulsor
loading, operating in uniform flow of inviscid fluid:
9
7
Table 6 Propulsor operation in uniform flow, inviscid fluid

KT KQ η Ja CTa CNa
[-] [-] [-] [-] [-] [-]
Propeller 0.2216 0.0375 - -
- -
Stator -0.0070 - - - - -
Propeller+Stator 0.2146 0.0375 0.811 0.8904 0.689 0.850
Equivalent propeller 0.2146 0.0389 0.781 0.8895 0.691 0.884
In the paper (Hooijmans et al 2010.) the Pre Swirl Stator is designed via Lifting Line Method,
which is applied for optimization of bound circulation distribution. For avoidance of flow separation
on the stator, local value of lift coefficient CL is kept below prescribed margin (precise value is not
given in the paper). For the design case a container vessel was adopted (vessel particulars not provided,
propeller diameter D = 8.75m). The model tests were conducted for Pre Swirl Stator of diameter 2Λ =
9.10m and two propellers and revealed following results:
Table 7 Results of Pre Swirl Stator application
Propeller D Z ΔPD ΔT nSBARE nSESD
[m] [-] [%] [%] [rpm] [rpm]
P1 8.75 6 -2.5% +10% 100.2 96.0
P2 9.00 4 -5.0% n/a 100.2 96.0
In the paper (Hou et al. 2015) a Post Swirl Stator is analysed. It is designed with use of Vortex
Lattice Model. The design criteria states that rotation of fluid passing the stator plane should be
cancelled, what (upon Stokes theorem) directly gives ratio between propeller and stator bound
circulation at corresponding radius. The stator diameter is smaller than the propeller’s, due to
contraction. The contraction ratio is determined directly from ideal propulsor model and axial induced
velocity.
As the design example a Container Ship was adopted:
Table 8 Particulars of Container Ship and its propeller
λ LPP LWL B T ∇ D
[-] [m] [m] [m] [m] [m3] [m]
Full scale 1 286.00 290.60 48.20 14.80 143390 9.60
Model 50.794 5.631 5.721 0.949 0.291 1.094 0.189
The required thrust force is defined as “roughly T = 2910416.35N”. The model tests revealed
following propulsor operation, for the design speed of VS = 22.0knots (11.32m/s, as given in the paper):
Table 9 Propulsor operation at V = 22.0knots
KT KQ η Ja CTa CNa
[-] [-] [-] [-] [-] [-]
Propeller 0.2231 0.0345 -
Stator 0.0163 - -
Propeller+Stator 0.2391 0.0345 0.797 0.7226 1.166 1.463
Equivalent propeller 0.2201 0.0344 0.736 0.7228 1.073 1.458
Declared difference between propeller-stator system and equivalent propeller from KT point of view is
probably result of altered thrust deduction factor.
The paper (Hyoun-Joon eta al 2013) begins with analysis of the efficiency losses existing on the
marine propeller:
9
8
Table 10 Particulars of analyzed propeller and its operation

D P0.7/D AE/A0 Z J KT ηi Axial losses Rotational losses CTa
[m] [-] [-] [-][-] [-] [-] [-] [-] [-]
0.40 0.1855 0.6038 0.3308 0.0355 2.952
10.0 0.714 0.471 4 0.45 0.1630 0.6675 0.2718 0.0312 2.050
0.50 0.1401 0.7285 0.2183 0.0268 1.427
Axial losses were determined via ideal propulsor theory. Rotational losses were determined as a
remaining factor between real propeller efficiency and the ηi determined via BEM calculations with no
viscosity effects included.
The design of a Pre Swirl Duct was realized via variational optimization by means of FVM
simulation. The vessel adopted for the design example was VLCC
Table 11 Particulars of VLCC and its propeller
λ LPP B T D
[-] [m] [m] [m] [m]
Full scale 1 322.00 60.00 21.00 10.00
Model 43 7.488 1.395 0.488 0.233
The model tests were conducted both for ‘bare hull’ conditions and with ESD installed:
Table 12 Model test results
VS = 15.6knots T n ηD KT CTa
[%] [%] [%] [-] [-]
No ESD 100.0 100.0 100.0 0.1630 2.050
PSD 99.4 97.3 105.1 0.1855 2.952
a – correlated with KT from table 10
Precise value of ηD was not given in the paper; only its increase of 5.1% was declared. Thus value of ηD
= 100% for ‘bare’ case does not mean that PD = PE, but only taking it as a reference level.
In the paper (Joon-Hyoung et al. 2017) the topic is combined ESD, consisting of Wavy Twisted
Rudder (WTR), Tip Raked Propeller (TPR) and pre swirl stator (PSS). The PSS design is conducted via
potential code in the initial stage and is finalized with FVM. The design criteria is to minimize its
resistance. The design case, adopted as the example, is the Kriso Container Ship (KCS).
Table 13 Particulars of Kriso Container Ship (KCS)
λ LPP LWL B T CB
[-] [m] [m] [m] [m] [-]
Ship 1 230.0 232.5 32.2 10.8
0.651
Model 39.5 5.82 5.89 0.81 0.27
The design speed is defined as VS = 24.00knots. In the first part of the paper, the model tests with various
rudder types are conducted to reveal delivered power dependency:
Table 14 Rudders influence on propeller delivered power (model scale values)
Rudder type 2πQMnM
[W]
Full Spade 80.40
Twisted 79.16
WTR 78.92
Upon these results the WTR was adopted for further work. Subsequent model tests were conducted with
each ESD (PSS, TPR and WTR) installed alone and with all of them together (Combined ESD):
9
9
Table 15 Model test results with subsequent ESDs:

RTM QM nM 2πQMnM
[N] [Nm] [rps] [W]
No ESD 45.45 1.170 10.90 80.40
PSS 46.30 1.155 10.65 77.29
TRP - 1.157 10.90 79.23
WTR 45.83 1.156 10.87 78.92
Combined 46.77 1.150 10.53 76.09
The paper (Moon-Chan et al. 2017) is focused not on the design process itself but in method of
model tests extrapolation for the vessels equipped with pre swirl ESDs. Main topic of the article is in
fact method for correct prediction of effective wake fraction coefficient for full scale vessel, that shall
be adopted for propulsion prognosis. Authors analyse the results of adopting already known methods
(ITTC’78, ITTC’99A and ITTC’99B) and their own-developed one. Main difference between ITTC
methods and the proposed one is to include fact that the pre swirl devices produce not only rotational
inflow in fact but they influence the axial flow also. The latter component should be treated separately
during scaling procedure.
Wide range of model test – derived data are given in the paper. Two vessels are taken into analysis:
KVLLC2 (equipped with Pre Swirl Duct) and KCS (equipped with Pre Swirl Stator). Vessels
particulars and model scale ratio are not provided in the paper.
Table 16 Model test results KVLLC2
VS = 15.5knots RTM TM QM nM 2πQMnM VX/V VT/V t
[N] [N] [Nm] [rps] [W] [-] [-] [-]
No ESD 64.08 42.27 1.120 8.61 54.1 0.954 0.243 0.234
PSD 65.36 43.31 1.130 8.45 53.1 0.952 0.203 0.172
Table 17 Model test results KCS

VS = 24.0knots RTM [N] TM QM nM 2πQMnM VX/V VT/V t
[N] [Nm] [rps] [W] [-] [-] [-]
No ESD 52.54 34.11 1.156 10.71 67.3 0.918 0.149 0.172
PSS 54.02 35.04 1.123 10.65 66.9 0.0871 0.062 0.174
Table 18 Full scale propulsion prognosis for KVLLC2

VS = 15.5knots wES PDS nS
[-] [kW] [rpm]
No ESD, ITTC’78 0.342 26226 71.73
PSD, ITTC’78 0.363 24959 70.42
PSD, ITTC’99 0.383 24384 69.53
PSD, Moon et al. 2017 0.367 24793 70.23
Table 19 Full scale propulsion prognosis for KCS

VS = 24.0knots wES PDS nS
[-] [kW] [rpm]
No ESD, ITTC’78 0.254 43.672 106.28
PSS, ITTC’78 0.282 42.857 105.19
PSS, ITTC’99 0.303 41.833 103.73
PSS, Moon et al. 2017 0.298 42.094 104.11
The paper (Kim et al. 2015) presents design example for two kinds of ESD: pre swirl stator (referred
to as “Wake Adapted Control Fin” – WAFon) and pre swirl duct (referred to as “WAFon with duct” –
WAFon-D). The design is realized via variational optimization by means of FVM simulation. The vessel
adopted for the design example was of following particulars:
9
10
Table 20 Particulars of analyzed vessel

λ LPP B T D
[-] [m] [m] [m] [m]
Ship 1 319.00 60.00 21.00 10.00
Model 45 7.089 1.333 0.467 0.222

VS = 16.21knots TM QM nM 2πQMnM
[N] [Nm] [rps] [W]
No ESD 26.67 0.6708 7.89 33.3
WAFon 27.39 0.6751 7.81 33.1
WAFon-D 27.10 0.6543 7.65 31.5
Table 22 Full scale prediction (WAFon)

VS=16.21knots t wTM wTS ηH ηR η0 ηD PD ΔnS
[-] [-] [-] [-] [-] [-] [-] [kW] [%]
No ESD 0.235 0.432 0.353 1.183 1.01 0.594 0.706 22.971 0.0
ITTC-78 0.253 0.473 0.379 1.203 1.022 0.581 0.712 22.785 -0.5
ITTC-99 0.253 0.473 0.395 1.234 1.022 0.575 0.724 22.394 -1.4
Mod. ITTC-78 0.253 0.473 0.400 1.246 1.022 0.572 0.728 22.274 -1.8
Direct 0.253 0.473 0.400 1.246 1.022 0.572 0.728 22.273 -1.8
Table 23 Full scale prediction (WAFon-D)

VS=16.21knots t wTM wTS ηH ηR η0 ηD PD ΔnS
[-] [-] [-] [-] [-] [-] [-] [kW] [%]
No ESD 0.235 0.432 0.353 1.183 1.01 0.594 0.706 22.971 0.0
ITTC-78 0.249 0.502 0.390 1.232 1.034 0.576 0.734 22.232 -1.4
ITTC-99 0.249 0.502 0.423 1.302 1.034 0.561 0.754 21.637 -2.5
Mod. ITTC-78 0.249 0.502 0.433 1.324 1.034 0.557 0.759 21.475 -3.0
Direct 0.249 0.502 0.427 1.311 1.034 0.559 0.756 21.574 -2.6
Table 24 Propeller loading – calculated based on data provided in original paper (model scale)
VS = 16.21knots VM JM KT KQ η0M CT CN
[m/s] [-] [-] [-] [-] [-] [-]
No ESD 0.4031 0.1757 0.0199 0.566 2.753 4.860
WAFon 1.243 0.3779 0.1841 0.0204 0.543 3.284 6.050
WAFon-D 0.3645 0.1899 0.0206 0.535 3.639 6.805
The paper (Kim et al. 2004) deals with pre swirl stator design, which is realized via Lifting Surface
Method. The stator is assumed to have elliptic loading distribution on each blade and to cancel around
50% of the propeller slipstream rotation. In order to avoid flow separation stator angle of attack is limited
to 15°. For the design example the KVLLC was adopted:
Table 25 Particulars of KVLLC and its propeller
λ LPP LOA B T CB
[-] [m] [m] [m] [m] [-]
Full scale 1 320.00 332.5 58.0 20.8
0.8101
Model 50 6.400 6.650 1.160 0.416
The design speed is defined as VS = 15.5knots. The model tests resulted in following propulsion
prediction:
Table 26 Full scale propulsion prognosis for KVLLC
9
11
VS = 15.5knots PDS nS
[kW] [rpm]
No ESD 29.441 75.06
PSS 27.787 70.91
180° turned PSS 29.224 70.72
The model tests were conducted for pre swirl stator turned by 180° also. The test revealed that it no
longer works as an effective ESD (due to non-optimum loading distribution) however still reduces
propeller rate of revolution.
The paper (Nielsen et al. 2019) addresses topic of Controllable Pre Swirl Fins (CPSF), having fixed
and controllable (“flap”) parts. The geometry of the blades is designed via variational optimization by
means of FVM simulations. What is great advantage of adopted solution – each blade can be adapted
for actual vessel loading conditions to preserve optimal propulsor operation – thanks to the controllable
part of the blade. For the design example a 32500DWT bulk carrier was selected.
Table 27 Particulars of 32500DWT Bulk Carrier and its propeller
λ LPP B T ∇ S D AE/A0 Z
[-] [m] [m] [m] [m3] [m2] [m] [-] [-]
Full scale 1 171.506 28.40 10.00 39961 7234.2 5.20
0.502 4
Model 22.317 7.685 1.273 0.448 3.595 14.524 0.233
The model tests were carried out for two draughts: “Design” (T = 10.0m, even keel) and “Ballast”
(TA = 7.0m, TF = 5.0). Model tests were realized both for propeller operating alone and with ESD
installed, for each draught. The flap angle was optimized before these tests.
Table 28 Model test results (Design draught)

VS=11.5knots t wTM J ηH ηR η0 ηD PDM nM
[-] [-] [-] [-] [-] [-] [-] [W] [rps]
No ESD 0.147 0.415 0.377 1.458 1.014 0.504 0.745 48.82 8.328
CPSF 0.164 0.488 0.334 1.631 1.013 0.467 0.771 47.61 8.104
Table 29 Model test results (Ballast draught)

VS=12.0knots t wTM J ηH ηR η0 ηD PDM nM
[-] [-] [-] [-] [-] [-] [-] [W] [rps]
No ESD 0.161 0.469 0.372 1.580 1.010 0.499 0.796 43.70 7.995
CPSF 0.185 0.555 0.325 1.832 1.010 0.451 0.835 41.58 7.685
Table 30 Propeller power loading coefficient (calculated based on data provided in original paper)
VS KQNo ESD KQESD CNNo ESD CNESD
[knots] [-] [-] [-] [-]
Design 11.5 0.0196 0.0208 5.858 8.915
Ballast 12.0 0.0198 0.0213 5.926 9.130
The paper (Krasilnikov et al. 2019) addresses the topic of pre swirl stator design. As in most cases
of nowadays researches – it is realized via optimization by means of FVM simulation. The vessel
adopted for the design example is the Chemical Tanker of following particulars:
Table 31 Particulars of Chemical Tanker and its propeller
λ LPP B T ∇ CB D P0.7/D AE/A0 Z
[-] [m] [m] [m] [m3] [-] [m] [-] [-] [-]
Ship 1 175.60 32.23 11.85 56148 6.50
0.812 0.821 0.520 4
Model 26 6.754 1.240 0.456 3.195 0.250
9
12
The design speed of the vessel is V = 14.0knots. Required thrust force is defined by thrust loading
coefficient equal CT = 1.57 (defined for case without PSS).
Table 32 Model test results, no PSS installed
VS=14.0knots t wTS ηH ηR ηD PDS nS CNa
[-] [-] [-] [-] [-] [kW] [rpm] [-]
No ESD 0.195 0.312 1.170 1.006 0.715 5597 86.40 2.703
Complete propulsion prediction (based on model scale experiment) after PSS installation is not
provided in the paper; only selected propulsive coefficients are compared for both cases:
Table 33 Model test results, no PSS installed
VS=14.0knots t wTS ηHSb ηR
[-] [-] [-] [-]
No ESD 0.195 0.312 1.170 1.006
PSS 0.209a 0.363 1.242 1.004
a – value determine via FVM simulation
b – value determined upon provided t and wTS values
In the dissertation (Król 2019) two propulsor systems are presented (along with model test results):
propeller CP745 with Pre Swirl Stator ST001 and propeller CP753 with Pre Swirl Stator ST002. The
latter propulsor system was designed with use of more mature version of the design algorithm; both
propulsor systems were designed via combined Lifting Line – Lifting Surface approach. Each PSS blade
is assumed to have elliptic loading distribution. The vessel adopted for the design case was Nawigator
XXI.
Table 34 Particulars Nawigator XXI
λ LPP B T ∇ S D AE/A0 Z
[-] [m] [m] [m] [m3] [m2] [m] [-] [-]
Full scale 1 171.506 28.40 10.00 39961 7234.2 5.20
0.502 4
Model 10 7.685 1.273 0.448 3.595 14.524 0.233
Table 35 Particulars of the propellers

D P0.7/D AE/A0 Z
[m] [-] [-] [-]
CP745 2.26 0.772 0.759 5
CP753 2.26 0.828 0.606 5
Model tests were conducted for four configurations:

VS=13.0knots TM QM nM 2πQMnM
[N] [Nm] [rps] [W]
CP745 120.49 3.583 15.06 339.04
CP745+ST001 121.48 3.505 14.65 322.63
CP753 119.71 3.649 14.09 323.05
CP753+ST002 122.44 3.580 13.79 310.19

VS=13.0knots JM t wTM ηHM ηR η0M
[-] [-] [-] [-] [-] [-]
CP745 0.377 0.232 0.393 1.265 1.044 0.437
CP745+ST001 0.352 0.239 0.449 1.381 1.058 0.414
CP753 0.418 0.227 0.370 1.227 1.008 0.490
CP753+ST002 0.384 0.245 0.435 1.336 1.037 0.455
9
13
Table 38 Propeller loading conditions (model scale)

VS=13.0knots JM KTM KQM CTM CNM
[-] [-] [-] [-] [-]
CP745 0.377 0.2039 0.0268 3.654 8.012
CP745+ST001 0.352 0.2173 0.0277 4.465 10.176
CP753 0.418 0.2315 0.0312 3.373 6.839
CP753+ST002 0.384 0.2472 0.0320 4.268 9.035
Table 39 Full scale prognosis

VS=13.0knots JS wTS ηHS η0S ηDS TS QS nS PDS
[-] [-] [-] [-] [-] [kN] [kNm] [rpm] [kW]
CP745 0.396 0.371 1.221 0.444 0.566 124 37.1 289.3 1123
CP745+ST001a 0.342 0.473 1.444 0.410 0.626 125 35.4 273.6 1014
CP745+ST001b 0.340 0.476 1.452 0.408 0.627 125 35.4 273.0 1011
CP753 0.427 0.351 1.191 0.497 0.597 125 37.7 269.8 1065
CP753+ST002a 0.367 0.468 1.419 0.438 0.645 126 36.5 257.7 986
CP753+ST002b 0.363 0.474 1.435 0.434 0.646 126 36.5 257.0 982
a – scaling method according to ITTC’99
b – scaling method proposed in paper (Moon-Chan et al. 2017)
Table 40 Propeller loading conditions (full scale)

VS=13.0knots JS KTS KQS CTS CNS
[-] [-] [-] [-] [-]
CP745 0.396 0.1993 0.0264 3.236 6.798
CP745+ST001a 0.342 0.2246 0.0281 4.890 11.258
CP745+ST001b 0.340 0.2256 0.0283 4.970 11.508
CP753 0.427 0.2310 0.0308 3.226 6.335
CP753+ST002a 0.367 0.2552 0.0327 4.825 10.589
CP753+ST002b 0.363 0.2566 0.0329 4.959 11.002
a – scaling method according to ITTC’99
b – scaling method proposed in paper (Moon-Chan et al. 2017)
Discussed propulsor systems were earlier presented in papers (Król et al. 2017), (Król, Bugalski
2018) and (Król, Tesch 2018). Data presented in (Król 2019) may differ slightly as were re-calculated
anew.
In the paper (Kwi-Joo et al. 2014) a “Crown Duct” (CD) ESD device is discussed. Unlike in case of
usual pre swirl ducts, the duct part is not located at the ESD’s tip but in midspan rather. Some of pre
swirl foils end at the duct, but some of the extend beyond it or even start at it. The Crown Duct design
is realized via variational optimization by means of FVM simulation. For the design example a vessel
of following particulars is adopted:
Table 41 Particulars of analyzed vessel and its propeller
λ LPP LWL B T D P0.75/D AE/A0 Z
[-] [m] [m] [m] [m] [m] [-] [-] [-]
Ship 1 174 178 32.2 11 6.1
0.715 0.53 4
Model 25.417 6.846 7.003 1.267 0.433 0.24
The design speed is defined as 7.72m/s = 15knots. The model tests were carried out both for “bare”
hull and with Crown Duct installed, for two draughts (TA, TF values not precisely defined). For full scale
prediction ITTC’99 procedure of effective wake coefficient scaling was adopted. Full scale ship speeds
for which the prognosis were prepared are not clearly defined in the paper.
9
14
Table 42 Model test results (Full load)

PES t wTS ηHa ηR η0S ηDS PDS CNb
[kW] [-] [-] [-] [-] [-] [-] [kW] [-]
No ESD 4783 0.220 0.358 1.215 1.018 0.543 0.681 6740 3.698
CD 4792 0.247 0.453 1.377 1.031 0.494 0.714 6444 5.716
a – value determined upon provided t and wTS values
b – calculated based on data provided in original paper, speed of VS = 15.0knots assumed
Table 43 Model test results (Ballast condition)

PES t wTS ηHa ηR η0 ηD PDS CNb
[kW] [-] [-] [-] [-] [-] [-] [kW] [-]
No ESD 4256 0.215 0.399 1.306 1.014 0.541 0.717 5724 3.828
CD 4243 0.233 0.490 1.504 1.030 0.493 0.764 5331 5.834
a – value determined upon provided t and wTS values
b – calculated based on data provided in original paper, speed of VS = 15.0knots assumed
The paper (Simonsen et al. 2012) is focused mainly on rudder type and its distance from the propeller
influence on propulsion characteristics. These are valuable data from this study point of view as the
rudder may be thought of as a kind of post swirls stator. Moreover the paper presents some results with
the pre swirl stator installed – however for this case only numerical results are available.
For the design example 180m Diamond 34 Bulk Carrier was adopted:
Table 44 Particulars of 180m Diamond 34 Bulk Carrier and its propeller

λ LPP B T S CB D P0.7/D Z
[-] [m] [m] [m] [m2] [-] [m] [-] [-]
Ship 1 176.76 30.0 9.75 8747 5.6
0.8 0.839 4
Model 23. 73 7.449 1.264 0.411 15.533 0.236
The design speed is defined as V = 14knots (Fr = 0.173). The model tests were conducted for two
rudder types (“ORIG” – original horn rudder and “NACA” – spade rudder based on NACA20 section),
for three values of propeller – rudder distances for each of them:
Table 45 Model test results for rudder ORIG
ΔX / D TM QM nM 2πQMnM K Ta K Qa
[-] [N] [Nm] [rps] [W] [-] [-]
0.455 42.78 1.333 9.42 78.92 0.1556 0.0205
0.371 42.22 1.318 9.32 77.18 0.1569 0.0208
0.286 42.97 1.331 9.32 77.96 0.1597 0.0210
Table 46 Model test results for rudder NACA
ΔX / D TM QM nM 2πQMnM K Ta K Qa
[-] [N] [Nm] [rps] [W] [-] [-]
0.441 41.66 1.330 9.39 78.47 0.1525 0.0206
0.357 41.75 1.323 9.35 77.68 0.1542 0.0207
0.272 41.95 1.319 9.29 76.98 0.1569 0.0209
The Pre Swirl Stator influence was analysed via FVM simulation only for configuration with ORIG
rudder:
9
15
Table 47 FVM simulation results for rudder ORIG

QM nM 2πnQ K Qa
[Nm] [rps] [W] [-]
No ESD 1.315 9.32 77.02 0.207
PSS 1.313 9.17 75.66 0.0214
Detailed design procedure for PSS was not provided in the paper. It is only stated that its blades are:
“angled based on the flow direction in the region where they operate in order not to introduce extreme
angles of attack and separation on the fins.” Figures included in the paper suggest symmetric section
profiles, however it is not stated directly. The must however be some non-zero camber and/or angle of
attack, as there is a lift force generated on the PSS blades and it introduces demanded counter-swirl, as
stated in the paper text.
The paper (Xing-Keading et al. 2015) describes the design procedure Pre Swirl Stator device and
analysis of its impact on economic aspect of the vessel exploitation. As stated in the paper’s text, the
design procedure includes following steps:
“1. RANS computation to obtain the wake field at the pre-selected PSS positions
2. BEM optimisation to obtain the optimal twist and camber of the PSS
3. Creation of parametric model for ESD
4. RANS Self-Propulsion computations/optimisations to evaluate the PSS design”.
For the design example the Bulk Carrier “VALVOLINE” was adopted.
Table 48 Particulars of Bulk Carrier “VALVOLINE” and its propeller
LPP B T D Z
[m] [m] [m] [m] [-]
182 32 11 5.8 4
The design speed of the vessel is defined as VS = 15.0knots at propeller rate of revolution of nS = 123rpm.
The FVM simulation predicted delivered power reduction by 5.3% due to PSS installation, meanwhile
sea trials revealed it to be 6.8%. Propeller rate of revolution was reduced by 5.2% due to PSS.
In the paper (Yong-Jin et al. 2015) approach to PSS design is based on circulation distribution,
replaced with equivalent angle of attack. Detailed criteria for selecting stator loading distribution and
magnitude is not given however. For the design example Kriso Container Ship (KCS) was adopted:
Table 49 Particulars of Kriso Container Ship (KCS)
λ LPP LWL B T CB
[-] [m] [m] [m] [m] [-]
Ship 1 230.0 232.5 32.2 10.8
0.651
Model 39.5 5.82 5.89 0.81 0.27
The design speed was defined as VS = 24.0knots. Numerical results presented in the paper refer to PSS
influence on the hull resistance (resistance tests, without operating propeller)
Table 50 Resistance test results
RTM PES
[N] [%]
No ESD 39.07 100
PSS 41.77 111
9
16
4. Conclusions
As one can imagine, wide scatter of data presented above does not allow to arrange all of them in
common base, that would make reliable reference level for power saving evaluation. Thus only selected
values from data presented above are taken for the analysis.
Table 51 Propeller loading and power saving (CT, CN taken for “No ESD” conditions)
Paper ESD CT CN CT x CN ΔηDS
[-] [-] [-] [-]
Coney 1989 PreSS 0.691 0.884 0.611 +0.030
Hou et al. 2015 PostSS 1.073 1.458 1.564 +0.061
Hyoun-Joon et al. 2013 PSD 2.050 n/a n/a +0.051
Kim et al. 2015 WAFon 2.753 4.860 13.380 +0.018
Kim et al. 2015 WAFon-D 2.753 4.860 13.380 +0.048
Nielsen et al. 2019 CPSF n/a 5.858 n/a +0.026
Nielsen et al. 2019 CPSF n/a 5.926 n/a +0.039
Król 2019 PSS 3.236 6.798 21.998 +0.060
Król 2019 PSS 3.226 6.335 20.437 +0.048
Kwi-Joo et al. 2014 CD n/a 3.698 n/a +0.033
CT, CN [-]
8
4 CT
CN
3
0
0 0.01 0.02 0.03 0.04 0.05 0.06 0.07
ΔηDS [-]
Figure 1 Propeller (“No ESD” conditions) versus declared efficiency increase
From Figure 1 it can be seen that higher efficiency gains are declared for cases with higher propeller
loading. The trend is however strongly irregular and seems to be highly case-dependent. Due to this it
seems that only general statement may be made, that efficiency increase level around 3-6% may be
expected in most application, due to pre swirl device. Precise value should however be still carefully
evaluated for each individually.
Further studies are planned to be carried out, in order to find out if increased number of analysed
cases would allow to construct more reliable trend.
9
17
5. References
1 Abbott I. H., Doenhoff A. E. “Theory of wing sections Including a summary of airfoil data”
Dover Publications INC., New York 1959
2 Celik F., Guner M. „Energy saving device of stator for marine propellers” Ocean Engineering 34
(2007) 850-855
3 Chen J.P., Su J., Jie J.W., Yang L. “Investigation on propulsion and flow field of ships with
energy saving devices using CFD predictions and model tests” The 12th International
Conference on Hydrodynamics, Egmond aan Zee,The Netherlands, September 2016
4 Coney B. W. „A method for the design of a class of optimum marine propulsors” Massachusetts
Institute of Technology 1989
5 Hooijmans P. M., Holtrop J., Windt J. “Refiting to save fuel and new approaches in the design of
newbuidlings” 11th International Symposium on Practical Design of Ships and Other Floating
Structures, 2010
6 Hou L. X., Wang C. H., Hu A. K., Han F. L. “Wake – adapted design of fixed guide vane type
energy saving device for marine propeller” Ocean Engineering 2015
7 1.5 Hyun-Joon S., Jong-Seung L., Jang-Hoon L., Myung-Ryun Han, Eui-Beom H., Sung-
Chul S.. “Numerical and experimental investigation of conventional and un-conventional
preswirl duct for VLCC” International Journal of Naval Architecture and Ocean Engineering,
Volume 5, Issue 3, 2013
8 Joon-Hyoung L., Moon-Chan K., Young-Jin S., Jin-Gu K. „Study on Performance of Combined
Energy Saving Devices For Container Ship by Experiments” SMP17
9 Kerwin J. E., Coney W.B., Hsin C-Y „Hydrodynamic Aspects of Propeller/Stator Design” The
Society of Naval Architects and Marine Engineers, 1988.
10 Kim J-H., Choi J-E., Choi B-J., Chung S-H., Seo H-W. „Development of energy-saving devices
for a full slow-speed ship through improving propulsion performance” International Journal of
Naval Architecture and Ocean Engineering 7:390-398, 2015
11 Kim M. C., Chun H. H., Kang Y. D. „Design and Experimental Study on a New Concept of
Preswirl Stator as an Efficient Energy-Saving Device for Slow Speed Full Body Ship” SNAME
2004
12 Kraslinikov V., Koushan K., Nataletti M., Sileo L., Spence S. “Design and Numerical and
Experimental Investigation of Pre-Swirl Stators PSS” Sixth Symposium on Marine Propulsors
SMP’19, Rome 2019
13 Król P. “Vortex Model of Ideal Guide Vane and its Application to the Real Guide Vane” –
konferencja międzynarodowa SMP’19
14 Król P. „A new design method of propulsor systems with optimum distribution of bound
circulation” Gdańsk University of Technology 2019
15 Król P., Bugalski T. „Application of vortex flow model in propeller - stator system design and
analysis” – w czasopiśmie Polish Maritime Research No 2(98) 2018 Vol 25 (JCR 20pkt.)
16 Król P., Bugalski T., Wawrzusiszyn M. „Development of numerical methods for marine
propeller - pre-swirl stator system design and analysis” – konferencja międzynarodowa SMP’17
17 Król P., Tesch K. "Pre-swirl energy saving device in marine application" XXIII Krajowa
Konferencja Mechaniki Płynów, Zawiercie 2018
18 Kwi-Joo L., Jung-Sun A., Han-Joung K. „Comparative Study between Results of Theoretical
Calculation and Model Test for Performance Confirmation of ‘Crown Duct’” Journal of Ocean
Engineering adn Technology 28(1), 1-5 2014
19 Minchev A., Schmidt M., Schnack S. „Contemporary Bulk Carrier Design to Meet IMO EEDI
Requirments” SMP13
20 Moon_Chan K., Yong-Jin S., Won-Joon L., Joon_hyoung L. „Study on Extrapolation Method for
Self-propulsion Test with Pre-Swirl Device” SMP17
21 Nielsen J. R., Jin W. „Pre-Swirl Fins Adapted to Different Operation Conditions” SMP19
9
18
22 Simonsen C. D., Nielsen C., Klimt-Møllenbach C., Holm C. R., Minchev A. “CFD based
investigation of potential power saving for different rudder types, positions and pre-swirl fins”
Siemens report, November 2012
23 Xing-Kaeding Y., Gatchell S., Streckwall H. “Towards Practical Design Optimization of Pre-
Swirl Device and its Life Cycle Assessment” Fourth International Symposium on Marine
Propulsors smp’15, Austin, Texas
24 Yong-Jin S., Moon-Chan K., Won-Joon L., Kyoung-Wan L., Joon-Hyoung L.“Numerical and
Experimental Investigation of Performance of the Asymmetric Pre-Swirl Stator for Container
Ship” Fourth International Symposium on Marine Propulsors smp’15, Austin, Texas, USA, June
2015
9
19
HYDRONAV
22nd International Symposium 2020 In Ship Design, Safety,
On Hydrodynamics
Sopot-Iława,
Manoeuvring19-21 May 2021
and Operation
PLASTEX Sprint Boats - Research Investigations 1998-2020

T. BUGALSKI1 2
1 Gdansk University of Physical Education andHYDRONAV 2020
Sport, [email protected]
Sopot-Iława, 19-21 May 2021
Water Sport Equipment Group -Warsaw, [email protected]
2
The manuscript presents independent and original achievement – the process of designing shapes and hydrodynamic
parameters of hulls of the innovative Sprint Boats – Kayak and Canadian canoe line carried out in 1998-2012 for PLASTEX
PLASTEX
Composite Sprint Boats
PPH (PLASTEX) and since- Research
2012 for WaterInvestigations 1998-2020
Sport Equipment Group -a manufacturer of sport boats. The
manuscript shows the evolution of hull shapes of sprint boats' over time from the Olympic Games in Berlin (1936) to the
Olympics in Rio (2016), the influence of using Experimental Fluid Dynamics (EFD) and Computational Fluid Dynamics
T. BUGALSKI
(CFD)
12
on the performance of contemporary original shapes of Canadian canoes and Kayaks, and the development of
knowledge of sprint boat hydrodynamics as a specific subgroup of semi-submersible (high speed displacement) vessels and
1 Gdansk University of Physical Education and Sport, [email protected]
methods of their design using EFD and CFD. The paperwork presents practical implementation of research project results
2 Water Sport Equipment Group -Warsaw, [email protected]
and medal results of competitors from many national teams, achieved on Canadian canoes and Kayaks with an innovative
line of hull shapes in the finals of five consecutive Olympic Games from 2000 to 2016.
The manuscript
Keywords: presents
sprint boats, kayak,independent andOlympic
Canadian canoe, original achievement
Games, EFD, CFD– the process of designing shapes and hydrodynamic
parameters of hulls of the innovative Sprint Boats – Kayak and Canadian canoe line carried out in 1998-2012 for PLASTEX
1. The PPH
Composite history of development
(PLASTEX) and since 2012 of canoeing
for Water Sportas an Olympic
Equipment discipline
Group -a manufacturer of sport boats. The
manuscript shows the evolution of hull shapes of sprint boats' over time from the Olympic Games in Berlin (1936) to the
Olympics in Rio (2016), the influence of using Experimental Fluid Dynamics (EFD) and Computational Fluid Dynamics
In classic
(CFD) canoeing, sometimes
on the performance of contemporarycalled lowland,
original shapesregatta or canoe
of Canadian sprint,
canoes the competition
and Kayaks, takes place
and the development of
at distances
knowledge of 200m,
of sprint 500m, 1000m,
boat hydrodynamics 10000m.
as a specific Canoeofsprint
subgroup are done (high
semi-submersible on single kayaks (K1),
speed displacement) double
vessels and
(K2)
methods and four-seat
of their design (K4), and and
using EFD on CFD.
singleThe(C1), double
paperwork (C2) practical
presents and four-seat (C4) Canadian
implementation of researchcanoes. In a
project results
and medal
kayak, theresults of competitors
competitor fromforward,
sits face many national
usingteams, achieved on Canadian
a double-sided paddle. Incanoes and Kayaks
a Canadian with ancanoe,
regatta innovative
the
line of hull shapes in the finals of five consecutive Olympic Games from 2000 to 2016.
competitor kneels on one knee and paddles with a single-side paddle. Canoe sports are subject to
international regulations
Keywords: sprint boats, andcanoe,
kayak, Canadian areOlympic
associated in an
Games, EFD, CFDinternational organization. This international
organization, originally called Internationale Repraesentschaft fuer Kanusport (IRK), and after World
War II renamed
1. The historytoofInternational
development Canoe Federationas(ICF),
of canoeing was established
an Olympic disciplineon 19th January, 1924, in
Copenhagen.
In classic
The shapescanoeing,
of boats sometimes
– Canadiancalled
canoeslowland, regatta
and kayaks fororsprint
canoecompetitions
sprint, the competition
have changed takes
overplace
the
years and several
at distances of 200m, periods
500m, (eras) can 10000m.
1000m, be distinguished in these
Canoe sprint are changes. The period
done on single kayaksof(K1),
shaping the
double
(K2) and four-seat
regulations and rules(K4),fromand theon single (C1),ofdouble
establishment IRK -(C2)
ICF and four-seat
to the Olympic(C4) Canadian
Games canoes.
in London In a
(1948).
The
kayak,eratheofcompetitor
round-bottomedsits faceCanadian
forward, canoes
using a and kayaks with
double-sided hullsIn made
paddle. of hot-formed
a Canadian wooden
regatta canoe, the
competitor
plywood onkneels on onewhich
metal forms, knee canandbepaddles withSamson
called the a single-side
era - frompaddle. Canoe of
the surname sports are subject
the designer to
of the
international
STRUER regulations
company - and theandera are associated
of slender canoesin made
an international organization.
of carbon fiber’s Thisthe
- the era after international
revolution
caused by innovative
organization, originally shapes
calledofInternationale
Canadian canoes and kayaks made
Repraesentschaft fuerbyKanusport
the PLASTEX(IRK),company.
and after World
WarSince
II renamed
1930, at to International
Technischen Canoe Federation
Hochschule Berlin(ICF),
(THB),waswork
established
on theonshape
19th January,
of regatta1924, in
boats,
including
Copenhagen. kayaks and canoes, for the Olympic Games in Berlin in 1936 was carried out. As part of this
Themany
work, shapesnewof shapes
boats –ofCanadian canoes
sprint boats wereand kayaksand
designed for pioneering
sprint competitions havestudies
comparative changedat aover the
towing
yearswere
tank and carried
several out periods
at the(eras) can be distinguished
Versuchsanstalt für Wasserbauin these
und changes.
SchiffbauThe period
(VWS) in of shaping
Berlin. the
At that
regulations and rules from the establishment of IRK - ICF to the Olympic Games in London (1948).
The era of round-bottomed Canadian canoes and kayaks with hulls made of hot-formed wooden
plywood on metal forms, which can be called the Samson era - from the surname of the designer of the
STRUER company - and the era of slender canoes made of carbon fiber’s - the era after the revolution
caused by innovative shapes of Canadian canoes and kayaks made by the PLASTEX company.
Since 1930, at Technischen Hochschule Berlin (THB), work on the shape of regatta boats,
including kayaks and canoes, for the Olympic Games in Berlin in 1936 was carried out. As part of this
work, many new shapes of sprint boats were designed and pioneering comparative studies at a towing
tank were carried out at the Versuchsanstalt für Wasserbau und Schiffbau (VWS) in Berlin. At that
20
time, shapes and proportions of hulls characteristic for the era of shaping the Olympic canoeing were
developed [1].
From the Olympic Games in Helsinki, the designs of the Danish ship builder J. Samson, working
for the STRUER company, had a significant impact on the development of the shapes of Canadian
regatta canoes. The era of Samson's projects lasted from 1948 (K1, gold –Karen Hoff) to 1996 (K1,
gold –Antonio Rossi). Several dozen different versions of sport boats came out of Samson's drawing
board [2-4]. It is worth noting the national attempt to design original kayaks for Polish representatives
during the preparations for the Olympics in Moscow (1980), with a partial success (bronze in K1W),
as part of the CTW PO-PROMOR order. The research part of this project was carried out in the CTO
towing tank. In the last decade of the 20th century, new shapes of kayaks and Canadian canoes were
designed according to the regulation of 6.2 International Canoe Federation (ICF), specifying the
minimum boat widths, different for each category.
Figure1 Canoe C1 DELTA (STRUER) photograph of the shape of the hull (PLASTEX museum)
In 1993, the Polish company Plastex Composite PPH (PLASTEX) purchased a kayak shape license
for the Eagle (Van Dusen), with the right to manufacture and modify it. This shape became the starting
point for subsequent modifications of boat hull designs. PLASTEX started active search for new
solutions and the first modifications aimed to improve the "rowing ergonomics", which was obtained
by modifying the deck shape - introducing hollows allowing the competitor to move the paddle closer
to the boat’s symmetry plane. The current development era of kayak and Canadian canoe shapes - the
"PLASTEX era" began with the World Cup in 1998. Canadian canoes - C1 Starlight, and kayaks - K1
Destroyer and K2 Tender- designed in accordance with innovative assumptions, where while shaping
the hull, contrary to the applicable width limits, the anatomical parameter - intertrochanteric diameter
(intertrochanterica distantia) characteristic of each individual competitor was taken into account,
conquered the finals of the World Championships and the Olympic Games.
Figure 2 Canoe C1 Starlight (PLASTEX) - photo of the shape of the hull (PLASTEX museum)
21
Based on the analytical and design works for the Olympic Games, innovative boats - C1 Futura, C2
Futura, K1 Destroyer, K2 Sting and K4 Arrow - were prepared. Competitors from many national
teams, competing in PLASTEX boats, achieved excellent results (19 medals). The development of an
innovative line of sprint boats and significant medal achievements of competitors using those boats
cast doubt on the correctness of the minimum width rule, and in 2000, ICF definitively gave up the
minimum width requirement in the technical regulations for sprint boats.
For the next, second Olympic Games in Athens (2004), C1 Olympia, C2 Olympia Maxima, K1
Olympic Warrior, K2 Eureka and K4 Furry hulls were prepared in the full design process and using
hydrodynamic analyses of experimental results of the M506 reference model and CFD calculations
(PANSHIP, Comet). The K2 project was a developed modification of the M541 model, also a former
object of experimental research. All projects were made and measured by the ICF technical committee
during the 2003 World Cup. PLASTEX sprint boats achieved 13 medal positions at the Athens
Games.
For the third Olympic Games in Beijing, had been designed a new innovative line of Olympic
kayaks as part of the PLASTEX-CTO research project. From the new line of kayaks developed in this
project (experimental research and CFD – Comet analyses with the CTO’s dynamic displacement
procedure), PLASTEX sent 5 types of boats from 8 prepared projects to the Olympic Games in
Beijing (2008). These were C1 Smart, C2 Smart, K2 Eureka2006, K4 Furry2008 and K1 Midas2.
Models K4 Furry2008 and K1 Midas2 were designed specifically for Polish representatives – custom
designs (competitors' weight and their positioning, paddling characteristics, tactics at a distance, ...).
The prototype of the K1M Midas2 model was used a Polish representative (Marek Twardowski) to
win gold at the World Championships in 2007, unfortunately he did not confirm it in Beijing. The K4
Furry2008 model was made in two copies. On a dedicated and technologically developed boat, Polish
competitors ended the final run far beyond the podium, watching the backs of their rivals.
Representatives of Belarus had the prototype boat from PLASTEX... and won gold. Silver was won by
Slovaks on the previous version of the PLASTEX boat (project from 2004), before the competition
they had claimed that their boat was perfect and did not need to be improved. Boats made according to
new projects won 9 medal positions.
For the next (fourth) Olympic Games in London (2012), C1 and C2 Canadian canoes and Fighter
line kayaks characterized by modified trapezoidal shaped frame sections had been designed. Hull
shapes of sprint boats had been optimize using the CFD -STAR CCM + package (CD-Adapco). The
implemented modifications were to prevent the hull sinkage, as this was the result of the natural
movements of canoeists, especially on Canadian canoes. The prototype model M761 -K1W Fighter
was used by the representative of Hungary (Natasha Janicz) to win gold (200m) and silver (500m) at
the World Championships in 2010. The prototype model M731 -K1M Fighter was used by the
representative of Poland (Piotr Siemionowski), to win gold at the World Championships in 2011.
Unfortunately, they did not confirm it in London. Ultimately, the achievements of those Games
concerned only "Canadian canoes" on which 6 medal positions were won.
Figure 3 K1, K2 and K4 canoes – free surface elevation
For the last fifth Olympic Games in Rio (2016), the C1 Canadian canoe, had been developed with
modifying the bow section. New boats, called Fighter Rio, were excellent at both distances (200m and
1000 meters). A German representative (Sebastian Brendel) won gold medals. In total, 4 medal
positions were won on those boats.
22
2. PLASTEX Sprint Boats -Original research and design achievements

The work on boat projects for PLASTEX in 1998 had started with a critical review of literature
and analyses of basic works by Jackson [2], Lazauskas & Tuck [3], Lazauskas & Winters [4]. Then,
measurements of hull shapes of the most popular kayaks and Canadian canoes used by the competitors
at that time were made. The boat shapes were introduced into the hull shape design systems: KORAB,
MAXSURF, FreeSHIP, NAPA (Naval Architecture Packet). In January and February 1999, two cycles
of boat resistance measurements of four kayaks selected by PLASTEX were made at the CTO towing
tank: K1 (CleaverX, Eagle, Sprint85, Destroyer85), kayak K2 PLASTEX and Canadian canoe C1
PLASTEX -Starlight.
The shapes of innovative boat hulls for the Olympic Games were designed in accordance with the
“forget about the width limits” guideline by Ryszard Seruga (PLASTEX). The design basics included
analyses of experimental results, hydrostatic calculations from NAPA, FreeShip and Maxsurf
packages, resistance calculations using the "slender body" and Savitzky’s method (for semi-sliding
and sliding units) and single calculations using the GODZILLA program (Lazauskas, publicly
available for objects up to 1m3 of displacement). When developing the model boat shapes C1M
(model 506) and K2M (model 541), modules of the PANSHIP system with the new waveform model
(prototype of the BOS-L package) were used (Dudziak [5] 293-296). The hulls of the M506 and M541
models were made in carbon fiber technology and from the half of 1999 they were subjected to
experimental research as part of the KBN research project. Due to the fact of using the measurement
results in terms of minimizing resistance, the tests in the towing tank were conducted in accordance
with the procedures for analysing the uncertainty of resistance measurements (Bugalski and Liźiński
[6]). The results of the KBN project research in the part referring to sprint boats (3 out of 5 tested
models) were used only at the next Olympic Games in Athens. This is due to the current regulation
that boats can be admitted to the Olympic Games if their prototypes are measured by the ICF technical
committee, at the latest at the World Championships preceding the Olympic Games. The innovative
hulls of sprint boats had been designed had the narrowest waterlines among boats available on the
market.
Figure 4 Canoe C1 M506 - photo of the shape of the hull (PLASTEX museum)
Figure 5 Tests of the M506 model in the towing tank (CTO, 2000)
In the spring of 2007, PLASTEX and CTO SA started joint activities [7-9] co-financed by the
Ministry of Science and Higher Education –(6ZR6 2006 C/06811). The works made CTO were
presented in publications and presentation [10] at Canoe/Kayak Coache's Symposium – Warsaw,
23
February-March, 2009. The process of searching for better boat shapes in terms of resistance began
with comparative studies of existing (used today) canoe shapes at the CTO towing tank. The surface
models of hull shapes of existing kayaks were obtained by optical scanning of a boat made with an
optical scanner and a photogrammetric system. The main objective of the work was to find the
dependence between the basic parameters of the shape and the output trim (trim at the bow) and the
performance of canoes. Although these studies provided a significant amount of reliable data in a short
time, they did not allow to fully explain the physical mechanisms linking the shape parameters to the
boat's performance.
Figure 6 The complex test of K1 kayak performance before Olympic Games in Beijing (CTO S.A., 2008)
Hence, the experimental research was supported by a wide-ranging numerical simulation program
(CFD, Comet program with CTO procedure including dynamic buoyancy), which allowed to analyse
the effect of the flow parameters, the shape of the wave system and the pressure distribution on the
hull in various variants of hull shapes. The project was completed by research on selected boat
prototypes and K4M field tests under the representation of Poland in OPO in Wałcz. The K1M
prototype was tested by the current world champion (Marek Twardowski) directly in the towing tank.
The work on new sprinter boat projects for the Olympic Games in London (2012) and in Rio
(2016) was based on methodology of resistance calculation, taking into account the phenomenon of
dynamic trimming using the STAR CCM+ solver (CD adapco). The obtained consistency of CFD
calculations with the results of experimental measurements of resistance in the towing tank allowed
for the reduction of costly tests. Only in cases where there was a concern that CFD methods cannot be
used to select a better shape, studies of models or finished boats in the towing tank were analysed.
Figure 7 Kayaks M937 K1M PLASTEX and M938 K1M Cinco NELO, Resistance tests, V=5.20m/s,
(Bugalski: CTO RH-2015/T-040E)
24
In 2015, had been conducted comparative tests of the performance of K1 kayaks from the
PLASTEX offer and its competitor (NELO) at the CTO towing tank. With additional analysis of
uncertainty according to ITTC procedures, the measurement results showed a slightly, but clearly
above the measurement resolution, better hydrodynamic parameters of the PLASTEX kayaks,
sufficient to achieve advantage from 1/3 to ½ of the boat's length.
3. Hydrodynamics aspects of sprint boats design

Kayaks and sprint canoes, from the point of view of the theory of ships Larsson and Raven [11],
(Faltinsen [12], Bertram [13]), belong to a group of semi-displacement units, obtain the Froude
numbers from 0.45 to 0.75 and even above, which means that they are placed between high-speed
displacement ships and sliding units. When analysing hydrodynamic parameters, it can be stated that
when designing such units, dynamic phenomena (the so-called dynamic buoyancy) must be taken into
account. It is extremely important to choose the right trim for the expected weather conditions (wind
and wave) and distribute crew weight in relation to the length of the boat. Currently, it is necessary to
personally shape the boat to a particular competitor with its mass and rowing characteristics, taking
into account the distance, expected wave and wind conditions.
Thus, while designing kayaks and canoes, the length of the hull is determined mainly by factors
that are not hydrodynamic in nature. Width B is one of the key factors to ensure adequate stability,
hence the minimum value of the B/T ratio needs to be maintained. Increasing B/T will increase the
resistance because the volume moves closer to the free surface, which increases the wave resistance.
The specificity of canoes and kayaks requires minimizing the width, and the limitations to this are the
anatomy of competitor and the "comfort" of rowing.
The minimum wetted surface for a given buoyancy is also sensitive to the B/T ratio, the optimal
value of which is around 3 for CB=0.50. However, the losses resulting from going away from this size
are not large. The effects of B/T ratio changes on wave resistance can be analysed based on the results
of model tests and CFD calculations. The fullness and longitudinal distribution of displacement are
determined by the curve of frame sections, which is an important parameter given by all CAD
programs for ship design. It is often desirable to start with a good curve of frame sections, and leave it
constant while optimizing the shape of the hull. The two main parameters known on the basis of the
frame section curve are the CP prismatic coefficient and the longitudinal position of the LCB
buoyancy centre. The optimal values of these parameters strongly depend on Froude’s number, and the
field curve should be designed in such a way to achieve these values. When increasing Froude’s
number, the optimal CP value decreases to a minimum within the limits of 0.55-0.6 in the range of
Froude's number of Fn=0.30-0.35, and then grows relatively quickly to 0.60-0.65 near the main ridge.
In terms of semi-slide, CP grows slowly to 0.65-0.70, and this increase continues in the area of the
slide. The optimal LCB position relative to the midship as a function of Froude’s number varies from
+3% to about -2% in the range of Froude's number of 0.15-0.25, reaching -6% when Froude’s number
is 1.0. The LCB distance is dimensioned by LPP, positive towards the bow. The reason for CP and
LCB changes is the relations between two main components of resistance: resistance of viscosity and
wave resistance. At low speed, viscosity resistance prevails, but its share decreases with increasing
speed.
Near the main hump of the wave resistance curve (Fn=~0.5), wave resistance is the dominant
component. For higher Froude’s numbers, both wave resistance and viscosity resistance are important.
However, it should be remembered here that focusing only on resistance can easily lead to a shape that
is not optimal from the point of view of drive force. To minimize the demanded force, the interaction
between the paddle and the hull, and the ability to pass force to the paddle, should be taken into
account.
In the range of speed of displacement vessels, hydrostatic forces dominate; in the area of full slide,
hydrodynamic forces take over the main part of the lifting force required to keep the weight of the
hull. In the intermediate region, both hydrostatic and hydrodynamic forces are important. The upper
limit of buoyancy swimming is slightly arbitrarily set at the level Fn= 0.5, while the lower limit of the
slide is defined as Fn=1.0. The intermediate range is defined as the semi-displacement or semi-slide
swimming speed range typical for sprint boats – Canadian canoes and kayaks.
25
Figure 8 Example of results of towing tank resistance tests K1 -CTO, Gdansk (Bugalski: CTO RH-2007/T-075,
CTO RH-2007/T-107, CTO RH-2015/T-040E)
Figure 9 Results of comparison resistance tests K1 -CTO -2015, Gdansk (Bugalski: CTO RH-2015/T-040E)
4. Summary
For hydrodynamic analyses of sprint boats – Canadian canoes and kayaks, the methods of
experimental and numerical hydrodynamics had been used, perfecting these tools in subsequent
Olympic cycles. Currently, CFD analyses and experimental tests of kayak performance are limited to
stationary analyses – the hull is towed at a constant speed with an unchanging centre of mass. This
simplifying assumption still allows to effectively "improve" the resistance of canoe hull shapes,
improving their resistance by 0.5-1%, seemingly a small difference, however, used by the competitor
it can bring the Olympic laurel or the bitterness of defeat. The use of CFD methods, taking into
account the phenomenon of dynamic buoyancy, allows to reduce costs by limiting the number of boat
26
projects and the need to produce many different hull shapes. What is more, understanding the flow
phenomena with CFD allows performing "optimizations" far faster than previously possible.
Currently, the shapes of sprint boats – regatta kayaks seem to be optimally from resistance point of
view. The significant progress in performance can probably take place by including dynamic
behaviour of the competitor-hull-environment system in the design process. To do this, the phenomena
occurring during the race – including cyclic movements and unsteady forces exerted on the kayak's
hull, aerodynamics of the crew's body parts – need to be taken into account.
Acknowledgement
This paper was prepared under the project POIR.01.01.01-00-0103/19 "Innowacyjne łodzie klasy
oilimpijskiej i paraolimpijskiej do sprintu kajakowego - Innovative Olympic and Para-Olympic class
boats for the canoe sprint" co-financed under the Smart Growth Operation Programme competition
2/1.1.1/2019.
References
[1] Weitbrecht H.M.: Über Formgebung von Rennbooten für Ruder und Paddel, Jahrbuch der
Schiffbautechnischen Gesellschaft, 235-259, Brand 38, 1937.
[2] Jackson P.S.: Performance Prediction for Olympic Kayaks, Journal of Sports Sciences, Vol.13, 239-245,
1995.
[3] Lazauskas L. and Tuck, E.O.: Low Drag Racing Kayaks, University of Adelaide, Dept. of Applied
Mathematics, Australia, 1996.
[4] Lazauskas L. and Winters J.: Hydromechanics Drag of Some Small Sprint Kayaks, University of
Adelaide, Dept. of Applied Mathematics, Australia, 1997.
[5] Dudziak J.: Teoria okrętu. Fundacja Promocji Przemysłu Okrętowego i Gospodarki Morskiej, 2008.
[6] Bugalski T. and Liziński L.; Analiza niepewności wyników eksperymentu w okrętowych badaniach
modelowych. Jubileuszowa Konferencja Naukowa CTO, cz. I, , 65-78, Gdańsk-Jurata 2001.
[7] Bugalski T.: Optimization of the canoes for the Olympic Games in Beijing, 11th Numerical Towing Tank
Symposium, France, 2008.
[8] Bugalski T. and Kraskowski M.: CTO and PLASTEX Paddle to Olympic Glory using CFD Simulation,
CD-adapco, e-dynamic, 2008.
[9] Bugalski T.: Development of the New Line of Sprint Canoes for the Olympic Games, Proceedings of 10th
Int. Conference of Fast Sea Transportation -FAST 2009, 1039-1049, Athens, 2009.
[10] Bugalski T.: Hydromechanics for development of sprint canoes for the Olympic games, Int. Canoe
Federation Symposium, Warsaw, 2009.
[11] Larsson L. and Raven H.: Ship resistance and flow, Society of Naval Architects and Marine Engineers,
2010.
[12] Faltinsen O. M.: Hydrodynamics of High-Speed marine vehicles, Cambridge Univ. Press, 2005.
[13] Bertram V.: Practical Ship Hydrodynamics, Butterworth Heinemann, 2000.
Selected Test Reports (B) and Technical Reports (T) about sprint boats ,CTO Gdansk
CTO RH-1980/Z-105 & Z-106 –The resistance test results for 10 different K1 kayaks,
CTO RH-1999/T-008 –The results of hull shape measurements and resistance tests for 3K1 (CleaverX,
Eagle, Sprint85) and K2 (Plastex) kayaks,
CTO RH-1999/T-042 –The results of hull shape measurements and resistance tests for K1 kayak
(Destroyer85) and C1 canoe (Starlight),
CTO RH-2007/T-047 CFD simulations of sprint canoes performance,
CTO RH-2007/T-075 The experimental test of 13 sprint model canoes,
CTO RH-2007/T-092 The hydrodynamic optimization of New line of Olympic sprint canoes using CFD
methods,
CTO RH-2007/T-106 & T-025 The execution of 8 new sprint canoe prototypes on CNC machine,
CTO RH-2007/T-107 The towing tank experimental test of new sprint canoe prototypes,
CTO RH-2008/T-101 The pitch and roll influence on performance of sprint canoes. The towing tank tests
of kayak powered by human,
CTO RH-2015/T-040E Sprint boats ICF K1 Olympic class. Results of experimental tests of boat resistance
in calm water. M937 Fighter Rio2016; M938 Cinco XXL; M939 Warrior 90-2013.
27
HYDRONAV Resistance
2020 and Propulsion
HYDRONAV 2020
Short review of actual ice class requirements
Sopot-Iława, for2021
19-21 May ship propulsion and its
implementation into PRS Rules
M.NAREWSKI1
Short
1
review of actual ice class requirements for ship propulsion and its
Polish Register of Shipping, e-mail: [email protected]
implementation into PRS Rules
Keywords: ice class, FSICR, IACS polar class, propulsion
M.NAREWSKI1
Abstract
1 Polish Register of Shipping, e-mail: [email protected]
Ships navigating in ice infested waters must have hull and machinery designed and
constructed
Keywords: to FSICR,
ice class, withstand iceclass,
IACS polar loads during navigation. The process of ship design includes the
propulsion
specific knowledge. PRS was involved recently in classification of new ship design where
Abstract with ice class requirements was deeply analysed. The paper is brief introduction
compliance
intoShips
machinery
navigating aspects related
in ice to verification
infested waters mustbyhaveclass society
hull of ship propulsion
and machinery designedmachinery
and
design for compliance with requirements of rules for classification
constructed to withstand ice loads during navigation. The process of ship design includes and construction of ships.
the
The positive
specific result ofPRS
knowledge. verification
was involvedby class society
recently is visible as relevant
in classification of new notation
ship designmark in a
where
symbol of ship
compliance with class certificate
ice class being assigned
requirements afteranalysed.
was deeply construction. Majority
The paper of ship
is brief designers
introduction
and shipping
into machinery professionals
aspects related aretonot very familiar
verification with origin
by class societyofofice class
ship requirements
propulsion and
machinery
application of ice requirements to propulsion machinery where ships are
design for compliance with requirements of rules for classification and construction of ships. intended to perform
other type service
The positive resultthan carrying cargo
of verification by classin winter
societytime in ice as
is visible infested waters.
relevant notation mark in a
symbol of ship class certificate being assigned after construction.
The need for development of ice class requirements resulted from winter navigation Majority of ship designers
and shipping
practice, professionals
research are not
and evolution of very familiar with
requirements originhaving
for ships of ice class requirements
both hull and
and machinery
application of ice requirements to propulsion machinery where ships
suitable for operation in ice infested waters. As a member of IACS, PRS takes active part inare intended to perform
other type service
development than carrying
and amendment cargo technical
of IACS in winter requirements
time in ice infested waters.
applicable to ships navigating
in ice
Thecovered
need forwaters. In line of
development withiceIACS
classprocedures,
requirements allresulted
mandatoryfromtechnical and safety
winter navigation
requirements
practice, research for ice andclassed shipsofarerequirements
evolution implementedfor into PRShaving
ships Rules both hull and machinery
suitable for operation in ice infested waters. As a member of IACS, PRS takes active part in
development and amendment of IACS technical requirements applicable to ships navigating
in ice covered waters. In line with IACS procedures, all mandatory technical and safety
requirements for ice classed ships are implemented into PRS Rules
28
1. Brief outline of origin of Ice Class Rules
Winter navigation was a great challenge for all ships in times before the first
requirements for ship design and construction were developed. Initially, each of major
classification societies developed own ice class requirements. Such approach lasted over
100 years of ships mechanical propulsion. In the end of XX century unification of
requirements was initiated – to harmonize and unify the requirements of various bodies in
respect of hull and propulsion machinery requirements of ships navigating ice infested
waters.
Historical development of Finish-Swedish Ice class Rules
Finish-Swedish Ice class Rules (FSICR-Baltic Class – TRAFI)
Actually - FSICR Rules apply to cargo ships having more than 2000 GRT navigating to finish ports in winter season.
1890 The first Finnish statutory regulation for ships navigating in ice
1920 The first Finnish ice class rules that included the fairway fees
1923 Circular about the classification of ice-strengthened ships was sent to the largest classification societies
1932 The Finnish ice class rules published in 1932 introduced the ice classes:
1A, 1B and 1C for ships strengthened for navigation in ice,
ice class 2 for ships classified for unrestricted service but not strengthened for navigation in ice and ice
class 3 for other vessels.
1960 The Finnish ice class rules included only minor modifications to the existing rules. The subdivision of ice
class 2 was abolished as radios had become more common
1965 Finnish ice class rules introduced a new ice class, 1A Super, which was considerably stronger than the
existing classes. Since ice class 1A was already exempted from the ice fees, the 50% reduction to
lighthouse fees was given to ships of the new ice class. However, the additional requirements of the
highest ice class were deemed excessive and only three such ships were built.
1971 Finnish-Swedish ice class rules. According to the new rules the requirements for a ship with ice class 1A,
deadweight tonnage of 3,500 tons and engine output of 1,500 hp were kept largely the same, but the
requirements for ships of the lowest ice class, 1C, were increased considerably as in the past such ships
had been essentially open-water vessels with a strengthened bow.
1985 The Finnish-Swedish Ice Class Rules edition 1986 introduced changes to the hull dimensioning. The
plastic deformation theory used in the previous rules was changed to elastic, and the load height was
changed to more realistic. The minimum engine power requirements were changed in 2002 to correspond
to the resistance of the ship in a brash ice channel, calculated as a function of ship size and hull geometry.
The rules were subject of revisions in 2002, 2008, 2010 as a result of number of research projects
completed with participation of Finnish companies and DNV.
2017 The Finnish-Swedish Ice Class Rules (TRAFI-2017), The current Finnish-Swedish ice class rules are
applicable to ships contracted for construction on or after 1 January 2019. However, the current ice class
rules become applicable retroactively also to ships of ice classes 1A and 1A Super laid down before 1
September 2003 in the beginning of the year when 20 years has passed from the delivery of the vessel and
ships failing to meet the requirements will have their ice classes downgraded.
IACS Polar Class unified requirements Rules - - Adopted by all IACS societies as part of their Polar Class Rules
UR’s apply to ships constructed of steel and intended for independent navigation in ice-infested polar waters.
2006/2007 UR I1 Polar Class Descriptions and Application
UR I2 Structural Requirements for Polar Class Ships
UR I3 Machinery Requirements for Polar Class Ships
2. FSICR-Baltic Class
Finnish-Swedish Ice Class Rules (FSICR) is an ice class assigned to a vessel operating in
first-year ice in the Baltic Sea and calling Finnish or Swedish ports. Ships are divided into six
29
ice classes based on requirements for hull structural design, engine output and performance in
ice according to the regulations issued by the Finnish Transport and Communications Agency
(Traficom) and the Swedish Maritime Administration. During the winter months, Finnish and
Swedish authorities may declare traffic restrictions in the Northern Baltic Sea in order to
ensure that ships operating in the region are capable of navigating in ice-covered waters safely
and efficiently. These restrictions, for example "ice class 1A, 2000 DWT", declare the
minimum ice class and other requirements for ships that may be provided with icebreaker
assistance. The Finnish fairway dues, a system of fees charged for the use of sea lanes to
cover the costs of management and icebreaker assistance, also depend on the vessels' ice
class. Since ships of lower ice classes generally require more assistance during the winter
months, their fairway dues are considerably higher than those of ships of the highest ice
classes. For this reason the majority of ships regularly calling Finnish ports are built to the
highest ice classes. In the beginning of 2008, 47% of the Finnish tonnage were of ice class 1A
Super.
3. Polar Class (PC)
Polar Class (PC) refers to the ice class assigned to a ship by a classification society based
on the Unified Requirements for Polar Class Ships developed by the International Association
of Classification Societies (IACS). The requirements of IACS UR I3 contain requirements
applicable to main propulsion, steering gear, emergency and essential auxiliary systems
essential for the safety of the ship and the survivability of the crew. Seven Polar Classes are
defined in the rules, ranging from PC 1 for year-round operation in all polar waters to PC 7
for summer and autumn operation in thin first-year ice. The two highest Finnish-Swedish ice
classes, 1A and 1A Super, are somewhat equivalent to the two lowest Polar Classes, PC 7 and
PC 6, respectively. The practice exists that ship for operation in polar areas can have separate
class notation assigned for hull and separate for ship machinery. The IACS Polar Class rules
should not be confused with International Code for Ships Operating in Polar Waters (Polar
Code) by the International Maritime Organization (IMO).
4. IMO's International Code for Ships Operating in Polar Waters (Polar Code)
IMO Polar Code is mandatory under both the International Convention for the Safety of Life
at Sea (SOLAS) and the International Convention for the Prevention of Pollution from Ships
(MARPOL). The Polar Code covers the full range of design, construction, equipment,
operational, training, search and rescue and environmental protection matters relevant to ships
operating in the inhospitable waters surrounding the two poles. The Polar Code entered into
force on 1st January 2017.
5. FSICR and IACS Polar Class implementation by PRS
For over 20 years PRS Rules contained ice class requirements machinery as developed
1971. The first major change were implemented in year 2006 when new edition of the Finish-
Swedish Ice class Rules was developed and was requested to be mandatory for IACS class
societies. Finish and Swedish Maritime Administration decided that ice class assigned by
classification societies can be accepted by the Administration without additional verification.
All revisions of FSICR released in 2010, 2012, 2017 and 2019 are being implemented by
PRS. The Table underneath provides concise summary and comparison of FSICR
requirements and PRS Rules and its relation to IMO Polar Code to understand what is
hidden under ice class notation mark.
30
PRS Ice Class notations in comparison to Finish-Swedish Ice Class, IACS Polar class
FSICR
PRS Ice breaking
PRS Polar Class/IACS Polar Class Baltic
Ice Class capability
Ice Class
PC1 Year round operation in all Polar Waters

Year round operation in moderate multi-year ice
PC2
conditions
Year round operation in second-year ice which
PC3
may include old ice inclusions
Year round operation in thick first-year ice which
PC4
may include old ice inclusions
Year round operation in medium first-year ice
PC5
which may include old ice inclusions For guidance only
Summer/Autumn operation in medium first-year
PC6 L1A I A Super 1 m first-year Ice
ice which may include old ice inclusions
Summer/Autumn operation in thin first-year ice
PC7 L1 IA 0,8 m first-year Ice
which may include old ice inclusions
L2 IB 0,6 m first-year Ice
L3 IC 0,4 m first-year Ice
L4 II
E
Note:
PC1 to PC6 may be assigned additional class notation
ICEBREAKER
6. Propulsion systems in ice
Offshore oil & gas projects in 1980ties brought interest in construction of ships prepared to
navigated in sea ice conditions. It was found that requirements for ice strengthening of
propulsion machinery were inadequate to perform reliable operation. As an examples
propeller blade scantlings in the Baltic and Canadian Rules were dependent upon design ice
torque rather than a direct expression of the ice load on a propeller blade., which can lead to
major blade deformation and breakage or fatigue failure. Another problems was identified
as cracking of blade tips and trailing edges of high skew CP-propeller blades.
In case of traditional propeller – ice impact loads will normally origin from either milling
action (the propeller blade edge is cutting through the ice), ice crushing (ice is pressed onto
propeller blade or another surface until the crushing pressure of the ice is reached) or
combination of the two. The direction of force from ice loading acting on propeller blade
depends on relative velocities between ice and propeller, so the ice loads will generate both
axial and transverse forces generating bending and spindle moments on the blades and
torque on the propeller. As a consequence, shafting system response is generated. The most
rules at that time focused on a propeller blades and a static pyramid strength of propeller
hub. IACS I3 and FSICR (2008) requested that CP propeller mechanism to be assessed both
under action of static and dynamic loads. While applicable loads and permissible stresses are
defined the strength assessment is left to the discretion of respective classification society.
Changes resulting from new approach for class verification of the propeller blades design
results in class society demand for additional documentation for completion of approval.
Additional calculation and usage of FEM methods calls for closer cooperation between
classification society and manufacturer of propulsion systems and ship designers.
31
7.Conclusions
Naval Architects and Ship designers must be aware of principles for use of actual ice class
rules and its applicability to ships that are expected to be assigned ice class notation.
Usage of traditional ice class rules in case of specialized ships like rescue ships, high speed
crafts or warships is to be carefully analysed due to different loads on hull and propulsion
system and propeller itself that may appear as consequence of operational requirements that
the said ships must comply with.
Introduction of requirements for polar class ships and icebreakers is an example of ongoing
work within class societies aimed on unification of requirements due to demand from other
shipowners interested in construction of ships for year round navigation in polar waters.
Gaps identified in case of application of FSICR ice class rules to smaller ships and crafts are
existing e.g in technical requirements for inland ships like river icebreakers. PRS will take
measures to review own rules and consequently to take decision to modify own requirements
in order to make them more suitable for potential users and actualized technical requirements.
7.References
[1] TRAFICOM - Guidelines For The Application Of The 2017 Finnish -Swedish Ice Class Rules;
https://www.traficom.fi › FSICR Guidelines 2019
[2] TRAFI - Act on the Ice Classes of Ships and Icebreaker Assistance (1121/2005),
TRAFI/494131/03.04.01.00/201, Date of Issue 14 Nov. 2017,
[3] IACS URI Polar Class: https://www.iacs.org.uk/publications/unified-requirements/ur-i/
[4] PRS Publication 122/P; Requirements for Baltic Ice Class and Polar Class for Ships under PRS
Supervision – April 2021: https://www.prs.pl
[5] L. Norhamo, G.M. Bakken, O. Deinboll, J. J. Iseskär: Challenges related to propulsor – ice
interaction in arctic waters: 1st International Symposium on Marine Propulsors, smp’09,
Trondheim, Norway, June 2009
[6] Internet – Public Domain
32
Ship
manoeuvrability
33
Ilawa Ship Handling Research and Training Centre:

The establishment of the centre, and its present day.
Jacek Nowicki
The Foundation for Safety of Navigation and Environment Protection

(Abstract)
Thirty years ago, in May 1991 the Foundation for Safety of Navigation and
Environment Protection was established. The founders of the Foundation are two universities
from the Tricity: Gdansk Technical University, Gdynia Maritime University and, due to local
interests the city of Ilawa. Administratively, the Foundation is located in Ilawa, due to the
location of the Ship Handling Research and Training Centre on the lake Silm in Kamionka
near Ilawa, belonging to the Foundation. Some laboratories are also located in Gdańsk. The
process of establishing the Foundation lasted about 15 years due to the political and economic
problems in Poland.
Professor Kobylinski from Gdansk Technical University and Professor Orszulok from
Gdynia Maritime University played a special role in establishing the Foundation. The
establishment of the Foundation was also favored by the successive authorities of both
universities.
Fig.1 The first training manned ship model representing a LCC during training on
Jeziorak Lake (1991)
The goal of the Foundation is to carry out activities aimed at increasing the safety of
navigation and work on the protection of the water environment. The activity on increasing
the safety of navigation is supposed to be based on conducting specialized training in the field
of ship handling using the methodology based on large manned models of ships equipped
with simulators of all necessary ship systems: propulsion steering system, anchors, e-nav,
ballast system, etc.
34
Fig.2 Damage ship drift model tests
The research programs concern mainly issues related to the dynamics of the ship. The
protection of the water environment includes works related to the design and construction of
ecological marinas, ecological ships and the design of water aeration devices.
the initial period of the functioning of the facility was not easy due to the necessity to
overcome many barriers arising from the past caused by a different economic system, as well
as the apparent lack of trust in the new research and training center located in Poland. The
world market, however, was the only possibility of the Foundation's development due to the
crisis of Polish ship owners and thus the almost complete lack of Poles willing to be trained in
ship handling.
Fig.3 Training in ship handling. Harbour maneuvers of a ULCS with the assistance of tug
The situation was getting better every year, mainly thanks to positive opinions about
the quality of the Ilawa centre's equipment and level of the staff preparing and conducting the
training in ship handling. The number of trained masters and pilots increases annually,
currently reaching about 220 people on average annually. Training participants come from
almost 50 countries from all continents and are half masters and half pilots. A few years ago
35
we started training tugmasters. A training completion certificate no. 5560 was issued a few
days ago.
On the Silm lake there are training areas with shallow and deep water areas. There are
built mock ups of ports and specialized terminals, as well as mock ups of canals and rivers
equipped with locks.
Practical trainings are carried out on 11 ship models representing various types of
ships: from 400m container vessels to escort tugs. It is possible to adapt the equipment of the
training models to the individual needs of the trainee, from the propulsion system, through the
steering system to the level of e-nav equipment. he Centre also has a FPSO and SBM models
for exercises in the maneuvering carried out in offshore industry, The proposed training
programs can also be adapted to the individual needs of the participants. It certainly requires
greater involvement of the foundation's staff, but it certainly has a positive effect. The opinion
of participants confirms this.
A few years ago, an RTK reference station was installed on the lake shore, enabling
the trajectory of training and research ship models to be measured with an accuracy of 1 cm.
This made it possible to introduce electronic navigation elements such as ECDIS and AIS. It
also extended the scope of research works that could be carried out in Ilawa to include issues
related to anti-collision and remote pilotage.
The currently renewed equipment of some ship models made it possible to deal with
the problems of autonomous ships navigation in terms of defined as level 4 according to
IMO. A year ago, the research works included in the AVAL project and focused on
autonomous ships, were completed.
Fig. 4 Aerial view of the Research and Training Centre located on Lake Silm
Other research projects also concern the issues related to the future of shipping: the
maneuvering qualities of ships equipped with azimuthing propulsion, or the safety of
navigation: the drift of damaged ships or the deep sea towing of large ships.
Currently, the Foundation for Safety of Navigation and Environment Protection is a

well-known and recognized training and research centre that uses technologies based on large
ship models. The contracts we have for research and training activities allow us to be
optimistic about the future.
36
“Towards the Maritime Autonomous Surface Ship (MASS”) A.Miller, A.Rak, M. Rybczak
Abstract
Maritime Autonomous Surface Ship (MASS) is the future vision of maritime transportation. Replacing
human crew with an autonomous system requires: adaptation of relevant legal documents, creation
of reliable safe trajectory determination systems, creation of automatic ship control algorithms,
automatically controlled actuators application and all these subsystems application on the specific
ship. In this publication all mentioned above subsystems are described in detail. Two exemplary ways
of ship automatic control are proposed. An autonomous training ships are also presented, on which
all subsystems were installed. Shown training MASS is equipped with a graphical user interface
enabling monitoring of individual motion parameters and remote control of the ship in case of a
breakdown.
37
Numerical assessment of hull hydrodynamic derivatives in ship

maneuvering prediction
Radosław Kołodziej
Maritime Advanced Research Centre CTO, Gdansk, Poland
Gdansk University of Technology, Poland
Paweł Hoffmann
Maritime Advanced Research Centre, Gdansk, Poland
ABSTRACT
Operating in crowded waterways pose a risk of accidents and disasters due to maneuvering limitations of the
ship. In order to predict ship’s maneuvering characteristics at the design stage, model tests are often executed as
the most accurate prediction tool. Two approaches can be distinguished here: free running model tests and
numerical simulations based on planar motion model with the use of hydrodynamic derivatives obtained from
captive model tests. However, although the model tests give very accurate prediction of ship’s maneuvering
performance, they are often not affordable both due to the cost and due to their duration, making them less
efficient in iterative design process. Thus, CFD simulations became one of the most commonly used research
methods in ship hydrodynamics, offering reasonable compromise between the accuracy and the cost. Along with
continuously improving computing power and increasingly accurate numerical methods, it became possible to
develop new ways of assessing ship maneuvering performance with the use state of the art CFD tools. The
purpose of this paper is to present one of such methods which is established for practical use in design process of
a ship. The analysis is focused on evaluation of hydrodynamic derivatives for the hull based on CFD simulation
equivalent to the rotating arm test - one of the variants of captive tests. The results are verified versus the results
of model tests carried out in Maritime Advanced Research Centre CTO S.A. with the use of planar motion
mechanism - PMM.
KEYWORDS: Captive model tests, CFD tools in maneuvering prediction, MOERI Container ship (KCS), Hull hydrodynamic derivatives,
Mathematical model of ship motion.
38
22nd International Symposium

HYDRONAV 2020 In Ship Design, Safety,
On Hydrodynamics
Influence of propulsion-steering system on the position of pivot point on a

large LNG carrier during standard manoeuvres
M.REICHEL1,2 HYDRONAV 2020
1 Ship Handling Research and Training Centre, Foundation for Safety of Navigation and Environment Protection, Iława,
Poland
2 Institute of Naval Architecture and Ocean Engineering, Gdańsk University of Technology, Gdańsk, Poland
Influence of propulsion-steering system on the position of pivot point on a

Among all the points on the centreline of a ship in planar motion, there is only one point at which the effects of sway and yaw
large LNG
completely cancel carrier during
each other, thus standard
making this point seem manoeuvres
stationary. All other points appear to be turning around this point.
This centre of rotation is called the pivot point. The information on pivot point position is extremely important for masters
M.REICHEL
and 1,2 carrying out various manoeuvres both at high and at low speed. The idea of presented research was to
pilots that are
check how different propulsion-steering systems influence the position of pivot point and thus indirectly the safety of
navigation. For comparison
1 Ship Handling Research and two sister ships
Training have
Centre, been taken
Foundation forinto consideration,
Safety of Navigationi.e. and
twin-pod and twin-propeller
Environment Protection, twin-rudder
Iława,
configurations
Poland of a large LNG carrier. To elaborate the investigated issue, experiments with free-sailing manned model have
been carried
2 Institute out. Standard
of Naval manoeuvres
Architecture and Oceandefined by International
Engineering, GdańskMaritime
UniversityOrganisation, i.e. Gdańsk,
of Technology, turning circle,
Polandzig-zag and spiral
test have been done during the test campaign. Additionally, for pod-driven version, various pod propulsor settings have been
checked to find the pod helm angle that gives similar pivot point position as for conventional twin-propeller twin-rudder
version.
Among all Thethepaper
pointspresents
on the the first part
centreline ofof the research,
a ship in planari.e. the turning
motion, there iscircle
onlytests.
one point at which the effects of sway and yaw
completely cancel each other, thus making this point seem stationary. All other points appear to be turning around this point.
Keywords: pod-driven ship, twin-propeller twin-rudder ship, manned models, free running model tests
This centre of rotation is called the pivot point. The information on pivot point position is extremely important for masters
and pilots that are carrying out various manoeuvres both at high and at low speed. The idea of presented research was to
1. Introduction
check how different propulsion-steering systems influence the position of pivot point and thus indirectly the safety of
navigation. For comparison two sister ships have been taken into consideration, i.e. twin-pod and twin-propeller twin-rudder
It is well of
configurations known, that behaviour
a large LNG of pod-driven
carrier. To elaborate ships issue,
the investigated is dramatically
experiments different than manned
with free-sailing the behaviour of
model have
been
shipscarried
without. Standard manoeuvres
conventional defined by International
propeller-rudder Maritime Organisation,
steering-propulsion i.e. turning
arrangement. circle, zig-zag
Propulsion and spiral
efficiency,
test have been done during the test campaign. Additionally, for pod-driven version, various pod propulsor settings have been
cavitation
checked andthevibrations
to find fromthat
pod helm angle hydrodynamic pointpoint
gives similar pivot of view
positionorasmaintenance,
for conventionalelectric and ventilation
twin-propeller twin-rudder
systemsThefrom
version. paperoperational point
presents the first oftheview
part of differ
research, significantly
i.e. the and forced naval architects and ship
turning circle tests.
engineers to elaborate new methods for design of pod-driven ships.
Frompod-driven
Keywords: the safety of navigation
ship, twin-propeller pointship,
twin-rudder of manned
view,models,
it is free
recognised, that
running model tests pod-driven
ships have also
significantly different manoeuvring abilities comparing to ships with conventional propeller-rudder
1. Introduction
system [1]. One of the reasons of so different manoeuvring abilities is the location of pivot point [2].
It is obvious that the location of pivot point is not constant and depends on many factors such as
It is well known, that behaviour of pod-driven ships is dramatically different than the behaviour of
use of tugs, bow thrusters, ship acceleration or deceleration, ship motion ahead or astern etc. The
ships with conventional propeller-rudder steering-propulsion arrangement. Propulsion efficiency,
position of pivot point is also determined by the amount of steering force that is applied on the
cavitation and vibrations from hydrodynamic point of view or maintenance, electric and ventilation
manoeuvring ship. The steering force depends of course on the type and capabilities of steering system
systems from operational point of view differ significantly and forced naval architects and ship
device.
engineers to elaborate new methods for design of pod-driven ships.
From the safety of navigation point of view, it is recognised, that pod-driven ships have also
significantly different manoeuvring abilities comparing to ships with conventional propeller-rudder
system [1]. One of the reasons of so different manoeuvring abilities is the location of pivot point [2].
It is obvious that the location of pivot point is not constant and depends on many factors such as
use of tugs, bow thrusters, ship acceleration or deceleration, ship motion ahead or astern etc. The
position of pivot point is also determined by the amount of steering force that is applied on the
manoeuvring ship. The steering force depends of course on the type and capabilities of steering system
device.
39
2. Model tests
A model of an LNG carrier built for ship handling training purposes was used for described
research. The model has been built to a scale 1:24 in three sister ship versions with replaceable stern
parts. The available versions are: single-propeller single-rudder, twin-propeller twin-rudder and twin-
pod with pulling propellers. During the presented research, those versions with two propellers have
been used for experiments.
During the tests, manned model fully equipped for ship handling training was used. It was
equipped with classic engine telegraph and wheel for twin-propeller twin-rudder version and
angle/thrust joysticks for twin-pod version.
2.1 Investigated ship

The investigated LNG carrier is in full scale about 270 m long and has the capacity of 160,000
cubic metres. Principal dimensions of the model versions are shown in Table 1. In Figures 1 and 2
shape of stern part of tested models is presented. It has to be noted that both versions have been
designed according to the best knowledge to achieve the compromise between propulsion and
manoeuvring abilities. As a consequence, the stern part of both ship versions differ slightly between
each other. Almost all of the differences are above the water, e.g. for accommodation of steering
gears. However, some minor odds might be seen on the underwater transom part including the length
of rudder head boxes. This may influence for instance roll response during manoeuvring, but should
not change the overall conclusions on the equivalent manoeuvres.
It should be mentioned that in case of twin-propeller twin-rudder configuration the aft
perpendicular is on the rudder stock and in case of twin-pod configuration on the propulsors vertical
axis.
Table 1. Principal particulars of tested ship versions
Twin-propeller twin-
Principal particular Twin-pod version
rudder version
Length overall (m) 11.55 11.35
Length between perpendiculars (m) 11.33 10.93
Beam (m) 1.80 1.80
Draft (m) 0.50 0.50
Displacement (t) 8.210 8.210
Block coefficient 0.790 0.804
Rudder area ratio (%) 1.81 -
Fig. 1. Twin-propeller twin-rudder version Fig. 2. Twin-pod version
2.1 Model test details

All tests for both propulsion arrangement configurations have been carried out for approach speed
of around 1.26 m/s, which corresponds to 12 knots in full scale. The tests have been carried at an
open-air station at Ship Handling Research and Training Centre near Ilawa in Poland. The lake is 59
ha large and provides unique opportunity to check manoeuvring abilities of ships both on deep and
shallow water.
40
Programme of model tests is shown in Table 2. Experiments covered tests carried out according to
standard IMO procedure like turning and zig-zag tests, together with additional tests like spiral test.
Turning and zig-zag tests have been carried out with various pod drive angles, while the twin-propeller
twin-rudder version only with IMO standard helm angles.
Table 2. Model tests programme
Test type Twin-propeller twin-rudder version Twin-pod version
Turning circle (°) 35 10, 15, 25, 35
Zig-zag (°) 10/10, 20/20 10/10, 15/15, 20/20
Spiral curve (°) ±35, ±25, ±15, ±10, ±5, ±1, 0 35, 25, 15, 10, 6, 4, 1, 0, -1;
-35, -25, -15, -10, -5, -2, -1, 0, 1
The tested models were equipped with GPS system, with RTK reference station owned by Ship
Handling Research and Training Centre, what gives nominal position monitoring accuracy of 10
millimetres. The heading was measured with the use of gyrocompass with 0.1° accuracy. It should be
noted, that both the approach speed, the course of the model and the course changes were set
manually, i.e. by an experienced helmsman.
2.2 IMO standard manoeuvres procedure

The Standards for ship manoeuvrability have been adopted by IMO [3] and are the governing
regulations for ships of all rudder and propulsion types, of 100 m in length and over, and chemical
tankers and gas carriers regardless of the length.
Regarding the test parameters that relate to steering device settings, the Standards say:
• Turning circle manoeuvre is the manoeuvre to be performed to both starboard and port
with 35° rudder angle or the maximum rudder angle permissible at the test speed,
following a steady approach with zero yaw rate.
• The 10°/10° zig-zag test is performed by turning the rudder alternately by 10° to either side
following a heading deviation of 10° from the original heading in accordance with the
following procedure:
1. after a steady approach with zero yaw rate, the rudder is put over to 10° to starboard
or port (first execute);
2. when the heading has changed to 10° off the original heading, the rudder is reversed
to 10° to port or starboard (second execute); and
3. after the rudder has been turned to port/starboard, the ship will continue turning in
the original direction with decreasing turning rate. In response to the rudder, the ship
should then turn to port/starboard. When the ship has reached a heading of 10° to
port/starboard of the original course the rudder is again reversed to 10° to
starboard/port (third execute).
• The 20°/20° zig-zag test is performed using the procedure given above using 20° rudder
angles and 20° change of heading, instead of 10° rudder angles and 10° change of heading,
respectively.
A direct spiral manoeuvre have been conducted using the following procedure:
1. the ship is brought to a steady course and speed according to the specific initial
condition;
2. the rudder is turned about 35 degrees and held until the yaw rate remains constant for
approximately one minute;
3. the rudder angle is then decreased in approximately 10 or 5 degree increments. At
each increment the rudder is held fixed until a steady yaw rate is obtained, measured
and then decreased again;
4. this is repeated for different rudder angles starting from large angles to both port and
starboard; and
5. when a sufficient number of points is defined, data recording stops.
41
3. Model test results

As mentioned before, this paper presents the first part of the analyses, i.e. results of turning circle
only. Figures 3-6 show the Pivot Point position in non-dimensional form following equation
for pod-driven ship version.
Fig. 3. Non-dimensional position of Pivot Point in 35° turn to SB (left) and PS (right)
42
Figure 7 shows the Pivot Point position in non-dimensional form following equation
for twin-propeller twin-rudder ship version.
Table 3 shows the Pivot Point position in non-dimensional values for steady turning for both tested
ship versions.
Table 3. Non-dimensional position of pivot point during turning circle

Helm angle (°) Twin-propeller twin-rudder version Twin-pod version
35 0,464 0,327
25 - 0,441
SB
15 - 0,621
10 - -
35 0,540 0,336
25 - 0,438
PS
15 - 0,686
10 - 0,556
3.1 Short results analysis

As it was expected the position of Pivot Point on the pod-driven ship during turning circle is much
closer to the midships than in case of twin-propeller twin-rudder version for the same 35° steering
helm.
Similar position of Pivot Point as for 35° steering helm on twin-propeller twin-rudder ship version
is estimated for around 23° (SB) and 21° (PS) pod deflection angles.
It is anticipated that for smaller ship speeds this relation remains basically similar, but to be sure
additional experiments are planned.
Such an information might be very useful for planning and prediction of manoeuvres especially in
restricted waters, as reaction of ship in terms of pivoting is crucial for safety of navigation.
4. Acknowledgments
Research presented in the paper has been financed by National Science Centre, Poland under the
grant 2017/01/X/ST8/00785 “Equivalent standard manoeuvres for pod-driven ships”.
5. References
[1] Reichel M. (2019). Equivalent standard manoeuvres for pod-driven ships. Ocean Engineering, 187,
106165, doi.org/10.1016/j.oceaneng.2019.106165
[2] Artyszuk J. (2010). Pivot point in ship manoeuvring. Scientific Journals of Maritime University of
Szczecin, 20(92) pp. 13–24
[3] International Maritime Organization (2002). Standards for Ship Manoeuvrability - Resolution MSC 137
(76)
43
Investigation of heading and manoeuvring properties of stern trawler B-280 with

Hubble type rudder blade
1)
mgr. inż. Paweł Dereszewski
mgr. inż. Włodzimierz Hebel 1)
dr. inż. Mirosław Dereszewski
dr. hab. inż. Paweł Dymarski
1) Mapalu Marine Innovations
Abstract
The course stability and turning parameters such as circulation time and radius of stern
trawler type B-280 were investigated, both analytically and practically. The paper consist of
description of stand experiments with the rudder blade model and particulars of sea test planning
and execution. New construction, replacing NACA profile blade, typical for that type of fishing
vessels, shall enhance manoeuvring properties. Paralelly, positive influence at fuel consumption was
expected. The goal of investigation, conducted at open sea, was to find out about potential
advantages and disadvantages of new blade in comparison to traditional one.
First part of the paper consist of description of general idea and technical particulars of
nozzle blade, as well as short explanation of experiments with the blade model, carried out in small
test tank. In general, one can say that mode of operation is based on flow of stream of water through
a nozzle, what creates force vector enhancing action of a rudder blade. Second part is dedicated to
description of the trawler B 280 technical particulars and details of sea trials plan and realisation.
Plan of the research included tests of circulation as well as heading stability for yawing level
determination. Third part of the paper presents obtained results concerning better manoeuvrability
due to significant turn radius reduction and improvement of course stabilisation what should have
influence at lower fuel consumption.
Key words: trawler, Hubble ruder, model calculation, sea trials
44
Advanced computational
and experimental
modeling
45
HYDRONAV 2020
22nd InternationalSopot
Symposium on14-16
- Iława, Hydrodynamics in Ship Design, Safety,
October 2020
Measurement Stand for Experimental Model Tests of the Dynamically
Positioned Vessels
MAREK KRASKOWSKI1
PIOTR GRYMAJŁO2
HYDRONAV 2020
1
Sopot
Maritime Advanced Research Centre (CTO- Iława, 14-16
S.A.) Gdańsk, Poland,October 2020
[email protected]
2 Gdańsk University of Technology, Gdańsk, Poland, [email protected]
Measurement Stand for Experimental Model Tests of the Dynamically

The paper presents the process of design, development and implementation of the measurement stand dedicated for
experimental model tests of the vessel’s dynamic Positioned
positioning (DP)Vessels
capability. The presented stand was developed using the
typical anchor handling tug/supply (AHTS) hull model as a representative testing object, characterized by the most common
MAREKofKRASKOWSKI
configuration 1
thrusters. The paper presents the details of the hardware developed to imitate the actual propulsion system
and control system of the vessel at model scale, as well as the details of adopted control strategy. The purpose of the
development of described measurement
PIOTR GRYMAJŁO 2 stand is primarily the validation and calibration of the computational model
developed in parallel, as well as model scale experimental investigation of the motion of dynamically positioned vessels. The
results of the first successful experiments are presented, including the response to irregular waves in DP mode, and response
to given thrust forcing signal, measured in order to identify the inertia and damping characteristics of the vessel. These
1 Maritime Advanced Research Centre (CTO S.A.) Gdańsk, Poland, [email protected]
parameters, used to tune the mathematical model of the vessel’s motion, are compared with approximate numerical
2 Gdańsk University of Technology, Gdańsk, Poland, [email protected]
predictions. Finally, the conclusions and directions of further work are presented. The experiments were carried out in Ship
Hydromechanics Division of the Maritime Advanced Research Centre CTO S.A.
TheKeywords:
paper presents
dynamicthe processmarine
positioning, of design,
control development and implementation
systems, scale model of the measurement stand dedicated for
tests, numerical simulation
experimental model tests of the vessel’s dynamic positioning (DP) capability. The presented stand was developed using the
typical anchor handling tug/supply (AHTS) hull model as a representative testing object, characterized by the most common
configuration of thrusters. The paper 1. Introduction
presents the details of the hardware developed to imitate the actual propulsion system
and control system of the vessel at model scale, as well as the details of adopted control strategy. The purpose of the
The floating
development object’s
of described ability to
measurement standmaintain its the
is primarily position andandheading
validation in ofadverse
calibration environmental
the computational model
developed in parallel,
conditions withoutasthe wellmooring
as model scale experimental
system, referredinvestigation
to as dynamicof the motion of dynamically
positioning (DP), ispositioned
nowadays vessels.
oneTheof
results
the of the firstfeatures
standard successfulofexperiments
many types are presented,
of vessels including the response
and offshore to irregular
objects, waves ine.g.
including DP mode, and response
offshore/diving
to given thrust forcing signal, measured in order to identify the inertia and damping characteristics of the vessel. These
support
parameters,vessels,
used todrill
tuneships, rescue vessels,
the mathematical modelwind
of the turbine
vessel’sinstallation
motion, areunits, FPSO’s
compared with and others. Meeting
approximate numerical
the requirements
predictions. Finally, the related to dynamic
conclusions positioning
and directions capability
of further work thus becomes
are presented. one of
The experiments thecarried
were design tasks,
out in Ship
consisting
Hydromechanics in optimal
Division ofselection of thruster
the Maritime Advanced types
Research and configuration,
Centre CTO S.A. as well as sizing of the thrusters
and the power plant capability; certain redundancy is often required, i.e. the vessel is required to
maintain itsdynamic
Keywords: position and heading
positioning, also
marine control in case
systems, scale of thruster
model or power
tests, numerical plant failure. The basic design
simulation
analysis of the DP system consists in proving that the thrusters are capable to balance the predicted
environmental forces due to 1. wind,
Introduction
wave and current, without exceeding their design power rating. In
order to take into account the natural unsteadiness of the environmental forces, their mean values are
The floating
multiplied object’s factor,
by a dynamic ability e.g.
to maintain
1.25 [1]. It itsshould
position be and
notedheading
that such in an
adverse environmental
analysis only allows
conditions without the mooring system, referred to as dynamic positioning
stating that the heading and position of the object at given power output can be maintained. However, (DP), is nowadays one of
the standard features of many types of vessels and offshore objects,
no information is provided about the accuracy of position and heading control, i.e. what is the including e.g. offshore/diving
support vessels, drill ships, rescue vessels, wind turbine installation units, FPSO’s and others. Meeting
the requirements related to dynamic positioning capability thus becomes one of the design tasks, 1
consisting in optimal selection of thruster types and configuration, as well as sizing of the thrusters
and the power plant capability; certain redundancy is often required, i.e. the vessel is required to
maintain its position and heading also in case of thruster or power plant failure. The basic design
analysis of the DP system consists in proving that the thrusters are capable to balance the predicted
environmental forces due to wind, wave and current, without exceeding their design power rating. In
order to take into account the natural unsteadiness of the environmental forces, their mean values are
multiplied by a dynamic factor, e.g. 1.25 [1]. It should be noted that such an analysis only allows
stating that the heading and position of the object at given power output can be maintained. However,
no information is provided about the accuracy of position and heading control, i.e. what is the
1
46
maximum deviation from the position and heading setpoints at given power output. As these
parameters are often important and are included in technical specifications of the vessel, the quasi-
steady analysis based on mean environmental forces becomes insufficient and next level of design
analysis is required, including complete dynamics of the object, the environment as well as the control
system. The presented work aims at the development of experimental and computational tools capable
of complete, fully coupled dynamic analysis of the dynamically positioned vessel, including actual
simulation of environmental forces and realistic control strategy. This paper focuses on the first stage
of this work, i.e. development of the measurement stand. This type of experimental tests presents
substantial challenge, both due to the difficulty of the control task, as well as due to the large
complexity of the model equipment, much more sophisticated than for any other tests conducted in the
towing tanks. For this reason, DP tests are not a standard and are currently carried out by only few
institutions in the world. In this paper, the adopted assumptions, design requirements and selection of
testing object are described at first; then, complete description of the control system is given. The
results of first successful tests are presented, and compared with the results of predictions based on
well established computational methods. Finally, conclusions and directions of further development
are presented.the concise survey of computation methods for added resistance in waves prediction in
comparison with experiments in particular for the following waves condition. The model scale
experimental methods for evaluation of ship’s resistance due to wave are the
2. Design requirements and assumptions.
The described measurement stand for model testing the dynamically positioned objects was
developed to be used in the auxiliary towing tank of Maritime Advanced Research Centre. The facility
is 60m long, 7m wide and 3m deep, equipped with computer-controlled wavemaker with active wave
damping (two identical sets of paddles are installed on both ends of the tank, one of them serving as
active damper, generating the waves in counter-phase).
The tank is also equipped with wind generator, capable of generating the specified wind spectrum,
installed on the separate carriage. In order to be able to control the position and heading of the vessel
model, accurate motion tracking system is necessary; while the full scale objects usually use differential
GPS system, at model scale the optical QUALISYS system is used. The motion tracking system is to be
installed also on the movable carriage, which allows simplified simulation of the current, i.e. motion of
the “earth” reference coordinate system is recognized by the control system as the current speed.
The most challenging requirement for the DP control system, related both to actual system and the
laboratory system, is filtering out the wave frequency motion from total motion of the object. An
attempt on preventing the wave frequency motion would lead to unacceptable power consumption
and wear of thrusters, so the dynamic positioning systems are designed so as to compensate the mean
wave drift and the slowly-varying drift forces while allowing the vessel to move freely with wave
frequency. Details are given in the section related to control system.
The measurement stand in its current version is designed for ship-type objects, equipped with two
tunnel thrusters in the bow and two azimuthing thrusters in the aft part. The control system is
required to provide independent speed control of all thrusters as well as independent azimuth angle
control for azimuthing thrusters.
The operator’s station should be located on the towing carriage, and provide full access to all
parameters to the control system; besides the automatic control, full manual control of the model
motion (e.g. with the joystick) should be possible. In order to assure sufficient accuracy of the
measurements, it is necessary to make the connection between the operator station and the model
completely wireless, due to the fact that the cable introduces the force of the same order as the
environmental forces or even larger. For that reason, battery power supply and wi-fi communication
should be adopted.
2
47
One additional requirement for the laboratory DP system is related to the control signals for the
thrusters; unlike full scale thrusters, characterized by significant inertia, the model scale thrusters are
capable of reaching full thrust almost immediately. For that reason, in order to mimic the actual
dynamics of the full scale propulsion system, a ramp must be applied in the control signal must me
introduced.
Size of the towing tank limits the model length to approximately 3 meters to prevent excessive
blockage when floating transverse to the tank axis. For the purpose of development of the
measurement stand for testing the DP systems, the typical anchor handling tug/supply vessel was
selected. The length of the full scale ship is 98.7m; the model scale for presented experiments is 1:36.
As mentioned above, it is equipped with two tunnel thrusters in the bow and two azimuthing thrusters
in the aft part.
3. DP control system
The general control strategy is based on the idea of nonlinear passive observer, presented by Thor I.
Fossen ([3], [4]). The block scheme of the applied DP control strategy is presented in Fig. 1 below.
Fig. 1. Block scheme of the DP control system

The reference model enables generating smooth reference trajectory. When the setpoint changes,
the theoretical trajectory for the vessel to follow becomes a unit step function. An attempt to follow
such a trajectory by the control system results in sudden thrust changes, unacceptable accelerations
and excessive thrusters wear. The trajectory generated with the use of the reference model takes into
account physical limitations of the vessel and the thrusters. This allows improving the working
conditions of the crew, improving the precision of DP control and reducing the thrusters wear.
The control law applied in the DP controller is similar to that used in PID controller; three
considered degrees of freedom of the vessel's motion are controlled independently. The difference
consists in the term related to the vessel's velocity: instead of differentiation of the error signal (like in
the PID controller), the term proportional to the velocity error signal is used. The velocity error is
defined as the difference between the set velocity resulting from the reference trajectory and the
estimation of the velocity provided by the estimator.
Thrust allocation means evaluation of the thrust for subsequent thrusters based on global steering
force and moment provided by the DP controller. In presented version of the control system, constant
values of the angles of azimuth thruster are assumed.
Measurement of the vessel's location and heading at model scale is realized by the optical system
QUALISYS. Using the filtering and evaluation of the signal quality allows preventing adverse influence
of e.g. artificial signal peaks or momentary loss of signal on the control quality.
3
48
The main task of the estimator in modern DP systems as well as in the discussed laboratory stand,
is to determine the slowly-varying component of position, heading and velocity of the model, which
makes it possible to eliminate oscillatory motion induced by the waves from the feedback loop. As a
result, the control energy is used to prevent the slowly-varying tendencies in the ship's motion, and it is
not wasted on preventing the inevitable wave-frequency oscillation around the operating point - this
reduces energy, extends the operating range of the DP system and prevents premature wear of the
thrusters. In addition, the estimator, which combines the measurements (often distorted, irregular or
even temporarily unavailable) with a mathematical model of the ship, allows to improve the quality of
measured signals, which translates into improved control quality (accuracy and ability to maintain
position and heading).
4. Vessel's response to irregular wave in DP mode
The first experiments carried out after completion of the measurement stand consisted in
qualitative verification of the response to irregular waves of different spectra, selected to achieve large
difference in steady drift force generated by the wave. Exact parameters of the wave spectra used in the
experiment are listed in Tab. 2.
Table 1.Parameters of the wave spectra
Spectrum
No. HS [m] TP [s]
1 1.44 9.60
2 1.44 5.52
The resulting response of the dynamically positioned vessel to the irregular wave defined by two
spectra specified above is presented in the following form:
− time history of revolution rate for all four thrusters
− time history of total force acting on the hull, computed on the basis of known characteristics of the
thrusters and their configuration
− time history of the vessel motion (surge motion and yaw angle)
Fig. 2. Thrusters revolution rate - dynamic positioning for low drift force
4
49
Fig. 3. Thrusters revolution rate - dynamic positioning for large drift force
Fig. 4. Mean drift force acting on the hull - measured (blue) vs. predicted mean value (orange) - low drift force
Fig. 5. Mean drift force acting on the hull - measured (blue) vs. predicted mean value (orange) - large drift force
Fig. 6. Surge motion - large drift force (black) vs low drift force (red)
5
50
Fig. 7. Yaw angle - large drift force (black) vs low drift force (red)
As it can be seen, the predicted large difference in mean drift force is very clearly visible in the
results of the experiment. For the case characterized by large drift force, the control signal for one of
the thruster and thus the resulting revolution rate even gets saturated (Fig.3). Also, reasonable
agreement between predicted mean drift force and actual force computed on the basis of thruster
characteristics was achieved (Fig. 4, Fig.5). At the moment, no hull-thruster interaction was taken into
account in evaluation of total force; further studies are required here.
5. Identification of the parameters of mathematical model of the vessel motion in DP

mode
As mentioned in the introduction, one of the main purposes of the development of the
measurement stand is to obtain the platform for validation of numerical model for simulating the
vessel motion in DP mode. Using the numerical simulations instead of experiments is obviously
preferred whenever possible, however, it must be proved that the applied numerical model provides
reliable results. The development of full simulation model with control loop is currently in progress;
the work done so far consists in selection of model motion and exact experimental identification of its
parameters.
The nature of motion in dynamic positioning mode allows making the following assumptions:
− planar motion (surge, sway, yaw) is a sufficient description of the actual motion - similarly as the
commonly accepted manoeuvrability models;
− due to low velocities (required values are close to zero), negligible coupling between considered
degrees of freedom occurs; the off-diagonal terms of the mass and damping matrices can thus be
neglected.
The mathematical model of motion in the following form was applied:
M ⋅ v + DL ⋅ vr + DNL (vr )vr = τ + wwnd (t ) + wwav (t ) + RT (ψ )b

η = R(ψ )v
where:
η = [x, y,ψ ]T - vector of position and heading

v = [u , v, r ] - velocity vector in the vessel coordinate system
T
vr = v − vc - relative velocity vector in the vessel coordinate system

vc - current velocity
M - generic mass matrix (hull+added mass)
DL - matrix of linear damping coefficients
DNL - matrix of nonlinear damping coefficients
6
51
τ - vector of global forces and moment generated by the thrusters

wwnd (t ) - wind force
wwav (t ) - wave force
b - vector of constant and slow-varying forces defined in Earth coordinate system; enables including
additional forces in the simulation.
The parameters to be identified are: added masses and damping coefficients. The experiment
conducted in order to identify these parameters consisted in recording the model trajectory and the
revolution rate of the thrusters in manually controlled motion of the model. The model trajectories
were selected to identify the characteristics for one selected degree of freedom in each case, i.e. the
analysed motions were: pure surge, pure sway and pure yaw. Knowing the characteristics of the
thrusters (i.e. thrust vs. rpm) allow to obtain the thruster forces as well as the resulting motion as a
direct result of the experiment. It should be noted than environmental forces are zero in analysed case.
Processing of the results consisted in iterative adjustment of total mass (i.e. hull plus added mass),
linear damping coefficient and nonlinear damping coefficient so as to achieve best possible match
between simulated trajectory and actual measured trajectory. The figures below show the results of
simulated motion with the use of most suitable coefficients for surge, sway and yaw, respectively.
Velocity - X
Velocity - Y
Angular velocity
Fig. 8. Actual and simulated motion: surge, sway and yaw
An approximate verification of the parameters obtained as presented above can be done using the
following data:
− actual mass of the hull model;
− added masses for surge, sway and yaw evaluated with the use of ANSYS AQWA software;
− hull drag coefficients obtained from the model scale experiments.
The drag coefficients for the vessel of very similar shape are available; they were obtained by means
of drag measurements in a wind tunnel, realized for double-body hull located in the middle of the
measurement chamber. It is a commonly accepted method for evaluation of drag coefficients for
7
52
submerged part of the hull, characterized by similar accuracy as the measurements in the towing tank
as well as very short execution time.
Evaluation of the drag coefficients in the wind tunnel at constant speed is based on the assumption
that the relation between the flow velocity and the drag is a quadratic function, which corresponds to
purely quadratic damping in assumed motion model. This assumption is quite close to obtained
results of the model identification, as the damping is dominated by quadratic damping. Such a
verification of damping coefficients is possible only for surge and sway, as it is not possible to obtain
the flow corresponding to pure yaw in the wind tunnel.
The tables below show the results of approximate verification of the identified parameters of the
mathematical model of motion.
Table 2. Identified mass parameters - verification
Hull mass+added mass
Experiment AQWA
Surge 250.0 kg 252.3 kg
Sway 390.0 kg 396.1 kg
Yaw 155.8 kg·m 2
131.9 kg·m2
Table 3. Identified damping coefficients - verification

Quadratic damping coefficient
Experiment Wind tunnel
Surge -10.1 Ns2/m2 -13.8 Ns2/m2
Sway -240.1 Ns2/m2 -247 Ns2/m2
As it can be seen, almost perfect math of total masses for surge and sway motions has been
achieved. For yaw motion, considerable difference is visible; which requires further explanation. One
possible reason is the fact that the moment of inertia for yaw motion was not measured directly as it
was technically impossible. It has been assumed that the moment of inertia in yaw motion is equal to
the moment of inertia in pitch motion.
6. Conclusions and further work
The described work resulted in development of fully functional measurement stand for direct
measurements of the motion and required power output for the ship-shaped object operating in
dynamic positioning mode. The conclusions from the experiments done so far are as follows:
• experience gained during the development of the hardware revealed absolute need of fully
wireless communication between the model and the operator's station;
• the predicted difference in mean drift force resulting from the change of the wave spectrum peak
period was fully confirmed during the experiment;
• exact identification of the mathematical motion model allowed for simulating the surge, sway and
yaw motions with good accuracy;
• during entire experiment, the equipment as well as the control system proves its robustness and
reliability.
Further works, already started, include:

• exact evaluation of the hull-thruster interaction;
8
53
• repeating the experiments for wide range of wave conditions to gain more validation data for
numerical predictions of mean drift force;
• development of numerical model of the hull motion in DP mode, including complete simulation
of the control system and the hull dynamics;
• experiments focused on detailed validation of the model;
7. Acknowledgement
The research was financed from the funds of the Ministry of Science and Higher Education of the
Republic of Poland, allocated for the statutory activity of Maritime Advanced Research Centre CTO
S.A.
8. References
1. DNV-GL standard No. DNVGL-ST-0111 "Assessment of station keeping capability of dynamic

positioning vessels", 2016
2. Faltinsen O.M., Sea Loads on Ships and Offshore Structures, Cambridge University Press, 1993
3. Fossen, T., "Marine control systems. Guidance, navigation and control of ships, rigs and
underwater vehicles.", Marine Cybernetics, Trondheim, 2002
4. Strand J.P., Fossen T.I. (1999) Nonlinear passive observer design for ships with adaptive wave
filtering. New Directions in nonlinear observer design. Lecture Notes in Control and Information
Sciences, vol 244. Springer, London.
9
54
HYDRONAV 2020
22nd International Symposium On 19-21
Sopot-Iława, Hydrodynamics
May 2021In Ship Design, Safety,
Digital twins concept in the activities of Classification Societies

Janusz Narewski, Polski Rejestr Statkow S.A.
Abstract:
HYDRONAV 2020
Until recently, simulation and computational models were mainly used to support the design and verification of
Sopot-Iława,
structures. For some time, advanced computational tools have 19-21 May
been adopted also 2021
as the equipment used during ship operation.
One of the most important and relatively new terms associated with this trend, popularized during the current digital
transformation of industry since beginning of the 21st century, is the concept of 'digital twin'.
The article is an overview of its significance for the maritime industry, indicating what should distinguish such
systems from classic simulation and computing environments and how this concept is shaped in the era of digital
Digital
transformation. twins
It also gives concept
examples in the
of such systems activities
developed and usedof Classification
in the Societies
activities of a classification society such as
Polish Register of Shipping. Janusz Narewski, Polski Rejestr Statkow S.A.
Key words:
digital twin, remote survey, ship automation, autonomous ship, smart shipping
Abstract:
Until recently, simulation and computational models were mainly used to support the design and verification of
General concept of digital twins
structures. For some time, advanced computational tools have been adopted also as the equipment used during ship operation.
One of the most important and relatively new terms associated with this trend, popularized during the current digital
transformation
The of industry
idea of “digital since(DT)
twins” beginning
has of the 21st
been century, is and
popularized the concept
growingof 'digital twin'. in manufacturing
intensively
The article is an overview of its significance for the maritime industry, indicating what should distinguish such
industry since 2002 [1].
systems from classic simulation and computing environments and how this concept is shaped in the era of digital
Due to the variety
transformation. It alsoofgives
possible applications,
examples it isdeveloped
of such systems difficultand
to used
get unique definition
in the activities of “digital society
of a classification twins”.
such as
One ofRegister
Polish the simplest is "model based definition" (MBE) which utilize 3D models and associated
of Shipping.
metadata
Key words: to work as digital functional clone. Definition with term “digital product definition” (DPD)
digital
is described in twin,
[2] . remote
Manysurvey, ship based
other are automation, autonomous ship,
on enumerations ofsmart shipping of DT. For example:
key features
• asset representation models (contains a description of the most important physical and
constructional features of an object). Usually, it is a documentation in a modern 3D model format
General concept of digital twins
containing both geometric features and technological features or material properties.
• behavioral models (contains models and data taking into account configuration and internal and /
Theor idea of “digital
external stability twins” (DT) has
depending been
on the popularized
loading and growing
condition), such modelintensively in manufacturing
can be used in computational
industry since
tools, e.g. 2002 [1].
stability calculators, strength calculators or FEA, CFD solvers.
• Due to the variety
conditions of possible applications,
and configurations data (data it reflecting
is difficultphysical
to get unique definition of changes)
and configuration “digital twins”.
One of the simplest is "model based definition" (MBE) which utilize 3D models and associated
metadata to work
What does as digital
digital twinfunctional
term mean clone. Definition
in practice? with term “digital
In shipbuilding product
industry, definition”
it could (DPD)
take a form of
is described in [2] . Many other are based on enumerations of key features
hybrid hardware-software solution arranged to increase intelligence and functionality common ship of DT. For example:
• asset representation
subsystem (engines, steering) models
or of(contains a description
an additional of the mostitsimportant
system controlling behaviour physical and or
e.g. strength
constructional
stability aspects. Infeatures of an object).
XXI century Usually,we
era of internet, it observe
is a documentation
explosion ofinmany
a modern 3D model
solutions aimedformat
at
containing both geometric features and technological features or material
fusion of information technologies with hardware systems to increase their autonomy, flexibility, properties.
• behavioral
safety and sometimesmodels to (contains models
creates quite newandadds. data
Intaking
many of intothese
account configuration
solutions, there areand internal
digital and /
models
or external stability depending on the loading condition),
used and solutions in many cases are described as “digital twins”. such model can be used in computational
tools, e.g. stability
Figure 1 presents calculators,
an example strength calculators
that utilizes digitalortwin
FEA, asCFD solvers.
computer system named loading and
• conditions and configurations data (data reflecting physical
stability calculator. This is one of the most popular and well known implementations. and configuration changes)
The example
What does digital twin term mean in practice? In shipbuilding industry, it could take a form of
hybrid hardware-software solution arranged to increase intelligence and functionality common ship
1subsystem (engines, steering) or of an additional system controlling its behaviour e.g. strength or
stability aspects. In XXI century era of internet, we observe explosion of many solutions aimed at
fusion of information technologies with hardware systems to increase their autonomy, flexibility,
safety and sometimes to creates quite new adds. In many of these solutions, there are digital models
used and solutions in many cases are described as “digital twins”.
Figure 1 presents an example that utilizes digital twin as computer system named loading and
stability calculator. This is one of the most popular and well known implementations. The example
1 55
describes an extended version with ability to use it remotely in Fleet Management Center on shore.
This allows to provide remote monitoring and control of ship condition. The most important benefit of
this solution is the ability to use it in emergency situations where external experts can help crew in
making optimal rescue decisions or in elaboration of crisis solving measures.
Ship command center Shore support center

• real ship data • twin remote copy
• twin ship data Communication link • experts
Batch data link
Figure 1 Basic concept of digital twin as digital functional copy of real construction.
From the perspective of classification societies, most important are digital systems supporting
maintenance and safety control. The most known and used by customers ships are software known as
stability calculators. PRS as one of leading classification societies actively participates in the
development of such solutions. This kind of solutions enables to support an additional service named
Emergency Resource Center (ERC) [3].
Digital twins need connection to dedicated infrastructure of sensors network to operate
efficiently. General tendency is to create systems that can use number of digital assets like computer
equipment integrated and named as 'smart ship infrastructure', which nowadays enables new
possibilities to ships functionality.
Smart ships infrastructure in the era of digital twins

Smart and intelligent ship infrastructure means equipment with dozens of
sensors connected by structural networks and smart data gateways/servers to
unified controlling and
steering centers. The Communication
difference in relation to links Batch data acquisition
previous generations of
such equipment is the use
in the background internet
technologies. This allows Legend:
sensors
to create modern,
control equip.
efficient and open to
gateways
digital revolution
smart tools
infrastructure. Typical
telekom links
elements of the smart ship
infrastructure are
presented in Figure 2
In this
architecture, key role is
played by ‘smart’ tools.
Figure 2 Smart ship infrastructure
One of the most
important is named ‘ship data center’ which may be interpreted as heart of ship digital intelligence
56 2
and center of integration. The infrastructure is complemented by

a module for external communication ensuring on-line ship
connectivity with the shore and world services.
For industry, important role is played by standards. The
industry follows general trends of so-called IoT (internet of
things) and in case of ship, a branch named IoS (Internet of
ship) must be prepared. Asia area and especially Japan industry
promote solutions oriented on ShipDC (Ship Data Center)
managed by JSMEA association. In the year 2021there is a
number of ships equipped with this kind of system, with their
quantity reported to over 200 vessels [4].
In modern ships, the importance of availability of
digital solutions may be emphasized by special notations in the
class marks describing ship digital equipment. For example,
ABS classification society assigns the following notations in the
class marks:
'SMART (INF)' meaning Data Infrastructure for Smart
Function implementation
'SMART (SHM)' meaning Structural Health Monitoring
Infrastructure.
Table 1 Example categories of smart vessel functions,
The growing market of smart solutions for shipbuilding is structures and systems from ABS (2020)
expressed by guidelines classifying these systems. Table 1 from
ABS [5] in 2020 classifies sixteen different functional types of smart systems grouped in three main
categories (health, efficiency, crew support).
This trend in industry needs standardization. Popular are:

• for ship equipment and sensors: SFI Coding and classification system and ISO
19848 standard
• for ship smart devices: ISO 19847 standard (Figure 4)
Figure 3 SFI Groups

codes
Figure 4 ISO19847 Architecture

• for complex digital solutions: Smart Ship Application Protocol initiative (SSAP)
3 57
• ship data centers and integrating

modules:
ShipDC standard [Figure 5] element of
SSAP initiative.
Smart devices are oriented on data

processing. In the digital world, the use of
digital twins takes special place. Smart
devices need smart solutions to cooperate
with the real systems. Arranging them in
SIL and HIL configurations appears one of
the best concepts.
Figure 5 ShipDC standard

Idea of hardware and software in the
loop (HIL & SIL)
Utilizing digital twins in applications using "hardware in the loop" (HIL) and "software in the
loop" (SIL) is a special concept. Basic idea is described in Figure 6 .
Two main steps can be realized: a) comparing signals from real systems (HIL) and digital twin
systems (SIL) and b) analyzing differences.
Smart Tool
Software system (SIL)

digital twin +
Validation
Hardware system (HIL)' -
real ship' Comparing unit
Figure 6 Validation of HIL and SIL solution scheme
The famous test developed in 1950 by Alan Turing,

pioneer of computer science, may show how powerful and
universal is this idea. The original test was developed for
checking if machine is able to emulate human intelligence, see
[6] and Figure 7 presents how to compare and classify two
variants of communications: C-B (human-human conversation)
and C-A (human-machine conversation).
HIL and SIL systems adopt Turing concept also in
shipbuilding for testing and verifying complex systems and they
may be found in many modern digital solutions increasing Figure 7 Turing Test example
(Wikipedia)
autonomy of ships. Some classification societies, (for example
Bureau Veritas in [7] , DNVGL in [8] ), publish guidelines on how to utilize theis HIL and SIL
solutions in real world.
458
Digital Twins as co-simulation approach

Implementing digital twins
should be conformant with appropriate
methodologies/ standards. One of them
is named co-simulation [5]. It allows to
prepare behavioral part of digital model
in a modular way with defined inputs,
outputs and internal state logic.
Many commercial and
academic tools offer stable standards
for that. One of the most popular
projects is based on the language
Modelica. Many other CAE/CAS tools
have dedicated modules allowing to
build prototypes with similar way (e.g.
ANSYS, Matlab ).
Figure 8 Open Modellica model - gas turbine generator
The relatively new Open
Simulation Platform (OSP) initiated in
2017 by DNVGL, Kongsberg
Maritime, SINTEF, NTNU is
interesting for shipbuilding industry
and now shared by several open
repository useful ship subsystem
models. Figure 9 shows an example -
ship DP dynamic positioning system.
Figure 9 Vessel DP positioning dynamic model – opensimulationplatform.com
Concept of fusion of the ship survey and stability data as the method to update state of
ERC digital twin
Classification society which maintains digital twin models (in PRS, ERC service) needs
effective methods to keep the data of the structure up to date also in the simulation process. Evolved
concept utilizing idea of HIL and SIL is presented in Figure 10.
The schema shows the evolution of two previously independent processes:
(a) utilizing simulations for R&D activity and
(b) processing data from ongoing inspections, and
(c) integrated process by fusion of tasks a+b.
5 59
Legend:
Simulation 1. Analyses, checks,
4 7 experience from
• Ship model
digital twin
• scenarios, cases simulations,
2. Analyses and
reports by human
Analyses, Checks, surveys
1 3. Data from
Queries,
monitoring systems
and sensors
a) Computational tool 4. Core simulation
model with
collections of cases,
Real Ship 5. Data for control and
Monitoring, 7 automation
2 Survey Reports, 6. Stream from ERC
team
7. Decisions,
b) Real Construction survey life Knowledge , R&D
cycle
+
Digital 4 7
Twin • Ship model
• scenarios
• dane+analizy
5 6 1
Analyses, Checks,
Answers
Real 2 Survey Reports
Twin
3 Monitoring,
c) Coupled system as digital and real twin with fusion of experience database
Figure 10 Fusion survey and stability model to update digital twin state
Review of ship applications having elements of digital twins
Digital twins for ship stability verification

This is one of the most popular type of “digital twins” evolving within software solutions for
more than twenty years. Specialized software enables computing and verifying stability margins in the
intact and damaged states of ship or other object. Based on hydrostatic data from hull documentation,
DT allows also to calculate bending moments and shear forces for general strength.
Due to the usefulness of these systems and the direct impact on ship safety control, maritime
regulators introduced requirements and rules for its functions, e.g. IMO IS CODE 2008 [10] and
IACS UR L5 [11] (last rev from 2017).
60 6
Computational ship models contain data of geometry

of hull and internal compartments. Based on the load data,
volumes, masses and many other quantities important for the
estimation of ship static and dynamic behaviour are calculated.
As basic result, we are able to get estimated, displacement,
draught, heel angles and inclined stability margin.
Computational accuracy is strictly dependent on the
similarity of the hull model shape and properties and actual
ship structure. This may vary with ship aging and conversion
or even with inaccuracy of input data. Therefore periodical
calibration or review is required.
IACS UR L5 [11] define official international margins
for this kind of software (see Figure 11 ).
The requirements define four levels of functionality for
these systems. In practice, type 4 is the most advanced one,
which can offer emergency support assessing the possibility to
safely reach the port (SRtP) decision.
More advanced installations allow also to:
- manage cargo, Figure 11 Required accuracy common ship stability
parameters in IACS UR L5
- integrate with ship sensors and gauge instruments
to work in more automatic modes,
- communicate with shore operational centers or emergency response centers (ERC)
Figure 12 Screens from PRS SA software StabPRS
PRS describes appropriate requirements in publications [12], [13]. PRS R&D Division also offers to
our customers this kind of tools (type 4) to be installed on shipboard. Based on this kind of software, it
is also possible to offer support and special advisory service, called Emergency Response Center,
allowing assistance in the case of accidents requiring advanced calculations.
Extending stability measured parameters by smart systems

Some important calculations cannot be done directly. Using sensors data and computing them in
dynamic conditions by smart systems enables to estimate a few more parameters:
• LCG, VCG, TCG, based on accelerations and angular velocities, apart from ship particular
parameters: [14]
• Longitudinal center of flotation (or buoyancy) (LCB) [15]
61
7
Digital Twins for controlling hull strength

For the largest ships, smart systems with embedded digital twins are used to control hull strength.
One of such solutions is offered by the Korean company Global Maritime Engineering [16]. Their
Figure 13 Global Maritime Engineering offered Hull Stress Monitoring System

system (see Figure 13) based on data from the set of sensors (strain gauge, accelerometers, pressure)
allows us to control:
- Ship bending moment;
- Slamming;
- Estimated fatigue life.
Summary:
Smart digital solutions in shipbuilding are changing the entire industry.
Digital twins are based on well-known technologies (simulation systems). Used onboard (or remotely)
real ships during its life are able to create new functional opportunities. Many different “digital twin”
configurations can be useful:
- for monitoring – allow to control state of the sensors by emulated functionality;
- for optimization - indicate appropriate solutions (e.g. load distribution), taking into
account many factors (weather, route);
- emergency and rescue (support to cope with deviating scenarios).
For quality, maintaining the “digital twin” in the up to date state is important (e.g. for cargo
calculators, the current model of solid masses or displacement calibration). Where necessary, smart
technologies should support diagnostic modes and be self-adaptive. It implies, in particular, the HIL
(Hardware In the Loop) and SIL (Software in the Loop) solutions, when we modify model state based
also on machine learning from sensors and other available sources.
New technologies have important influence on maintenance procedures and service
inspections changes. The industry evolved from the "corrective maintenance" approach (the simplest
but the most expensive strategy of ad hoc interventions after a failure) to active supervision methods
based on data analysis and device operation history, called preventive maintenance. However, they
also usually use hardset surveying standards. The most advanced solutions are called predictive
maintenance methodology (e.g. [17]) and are using data fusion techniques based on information
collected on-line from dedicated sensors and smart modules, to compute permissible operating
regimes. Stability systems, strength controls systems and other are used in this way.
62
8
Digitalization of shipping is bringing many changes for services performed by class societies. Also
introduction of maritime broadband communication solutions enables allows on-line connection of
ship systems to world internet infrastructure. This opens a gate to ship and ships systems remote
monitoring (shore control centers) and remote inspections (remote services and surveys) - a new
world, where classification society must also change its traditional way of operation and services, to
be able to support safe navigation and shipping activities as requested by the whole maritime and
shipping community.
References:
[1] „Wikipedia ENG: Digital Twin description,” 2021.
[2] „Wikipedia ENG: Model-based definition(MBD) , digital product definition (DPD),” 2021.
[3] „PRS: Emergency Response Center (ERC) Service https://www.prs.pl/offer-16/ship-surveys-1/offer-
2/emergency-response-center-erc-3,” 2016.
[4] „NYK expands IoS-OP connected vessel fleet to 200,” 2021.
[5] „ABS: Smart functions for marine vessels and offshore units,” 2019.
[6] „Wikipedia ENG: Turing Test description,” 2021.
[7] „BV: NR632 Hardware-in-the-Loop Testing,” 2016.
[8] „DNVGL: HIL Testing Concept Explanation https://www.dnvgl.no/services/hil-testing-concept-
explanation--83385”.
[9] „Wikipedia ENG: Co-simulation,” 2021.
[10] „IMO RESOLUTION MSC.267(85) ADOPTION OF THE INTERNATIONAL CODE ON INTACT
STABILITY, 2008,” 2008.
[11] „IACS: UR L5 - Onboard Computers for Stability Calculations, Requirements concerning LOAD LINE ,
2006,” 2006.
[12] PRS, „PRS Publikacja nr 16P Środki kontroli obciążenia statku, , Gdańsk, rev 2011.,” Gdańsk, 2011.
[13] „PRS: Publikacja nr 6/P Zastosowanie na statkach programów komputerowych do obliczeń stateczności,
Gdańsk, 2005,” PRS, Gdańsk, 2005.
[14] „Approach to estimate the ship center of gravity based on accelerations and angular velocities without ship
parameters,” 2019.
[15] „IEEE Estimating the Longitudinal Centre of Flotation of a Vessel in Waves using Acceleration
Measurements , Nana O. Abankwa, James Bowker, Steven J. Johnston, Mark Scott, and Simon J. Cox,”
2018.
[16] „Global Maritime Engineering: HULL STRESS MONITORING SYSTEM (catalog)”.
[17] "Condition Based Maintenance Plus DoD Guidebook," 2008.
[18] „Investigations of Marine Safety Improvements by Structural Health Monitoring Systems Transnav,”
2012.
[19] „A State-of-the-Art Survey of Digital Twin: Techniques, Engineering Product Lifecycle Management and
Business Innovation Perspectives, Lim Yan Hong Kendrik, Pai Zheng, and Chu,” 2020.
[20] „IMO:Guidelines for the approval of stability instruments, MSC.1/Circ.1229, London 2006.,” 2006.
[21] „IMO: Guidelines for shipboard loading and stability computer programs, MSC/Circ.854, London 1998,”
1998.
[22] „IMO: Guidelines for the on-board use and application of computers, MSC/Circ.891, London 1998.,”
1998.
[23] IACS, „IACS: Recommendations on Loading Instruments, Rec. No.48, 1997],” 1997.
[24] IMO, „IMO: Recommendation on loading instruments, Resolution 5 of the 1997 SOLAS Conference,
London 1997,” 1997.
[25] SSC, „SSC-468 Development of Structural Health Monitoring Prototype for ship structures,” 2013.
[26] „Compit 2019: A Hitchhiker’s Guide to the Galaxy of Maritime e-Learning, V.Bertram,T.Plowman,”
2019.
[27] „COMPIT 2018: Digital Model or Digital Twin ; Cabos , Rostock,” Rostock, 2018.
9 63
[28] „IACS: UR S1 Requirements for Loading Conditions, Loading Manuals and Loading Instruments,
Resolution, 2010,” 2010.
[29] „JSMEA: Smart Ship Application Platform,” 2018.
[31] „Estimating the Longitudinal Center of Flotation of a Vessel in Waves Using Acceleration
Measurements,” 2018.
[32] „Open Simulation Platform - opensimulationplatform.com,” 2017.
64 10
Quasi-real verification of autonomous navigation.
Zbigniew Pietrzykowski 1*, Piotr Wolejsza 1
1 Faculty of Computer Science and Telecommunication, Maritime University of Szczecin, Szczecin 70-500, Poland;
[email protected], [email protected];
* Correspondence: [email protected]; Tel.: +48-505-726-464 (Z.P.)
Keywords
anti-collision trajectories; navigational decision support system; autonomous ship; quasi-real condition;
Abstract
Works to build autonomous ships have been undertaken in many countries and for various reasons. The most
important are safe and efficient navigation.
The article presents the results of the implementation of autonomous navigation in quasi-real conditions, i.e. at
the Ship Handling Research and Training Center in Iława, Poland.
To determine a safe trajectory in collision situations several factors are taken into account: available water area,
navigational obstacles, ships’ parameters (size, speed), ships’ manoeuvrability under different
hydrometeorological conditions (wind direction, wind speed, current direction, current speed). The optimization
task is to pass safely all objects along a shortest possible trajectory. For this purpose, algorithms and methods of
information acquisition and integration, situation assessment, decision making, decision execution are
consistently developed and improved. The main goal of our research carried out in quasi-real conditions was to
identify potential gaps in the algorithms developed and tested in laboratory conditions before tests in real
conditions at sea.
The authors have prepared over 80 different scenarios, each involving four ships. Two of them were autonomous,
while the other two were manned. The goal of autonomous ships was to pass from point A to B and avoid
collisions with all targets, including the coastline, while the manned ships were trying to disturb and obstruct the
movement of autonomous ships. The results of the verification in quasi-real conditions were presented in this
paper.
1
65
22nd International Symposium On Hydrodynamics In Ship Design, Safety,

Manoeuvring and
Advanced Operation and experimental modeling
computational
HYDRONAV 2021
HYDRONAV 2021
Mesh dependence study for numerical assessment of hydrodynamic
characteristics of windsurfing fin
Mesh
H. dependence
PRUSZKO1 study for numerical assessment of hydrodynamic
characteristics of windsurfing fin
1 Gdańsk University of Technology, Faculty of Ocean Engineering and Ship Technology, Narutowicza 11/12 80-233 Gdańsk,
Poland, e-mail: [email protected]

H. PRUSZKO 1
1Gdańsk University of Technology, Faculty of Ocean Engineering and Ship Technology, Narutowicza 11/12 80-233 Gdańsk,
Keywords:
Poland, e-mail: [email protected]
1. Introduction
Keywords:
The development of hi-tech materials allowed for introducing into the real life ideas that had been
evaluated many years ago, but had been limited by the practical issues in that moment. A very good
1. Introduction
example for it is the use of composite materials in high-performance sailing for manufacturing of
hydrofoils.
The developmentHowever, it brought
of hi-tech also novel
materials allowed scientific questions,into
for introducing as the
understanding
real life ideas of the
thatcomposite
had been
materials is
evaluated manya challenge.
years ago,Good but had example confirming
been limited by thethis statement
practical issuesareinobserved
that moment. differences
A veryingood the
performance
example for itof ismonotype
the use of fins for Olympic
composite (untilin2021)
materials windsurfing RS:X
high-performance sailingclass. For strong wind
for manufacturing of
conditions the fin is the only control surface responsible for generation
hydrofoils. However, it brought also novel scientific questions, as understanding of the compositeof the side force that allows to
keep the course of board. This side force balances the sail aerodynamic
materials is a challenge. Good example confirming this statement are observed differences in the force. It has already been
proven that use
performance of of laminate fins
monotype allows fortoOlympic
tailor the(until
performance of the liftingRS:X
2021) windsurfing surface to the
class. Forrequired level
strong wind
(Marimon Giovannetti et al., 2018), however, there are still some uncertainties
conditions the fin is the only control surface responsible for generation of the side force that allows to and unknowns.
Utilization
keep of composite
the course of board. material
This side for manufacturing
force balances of thefins
sailmakes the construction
aerodynamic force. It more flexiblebeen
has already and
prone
proventothat
deformation.
use of laminateSeveral studies
allows to have
tailorbeen carried out toofinvestigate
the performance the lifting into the to
surface relation betweenlevel
the required the
lay-up of composite hydrofoil and the hydrodynamic performance
(Marimon Giovannetti et al., 2018), however, there are still some uncertainties and unknowns. (Liao et al., 2019; Zarruk et al.,
2014; Marimon Giovannetti et al., 2016). It shows that the dependence
Utilization of composite material for manufacturing of fins makes the construction more flexible and of the hydrodynamic
performance on the internal
prone to deformation. Several structure
studies ishaveverybeen
strong.
carried out to investigate into the relation between the
This article aims to present the results
lay-up of composite hydrofoil and the hydrodynamic of numerical investigation
performance into(Liao
the hydro
et al.,mechanical
2019; Zarruk properties
et al.,
of RS:X
2014; fins andGiovannetti
Marimon there are two et practical
al., 2016).reasons why the
It shows accurate
that assessmentofof the
the dependence the hydrodynamic
hydrodynamic
load is crucial.
performance on First of themstructure
the internal is occurrence
is very of the phenomena called ‘spin out’. From the windsurfer
strong.
point of view spin out is accompanied
This article aims to present the results of numerical with the feeling of suddeninto
investigation lossthe
of hydro
coursemechanical
stability. Itproperties
is caused
by reaching the critical angle of attack, separation and stall (Gourlay &
of RS:X fins and there are two practical reasons why the accurate assessment of the hydrodynamic Martellotta, 2011).
Second of all, itFirst
load is crucial. wasofobserved
them is that the windsurfing
occurrence fins are greatly
of the phenomena deformed
called ‘spin out’. during
From the thewindsurfer
sailing in
heavy
point of view spin out is accompanied with the feeling of sudden loss of course stability. It from
wind conditions. Due to the strong dependence of the composite material the
is caused
manufacturing
by reaching the conditions
critical angle actual deformations
of attack, separation forand
different fins might
stall (Gourlay significantly
& Martellotta, vary from each
2011).
other.
SecondAccording
of all, it to was(Young
observedet al.,that
2018)
the for compositefins
windsurfing hydrofoils the strong
are greatly bend-twist
deformed duringcoupling could
the sailing in
be observed. It is expected that the differing material properties
heavy wind conditions. Due to the strong dependence of the composite material from the of single fin would result in varying
strains – bend deformations
manufacturing conditions actual and that are accompanied
deformations with fins
for different localmight
changes of angle of
significantly attack.
vary fromIteach
is a
motivation for carrying out the FSI analysis. That is why proper assessment
other. According to (Young et al., 2018) for composite hydrofoils the strong bend-twist coupling could of the hydrodynamic
pressure loadsIt isis of
be observed. a great that
expected interest. The numerical
the differing materialmodel verified
properties and validated
of single fin would willresult
be used in the
in varying
future
strainsfor the Fluid
– bend Structure Interaction
deformations and that arecalculations
accompanied of RS:X fin. changes of angle of attack. It is a
with local
motivation for carrying out the FSI analysis. That is why proper assessment of the hydrodynamic
pressure loads is of a great interest. The numerical model verified and validated will be used in the
future for the Fluid Structure Interaction calculations of RS:X fin.
66
At this stage of investigation the main focus was put into the study with the use of numerical methods.
In the linear zone of lift curve the accuracy of Computational Fluid Dynamics tools is well
documented and level of trust is high. However, accurate prediction of stall angle and forces that are
generated for the critical angle of attack still remains a challenge. Therefore, the part that requires
more attention was the assessment of the angle of attack for which the stall occurs. That is why the
research will include the numerical study of mesh dependence to assess the relation between the grid
resolution and critical angle of incidence value. At the same time two different turbulence models was
compared.
2. Aim and Scope

Aim of presented research was the numerical assessment of the drag coefficient and lift coefficient
versus angle of attack curves for windsurfing fin. The special attention in the research was be paid to
the evaluation of the stall angle value. The computations were performed for meshes with various
resolution and two turbulence models
The scope of the research included the validation and verification of numerical model. In particular
the study covered the generation and running the numerical simulation of two models of fins by means
of unsteady RANS solver. Simulations of RS:X fin were performed for the range of angle of attack
from 0 to 12 degrees and speed of 10 m/s. The mesh resolution sensitivity analysis was done by
systematically decreasing the value of y+. The influence of the mesh resolution and turbulence model
on the value of stall angle as well as curves of the drag and lift was checked. The validation study was
carried out in order to verify the numerical results with the experimental findings. For the validation
study the model of trapezoidal hydrofoil with NACA 0009 was generated. Numerical results were
compared with the experimental findings presented by Zarruk et al., (2014).
3. Numerical model
The first part of the research included creating computer files with fin 3D geometry. The model for
validation study was generated based on the description in (Zarruk et al., 2014). The RS:X fin model
was reconstructed by means of reverse engineering tools. The laser scanning head was used to create
the cloud of points. It was later transferred into final geometry of the fin that was used for
computations. Both models were presented in the Figure 1.
Fig. 1. 3D CAD models of windsurfing fin (on the left) and NACA 0009 (on the right)
The numerical simulations were performed by means unsteady RANS solver. Flow around the fin was
modeled as a single-phase, turbulent, viscous and incompressible. Two turbulence models were used
for the computations: k-ε model and k-ω SST. It allowed to investigate the influence of the turbulence
model on the value of the stall angle.
According to the ITTC guidelines (ITTC, 2011) if the significant lift force is generated by the lifting
surface then the upstream boundary plane should be placed at least 10 chord lengths in front of the
67
body and downstream boundary 20 chord lengths behind it. This rule was used for the simulations, so
the length of the domain was equal to 6 meters. Side boundaries were placed 1 meter from the fin. The
top boundary was overlapping with the head of the fin and the bottom boundary was placed 1 meter
below the tip. For the upstream and bottom boundaries the velocity inlet condition was selected. The
downstream boundary had assigned the pressure outlet condition. The symmetry condition was
assigned to the top and side boundaries.
The implicit unsteady solver was used. The time step was selected so that the Courant number was
kept below 5 for majority of the cells around the fin. Resulting time step was equal to 0.005 s
The mesh resolution had to be high enough to accurately represent the geometry, especially leading
and trailing edges as well as capture all the relevant features of the flow. On the other hand it should
not be too expensive computationally. Mesh refinement were applied on the tip and nose of the fin,
close proximity of the model and in the wake region including the expected area of tip vortex
generation. The mesh coarsening was applied on the outside domain boundaries, where high number
of cells is unnecessary
The mesh sensitivity analysis was performed varying the value of y+, that is defined as:
(1)
In this formula u* stands for the friction velocity, y is the distance to the wall and υ is the kinematic
viscosity. For obtaining the desired y+ value the y need to be estimated. The STAR CCM+ software
was used for computation. Defining the mesh topology with the use of so-called prism layers allows
for accurate resolving of the boundary layer and saving of computational resources. It is realized by
applying the cells whose thickness is significantly smaller than remaining dimensions in the closest
proximity to the body. For defining the mesh resolution with prism layers their thickness, number of
layers and thickness growth ratio (stretch factor) has to be specified. The thickness of the first layer
can be specified using the formula for the first term of geometric series:
(2)
Where:
δ – thickness of a boundary layer
r- common ratio – stretch factor
n – number of terms – number of prism layers
Length- dependent thickness of the boundary layer was estimated according to the formula:
(3)
Where:
x- position in x-direction ( length of the body)
Re X – Reynolds number
The Reynolds number for mean chord of the fin was equal to 1.12*106, whereas for the validation
study it was equal to 0.6*106.
It can be noticed then that there are three ways for controlling the y+ in numerical simulations: reduce
the thickness of the prism layers, increase the number of prism layers or increase the stretch factor.
It was observed that too high stretch factor might cause the simulation divergence. What is more the
ITTC recommends the 1.2 as the value of the stretch factor. It was decided that the parameter that will
be controlling the value of y+ would be the number of prism layers.
Three different resolutions meshes were created and they were numbered from 1 to 3 with increasing
number of prism layers. For M1 the number of prism layers was equal to 8, the target value of y+ was
above 30 and viscous sublayer was not resolved directly,. For the remaining two meshes the number of
prism layers was equal to 14 and 20 for mesh M2, M3 respectively. What is more due to convergence
issues the M3 had the stretch factor equal to 1.25, whereas remaining meshes 1.3
The surface averaged y+ values were equal to 40, 3.8 and 1.8 for M1, M2 and M3 respectively and
final number of cells was between 5.4 million cells and 7.4 million cells.
68
4. Results
The results of simulations were analyzed and presented in form of the plots – the curves of lift and
drag coefficient against the angle of attack in the Figure 2.
Fig. 2. Comparison of lift and drag coefficient
First of all, it could be noticed that in the linear region both turbulence models and all meshes provides
with similar results and significant differences starts to occur above 6 degrees angle of incidence.
The k-ω SST for M3 predicts the value of stall angle equal to 7 degrees, whereas for M2 and M1 this
is equal to 8 degrees and 9 degrees respectively. It could be noticed that for this turbulence model above
the critical angle of attack increasing resolution results in systematic increase in drag and decrease of
lift at the same time.
For k-ε turbulence model for the M1 the stall angle could not be observed at all, whereas for M2 and
M3 it was equal to 8 degrees. Moreover, in linear region the solution seems to be almost y+
independent. Above the stall angle M2 and M3 predicts the values of drag and lift coefficient with
very similar accuracy.
5. Validation study
The validation study was performed for the object similar to the RS:X windsurfing fin. The
requirements for the validation object was that it has to be the tapered airfoil or hydrofoil with
symmetrical profile and similar aspect ratio. Additional requirements included the experimental
conditions – the hydrofoil had to be tested at high Reynolds number and it needed to be placed inside
the measurement zone in the way that only the top edge of the object is attached to the channel wall,
so that the formation of tip vortices and were possible. The experiments described by the Zarruk et al.,
(2014) fulfils all the requirements above.
The numerical mesh used for the validation study was similar to the one described in the Section 3
with resolution of prism layers identical as M2. The computations were performed for the range of
angle of attack from 9 degrees to 12 degrees, which corresponds to close to stall conditions. They are
considered to be the most demanding and numerically difficult to be accurately resolved. In the
Figure 3 was presented the plots of drag and lift coefficient versus the angle of attack for M2 for two
turbulence models compared with experimental results.
69
Fig. 3. The validation study results
It could be noticed that for the drag coefficient curves (black lines) the k-ω SST model has the same
trend as it was observed for experimental data. The difference in results varied from 15% to -2.7% and
was the highest for 10.5 degrees angle of incidence. For lift coefficient (red lines) the k-ω SST model
over predicts the CL below the stall and under predicts for higher angles of incidents. The difference
was between 5.9% up to even -23% . Nevertheless, the value of stall angle was captured correctly. For
the k-ε turbulence model the stall angle was not captured at all, both the drag and lift coefficient raised
almost linearly. It resulted in differences up to -44.86% in drag coefficient and 29% in lift coefficient.
From the validation study it was concluded that the k-ω SST model was much more accurate in
predicting the hydrodynamic characteristics of NACA 0009 hydrofoil.
6. Discussion
It is clear that the mesh resolution in the region of boundary layer has a major influence on the
hydrodynamic characteristics of the hydrofoil. Based on the Figure 2 it could be easily concluded that
non-direct resolving of the viscous sublayer prevents from accurate prediction of the stall angle. As a
consequence the stall angle is delayed compared to high fidelity simulations. At the same time the lift
is greatly over estimated and drag is under estimated. For better understanding of those differences the
turbulent kinematic energy fields were compared for all three meshes and both turbulence models. The
results were presented in the Figure 4. The upper row corresponds to the k-ω SST model and lower
row to k- ε with increasing mesh resolution from left to right (M1, M2 and M3 respectively).
Fig. 4. Turbulence kinetic energy field for k-ω SST model (top) and k- ε model (bottom) and y+ = 40 (left), y+ = 3.8 (middle)
and y+ = 1.8 ( right)
70
The color scale was the same for all figures and red color corresponds to higher values of turbulence
kinetic energy. It could be easily noticed that the turbulence intensity is the highest for the finest mesh
for k-ω SST model. It could explain the differences in values of lift and drag for this angle of
incidence. The finest mesh and k-ω SST model allows for accurate capturing the changes in the
features of the flow resulting from increasing angle of incidence, separation and finally a stall. What is
more, comparing the results of mesh sensitivity study in particular the differences between M2 and M3
for k- ω SST model it seems that the best agreement with the experimental results would be reached
for M3. The M3 model predicted lower lift in around stall, higher lift past stall and higher drag than
M2. If the same relations would occur for NACA 0009 it would corresponds better to the experimental
results.
7. Conclusions
This paper presents the results of mesh sensitivity study of influence the y+ value on drag and lift
coefficients. The numerical simulations were performed for the fin of the Olympic windsurfing class
RS:X. The calculations were done for the angle of attack for 0 to 12 degrees and speed of 10 m/s.
Three meshes and two turbulence models were compared in simulations. Additionally, the validation
study was performed. The calculations were carried out for hydrofoil with NACA 0009 profile and the
results of computations were compared with experimental data presented by Zarruk et al., (2014).
Based on the obtained results following conclusions were formulated. The k-ω SST model in general
predicted the stall angle with much greater accuracy. What is more it shows higher sensitivity of
values of the y+ on the numerical results. However, it is computationally more expensive. On the top
of that is the fact that in the linear region the k-ε performs equally good. It was observed that for the
high y+ meshes both turbulence models failed to predict stall angle significantly delaying it. Therefore
in the linear region for low angles if attack the k-e model with high y+ could be sensible choice.
Nevertheless, for higher angles of attack the k-w SST model with low y+ value seems to be the only
acceptable solution. Therefore in the future for the close to stall conditions the M3 with k-ω SST
model will be used for calculations.
This study will be further continued by performing the FSI simulations to obtain the wide
understanding of the relation between mechanical and hydro mechanical properties of composite RS:X
fins. The study presented in this article allowed to assess the suitable CFD model for further
computations.
8. References
[1] Gourlay, T., & Martellotta, J. (2011). Aero-Hydrodynamics of an RS: X Olympic Racing Sailboard.
Cmst.Curtin.Edu.Au.
[2] ITTC. (2011). Practical Guidelines for Ship CFD Applications - 7.5-03-02-03. ITTC – Recommended
Procedures and Guidelines, 1–18.
[3] Liao, Y., Garg, N., Martins, J. R. R. A., & Young, Y. L. (2019). Viscous fluid–structure interaction
response of composite hydrofoils. Composite Structures, 212, 571–585.
https://doi.org/10.1016/j.compstruct.2019.01.043
[4] Marimon Giovannetti, L., Banks, J., Boyd, S. W., & Turnock, S. R. (2016). Developing tools for assessing
the fluid structure interaction of passive adaptive composite foils. Insights and Innovations in Structural
Engineering, Mechanics and Computation - Proceedings of the 6th International Conference on Structural
Engineering, Mechanics and Computation, SEMC 2016, 586–591.
https://doi.org/10.1201/9781315641645-97
[5] Marimon Giovannetti, L., Banks, J., Ledri, M., Turnock, S. R., & Boyd, S. W. (2018). Toward the
development of a hydrofoil tailored to passively reduce its lift response to fluid load. Ocean Engineering,
167, 1–10. https://doi.org/10.1016/j.oceaneng.2018.08.018
[6] Young, Y. L., Garg, N., Brandner, P. A., Pearce, B. W., Butler, D., Clarke, D., & Phillips, A. W. (2018).
Load-dependent bend-twist coupling effects on the steady-state hydroelastic response of composite
hydrofoils. Composite Structures, 189, 398–418. https://doi.org/10.1016/j.compstruct.2017.09.112
[7] Zarruk, G. A., Brandner, P. A., Pearce, B. W., & Phillips, A. W. (2014). Experimental study of the steady
fluid-structure interaction of flexible hydrofoils. Journal of Fluids and Structures.
https://doi.org/10.1016/j.jfluidstructs.2014.09.009
71
Ship
seakeeping
72
Ship seakeeping

HYDRONAV 2020
Simulation of fishing vessel motion for seakeeping

HYDRONAV 2020 analysis
M. WARMOWSKA-GADZIŃSKA AND A. LASKOWSKI , H. PUROWSKI , M. MALINGA(11
1 1 1
POINT TIMES NEW ROMAN FONT IN 1.0 LINE SPACING, UPPERCASE LETTERS)
1 Polski Rejestr Statków S.A., [email protected]
Simulation of fishing vessel motion for seakeeping analysis
Within International Maritime Organization1IMO,
M. WARMOWSKA-GADZIŃSKA ANDforA.many years, work have
LASKOWSKI 1
, H.been conducted to1,prepare
PUROWSKI rules governing the
M. MALINGA(11
stability requirements for ships, considering particularly dangerous phenomena. The IMO expert group has developed
POINT TIMES NEW ROMAN FONT IN 1.0 LINE SPACING, UPPERCASE LETTERS)
recommendations concerning direct assessment of ship stability, guidance on operational limitations, as well as the second
generation intact stability criteria.
1 Polski Rejestr Statków S.A., [email protected]
Polski Rejestr Statków (PRS) is ready to implement the guidelines proposed as the future IMO documents in the process of
ensuring navigation safety. Prepared and verified tools for the simulation of ship’s behaviour on regular or irregular waves
provide the possibility
Within International to perform
Maritime calculations
Organization IMO,and
forobservations,
many years, among the been
work have others, for phenomena
conducted such
to prepare as governing
rules surf-ridingthe
or
broaching.
stability requirements for ships, considering particularly dangerous phenomena. The IMO expert group has developed
recommendations concerning direct assessment of ship stability, guidance on operational limitations, as well as the second
The paper presents examples of numerical analyses defining the safe operation conditions for fishing vessels.
generation intact stability criteria.
Keywords:
Polski simulation
Rejestr of ship
Statków motion,
(PRS) ship safety,
is ready seakeeping,
to implement theship stability requirements
guidelines proposed as the future IMO documents in the process of
ensuring navigation safety. Prepared and verified tools for the simulation of ship’s behaviour on regular or irregular waves
1. Introduction
provide the possibility to perform calculations and observations, among the others, for phenomena such as surf-riding or
broaching.
TheThe
paperissue of examples
presents the stability of fishing
of numerical boats
analyses has been
defining studied
the safe for conditions
operation many years by a specialized
for fishing vessels. working
group of IMO: Sub-Committee on Stability and Load Lines and on Fishing Vessels Safety (SLF) who
Keywords:
has simulation
developed a of ship motion,
number ship safety, seakeeping,
of documents on the ship stability
safety requirements
of fishing vessels. The information on the intact
stability of the ship in various hazardous situations was published by SLF in [1]. This document is a
1. Introduction
handbook for skippers that indicates the phenomena that may lead to stability loss [2].
IMO Sub-Committee on Ship Design and Construction (SDC) was recently focused on preparing
The issue of the stability of fishing boats has been studied for many years by a specialized working
a guideline on the vulnerability criteria for the second generation of intact stability criteria. As a
group of IMO: Sub-Committee on Stability and Load Lines and on Fishing Vessels Safety (SLF) who
result of this inter area work five failure modes have been indicated for seakeeping analysis [3]:
has developed a number of documents on the safety of fishing vessels. The information on the intact
1. dead ship condition,
stability of the ship in various hazardous situations was published by SLF in [1]. This document is a
2. excessive acceleration,
handbook for skippers that indicates the phenomena that may lead to stability loss [2].
3. pure loss of stability,
IMO Sub-Committee on Ship Design and Construction (SDC) was recently focused on preparing
4. parametric rolling,
a guideline on the vulnerability criteria for the second generation of intact stability criteria. As a
5. surf-riding/broaching.
result of this inter area work five failure modes have been indicated for seakeeping analysis [3]:
IMO
1. encourages application of the new guideline by shipbuilders, shipmasters, shipowners, ship
dead ship condition,
operators
2. and shipping companies and reporting their experience gained through the trial use of the
excessive acceleration,
guidelines
3. to aloss
pure database. After appropriate analyses the collected data can be implemented in the
of stability,
development
4. of the new
parametric requirements. PRS is interested in adopting the new guidelines in its in-house
rolling,
programs
5. developed for ship safety verification.
surf-riding/broaching.
IMO encourages application of the new guideline by shipbuilders, shipmasters, shipowners, ship
operators and shipping companies and reporting their experience gained through the trial use of the
guidelines to a database. After appropriate analyses the collected data can be implemented in the
development of the new requirements. PRS is interested in adopting the new guidelines in its in-house
programs developed for ship safety verification.
73
Ship seakeeping
2. In-house programs
The task of PRS, as a classification institution, is to confirm compliance with the technical
requirements authorizing the vessel to navigate in specific regions, on the basis of the technical
documentation of the supervised units. Moreover, by employing in-house developed software
specifically tailored for the strict safety requirements, it can perform additional calculations and
simulations to confirm that the vessels meet the stability conditions.
The development of numerical methods and adequately efficient computers makes it possible to
simulate numerical tests of the ship on wave, that are similar to the model tests which the designed
ships are subjected to. As a result of the extensive scientific research, more and more accurate physical
and numerical models describing reality are being developed. Conducting numerical simulations is
one of the methods to predict dangerous situations of the stability loss. In the past, fishing boats were
built based on the builders’ experience. Currently, it is possible to perform many numerical tests for
the wave occurring in the shipping region that take into account the area of the ship's operation, as
well as, the ship’s structure and loading condition. Thanks to these calculations wrong design solutions
can be avoided and the safety is increased.
2.1 WinSea
PRS stability specialists work with the proprietary WinSea program, which is designed to
determine the ship's stability parameters. On the basis of the obtained results, an independent PRS
opinion is issued on the stability safety of the classified vessel. Based on the description of the ship
shape, holds, compartments and selected loading conditions, the program verifies the value of the
chosen stability parameters with the rule required values.
Fig. 1 presents the results of stability calculation obtained with the use of WinSea software. For the
given condition the ship does not meet the required stability parameters such as the angle for GZmax
arm and the range of the righting arms.
Fig. 1 Verification results by WinSea software (in polish)
74
Ship seakeeping
2.2 NakresSztormowy
A software NakresSztormowy, developed on the basis of the MSC.1 / Circ.1228 [1], is designed
to define dangerous navigation situations for the ship,see Fig. 2. This program can be a helpful tool for
ship captains, as it indicates the risks associated with the loss of ship stability and suggest the
possibility of changing speed or ship’s course in order to avoid it. The program takes into
consideration phenomena such as risk of surf-riding, reduction of intact stability and successive high-
wave attack. Before preparing simulation of ship’s behaviour the software allows to predict some risk
situation for the vessel.
Fig. 2 View of the NakresSztormowy application.
2.3 Software modelling ship’s motion

For many years, PRS has been working on the development of numerical methods describing
phenomena such as wave motion or motion of a ship on a wave [4], [5], [6]. A program modelling the
vessel's behaviour on an regular/irregular wave has been developed. The simulation results were
implemented to conduct research on the safety of bulk carriers, container ships and fishing vessels [7].
The wave motion, in the simulation programs developed in PRS, are based on the equations of
motion for an irregular wave, describing the ship's motion for six degrees of freedom[4]. It is assumed
that the main hydrodynamic forces acting on the vessel can be split into Froude-Krylov forces,
diffraction and radiation forces, other forces including those induced by water on deck, rudder forces
and nonlinear damping. In the case of Froud-Krylov forces, the deformation of the free surface is
75
Ship seakeeping
taken into account, which is important in the case of fishing vessels, when the draught of the ship is
comparable with the wave height [8].
Moreover, PRS has developed methods of describing the behaviour of water on board taking into
account phenomena such as the deck in the water, the water flowing over the sides on the deck, and
the water flowing through the openings in the side [9].
The hydrodynamic forces and moments defining the equations are determined in each time step.
The accuracy of the simulation depends on the accuracy of calculation of the hydrodynamic forces and
moments induced by the waves.
The analyses of the ship's behaviour in a regular wave were carried out. Table 1 shows examples of
ship and wave motion parameters as well as the obtained information about the threats to stability. The
following calculation parameters were adopted in the simulation: length of ship L equals to 23 m,
draught of ship d is 2,7 m, natural period of roll of ship Tr is equal to 3,961 s.
The sample sea conditions with types of risk are given in the table 1.
Table 1. Parameters of ship simulations
u Tw A ω λ c λ/L Te Risk A Risk B Risk C
[knots] [s] [m] [rad/s] [m] [knots] [s]
9 9 2 0.70 126.47 27.34 5.50 13.77 Yes No Yes
9 6 2 1,05 56,21 18,23 2,44 12,49 Yes No Yes
9 5 2 1,26 39,03 15,19 1,70 13,29 Yes Yes Yes
9 4 2 1,57 24,98 12,15 1,09 18,15 Yes Yes Yes
8 4 2 1,57 24,98 12,15 1,09 13,03 No Yes Yes
7 4 2 1,57 24,98 12,15 1,09 10,16 No Yes Yes
6 4 2 1,57 24,98 12,15 1,09 8,33 No Yes Yes
10 3 1.2 2,09 14,05 9,11 0,61 -19,35 Yes Yes No
9 3 1.2 2,09 14,05 9,11 0,61 -75,93 Yes Yes No
8 3 1.2 2,09 14,05 9,11 0,61 39,48 No Yes No
7 3 1.2 2,09 14,05 9,11 0,61 15,67 No Yes No
where: u – ship speed; Tw – wave period; A – amplitude of wave; ω – frequency of wave; λ– length
of wave; c – phase velocity of wave; L – ship length; Te – wave encounter period;
Risk A – Risk of surf-riding in following sea;
Risk B – Risk of intact stability reduction;
Risk C – Risk of successive high wave attack in following sea.
2.3.1 Case 1
The calculations were made for two cases of regular wave and following sea.The regular wave for
following sea was assumed during simulation, see Fig. 3.
Fig. 3 Wave amplitude of the harmonic wave during simulation – Tw = 9 s, A = 2 m, u = 9 knots.
76
Ship seakeeping
The equations of motion were solved over time. In the initial phase of the simulation, as presented
in Fig. 4, it was possible to observe how the system aims to achieve repeatable stabilization.
Fig. 4 Components of ship movement at the beginning of simulation – Tw = 9 s, A = 2 m, u = 9 knots.
The repeatable stabilization is shown in Fig. 5.
Fig. 5 Components of ship motion during simulation – Tw = 9s, A = 2m, u = 9knots.
The simulation of vessel’s motions in waves is based on numerical solutions of non-linear

equations of motion (non-linear model), hence the ship's response to harmonic inputs, Fig. 3, is not
always a harmonic function, Fig. 6.
Fig. 6 Component F3 of Froud–Krylov forces acting on ship,Tw = 9 s, A = 2 m, u = 9 knots.
77
Ship seakeeping
During the simulation, the following parameters were observed:

ship motion, force components acting on it, see Fig. 5, Fig. 6,
the position of the stern in relation to the undisturbed surface, see Fig. 9, as well as
the change of the wetted surface area, see Fig. 7.
The draught of the ship T is equal to 2,7 m and it corresponds to an average wetted area equal to
199,6m2. The value 175,6 m2 of wetted area corresponds to 2,51 m of ship’s draught, and 205 m2
corresponds to T = 2,74 m.
Fig. 7 The area of wetted surface As changing during simulation – Tw = 9s, A = 2 m, u = 9 knots.
The Fig. 8 shows the vessel's behaviour on the following wave in the successive seconds of the
simulation. The inertial system related to the average progressive speed of the ship is marked in green.
The layout related to the ship is marked in blue, while the surface of the wave is marked in red.
t = 617,4 s – the wave crest t = 619,8 s t = 622,2 s
t = 624,6 s – the wave trough t = 627,0 s t = 629,4 s
Fig. 8 Simulation of the ship’s position – Tw = 9 s, A = 2 m, u= 9 knots.
2.3.2 Case 2
Another simulation was carried out for the wave parameters: Tw = 5 s, A = 2 m. The ship was
moving at a speed of u = 9 knots. Fig. 9 shows the change in time of the vertical position of the stern
point h_stern relative to the undisturbed level of sea, as well as of the vertical position of the free
surface of wave h_wave above/below this point. According to Table 1, the wave encounter period Te
is equal to 13.29 seconds. In this case, the ratio of the wave length λ to the ship length L equals to 1.7
(Table 1).
78
Ship seakeeping
Fig. 9 Vertical position of stern h_stern and vertical position of wave h_wave (above or below the stern), Tw =5s, A = 2 m,
u=9 knots.
The values of sway (Fig. 10) corresponds to the slope of the wave surface and they are twice
greater than in the previous simulation (Fig. 5). In the time t from 611 s to 618 s, the stern is above the
free surface of wave, then the stern submerges quite rapidly into the wave and the water freely flows
over the deck, changing the stability of the fishing vessel (Fig. 11). This dangerous situation can be
avoided by changing the speed or the course of the vessel.
Fig. 10 Components of ship motion during simulation – Tw = 5s, A = 2m, u = 9 knots.
t = 613,4 s t = 616,2 s – the wave crest t = 618,6 s – dangerous situation
t = 621 s – the wave trough t = 623,4 s t = 625,8 s
Fig. 11 Simulation of the ship’s position – Tw = 5 s, A = 2 m, u= 9 knots.
79
Ship seakeeping
3. Summary
In this paper three different computer programs developed in PRS were presented. They help PRS’
surveyors to verify stability requirements. The WinSea software gives PRS the opportunity to check
the compliance with the stability of the selected ship parameters: shape and loading conditions. The
software NakresSztormowy is a tool that shows wave parameters. The program indicates the wave
parameters for which there may be a risk of loss of the ship' stability, according to document [1]. The
ShipSimulation software enables efficient simulation of ship’s motion on the waves with focus on
parameters of motion, forces acting on ship.
The next step of improving ship’ safety is the development of II generation stability regulations,
based on the guidelines developed by the SDC Sub-Committee [3]. The development of a new
software that simulates the behaviour of the ship is necessary to complete these guidelines. PRS has
extensive experience in modelling hydrodynamic problems including wave motion and ship motion,
description of fluid motion including deformable free surface.
The guidelines developed by SDC will soon be implemented by IMO. PRS is preparing to
participate in the process of verification of these guidelines and their implementation in the
verification process.
4. References
[1] IMO MSC.1/Circ.1228 11 January 2007 Revised guidance to the master for avoiding dangerous
situations in adverse weather and sea conditions, 2007;
[2] Jurdziński M., Unikanie niebezpiecznych sytuacji w złych warunkach pogodowych w ruchu statku na
fali nadążającej, Akademia Morska w Gdyni;
[3] IMO Sub-Committee on Ship Design And Construction Finalization of second generation intact
stability criteria, October 2019;
[4] Pawłowski M., Liniowy model ruchów statku na fali nieregularnej, Raport Techniczny PRS nr 41,
Gdańsk 2001;
[5] Pawłowski M., Równania ruchu okrętu, Raport Techniczny PRS nr 47, Gdańsk, 2002;
[6] Jankowski J., Statek wobec działania fali, Raport Techniczny PRS nr 52, Gdańsk,2007;
[7] Jankowski J., Laskowski A., Capsizing of small vessel due to waves and water trapped on deck,
Proceedings of the 9th International Conference on Stability of Ships and Ocean Vehicle, 2006;
[8] Warmowska M., Modelling of water flow with free surface, V International Conference on
Computational Methods in Marine Engineering, Marine 2013, Hamburg, 2013;
[9] Warmowska M., Jankowski J., Problem of water flow on deck of small vessel, 18th International
Conference on Hydrodynamics in Ship Design, Safety and Operation -HYDRONAV, 2010.
80
Ship seakeeping

HYDRONAV
Manoeuvring 2020
and Operation
Sopot-Iława, 14-16 October 2020
Prediction of Ship Motions in Irregular Waves

Based on RAO Evaluated Experimentally in Noise Wave
HYDRONAV 2020
SEBASTIAN BIELICKI1
Sopot-Iława, 14-16 October 2020
1 Maritime Advanced Research Centre, CTO S.A., [email protected]
Prediction of Ship Motions in Irregular Waves

The most common methods for predicting ship roll motions in specified sea state are direct measurements of motions in
representative irregular wave realization (time domain) or calculations of motions from response amplitude operators (RAO)
in frequency domain. Based onofRAO
The result the firstEvaluated
method is then Experimentally
valid only for tested seain Noise
state, whilstWave
the second method is more
flexible in that matter, but less accurate. The RAO-based predictions are calculated assuming linear model of ship motions in
waves. RAO functions are usually
SEBASTIAN BIELICKI1 evaluated by means of tests in regular waves for limited number of frequencies and constant
wave amplitude. This approach is time-consuming, and discrete form of RAO function obtained for limited number of
frequencies may lead to discrepancies in correct predictions of ship's seakeeping and often does not allow determining actual
1 Maritime Advanced Research Centre, CTO S.A., [email protected]
amplitude of response in resonant frequency. Another challenging issue is the appropriate selection of wave amplitude for tests
due to considerable influence of viscous damping on roll response in irregular sea waves. There are alternative methods for
experimental determination of RAO function and one of them is presented in this paper. The presented approach allows
The most common
evaluating methods
RAO function for run
in one predicting ship roll
by generation of motions
irregularinwave
specified sea statebyare
characterized direct
white measurements
or colored of motions
noise spectrum. in
This
representative irregular wave realization (time domain) or calculations of motions from response amplitude
method reduces the experiment duration, and almost continuous RAO characteristic are obtained in result. The flat (white operators (RAO)
in frequency
noise) domain.
and linear The result
(colored noise)ofwave
the first method
spectral is then
energy valid only for
characteristic weretested sea state,
considered in whilst the secondand
the experiment, method is more
the obtained
flexible in that
predictions are matter,
compared butwith
less the
accurate.
resultsThe RAO-based
of accurate predictionsinare
measurements calculated
irregular assuming linear model of ship motions in
waves.
waves. RAO functions are usually evaluated by means of tests in regular waves for limited number of frequencies and constant
wave amplitude.
Keywords: This
seakeeping, rollapproach is time-consuming,
motions, noise, and discrete form of RAO function obtained for limited number of
ship motions, experiments
frequencies may lead to discrepancies in correct predictions of ship's seakeeping and often does not allow determining actual
amplitude of response in resonant frequency. Another challenging issue is the appropriate selection of wave amplitude for tests
due1.toIntroduction
considerable influence of viscous damping on roll response in irregular sea waves. There are alternative methods for
experimental determination of RAO function and one of them is presented in this paper. The presented approach allows
Roll motion
evaluating predictions
RAO function in one run play very important
by generation rolewave
of irregular in characterized
an assessment of ship
by white operability
or colored as wellThis
noise spectrum. as
method
influence significantly on ship safety during operations in severe seas. These motions areflatmostly
reduces the experiment duration, and almost continuous RAO characteristic are obtained in result. The (white
noise) and linear (colored noise) wave spectral energy characteristic were considered in the experiment, and the obtained
important during the station keeping vessels in offshore industry during operations in vicinity of
predictions are compared with the results of accurate measurements in irregular waves.
platforms in open seas.
Roll motion
Keywords: isroll
seakeeping, themotions,
responsenoise,motion of vessel
ship motions, exposed to the sea
experiments waves. The most appropriate method
for prediction of ship response in a sea way is to carry out an experiment in irregular waves representing
1. Introduction
considered sea conditions. However this method is usually limited to couple of representative sea states.
This approach is applicable if the operability is not the main goal of the analysis, what means the ship's
seakeeping characteristics
Roll motion predictionsare
playregarded as long role
very important term inand
an over long distances.
assessment For this purpose
of ship operability as wellthe
as
assumptionsignificantly
influence of linear response
on shipof safety
marineduring
systemoperations
to random in excitation was developed
severe seas. in the first
These motions half of
are mostly
twentieth century.
important during the station keeping vessels in offshore industry during operations in vicinity of
platforms in open seas.
Roll motion is the response motion of vessel exposed to the sea waves. The most appropriate method
for prediction of ship response in a sea way is to carry out an experiment in irregular waves representing
considered sea conditions. However this method is usually limited to couple of representative sea states.
This approach is applicable if the operability is not the main goal of the analysis, what means the ship's
seakeeping characteristics are regarded as long term and over long distances. For this purpose the
assumption of linear response of marine system to random excitation was developed in the first half of
twentieth century.
81
Ship seakeeping
2. Theoretical Background
Ship’s motions being the response to ocean waves are considered as the input/output system with
known linear characteristic (Fig. 1). This characteristic is called response amplitude operator (RAO) and
is the function of wave frequency [1]. The response motions are irregular as the sea waves are. Taking
known wave energy spectrum into consideration with known ship’s response frequency characteristic
(RAO) the response spectra can be calculated. With response spectra the statistical properties of this
response can be found.
Fig. 1 Linear relation between wave (input) and motions (response)
Basing on an ocean wave representation in spectral form, the wave energy in frequency domain is
[2]:
1
𝑆𝑆𝑆𝑆𝜁𝜁𝜁𝜁 (𝜔𝜔𝜔𝜔)𝑑𝑑𝑑𝑑𝜔𝜔𝜔𝜔 = 𝜁𝜁𝜁𝜁𝐴𝐴𝐴𝐴2 (𝜔𝜔𝜔𝜔) 2
(1)
And by analogy the energy spectrum of the response is:

1
𝑆𝑆𝑆𝑆𝜁𝜁𝜁𝜁𝜁𝜁𝜁𝜁 (𝜔𝜔𝜔𝜔)𝑑𝑑𝑑𝑑𝜔𝜔𝜔𝜔 = 𝜂𝜂𝜂𝜂𝐴𝐴𝐴𝐴2 (𝜔𝜔𝜔𝜔) (2)
2
Linear response characteristic called RAO is:

𝜁𝜁𝜁𝜁 (𝜔𝜔𝜔𝜔)
𝑌𝑌𝑌𝑌𝜁𝜁𝜁𝜁 (𝜔𝜔𝜔𝜔) = 𝐴𝐴𝐴𝐴 (3)
(𝜔𝜔𝜔𝜔) 𝜁𝜁𝜁𝜁𝐴𝐴𝐴𝐴
Rearranging formula (3) the response amplitude is:

𝜂𝜂𝜂𝜂𝐴𝐴𝐴𝐴 (𝜔𝜔𝜔𝜔) = 𝜁𝜁𝜁𝜁𝐴𝐴𝐴𝐴 (𝜔𝜔𝜔𝜔)𝑌𝑌𝑌𝑌𝜁𝜁𝜁𝜁 (𝜔𝜔𝜔𝜔) (4)
And taking formula (4) to (2) the response spectrum may be shown as:
1
𝑆𝑆𝑆𝑆𝜁𝜁𝜁𝜁𝜁𝜁𝜁𝜁 (𝜔𝜔𝜔𝜔)𝑑𝑑𝑑𝑑𝜔𝜔𝜔𝜔 = 𝜁𝜁𝜁𝜁𝐴𝐴𝐴𝐴2 (𝜔𝜔𝜔𝜔)𝑌𝑌𝑌𝑌𝜁𝜁𝜁𝜁2 (𝜔𝜔𝜔𝜔) (5)
2
Further taking into account formula (1) the response spectrum is:
𝑆𝑆𝑆𝑆𝜁𝜁𝜁𝜁𝜁𝜁𝜁𝜁 (𝜔𝜔𝜔𝜔)𝑑𝑑𝑑𝑑𝜔𝜔𝜔𝜔 = 𝑆𝑆𝑆𝑆𝜁𝜁𝜁𝜁 (𝜔𝜔𝜔𝜔)𝑑𝑑𝑑𝑑𝜔𝜔𝜔𝜔𝑌𝑌𝑌𝑌𝜁𝜁𝜁𝜁2 (𝜔𝜔𝜔𝜔) (6)
The wave as well as ship’s response are random processes with Gaussian distribution function around
zero mean [3], thus the root mean square value of signal is expressed by:
𝜂𝜂𝜂𝜂𝑅𝑅𝑅𝑅𝑅𝑅𝑅𝑅𝑅𝑅𝑅𝑅 = �𝑚𝑚𝑚𝑚0𝜁𝜁𝜁𝜁 (7)
Where the zeroth moment in formula (7) is:
∞
𝑚𝑚𝑚𝑚0𝜁𝜁𝜁𝜁 = ∫0 𝑆𝑆𝑆𝑆𝜁𝜁𝜁𝜁𝜁𝜁𝜁𝜁 (𝜔𝜔𝜔𝜔)𝑑𝑑𝑑𝑑𝜔𝜔𝜔𝜔 (8)
The above is valid as the response amplitude 𝜂𝜂𝜂𝜂𝐴𝐴𝐴𝐴 is proportional to wave amplitude 𝜁𝜁𝜁𝜁𝐴𝐴𝐴𝐴 .
82
Ship seakeeping
The linearized equation of response motion is [2]:

·· ·
(𝑀𝑀𝑀𝑀 + 𝐴𝐴𝐴𝐴)𝜂𝜂𝜂𝜂 + 𝐵𝐵𝐵𝐵𝜂𝜂𝜂𝜂 + 𝐶𝐶𝐶𝐶𝜂𝜂𝜂𝜂 = 𝐹𝐹𝐹𝐹𝑊𝑊𝑊𝑊 (9)
where for the linear motions the 𝐹𝐹𝐹𝐹𝑊𝑊𝑊𝑊 is wave force acting on particular direction, and for the angular
motions 𝐹𝐹𝐹𝐹𝑊𝑊𝑊𝑊 is moment about the axis of rotation. The 𝑀𝑀𝑀𝑀 + 𝐴𝐴𝐴𝐴 component is the sum of inertia properties
such as mass for linear motions and moment of inertia for rotations and added mass 𝐴𝐴𝐴𝐴. 𝐵𝐵𝐵𝐵 and 𝐶𝐶𝐶𝐶
components are hydrodynamic damping and restoring coefficients respectively. Further the equation (9)
may be normalized by 𝑀𝑀𝑀𝑀 + 𝐴𝐴𝐴𝐴 and then:
𝜂𝜂𝜂𝜂̈ + 𝜈𝜈𝜈𝜈𝜑𝜑𝜑𝜑 𝜂𝜂𝜂𝜂̇ + 𝜔𝜔𝜔𝜔𝜁𝜁𝜁𝜁2 𝜂𝜂𝜂𝜂 = 𝑓𝑓𝑓𝑓𝑊𝑊𝑊𝑊 (10)
Where 𝜈𝜈𝜈𝜈𝜁𝜁𝜁𝜁 and 𝜔𝜔𝜔𝜔𝜁𝜁𝜁𝜁 are dimensionless damping coefficient and natural frequency respectively. The
damping coefficient must have linear character and then the assumption of proportionality of response
and excitation is valid. This assumption gives good approximation for all motions except the roll when
the nonlinear viscous damping is significantly higher than damping occurring from the generation of
waves. However the roll motion damping nonlinearity may be included in the equation of motion as
long as it depends only on the amplitude of resulting roll motion.
The formula (6) is the key to predict roll motions in considered sea state with specified spectral
energy density 𝑆𝑆𝑆𝑆𝜁𝜁𝜁𝜁 (𝜔𝜔𝜔𝜔). The transfer function of roll motions 𝑌𝑌𝑌𝑌𝜁𝜁𝜁𝜁𝜑𝜑𝜑𝜑 is usually established by experiments
in regular waves with constant amplitude or slope. Whilst the RAO of roll motion is narrow band the
tests are carried out for multiple frequencies 𝜔𝜔𝜔𝜔 within the certain range around the resonant frequency
𝜔𝜔𝜔𝜔𝜑𝜑𝜑𝜑. Due to the fact that viscous damping phenomena depends on the level of roll amplitude, the constant
wave amplitude or slope based RAO must be corrected to the expected response in considered sea state.
It means that predictions from regular waves experiment should be complemented by characteristics of
roll damping established either by roll decay test or other ones. This approach is thus rather time
consuming, so this paper presents alternative, less time consuming method for RAO evaluation.
The formula (6) may be modified to the form:
𝑆𝑆𝑆𝑆𝑁𝑁𝑁𝑁𝜑𝜑𝜑𝜑 (𝜔𝜔𝜔𝜔𝑖𝑖𝑖𝑖 ) = 𝑆𝑆𝑆𝑆𝑁𝑁𝑁𝑁 (𝜔𝜔𝜔𝜔𝑖𝑖𝑖𝑖 )𝑌𝑌𝑌𝑌𝜑𝜑𝜑𝜑2 (𝜔𝜔𝜔𝜔𝑖𝑖𝑖𝑖 ) (11)
the input to the system here is the noise energy density spectrum 𝑆𝑆𝑆𝑆𝑁𝑁𝑁𝑁 (𝜔𝜔𝜔𝜔) which may have constant value
over the frequencies 𝜔𝜔𝜔𝜔 (white noise - WN), may proportionally increase (blue noise - BN) or decrease
(pink noise - PN) with the frequencies 𝜔𝜔𝜔𝜔. The 𝑑𝑑𝑑𝑑𝜔𝜔𝜔𝜔 part of formula is always greater than zero due the
fact we are working with discrete signals and might be deleted. Making transformations of formula (10)
the RAO might be found as:
𝑅𝑅𝑅𝑅𝑁𝑁𝑁𝑁𝑁𝑁𝑁𝑁 (𝜔𝜔𝜔𝜔𝑖𝑖𝑖𝑖 )
𝑌𝑌𝑌𝑌𝜑𝜑𝜑𝜑 (𝜔𝜔𝜔𝜔𝑖𝑖𝑖𝑖 ) = �
𝑅𝑅𝑅𝑅𝑁𝑁𝑁𝑁 (𝜔𝜔𝜔𝜔𝑖𝑖𝑖𝑖 )
(12)
The discrete roll motions RAO, in dimensionless form where the roll angle is related to wave slope
angle 𝛼𝛼𝛼𝛼𝐴𝐴𝐴𝐴 = 𝑘𝑘𝑘𝑘𝜁𝜁𝜁𝜁𝐴𝐴𝐴𝐴 , might be further approximated by the formula [3]:
𝜅𝜅𝜅𝜅𝑁𝑁𝑁𝑁
𝑌𝑌𝑌𝑌𝛼𝛼𝛼𝛼𝜑𝜑𝜑𝜑 (𝜔𝜔𝜔𝜔) = 2 2
(13)
2
��1− 𝜔𝜔𝜔𝜔2 � +�2𝜈𝜈𝜈𝜈𝑁𝑁𝑁𝑁 𝜔𝜔𝜔𝜔2 �
𝜔𝜔𝜔𝜔𝑁𝑁𝑁𝑁 𝜔𝜔𝜔𝜔𝑁𝑁𝑁𝑁
𝐵𝐵𝐵𝐵
where dimensionless damping is found as 𝜈𝜈𝜈𝜈𝜑𝜑𝜑𝜑 =
(𝑅𝑅𝑅𝑅+𝐴𝐴𝐴𝐴)
Formula (13) simplifies ship roll response because sway and yaw coupling is neglected and the
contribution of diffraction component might be compensated by modified Froude-Krylov heeling
moment reduction factor:
2
𝜅𝜅𝜅𝜅𝜑𝜑𝜑𝜑 (𝜔𝜔𝜔𝜔) = 𝛽𝛽𝛽𝛽0 𝑒𝑒𝑒𝑒 −𝛽𝛽𝛽𝛽1 𝜔𝜔𝜔𝜔 (14)
The coefficients 𝛽𝛽𝛽𝛽0 and 𝛽𝛽𝛽𝛽1 in function 𝜅𝜅𝜅𝜅𝜑𝜑𝜑𝜑 are determined by least square approximation of RAO
evaluated by noise measurements according to the formula (13). Further it is showed, that damping of
roll motions in noise waves is equal to damping in demanded sea state with the same total energy, what
means the RMS value of input waves (noise and wind wave spectral form) are equal. The next key
feature of the proposed method is that changes in damping is assumed linear related to RMS of generated
noise signal. It means that to find damping coefficient for prediction of roll motions the RMS of
83
Ship seakeeping
generated noise waves should lay within the RMS values of demanded sea states. Then the damping
coefficient can be read from linear characteristic of RMS values as it is shown in below sketch.
Fig. 2 Linear characteristic of roll damping coefficient

As far as we have linear characteristic of damping coefficients from noise measurements the number
of runs may be limited.
3. Model Tests
The model tests were carried out for three models of vessels. The tests were carried out in CTO
towing tank no.2 equipped with flap type wave maker and active wave absorber. The rectangular shape
tank has length 50 m, breadth 7 m and adjustable depth, which for the purpose of current analyses was
set to represent deep water conditions. The models were tested only in beam seas. The main dimensions
of hull models are presented in the Table 1.
Table 1 Main particulars of tested hull models
HM#01 HM#02 HM#03
𝐿𝐿𝐿𝐿𝑃𝑃𝑃𝑃𝑃𝑃𝑃𝑃 [m] 3.585 3.553 3.212
𝐵𝐵𝐵𝐵 [m] 0.708 0.770 0.952
𝑇𝑇𝑇𝑇 [m] 0.196 0.205 0.363
𝐶𝐶𝐶𝐶𝐵𝐵𝐵𝐵 0.92 0.78 0.70
𝑇𝑇𝑇𝑇𝜑𝜑𝜑𝜑 [s] 2.47 2.16 3.02
Scale factor 53.0 24.0 17.0
The models were tested in different irregular waves representing demanded sea states and appropriate
noise waves realizations. One of the models was tested in regular waves. The detailed configuration of
test cases is presented in Table 2.
84
Ship seakeeping
Table 2 Tests configurations

Wave RMS Wave RMS Wave RMS
(HM#01) [mm] (HM#02) [mm] (HM#03) [mm]
(ITTC) (JONSWAP) (JONSWAP)
I1.1 12.8 I1.2 28.8 I1.3 16.8
I2.1 27.2 I2.2 46.5 I2.2 36.0
I3.1 36.0 I3.2 55.9 I3.2 48.4
I4.2 89.9
(Noise) (Noise) (Noise)
WN1.1 21.3 WN1.2 37.6 WN1.2 24.4
WN2.1 25.9 WN2.2 46.3 WN2.2 36.0
WN3.1 9.9 WN3.2 23.8 WN3.2 48.4
WN4.1 30.2 PN1.2 38.4
PN2.2 47.0
BN1.2 36.7
BN2.2 23.1
(Regular)
R0.1 25.3
The tests in regular waves were carried out in constant regular wave double amplitude being 1/50 of
model length. This level of regular wave amplitude is recommended by ITTC due to the practical use -
measured response as well as wave height has magnitude sufficient to keep reasonable accuracy of
measured signals. The RAO function evaluated by measurements in regular waves is presented in Fig.
7.
The irregular waves were set to represent wide variety of irregular waves with the most popular wave
spectra such as ITTC and JONSWAP. The general statistic values representing generated wave spectra
are presented in the Table 2 and an example of measured wave spectra and its theoretical representation
is presented in Fig. 3.
Fig. 3 Spectra of wind waves and roll motions response (top - I2.1 ITTC, bottom – I3.2 JONSWAP)
85
Ship seakeeping
The noise spectra has been prepared to get as narrow band width as the expected RAO will be within
the range of wave frequencies. Besides the amplitude of signal was tuned to the energy of noise within
the range of demanded sea energy. The examples of generated noise amplitude spectrum with certain
RMS values are presented in the Fig. 4 and Fig. 5.
Fig. 4 Comparison of noise measurements – white, HM#02 (left) vs. blue (right), HM#02
86
Ship seakeeping
Fig. 5 Comparison of noise measurements – blue, HM#02 (left) vs. white (right), HM#03
The Fig. 6 presents the relation between RMS values of generated noise waves and damping
coefficients achieved from approximation of RAOs by formula (13).
87
Ship seakeeping
Fig. 6 Relation between damping coefficient and wave RMS
88
Ship seakeeping
The next figures (7 to 9) present the RAO functions evaluated from results of measurements in noise
waves
Fig. 7 HM#01, Approximated transfer functions
89
Ship seakeeping
4. Results of analyses
The approximation formula (13) with modified Froude-Krylov coefficients fits very well to RAOs
evaluated from noise wave experiments. The values of damping coefficients lays on straight line in all
considered configurations of tests (Fig. 6). The Table 3 presents results of predictions from direct
measurements in modelled sea waves compared to the predictions calculated from RAOs calculated
here. The predicted and measured values of RMS were then plotted against each other (Fig. 10). The
straight red line depicts ideal correlation between prediction and measurements and grey lines represents
±5% margin of actual values.
Table 3 Comparison of roll motions RMS values
(measurement in irregular waves and noise RAO based predictions)
Noise Irregular Regular Description
I1.1 2.42 2.23 1.67 HM#01
I2.1 3.63 3.36 3.57
I3.1 4.15 4.13 4.75
I1.2 0.95 1.01 N/A HM#02.BN
I2.2 5.21 5.01 N/A
I3.2 7.46 7.14 N/A
I4,2 8.62 8.75 N/A
I1.2 0.92 1.01 N/A HM#02.PN
I2.2 4.99 5.01 N/A
I3.2 7.15 7.14 N/A
I4,2 8.47 8.75 N/A
I1.2 0.90 1.01 N/A HM#02.WN
I2.2 5.04 5.01 N/A
I3.2 7.34 7.14 N/A
I4,2 8.92 8.75 N/A
I1.3 0.41 0.34 N/A HM#03
I2.3 4.77 4.54 N/A
I3.3 6.51 7.25 N/A
90
Ship seakeeping
Predicted and measured values were analysed by means of Pearson’s correlation factor 𝑅𝑅𝑅𝑅𝑓𝑓𝑓𝑓𝑓𝑓𝑓𝑓 and root
mean square error 𝑅𝑅𝑅𝑅𝑀𝑀𝑀𝑀𝑀𝑀𝑀𝑀𝑅𝑅𝑅𝑅 given by formulas [4]:
1
𝑅𝑅𝑅𝑅𝑓𝑓𝑓𝑓𝑓𝑓𝑓𝑓 = ∑𝑁𝑁𝑁𝑁
𝑖𝑖𝑖𝑖 𝑦𝑦𝑦𝑦𝑓𝑓𝑓𝑓𝑖𝑖𝑖𝑖 𝑦𝑦𝑦𝑦𝑓𝑓𝑓𝑓𝑓𝑓𝑓𝑓 (15)
𝑁𝑁𝑁𝑁
1 2
𝑅𝑅𝑅𝑅𝑀𝑀𝑀𝑀𝑀𝑀𝑀𝑀𝑅𝑅𝑅𝑅𝑓𝑓𝑓𝑓𝑓𝑓𝑓𝑓 = � ∑𝑁𝑁𝑁𝑁
𝑖𝑖𝑖𝑖 �𝑌𝑌𝑌𝑌𝑓𝑓𝑓𝑓𝑖𝑖𝑖𝑖 − 𝑌𝑌𝑌𝑌𝑓𝑓𝑓𝑓𝑓𝑓𝑓𝑓 � (16)
𝑁𝑁𝑁𝑁
where:
𝑌𝑌𝑌𝑌𝑓𝑓𝑓𝑓𝑖𝑖𝑖𝑖 −𝜇𝜇𝜇𝜇𝑓𝑓𝑓𝑓 𝑌𝑌𝑌𝑌𝑜𝑜𝑜𝑜𝑖𝑖𝑖𝑖 −𝜇𝜇𝜇𝜇𝑜𝑜𝑜𝑜
𝑦𝑦𝑦𝑦𝑓𝑓𝑓𝑓𝑖𝑖𝑖𝑖 = , 𝑦𝑦𝑦𝑦𝑓𝑓𝑓𝑓𝑓𝑓𝑓𝑓 = (17)
𝜎𝜎𝜎𝜎𝑓𝑓𝑓𝑓 𝜎𝜎𝜎𝜎𝑜𝑜𝑜𝑜
The 𝑅𝑅𝑅𝑅𝑓𝑓𝑓𝑓𝑓𝑓𝑓𝑓 correlation coefficient quantifies statistically linear correlation between observations
(measurements) 𝑌𝑌𝑌𝑌𝑓𝑓𝑓𝑓 and predictions 𝑌𝑌𝑌𝑌𝑓𝑓𝑓𝑓 . Squared value of 𝑅𝑅𝑅𝑅𝑓𝑓𝑓𝑓𝑓𝑓𝑓𝑓 is varying within range 0 to 1.
Fig. 10 Correlation plot - measurements against prediction
5. Conclusions
The method presented in this paper may be the alternative experimental approach to evaluate RAO.
The experiments, were carried out in irregular waves with wide bandwidth spectra of colour noise
character. This approach is quite similar to proposed in the past by author (Bielicki, Bednarek,
Kraskowski [6]). The main differences are introduced here:
1) Modified Froude-Krylov coefficient 𝜅𝜅𝜅𝜅𝜑𝜑𝜑𝜑 - equation (14), arises from measurements.
2) Damping coefficient 𝜈𝜈𝜈𝜈𝜑𝜑𝜑𝜑 is found from least square approximation of equation (13).
3) Exact damping coefficient applied in predictions for arbitrary sea condition is found by linear
regression of generated waves RMS value.
Predicted roll motions in multiple sea states conditions fits quite well with measurements in selected
irregular waves (sea conditions) what is confirmed by squared Pearson’s correlation above 0.99 and root
mean square error below 0.25 [deg]. The noise waves based measurement may be extended for other
motions, like heave and pitch.
91
Ship seakeeping
6. Acknowledgements
The research was financed by the Ministry of Science and Higher Education of the Republic of
Poland and by Maritime Advanced Research Centre CTO S.A.
7. References
[1] M. St. Denis i W. Pierson, „On the motion of ships in confused seas,” Soc. Nav. Archit. Mar. Eng., tom Trans.
61, pp. 280-354, 1953.
[2] J. Journee and W. Massie, Introduction in Offshore Hydromechanics, Delft: Delft University of Technology,
2001.
[3] O. Rice, “Mathematical Analyses of Random Noise,” Bell System Technical Journal, Vols. Vols. 23, 24,
1945.
[4] J. Dudziak, Teoria Okrętu, Gdańsk: Fundacja Promocji Przemysłu Okrętowego i Gospodarki Morskiej, 2008.
[5] A. Barnston, “Correspondence Among the Correlation, RMSE and Heidke Forecast Verification Measures,”
American Meteorological Society, Climate Analysis Center, NMC/NWS/NOAA, Washingto D.C., 1992.
[6] S. Bielicki, A. Bednarek and M. Kraskowski, “Evaluation of Response Amplitude Operator of Ship Roll
Motions,” in Proceedings of the ASME 2017 36th International Conference on Ocean, Offshore and Arctic
Engineering, Trondheim, 2017.
92
Ship seakeeping

HYDRONAV 2020
Sopot - Iława, 14-16 October 2020
Review of Selected Methods for Prediction of Added Resistance

inHYDRONAV
Following Waves
2020
Sopot - Iława, 14-16 October 2020
MICHAŁ KALBARCZYK1
Review of Selected Methods for Prediction of Added Resistance
SEBASTIAN BIELICKI2 in Following Waves
1 Maritime Advanced Research Centre (CTO S.A.) Gdańsk, Poland, [email protected]

MICHAŁ
2 KALBARCZYK
Maritime Advanced
1
Research Centre (CTO S.A.) Gdańsk, Poland, [email protected]
SEBASTIAN BIELICKI2
The added resistance in waves is a mean value of non-linear, second order reaction of a ship to incoming waves. In the beginning
of the 20th century,
1 Maritime the experimental
Advanced Research Centremethods
(CTOforS.A.)
investigation of ship hydrodynamics
Gdańsk, Poland, at model scale were developed. They
[email protected]
allowed the evaluation
2 Maritime Advanced of added
Researchresistance
Centre by measurements
(CTO S.A.) Gdańsk,in irregular
Poland, waves (directly) or by measurements in regular waves
[email protected]
(in-direct method). The main goal was to find more precise information about the margin for on-board installed engine power
to fulfil operational demands of designed vessel. In the early sixties of the 20th century, numerical methods appeared, enabling
the seakeeping analyses in frequency domain, and including the methods for determination of added resistance due to head
waves. These computational methods, based on gravity character of added resistance, applied the potential flow theory and
The
wereadded
mainlyresistance in waves
determined is a mean
to cope valuewaves.
with head of non-linear, second
However, the order
issue reaction
of routeofoptimization
a ship to incoming
as wellwaves.
as theInphenomena
the beginningof
of
broaching and surf-riding of vessels going through following and aft-quartering waves gained more attention recently, They
the 20th century, the experimental methods for investigation of ship hydrodynamics at model scale were developed. what
allowed
demandsthe evaluation
accurate of added
prediction resistance
methods by measurements
applicable in irregular
also for these waves
conditions. To(directly) or by measurements
avoid dangerous behaviour ofin regular waves
a vessel, it is
(in-direct method). The main goal was to find more precise information about the margin for on-board
necessary to increase the speed, which means that the ship has to maintain a power margin for this operation. The level installed engine power of
to fulfilpower
added operational
can bedemands
assessedofbasing
designed vessel. In the early
on computations sixties
of added of the 20th
resistance century,
in waves numerical
coming frommethods
mentionedappeared, enabling
directions. The
the seakeeping
following paperanalyses
presentsinthe
frequency
comparisondomain, and
of the including
results the methods
obtained for determination
from selected computational of added resistance
methods, duepotential
including to head
waves. These computational methods, based on gravity character of added resistance,
methods, semi-empirical formulae and RANS-CFD, with the results of dedicated model scale experiment. applied the potential flow theory and
were mainly determined to cope with head waves. However, the issue of route optimization as well as the phenomena of
broaching and surf-riding of vessels going through following and aft-quartering waves gained more attention recently, what
Keywords:
demands addedprediction
accurate resistance, following
methods waves, KCS,also
applicable panelfor
method,
thesestrip theory, SNNM
conditions. formula
To avoid dangerous behaviour of a vessel, it is
necessary to increase the speed, which means that the ship has to maintain a power margin for this operation. The level of
added power can be assessed basing on computations of added resistance in waves coming from mentioned directions. The
following paper presents the comparison of the results obtained from selected computational methods, including potential
1. Introduction
methods, semi-empirical formulae and RANS-CFD, with the results of dedicated model scale experiment.
The paper
Keywords: presents
added resistance, the concise
following waves, survey
KCS, panelofmethod,
computation methods
strip theory, SNNM for added resistance in waves
formula
prediction in comparison with experiments in particular for the following waves condition. The model
scale experimental methods for evaluation of ship’s resistance due to wave are the most common and
crucial these days. Along the model tests the computational methods for added resistance due to sea
1. Introduction
waves have been developed since decades. The most common are based on potential theory as the ship
motions contribute
The paper mostly
presents thetoconcise
the energy lost by
survey of ship moving inmethods
computation waves. These
for addedmotions are caused
resistance by the
in waves
inertial and wave phenomena and mainly the wave influence may be
prediction in comparison with experiments in particular for the following waves condition. The model considered by potential
hydrodynamic
scale methods.
experimental methods Thisfor
simplification
evaluation ofis ship’s
mostlyresistance
valid when duevessel is running
to wave are the in headcommon
most waves with
and
crucial these days. Along the model tests the computational methods for added resistance due to sea
waves have been developed since decades. The most common are based on potential theory as the ship1
motions contribute mostly to the energy lost by ship moving in waves. These motions are caused by the
inertial and wave phenomena and mainly the wave influence may be considered by potential
hydrodynamic methods. This simplification is mostly valid when vessel is running in head waves with
93
Ship seakeeping
the speed when the heave and pitch motions are significant and waves reflection of shorter wave
components can be modelled well by potential theory. This approach is usually not sufficient when the
waves are going in the direction of ship's speed and the surf riding and broaching phenomena may occur.
2. Added resistance due to waves
Mainly added resistance due to waves is the result of dissipated energy caused by ship's pitch and
heave motions with some contribution of reflected waves. This phenomena was considered by Havelock
(1945) in his research work and developed further by Maruo [1] (1957). The fundamental outcome from
Maruo work was the idea of proportionality of added resistance to square of wave height. Moreover
Maruo noticed that additional wave resistance was independent of still water resistance. Basing on the
last assumption, the prediction of total resistance is calculated as a sum of resistance in calm water and
the part of resistance caused mainly by ship operating in waves (1)
𝑅𝑅𝑇𝑇𝑇𝑇 (𝑉𝑉) = 𝑅𝑅𝑇𝑇0 (𝑉𝑉) + 𝑅𝑅𝐴𝐴𝐴𝐴 (𝑉𝑉) (1)
where R T0 is the resistance of ship moving with speed 𝑉𝑉 and 𝑅𝑅𝐴𝐴𝐴𝐴 is the mean value of second order
longitudinal wave reaction (drift force) on ship moving in waves. The last component of formula (1)
takes the form:
∞
𝑅𝑅𝐴𝐴𝐴𝐴 = ∫ 𝑅𝑅𝐴𝐴𝐴𝐴 (𝜔𝜔)𝑆𝑆𝜁𝜁 (𝜔𝜔)𝑑𝑑𝑑𝑑
0
The first assumption considered for the mean added resistance in regular waves follows to relation
𝛿𝛿𝛿𝛿(𝜔𝜔)
𝑅𝑅𝐴𝐴𝐴𝐴 (𝜔𝜔) = 2
𝜁𝜁𝐴𝐴 (𝜔𝜔)
and is called quadratic transfer function (QTF). The 𝛿𝛿𝛿𝛿(𝜔𝜔) is the mean excess resistance in regular
wave of frequency 𝜔𝜔 and amplitude 𝜁𝜁𝐴𝐴 .
3. Some methods to obtain the added resistance in regular waves
Added resistance in following waves can be calculated applying different methods and formulas.
There are three computational software used in analyses. First one is the strip theory solver with few
methods for evaluation of added resistance in regular waves implemented in the code. The second is a
brand new method called SNNM and it is developed by approximation of statistical data of added
resistance values measured experimentally in variety of wave conditions and ship types. The last is the
potential based three dimensional radiation/diffraction panel code where the second order drift forces
(added resistance among them) are calculated by far-field or near-field solutions.
The strip theory solver and potential panel code may conduct calculation on detailed hull geometry
imported to the program in the form of CAD file, whilst the geometry of considered vessel in the form
of main dimensions, and some coefficients is sufficient to evaluate added resistance by SNNM method.
3.1 Strip theory code - Maxsurf Seakeeper

Seakeeper is the seakeeping analysis package of the Maxsurf software suite. It uses four alternative
methods developed by Salvesen, Havelock and Gerritsma and Beukelman (I and II) and all of them are
based on the outcome (damping and added mass) from strip theory calculations of motions with
geometry of ship’s hull represented by conformal mapping of adjustable level of approximation.
The added resistance based on Salvesen method [2] was calculated by formula:
𝑖𝑖𝑖𝑖
̂3 + 𝜂𝜂5 𝐹𝐹
𝑅𝑅𝐴𝐴𝐴𝐴 = (𝜂𝜂3 𝐹𝐹 ̂5 ) + 𝑅𝑅7 (2)
2
where:
̂3 = 𝜁𝜁 ∫ 𝑒𝑒 −𝑖𝑖𝑖𝑖𝑖𝑖 𝑒𝑒 −𝑘𝑘𝑘𝑘 [𝑐𝑐(𝜉𝜉) − 𝜔𝜔0 (𝜔𝜔𝑒𝑒 𝑎𝑎33 (𝜉𝜉) − 𝑖𝑖𝑏𝑏33 (𝜉𝜉))]𝑑𝑑𝑑𝑑
𝐹𝐹 (3)
𝐿𝐿
̂5 = −𝜁𝜁 ∫ 𝑒𝑒 −𝑖𝑖𝑖𝑖ϛ 𝑒𝑒 −𝑘𝑘𝑘𝑘 [𝑐𝑐(𝜉𝜉) − 𝜔𝜔0 (𝜉𝜉 + 𝑖𝑖𝑖𝑖 ) (𝜔𝜔𝑒𝑒 𝑎𝑎33 (𝜉𝜉) − 𝑖𝑖𝑏𝑏33 (𝜉𝜉))] 𝑑𝑑𝑑𝑑
𝐹𝐹 (4)
𝐿𝐿 𝜔𝜔𝑒𝑒
and
2
94
Ship seakeeping
𝜁𝜁 2 𝑘𝑘𝜔𝜔02
𝑅𝑅7 =
2𝜔𝜔𝑒𝑒 𝐿𝐿
∫ 𝑒𝑒 −2𝑘𝑘𝑘𝑘 𝑏𝑏33 (𝜉𝜉) 𝑑𝑑𝑑𝑑 (5)
The 𝜂𝜂3 and 𝜂𝜂5 are the complex heave and pitch motions respectively. As it is shown in equation (2)
the method relates the global wave excitation to the sectional added mass and damping coefficients. The
Salvensen work is limited to head waves only and should be applied for normal section shapes (no
bulbous bow, aft bulbs etc.) thus the results were not taken in consideration here.
Gerritsma and Beukelman [3] found the added resistance might be evaluated by relative vertical
velocity of the vessel in centre of gravity and modified damping coefficient of heave motion:
𝑘𝑘 ∗ 2
𝑅𝑅𝐴𝐴𝐴𝐴 = ∫𝐿𝐿 𝑏𝑏33 𝑉𝑉𝑧𝑧𝑧𝑧 𝑑𝑑𝑑𝑑 (6)
2𝜔𝜔𝑒𝑒
∗ 𝑑𝑑𝑎𝑎33
where modified section damping 𝑏𝑏33 = 𝑏𝑏33 − 𝑈𝑈 and relative vertical velocity is given as:
𝑑𝑑𝑑𝑑
𝑉𝑉𝑍𝑍𝐴𝐴 = 𝜂𝜂3̇ − 𝜁𝜁𝜂𝜂5̇ + 𝑈𝑈𝜂𝜂3 − ϛ̇∗ (7)
and ϛ̇∗ = −𝜔𝜔ϛ∗ in the I version of method or ϛ̇∗ = −𝜔𝜔𝑒𝑒 ϛ∗ for the II version. Similarly to the Salvensen
formulae the 𝜂𝜂3 , 𝜂𝜂5 , ϛ∗ are the complex heave, pitch and local relative vertical velocity.
Finally the method selected in this paper is the one proposed by Havelock [4]. Added resistance is
given by formula:
𝑘𝑘
𝑅𝑅𝐴𝐴𝐴𝐴 = − 𝐹𝐹3 𝜂𝜂3 𝑠𝑠𝑠𝑠𝑠𝑠𝜀𝜀3 + 𝐹𝐹5 𝜂𝜂5 𝑠𝑠𝑠𝑠𝑠𝑠𝜀𝜀5 (8)
2
where:
𝑘𝑘 wave number for unlimited water depth;
𝐹𝐹3 , 𝐹𝐹5 are amplitudes of the heave and pitch excitation force and moment respectively;
𝜂𝜂3 , 𝜂𝜂5 are the amplitudes of heave ant pitch motions respectively;
𝜀𝜀3 , 𝜀𝜀5 are the phase differences of the heave and pitch motions with the corresponding excitation
force moment.
This formula is very simple and is developed basing on the assumption that the main source of energy
dissipation are pitch and heave motions [4] (with the significant predominance of the first component).
The reflected waves are neglected here at all however taking into account that wave reflection
phenomena occurs mainly in short waves, and the ship running in stern waves usually is exposed to long
waves with small encounter frequency this simplification seems reasonable.
3.2 Semi-empirical code based on SNNM method

The semi-empirical formula SNNM [5] is a method of approximating added resistance of the ship in
waves. The input data in this formula are the main dimensions of the ship, speed and wave conditions.
This leads to an approximate quadratic transfer function for the average increase in resistance 𝑅𝑅𝐴𝐴𝐴𝐴 in
regular waves of amplitude ζ and any direction (form heading to following wave). The formula has been
refined with the continued growth of the experimental background database. Taking into account the
majority of the ships and models in the SNNM database [6] the limitation to this method were settled as
𝐿𝐿 𝐵𝐵
follows: 90𝑚𝑚 < 𝐿𝐿𝑃𝑃𝑃𝑃 < 380𝑚𝑚, 5.0 < < 7.5, 2.0 < < 8.0, 0.54 < 𝐶𝐶𝐵𝐵 < 0.87, 0.0 < 𝐹𝐹𝐹𝐹 < 0.3.
𝐵𝐵 𝑇𝑇
Besides attention should be paid when the formula is applied to a ship type, which is not covered by the
database and may have unique design features.
3
95
Ship seakeeping
Fig. 1. Sketch of the waterline profile of a ship and related definitions
Quadratic Transfer Function (QTF) of added resistance in regular waves 𝑅𝑅𝐴𝐴𝐴𝐴 can be calculated as
the sum of 𝐾𝐾𝐴𝐴𝐴𝐴𝐴𝐴 , the component of added resistance due to motion (radiation) effect and 𝐾𝐾𝐴𝐴𝐴𝐴𝐴𝐴 , the
component of added resistance due to reflection (diffraction) effect in regular wave is given by Equation
(7).
𝐵𝐵2
𝑅𝑅𝐴𝐴𝐴𝐴 = 4𝜌𝜌𝜌𝜌𝜁𝜁𝐴𝐴2 ( ) (𝐾𝐾𝐴𝐴𝐴𝐴𝐴𝐴 + 𝐾𝐾𝐴𝐴𝐴𝐴𝐴𝐴 ) (9)
𝐿𝐿
The expression of 𝐾𝐾𝐴𝐴𝐴𝐴𝐴𝐴 is given by Equation (8):

𝑏𝑏1
̅ 𝑑𝑑1 )
(1− 𝜔𝜔
̅ −1 𝑒𝑒 𝑑𝑑1
𝐾𝐾𝐴𝐴𝐴𝐴𝐴𝐴 = 𝑎𝑎1 𝑎𝑎2 𝑎𝑎3 𝜔𝜔 (10)
where:
3 𝐿𝐿𝑃𝑃𝑃𝑃 0.111 𝐵𝐵 𝐶𝐶𝐵𝐵 0.17 0.618(13 + 𝑐𝑐𝑐𝑐𝑐𝑐2𝛼𝛼)

̅ = 2.142 √𝑘𝑘𝑦𝑦𝑦𝑦 √
𝜔𝜔 [1 − (𝑙𝑙𝑙𝑙 − 𝑙𝑙𝑙𝑙2.75)] ( ) [(−1.377𝐹𝐹𝐹𝐹 2 + 1.1157𝐹𝐹𝐹𝐹)|𝑐𝑐𝑐𝑐𝑐𝑐𝑐𝑐| + ]
𝜆𝜆 𝐶𝐶𝐵𝐵 𝑇𝑇𝑚𝑚𝑚𝑚𝑚𝑚 0.65 14
2 0.87 −(1+𝐹𝐹𝐹𝐹)𝑐𝑐𝑐𝑐𝑐𝑐𝑐𝑐 𝐵𝐵 −1 1 − 2𝑐𝑐𝑐𝑐𝑐𝑐𝑐𝑐 𝜋𝜋

60.3𝐶𝐶𝐵𝐵1.34 (4𝑘𝑘𝑦𝑦𝑦𝑦 ) ( ) (𝑙𝑙𝑙𝑙 ) 𝑓𝑓𝑓𝑓𝑓𝑓 ≤ 𝛼𝛼 ≤ 𝜋𝜋
𝐶𝐶𝐵𝐵 𝑇𝑇𝑚𝑚𝑚𝑚𝑚𝑚 2 2
𝜋𝜋
𝑎𝑎1 = 𝑖𝑖𝑖𝑖𝑖𝑖𝑖𝑖𝑖𝑖𝑖𝑖𝑖𝑖𝑖𝑖𝑖𝑖𝑖𝑖𝑖𝑖𝑖𝑖𝑖𝑖 𝑏𝑏𝑏𝑏𝑏𝑏𝑏𝑏𝑏𝑏𝑏𝑏𝑏𝑏 𝑏𝑏𝑏𝑏𝑏𝑏𝑏𝑏 𝑎𝑎𝑎𝑎𝑎𝑎 𝑓𝑓𝑓𝑓𝑓𝑓𝑓𝑓𝑓𝑓𝑓𝑓𝑓𝑓𝑓𝑓𝑓𝑓 𝑤𝑤𝑤𝑤𝑤𝑤𝑤𝑤 𝑐𝑐𝑐𝑐𝑐𝑐𝑐𝑐𝑐𝑐 𝑓𝑓𝑓𝑓𝑓𝑓 0 < 𝛼𝛼 <
2
−1
2 𝐵𝐵
60.3𝐶𝐶𝐵𝐵1.34 (4𝑘𝑘𝑦𝑦𝑦𝑦 ) (𝑙𝑙𝑙𝑙 ) 𝑓𝑓𝑓𝑓𝑓𝑓 𝛼𝛼 = 0
{ 𝑇𝑇𝑚𝑚𝑚𝑚𝑚𝑚
𝑔𝑔𝑔𝑔
𝑉𝑉𝐶𝐶 =
2𝜋𝜋
(𝑉𝑉 − 0.5𝑉𝑉𝐶𝐶 )
𝐹𝐹𝑛𝑛𝑛𝑛 =
√𝑔𝑔𝑔𝑔
4𝑉𝑉 0.87
−0.0072 + 0.0072 ( ) 𝑓𝑓𝑓𝑓𝑓𝑓 𝑉𝑉 < 𝑉𝑉𝐶𝐶 /2
𝑉𝑉𝐶𝐶 𝐶𝐶𝐵𝐵
0.87
𝑎𝑎2 = 0.0072 + 0.1676𝐹𝐹𝑛𝑛𝑛𝑛 ( ) 𝑓𝑓𝑜𝑜𝑜𝑜 𝐹𝐹𝑛𝑛𝑛𝑛 ≤ 0.12
𝐶𝐶𝐵𝐵
1.5
0.87 (1+𝐹𝐹𝑛𝑛𝑛𝑛 )
𝐹𝐹𝑛𝑛𝑛𝑛 exp(−3.5𝐹𝐹𝑛𝑛𝑛𝑛 ) ( ) 𝑓𝑓𝑓𝑓𝑓𝑓 𝐹𝐹𝑛𝑛𝑛𝑛 > 0.12
{ 𝐶𝐶𝐵𝐵
|𝑇𝑇𝑎𝑎 − 𝑇𝑇𝑓𝑓 |
𝑎𝑎3 = 1.0 + 28.7𝑎𝑎𝑎𝑎𝑎𝑎𝑎𝑎
𝐿𝐿𝑃𝑃𝑃𝑃
11.0 𝑓𝑓𝑓𝑓𝑓𝑓 𝜔𝜔
̅<1
𝑏𝑏1 = {
−8.5 𝑒𝑒𝑒𝑒𝑒𝑒𝑒𝑒𝑒𝑒ℎ𝑒𝑒𝑒𝑒𝑒𝑒
4
96
Ship seakeeping
𝐿𝐿𝑃𝑃𝑃𝑃 𝐶𝐶𝐵𝐵 −2.66

566 ( ) 𝑓𝑓𝑓𝑓𝑓𝑓 𝜔𝜔
̅<1
𝐵𝐵
𝑑𝑑1 =
𝐿𝐿𝑃𝑃𝑃𝑃 −2.66 |𝑇𝑇𝑎𝑎 − 𝑇𝑇𝑓𝑓 |
−566 ( ) (4 − 125𝑎𝑎𝑎𝑎𝑎𝑎𝑎𝑎 ) 𝑓𝑓𝑓𝑓𝑓𝑓 𝜔𝜔
̅≥1
{ 𝐵𝐵 𝐿𝐿𝑃𝑃𝑃𝑃
The expression of 𝐾𝐾𝐴𝐴𝐴𝐴𝑅𝑅 is given by Equation (9):

4
𝐾𝐾𝐴𝐴𝐴𝐴𝐴𝐴 = ∑ 𝐾𝐾𝐴𝐴𝐴𝐴𝐴𝐴 𝑖𝑖 (9)

𝑖𝑖=1
when 𝐸𝐸1 ≤ 𝛼𝛼 ≤ 𝜋𝜋
1 2.25 2 (𝐸𝐸
2𝜔𝜔0 𝑈𝑈 0.87 (1+4√𝐹𝐹𝐹𝐹)𝑓𝑓(𝛼𝛼)
𝐾𝐾𝐴𝐴𝐴𝐴𝐴𝐴 = 𝜌𝜌𝜌𝜌𝜌𝜌𝛼𝛼 𝑇𝑇 {sin 1 − 𝛼𝛼) +
∗ [𝑐𝑐𝑐𝑐𝑐𝑐𝐸𝐸1 𝑐𝑐𝑐𝑐𝑐𝑐(𝐸𝐸1 − 𝛼𝛼) − 𝑐𝑐𝑐𝑐𝑐𝑐𝑐𝑐]} ( )
4 𝑔𝑔 𝐶𝐶𝐵𝐵
when 𝜋𝜋 − 𝐸𝐸1 ≤ 𝛼𝛼 ≤ 𝜋𝜋
2 2.25 2
2𝜔𝜔0 𝑈𝑈 0.87 (1+4√𝐹𝐹𝐹𝐹)𝑓𝑓(𝛼𝛼)
𝐾𝐾𝐴𝐴𝐴𝐴𝐴𝐴 = 𝜌𝜌𝜌𝜌𝜌𝜌𝛼𝛼 𝑇𝑇 ∗ {sin (𝐸𝐸1 + 𝛼𝛼) + [𝑐𝑐𝑐𝑐𝑐𝑐𝐸𝐸1 𝑐𝑐𝑐𝑐𝑐𝑐(𝐸𝐸1 + 𝛼𝛼) − 𝑐𝑐𝑐𝑐𝑐𝑐𝑐𝑐]} ( )
4 𝑔𝑔 𝐶𝐶𝐵𝐵
when 0 ≤ 𝛼𝛼 ≤ 𝜋𝜋 − 𝐸𝐸2
2.25 2𝜔𝜔0 𝑈𝑈
𝐾𝐾𝐴𝐴𝐴𝐴𝐴𝐴 3 = − 𝜌𝜌𝜌𝜌𝜌𝜌𝛼𝛼 𝑇𝑇 ∗ {sin2 (𝐸𝐸2 + 𝛼𝛼) + [𝑐𝑐𝑐𝑐𝑐𝑐𝐸𝐸2 𝑐𝑐𝑐𝑐𝑐𝑐(𝐸𝐸2 + 𝛼𝛼) − 𝑐𝑐𝑐𝑐𝑐𝑐𝑐𝑐]}
4 𝑔𝑔
when 0 ≤ 𝛼𝛼 ≤ 𝐸𝐸2
2.25 2𝜔𝜔0 𝑈𝑈
𝐾𝐾𝐴𝐴𝐴𝐴𝐴𝐴 4 = − 𝜌𝜌𝜌𝜌𝜌𝜌𝛼𝛼 𝑇𝑇 ∗ {sin2 (𝐸𝐸2 − 𝛼𝛼) + [𝑐𝑐𝑐𝑐𝑐𝑐𝐸𝐸2 𝑐𝑐𝑐𝑐𝑐𝑐(𝐸𝐸2 − 𝛼𝛼) − 𝑐𝑐𝑐𝑐𝑐𝑐𝑐𝑐]}
4 𝑔𝑔
where:
𝜔𝜔0 – is the natural frequency of the incident wave
𝛼𝛼 𝑇𝑇 ∗ - is the draft coefficient
−𝑐𝑐𝑐𝑐𝑐𝑐𝑐𝑐 𝜋𝜋 − 𝐸𝐸1 ≤ 𝛼𝛼 ≤ 𝜋𝜋
𝑓𝑓(𝛼𝛼) = {
0 𝛽𝛽 < 𝜋𝜋 − 𝐸𝐸1
𝑇𝑇 ∗ 𝑇𝑇 ∗
(−4𝜋𝜋( − )
2.5𝐿𝐿
𝛼𝛼 𝑇𝑇 ∗ = {1 − 𝑒𝑒 𝜆𝜆 𝑃𝑃𝑃𝑃 𝜆𝜆/𝐿𝐿𝑃𝑃𝑃𝑃 ≤ 2.5
0 𝜆𝜆/𝐿𝐿𝑃𝑃𝑃𝑃 > 2.5
where for 𝑆𝑆1 and 𝑆𝑆2 segments:
𝑇𝑇 ∗ = 𝑇𝑇𝑚𝑚𝑚𝑚𝑚𝑚
and for 𝑆𝑆3 i 𝑆𝑆4 segments:
𝑇𝑇𝑚𝑚𝑚𝑚𝑚𝑚 (4 + √|𝑐𝑐𝑐𝑐𝑐𝑐𝑐𝑐|) /5 𝐶𝐶𝐵𝐵 ≤ 0.75

𝑇𝑇 ∗ = {
𝑇𝑇𝑚𝑚𝑚𝑚𝑚𝑚 (2 + √|𝑐𝑐𝑐𝑐𝑐𝑐𝑐𝑐|) /3 𝐶𝐶𝐵𝐵 > 0.75
5
97
Ship seakeeping
3.3 ANSYS AQWA

The AQWA is the seakeeping package of ANSYS CAD/CAM system. The computations are carried
out basing on three-dimensional radiation/diffraction theory without viscous terms, but featured with
the manual application of external models (e.g.: additional viscous damping of roll motions). The added
resistance, as the second order mean force in horizontal plane along the longitudinal axis, is calculated
by the formula of the far field solution [7] of the second order mean wave drift forces coefficients under
unidirectional waves. Since the second order potential does not contribute to them, the fluid potential
includes the first order components only. The time averaged over a period of the incident wave the
horizontal mean drift force along the longitudinal axis is given by equation [7]:
𝜌𝜌 𝜕𝜕𝜕𝜕 2 𝜕𝜕𝜕𝜕 𝜕𝜕𝜕𝜕 1
𝐹𝐹1 = − ∫ ( ) 𝑛𝑛1 𝑑𝑑𝑑𝑑 − 𝜌𝜌 ∬𝑆𝑆 [ − ∇𝜙𝜙 ∙ ∇𝜙𝜙𝑛𝑛1 ] 𝑑𝑑𝑑𝑑 (11)
2𝑔𝑔 𝑙𝑙 𝜕𝜕𝜕𝜕 𝐶𝐶𝜕𝜕𝜕𝜕 𝜕𝜕𝜕𝜕 2
where 𝐹𝐹1 is the added resistance 𝑅𝑅𝐴𝐴𝐴𝐴 here. The 𝑙𝑙 is the intersection between control surface (domain)
𝑆𝑆𝐶𝐶 and the mean free surface, 𝜙𝜙 is the fluid potential and 𝑛𝑛 is the normal vector.
The AQWA code solves the equations of motions in frequency domain and outputs quadratic transfer
function (QTF) matrix of second order wave exciting forces due to the first order waves and motions
responses, as a pair of regular incident waves with (𝑎𝑎𝑗𝑗𝑗𝑗 , 𝜔𝜔𝑗𝑗𝑗𝑗 , 𝜒𝜒𝑗𝑗𝑗𝑗 , 𝛼𝛼𝑗𝑗𝑗𝑗 ) and (𝑎𝑎𝑘𝑘𝑘𝑘 , 𝜔𝜔𝑘𝑘𝑘𝑘 , 𝜒𝜒𝑘𝑘𝑘𝑘 , 𝛼𝛼𝑘𝑘𝑘𝑘 ):
(2)
𝐹𝐹⃗ = 𝑎𝑎𝑗𝑗𝑗𝑗 𝑎𝑎𝑘𝑘𝑘𝑘 {𝑃𝑃⃗⃗𝑗𝑗𝑗𝑗𝑗𝑗𝑗𝑗
𝑗𝑗𝑗𝑗𝑗𝑗𝑗𝑗
+
𝑐𝑐𝑐𝑐𝑐𝑐[(𝜔𝜔𝑗𝑗𝑗𝑗 + 𝜔𝜔𝑘𝑘𝑘𝑘 )𝑡𝑡 − (𝛼𝛼𝑗𝑗𝑗𝑗 + 𝛼𝛼𝑘𝑘𝑘𝑘 )]
⃗⃗𝑗𝑗𝑗𝑗𝑗𝑗𝑗𝑗
+ 𝑄𝑄 +
𝑠𝑠𝑠𝑠𝑠𝑠[(𝜔𝜔𝑗𝑗𝑗𝑗 + 𝜔𝜔𝑘𝑘𝑘𝑘 )𝑡𝑡 − (𝛼𝛼𝑗𝑗𝑗𝑗 + 𝛼𝛼𝑘𝑘𝑘𝑘 )]
(12)
+ 𝑃𝑃⃗⃗𝑗𝑗𝑗𝑗𝑗𝑗𝑗𝑗 𝑐𝑐𝑐𝑐𝑐𝑐[(𝜔𝜔𝑗𝑗𝑗𝑗 − 𝜔𝜔𝑘𝑘𝑘𝑘 )𝑡𝑡 − (𝛼𝛼𝑗𝑗𝑗𝑗 − 𝛼𝛼𝑘𝑘𝑘𝑘 )]
−
+ 𝑄𝑄 −
𝑠𝑠𝑠𝑠𝑠𝑠[(𝜔𝜔𝑗𝑗𝑗𝑗 − 𝜔𝜔𝑘𝑘𝑘𝑘 )𝑡𝑡 − (𝛼𝛼𝑗𝑗𝑗𝑗 − 𝛼𝛼𝑘𝑘𝑘𝑘 )]}
where the coefficients for the second order wave forces are
1 1
(𝑃𝑃⃗⃗𝑗𝑗𝑗𝑗𝑗𝑗𝑗𝑗
+ ⃗⃗𝑗𝑗𝑗𝑗𝑗𝑗𝑗𝑗
, 𝑄𝑄 +
) = − 𝜌𝜌𝜌𝜌 ∮ 𝜁𝜁𝑟𝑟𝑟𝑟𝑟𝑟′ ′
∙ 𝜁𝜁𝑟𝑟𝑟𝑟𝑟𝑟 ′
𝑛𝑛⃗⃗ 𝑑𝑑𝑑𝑑 + 𝜌𝜌 ∬[𝛻𝛻𝛷𝛷𝑗𝑗𝑗𝑗 ′
∙ 𝛻𝛻𝛻𝛻𝑘𝑘𝑘𝑘 ] 𝑛𝑛⃗⃗ 𝑑𝑑𝑑𝑑
4 4
𝑊𝑊𝑊𝑊 𝑆𝑆0
′
1 𝑑𝑑𝑑𝑑𝑘𝑘𝑘𝑘 1 ′
+ 𝜌𝜌 ∬ [𝑋𝑋⃗𝑗𝑗𝑗𝑗
′
∙ 𝛻𝛻 ] 𝑛𝑛⃗⃗ 𝑑𝑑𝑑𝑑 + 𝛼𝛼⃗𝑗𝑗𝑗𝑗 × 𝐹𝐹⃗𝑗𝑗𝑗𝑗
′
2 𝑑𝑑𝑑𝑑 2
𝑆𝑆0
1 ′ ∗
1 ′ ∗
(𝑃𝑃⃗⃗𝑗𝑗𝑗𝑗𝑗𝑗𝑗𝑗
− ⃗⃗𝑗𝑗𝑗𝑗𝑗𝑗𝑗𝑗
, 𝑄𝑄 − ′
) = − 𝜌𝜌𝜌𝜌 ∮ 𝜁𝜁𝑟𝑟𝑟𝑟𝑟𝑟 ∙ 𝜁𝜁𝑟𝑟𝑟𝑟𝑟𝑟 ′
𝑛𝑛⃗⃗ 𝑑𝑑𝑑𝑑 + 𝜌𝜌 ∬[𝛻𝛻𝛷𝛷𝑗𝑗𝑗𝑗 ∙ 𝛻𝛻𝛻𝛻𝑘𝑘𝑘𝑘 ] 𝑛𝑛⃗⃗ 𝑑𝑑𝑑𝑑
4 4
𝑊𝑊𝑊𝑊 𝑆𝑆0
′ ∗
1 𝑑𝑑𝑑𝑑𝑘𝑘𝑘𝑘 1 ′ ′ ∗
+ 𝜌𝜌 ∬ [𝑋𝑋⃗𝑗𝑗𝑗𝑗
′
∙ 𝛻𝛻 ] 𝑛𝑛⃗⃗ 𝑑𝑑𝑑𝑑 + 𝛼𝛼⃗𝑗𝑗𝑗𝑗 × 𝐹𝐹⃗𝑗𝑗𝑗𝑗
2 𝑑𝑑𝑑𝑑 2
𝑆𝑆0
𝑃𝑃⃗⃗𝑗𝑗𝑗𝑗𝑗𝑗𝑗𝑗
+ ⃗⃗𝑗𝑗𝑗𝑗𝑗𝑗𝑗𝑗
, 𝑄𝑄 +
are used as the sum frequency force components, while 𝑃𝑃⃗⃗𝑗𝑗𝑗𝑗𝑗𝑗𝑗𝑗
− ⃗⃗𝑗𝑗𝑗𝑗𝑗𝑗𝑗𝑗
, 𝑄𝑄 −
contribute only
in the difference frequency force components. Because the equation (12), known as far field solution,
has some limitations the near field solution is introduced in AQWA. Based on the mean wetted surface
integration approach the wave drift force can be given as a special case in (12) where 𝑗𝑗 = 𝑘𝑘 and the
mean drift force is expressed as the triple summation:
𝑁𝑁𝑑𝑑 𝑁𝑁𝑑𝑑 𝑁𝑁𝑑𝑑
−′ −′
𝐹𝐹⃗ (2) = ∑ ∑ ∑ 𝑎𝑎𝑗𝑗𝑗𝑗 𝑎𝑎𝑗𝑗𝑗𝑗 {𝑃𝑃⃗⃗𝑗𝑗𝑗𝑗𝑗𝑗𝑗𝑗 cos(𝛼𝛼𝑗𝑗𝑗𝑗 − 𝛼𝛼𝑘𝑘𝑘𝑘 ) − 𝑄𝑄 sin(𝛼𝛼𝑗𝑗𝑗𝑗 − 𝛼𝛼𝑘𝑘𝑘𝑘 )}
𝑚𝑚=1 𝑛𝑛=1 𝑗𝑗=1
𝑁𝑁𝑑𝑑 𝑁𝑁𝑑𝑑 𝑁𝑁𝑑𝑑 (13)
̅̅̅̅̅
= ∑ ∑ ∑ 𝑓𝑓⃗(2) (𝜔𝜔𝑗𝑗 ; 𝛽𝛽𝑚𝑚 ; 𝛽𝛽𝑛𝑛 )
𝑚𝑚=1 𝑛𝑛=1 𝑗𝑗=1
where the numbers of wave components of every individual wave directions are the same. As far as this
paper is focused only on long crested wave case (𝑁𝑁𝑑𝑑 = 1) the mean drift force, i.e. added resistance in
regular waves is reduced to:
(2) −′
𝐹𝐹⃗𝑗𝑗𝑗𝑗11 (𝜔𝜔𝑗𝑗 ) = 𝑅𝑅𝐴𝐴𝐴𝐴(𝜔𝜔𝑗𝑗) = 𝑎𝑎𝑗𝑗2 𝑃𝑃⃗⃗𝑗𝑗𝑗𝑗11 (𝜔𝜔𝑗𝑗 )
6
98
Ship seakeeping
4. Model Tests
Model tests were carried out in CTO, deep water towing tank sized 260m x 12m x 6m in (length,
breadth and depth respectively). For all tests scaled model of KCS (Korean Container ship) was used.
The main parameters of the ship are shown in Table 1 and the hull form is presented in Fig. 1.
Table 1 Main parameters
Full scale Model Scale
Length between perpendiculars LPP [m] 230.0 7.55
Breadth B [m] 32.3 1.06
Mean draught T [m] 10.8 0.355
Block coefficient CB [-] 0.651 0.651
Vertical centre of gravity KG [m] 9.38 0.308
Pitch radius of gyration kyy [m/LPP] 0.226 0.226
Scale ratio λ – 30.45
Fig. 1 KCS Hull Form

Prior to the tests, the model was balanced statically and dynamically for one draught (T=10.8m in
full scale). The scope of tests comprise two speeds (12 and 20 knots), several regular waves with double
amplitude of 3.0m and two irregular waves with significant wave heights 3.0m and 6.0m. The full scale
waves spectra are pretended in Fig. 3.
4.1 Added resistance in regular waves – Quadratic Transfer Functions

The test were carried out in calm water and series of regular following waves in order to obtain
Quadratic Transfer Function of added resistance. The added resistance of subsequent wave is
calculated as a difference of measured mean longitudinal surge force in waves and referenced
resistance in still water. The QTF was calculated then by:
𝑅𝑅 (𝜔𝜔)−𝑅𝑅
𝑅𝑅𝐴𝐴𝐴𝐴 (𝜔𝜔) = 𝑇𝑇𝑇𝑇 2 𝑇𝑇0
𝜁𝜁
(14)
𝐴𝐴
where:
𝑅𝑅𝑇𝑇𝑇𝑇 – resistance in regular wave with frequency 𝜔𝜔
𝑅𝑅0 – reference resistance in calm water
𝜁𝜁𝐴𝐴 – amplitude of regular wave
The resultant QTFs from experiments in regular waves are presented in Fig. 2
7
99
Ship seakeeping
Fig. 2 Quadratic Transfer Functions of added resistance in regular waves
4.2 Added resistance in irregular waves – direct measurements

The irregular waves was determine by means of JONSWAP spectral density function of wave
energy with the statistical parameters such as significant wave height 𝐻𝐻𝑆𝑆 and peak period 𝑇𝑇𝑃𝑃 shown in
Table 2.
Table 2 Wave parameters
Wave No. Hs [m] Tp [s] γ [-]
1 3.00 11.00 3.3
2 6.00 13.00 3.3
Fig. 3 Irregular waves spectra

The added resistance in irregular waves was evaluated as a difference between measured longitudinal
surge force, averaged within one repeat period, and reference resistance in calm water. The irregular as
well as calm water forces were measured at the same model speed.
5. Comparison of measured and predicted forces
The first comparison of results is presented in Fig. 4. This is the QTF curves plotted against wave
frequency 𝜔𝜔 in the full scale. The circles on the plots represents experimental data whilst curves are the
results of computational analysis with the use of methods considered here.
8
100
Ship seakeeping
Fig. 4 Comparison of added resistance QTF functions of selected methods

As it can be seen in the plots the strip theory method overestimates forces significantly while the
potential code fits the measured curves only at the lowest and highest wave frequencies. The level of
forces in the middle of the considered wave frequencies is underestimated here. Only the semi-empirical
SNNM method gives reasonable level of discrepancies in relation to experimental data.
The second comparison was focused on final outcome from experiments in waves i.e. the predicted
resistance increase in considered sea states. The results of an analyses are presented in the Table 3 and
Fig. 5.
Table 3 Predictions of added resistance in following waves
Unit Wave1 Wave 2

Ship’s speed knots 12.0 20.0 12.0 20.0
Experiment – irregular wave kN 11.4 32.6 72.3 ------
Experiment – regular wave kN 17.2 34.5 79.5 160.7
Seakeeper – Havelock method kN 67.8 48.6 508.0 419.1
AQWA kN 7.8 13.5 34.5 52.1
SNNM kN 11.2 48.6 75.2 159.4
Fig. 5 Predictions of added resistance for two selected irregular waves
9
101
Ship seakeeping
6. Conclusions
The application of three computational approaches to the problem of added resistance in waves was
analysed in this paper. The first is the most common method based on strip theory with the second order
problem related mainly to the pitch and heave motions. Due to the fact that presented methods are mainly
developed for head waves the results of predictions in following waves are overestimated. The second
method reviewed here is the potential code based solver which computes the second order wave forces
via quadratic transfer functions, where the near filed solutions is used to find mean added resistance
(second order surge force). This method in comparison to experimental data underestimates predictions
in following waves. The last computational method is based on completely different kind of approach
to added resistance problem. It is developed basing on statistical database of measured ship’s resistance
in regular waves for wide variety of ship types and sizes. Predictions calculated from QTFs evaluated
by SNNM method agrees very well with model scale experiments for highest waves and 20 knots and
lower waves and 12 knots conditions. For the remaining considered conditions the discrepancies are
much higher but still at the acceptable level.
Fig. 6 Ratio of added resistance in waves to calm water resistance

This assessment took into account comparison of absolute values only. However if we look at the
added resistance in waves predicted by all methods used in this paper normalized by total resistance in
still water we can set the more general conclusion about all predictions. The added resistance in
following waves reaches up to 20% of resistance in calm water for the higher of considered sea
conditions and it is quite significant from the point of view of ship's operability.
7. Acknowledgements
Authors would like to especially thank to Dr Shukui Liu for the support in application of SNNM
method in analyses.
This work was fully funded by the Maritime Advance Centre (CTO S.A.) own research fund.
10
102
Ship seakeeping
8. References
[1] H. Maruo, „The Excess Resistance of a Ship in Rough Seas,” International Shipbuilding
Progress, Vol. 4, 1957.
[2] N. Salvesen, „Added resistance of ships in waves,” Journal of Hydronautics, No. 12, pp. 24-
34, 1978.
[3] J. Gerritsma i W. Beukelman, „Analysis of the resistance increase in waves of a fast cargo
ship,” International Shipbuilding Progress, 1972.
[4] T. Havelock, „Drifting Force on a Ship among Waves,” Philosophical Magazine, Vol. 33,
1942.
[5] S. Liu, A. Papanikolaou i P. Feng, „The semi-empirical SNNM formula for the
approximation of added resistance in waves,” w The 4th meeting of ITTC "Specialist Committee
on Operation of Ships at Sea, Daejon, Korea, 2020.
[6] S. Liu i A. Papanikolaou, „Fast Approach to the Estimation of the Added Resistance of
Ships,” Ocean Engineering, 112, p. 211–225, 2016.
[7] J. Newman, „The drift force and moment on ships in waves,” Journal of Ship Research,
Vol.11, 1967.
[8] J. Pinkster, Low frequency second order wave exciting forces on floating structures, Delft,
Netherland: T.U. Delft, 1980.
11
103
Ship seakeeping

HYDRONAV
Manoeuvring 2021
and Operation
Analysis of lifeboat emergency launching from a passenger ro-ro ferry

during its motions in stochastic waves
HYDRONAV 2021
A. JANKOWSKI1 Sopot-Iława, 19-21 May 2021
1 Polski Rejestr Statków, [email protected]
Analysis
According of LSA-Code
to the lifeboatcriterion
emergency launching
4.4.1.7 (resolution from
MSC.48(66)), eacha fully
passenger ro-ro
loaded lifeboat whenferry
launched should
withstand a lateral impact against the ship's side at an impact velocity of at least 3.5 m/s. Therefore, it is particularly important
during
to determineits motions
the impact velocityin stochastic
at which the launchedwaves
lifeboat with people on board hits the ship's side. The motion of launched
lifeboat, attached to davits which are moving along with the ro-ro ship and inducing the lifeboat motion, was modelled as a
pendulum
A. with uniformly
JANKOWSKI 1 increasing length of the ropes on which the boat is lowered. It was assumed that the ship motion is
induced by stochastic beam seas. The numerical method was used to solve the differential equations describing the ship's
1motions in waves in the first step, and then, the motions of the pendulum induced by ship motions, in the second step. The
Polski Rejestr Statków, [email protected]
numerical computations were performed for one loading condition and for four sea states. The purpose of this study is to
determine the impact velocities at which the launched fully loaded lifeboat hits the ro-ro ship's side in the assumed sea states.
According to the LSA-Code criterion 4.4.1.7 (resolution MSC.48(66)), each fully loaded lifeboat when launched should
Keywords: LSA-Code, lifeboat, impact velocity, differential equations, ship motion, pendulum
withstand a lateral impact against the ship's side at an impact velocity of at least 3.5 m/s. Therefore, it is particularly important
to determine the impact velocity at which the launched lifeboat with people on board hits the ship's side. The motion of launched
1. Introduction
lifeboat, attached to davits which are moving along with the ro-ro ship and inducing the lifeboat motion, was modelled as a
pendulum with uniformly increasing length of the ropes on which the boat is lowered. It was assumed that the ship motion is
The purpose
induced of the
by stochastic beamstudy
seas. presented is to
The numerical determine
method thetomotion
was used solve theofdifferential
the emergency
equationslifeboat launching
describing the ship's
motions in waves in the first step, and then, the motions of the pendulum induced by ship motions,
from the ship. The motion of the boat is induced by moving ship in waves. It is particularly important in the second step. The
numerical computations were performed for one loading condition and for four sea states. The purpose of this study is to
to determine the speed at which the lifeboat hits the ship's side.
determine the impact velocities at which the launched fully loaded lifeboat hits the ro-ro ship's side in the assumed sea states.
The physics of the problem considered is quite complex, therefore, simplifying assumptions need to
be introduced
Keywords: in order
LSA-Code, toimpact
lifeboat, develop
velocity,an differential
appropriate model
equations, shipdetermining the motion of the lifeboat. However,
motion, pendulum
the model developed (differential equation) should embrace essential elements of the physics of the
1. Introduction
launching lifeboat.
In the study, the vessel considered is Ro-Ro/Pass Ferry which operates in short international voyages
The1518
with purpose of the
persons study presented
on board. This vesseliscomplies
to determine with the motion Agreement
Stockholm of the emergency lifeboat
for Signiant Wave launching
Height
up to 4.00[m]. The vessel length between perpendiculars is 150[m], breadth is 25.60[m] and theimportant
from the ship. The motion of the boat is induced by moving ship in waves. It is particularly summer
to determine
draught the speed at which the lifeboat hits the ship's side.
is 5.50[m].
The physics
The maximum of the problem
loading considered
condition, is quite𝑇𝑇𝑇𝑇 complex,
for which = 5.50[𝑚𝑚𝑚𝑚]therefore,
and foursimplifying
stochastic assumptions
sea states, givenneed in
to
be introduced in order to develop an appropriate model determining the motion
Section 4, inducing the motion of launched emergency lifeboat, were taken into account in the study. of the lifeboat. However,
the model
The developed
significant (differential
wave height 𝐻𝐻𝐻𝐻𝑠𝑠𝑠𝑠 andequation)
the mean shouldzero-upembrace
crossingessential
period 𝑇𝑇𝑇𝑇𝑧𝑧𝑧𝑧elements
determine ofthe
thespectrum
physics of the
of the
launching lifeboat.
waves considered.
InThe
thelifeboats
study, thearevessel considered
installed at Deckis9 Ro-Ro/Pass Ferry which
at port and starboard side.operates
For emergencyin short lifeboat
international
launchingvoyages
the
with
gravity roller type davit is used. The lowering velocity of the lifeboat is taken to be 𝑉𝑉𝑉𝑉𝑦𝑦𝑦𝑦 = 1[ ⁄𝑠𝑠𝑠𝑠],Height
1518 persons on board. This vessel complies with Stockholm Agreement for Signiant Wave
𝑚𝑚𝑚𝑚 while
up to 4.00[m]. The vessel length between perpendiculars is 150[m], breadth is 25.60[m] and the summer
the distance between the Deck 9 and the wave surface is not greater than ℎ = 20[𝑚𝑚𝑚𝑚]. The distance
draught is 5.50[m].
results from the loading conditions.
The maximum loading condition, for which 𝑇𝑇𝑇𝑇 = 5.50[𝑚𝑚𝑚𝑚] and four stochastic sea states, given in
Section 4, inducing the motion of launched emergency lifeboat, were taken into account in the study.
The significant wave height 𝐻𝐻𝐻𝐻𝑠𝑠𝑠𝑠 and the mean zero-up crossing period 𝑇𝑇𝑇𝑇𝑧𝑧𝑧𝑧 determine the spectrum of the
waves considered.
The lifeboats are installed at Deck 9 at port and starboard side. For emergency lifeboat launching the
gravity roller type davit is used. The lowering velocity of the lifeboat is taken to be 𝑉𝑉𝑉𝑉𝑦𝑦𝑦𝑦 = 1[𝑚𝑚𝑚𝑚⁄𝑠𝑠𝑠𝑠], while
the distance between the Deck 9 and the wave surface is not greater than ℎ = 20[𝑚𝑚𝑚𝑚]. The distance
results from the loading conditions.
104
Ship seakeeping
2. Differential equation of motion of launching emergency lifeboat

In order to determine the motion of launched emergency lifeboat it is necessary to develop equations
of motion of davit crane in the first step and then the motion of launched and oscillating lifeboat (Fig.
1). To do this it necessary to determine accelerations 𝜉𝜉𝜉𝜉S̈ , 𝜉𝜉𝜉𝜉𝐻𝐻𝐻𝐻
̈ , 𝜉𝜉𝜉𝜉𝑅𝑅𝑅𝑅̈ , roll velocity 𝜉𝜉𝜉𝜉𝑅𝑅𝑅𝑅̇ and the roll angle 𝜉𝜉𝜉𝜉𝑅𝑅𝑅𝑅 .
Fig. 1. Coordinate system and symbols used to derive motion equations of launched emergency lifeboat
For the case where the vessel is at beam sea the following system of equations of ship motion in irregular
waves are used:
(𝑚𝑚𝑚𝑚 + 𝑚𝑚𝑚𝑚22 )𝑢𝑢𝑢𝑢𝑢2 − 𝑚𝑚𝑚𝑚𝜔𝜔𝜔𝜔1 𝑢𝑢𝑢𝑢2 + 𝑚𝑚𝑚𝑚24 𝜔𝜔𝜔𝜔𝜔1 − 𝑚𝑚𝑚𝑚33 𝑢𝑢𝑢𝑢3 𝜔𝜔𝜔𝜔1 = 𝐹𝐹𝐹𝐹𝐹𝐹𝐹𝐹𝐹𝐹𝐹𝐹2 + 𝐹𝐹𝐹𝐹𝐷𝐷𝐷𝐷2 + 𝐹𝐹𝐹𝐹𝑅𝑅𝑅𝑅2 + 𝑚𝑚𝑚𝑚𝑚𝑚𝑚𝑚 sin 𝜃𝜃𝜃𝜃
(𝑚𝑚𝑚𝑚 + 𝑚𝑚𝑚𝑚33 )𝑢𝑢𝑢𝑢𝑢3 + 𝑚𝑚𝑚𝑚𝑚𝑚𝑚𝑚1 𝑢𝑢𝑢𝑢2 − 𝑚𝑚𝑚𝑚22 𝜔𝜔𝜔𝜔1 𝑢𝑢𝑢𝑢2 + 𝑚𝑚𝑚𝑚24 𝜔𝜔𝜔𝜔12 − 𝑚𝑚𝑚𝑚15 𝜔𝜔𝜔𝜔12 = 𝐹𝐹𝐹𝐹𝐹𝐹𝐹𝐹𝐹𝐹𝐹𝐹3 + 𝐹𝐹𝐹𝐹𝐷𝐷𝐷𝐷𝐷 + 𝐹𝐹𝐹𝐹𝑅𝑅𝑅𝑅𝑅 − 𝑚𝑚𝑚𝑚𝑚𝑚𝑚𝑚 cos 𝛷𝛷𝛷𝛷
(𝐽𝐽𝐽𝐽1 + 𝑚𝑚𝑚𝑚44 )𝜔𝜔𝜔𝜔̇1 + 𝑚𝑚𝑚𝑚42 𝑢𝑢𝑢𝑢̇2 = 𝐹𝐹𝐹𝐹𝐹𝐹𝐹𝐹𝐹𝐹𝐹𝐹4 + 𝐹𝐹𝐹𝐹𝐷𝐷𝐷𝐷𝐷 + 𝐹𝐹𝐹𝐹𝑅𝑅𝑅𝑅𝑅 (1)
𝜉𝜉𝜉𝜉𝑆𝑆𝑆𝑆̇ = 𝑢𝑢𝑢𝑢2 + 𝜔𝜔𝜔𝜔1 𝜉𝜉𝜉𝜉𝑆𝑆𝑆𝑆
̇ = 𝑢𝑢𝑢𝑢3 − 𝜔𝜔𝜔𝜔1 𝜉𝜉𝜉𝜉𝐻𝐻𝐻𝐻
𝜉𝜉𝜉𝜉𝐻𝐻𝐻𝐻
𝜉𝜉𝜉𝜉𝑅𝑅𝑅𝑅̇ = 𝜔𝜔𝜔𝜔1
where 𝑚𝑚𝑚𝑚 is the mass of the ship, 𝑚𝑚𝑚𝑚22 , 𝑚𝑚𝑚𝑚33, 𝑚𝑚𝑚𝑚24 , 𝑚𝑚𝑚𝑚44 are the added masses of water, 𝑢𝑢𝑢𝑢2 , 𝑢𝑢𝑢𝑢3 are the
velocities of the motion of the ships center of mass, 𝜔𝜔𝜔𝜔1 is the angular velocity around the 𝑥𝑥𝑥𝑥 axis, 𝐽𝐽𝐽𝐽 is the
moment of mass inertia about the 𝑥𝑥𝑥𝑥 axis, 𝜉𝜉𝜉𝜉𝑆𝑆𝑆𝑆 , 𝜉𝜉𝜉𝜉𝐻𝐻𝐻𝐻 , 𝜉𝜉𝜉𝜉𝑅𝑅𝑅𝑅 are the ship's sway, heave and roll motions, 𝐹𝐹𝐹𝐹𝐹𝐹𝐹𝐹𝐹𝐹𝐹𝐹𝐹𝐹𝐹𝐹 , 𝐹𝐹𝐹𝐹𝐷𝐷𝐷𝐷𝐹𝐹𝐹𝐹
and 𝐹𝐹𝐹𝐹𝑅𝑅𝑅𝑅𝐹𝐹𝐹𝐹 respectively are Froude-Krylov, diffraction and radiation forces and 𝑚𝑚𝑚𝑚 is gravitational
acceleration. Equations (1) determine the necessary accelerations, velocities and displacement: 𝜉𝜉𝜉𝜉S̈ , 𝜉𝜉𝜉𝜉𝐻𝐻𝐻𝐻 ̈ ,
𝜉𝜉𝜉𝜉𝑅𝑅𝑅𝑅̈ , 𝜉𝜉𝜉𝜉𝑅𝑅𝑅𝑅̇ , 𝜉𝜉𝜉𝜉𝑅𝑅𝑅𝑅 , used to describe the launched emergency lifeboat motions.
Fig. 2a. Ship accelerations: sway 𝜉𝜉𝜉𝜉𝑆𝑆𝑆𝑆̈ , heave 𝜉𝜉𝜉𝜉𝐻𝐻𝐻𝐻

̈ and roll 𝜉𝜉𝜉𝜉𝑅𝑅𝑅𝑅̈ for sea state 𝐻𝐻𝐻𝐻𝑠𝑠𝑠𝑠 = 2.5[𝑚𝑚𝑚𝑚] , 𝑇𝑇𝑇𝑇𝑧𝑧𝑧𝑧 = 5.5[𝑠𝑠𝑠𝑠]
105
Ship seakeeping
Fig. 2b. Roll velocity 𝜉𝜉𝜉𝜉𝑅𝑅𝑅𝑅̇ and roll angle 𝜉𝜉𝜉𝜉𝑅𝑅𝑅𝑅
In turn, the launched emergency lifeboat motions is approximated by pendulum with uniformly
increasing length of the rope. Then:
- applying the Euler-Lagrange equation to the energy of rolling ship inducing the motion of the
pendulum (launching emergency lifeboat);
- neglecting small values of ship rotation: sin 𝜉𝜉𝜉𝜉𝑅𝑅𝑅𝑅 ≈ 𝜉𝜉𝜉𝜉𝑅𝑅𝑅𝑅 , cos 𝜉𝜉𝜉𝜉𝑅𝑅𝑅𝑅 ≈ 1,
the oscillator problem is obtained.
So, the motion of launched emergency lifeboat, attached to davits which are moving along with the
ro-ro vessel and inducing the lifeboat motion, was modelled as a pendulum with uniformly increasing
length of the ropes on which the boat is lowered.
The motion of two bodies (ship and boat) is described by the following equations (see Fig.1):
𝑦𝑦𝑦𝑦 (𝑡𝑡𝑡𝑡) = 𝑦𝑦𝑦𝑦1 (𝑡𝑡𝑡𝑡) + 𝑙𝑙𝑙𝑙(𝑡𝑡𝑡𝑡) sin 𝜑𝜑𝜑𝜑(𝑡𝑡𝑡𝑡)
� 2 (2)
𝑧𝑧𝑧𝑧2 (𝑡𝑡𝑡𝑡) = 𝑧𝑧𝑧𝑧1 (𝑡𝑡𝑡𝑡) − 𝑙𝑙𝑙𝑙(𝑡𝑡𝑡𝑡) cos 𝜑𝜑𝜑𝜑(𝑡𝑡𝑡𝑡)
where 𝑙𝑙𝑙𝑙(𝑡𝑡𝑡𝑡) = 𝑣𝑣𝑣𝑣𝑦𝑦𝑦𝑦 𝑡𝑡𝑡𝑡 (𝑣𝑣𝑣𝑣𝑦𝑦𝑦𝑦 - is the lowering velocity of lifeboat) and
𝜉𝜉𝜉𝜉𝑅𝑅𝑅𝑅 (𝑡𝑡𝑡𝑡) + 2𝛼𝛼𝛼𝛼
⎧𝑦𝑦𝑦𝑦1 (𝑡𝑡𝑡𝑡) = 𝑟𝑟𝑟𝑟𝑟𝑟𝑟𝑟𝑅𝑅𝑅𝑅 (𝑡𝑡𝑡𝑡) cos � � + 𝜉𝜉𝜉𝜉𝑆𝑆𝑆𝑆 (𝑡𝑡𝑡𝑡)
⎪ 2
, (3)
⎨ 𝜉𝜉𝜉𝜉𝑅𝑅𝑅𝑅 (𝑡𝑡𝑡𝑡) + 2𝛼𝛼𝛼𝛼
⎪ 𝑧𝑧𝑧𝑧1 (𝑡𝑡𝑡𝑡) = 𝑟𝑟𝑟𝑟𝑟𝑟𝑟𝑟𝑅𝑅𝑅𝑅 (𝑡𝑡𝑡𝑡) sin � � + 𝜉𝜉𝜉𝜉𝐻𝐻𝐻𝐻 (𝑡𝑡𝑡𝑡)
⎩ 2
where 𝑦𝑦𝑦𝑦1 and 𝑧𝑧𝑧𝑧1 describes the motion of davit crane.
For the sea states in which ship operates 𝜉𝜉𝜉𝜉𝑅𝑅𝑅𝑅 (𝑡𝑡𝑡𝑡) are small, therefore, it can be assumed that sin 𝜉𝜉𝜉𝜉𝑅𝑅𝑅𝑅 ≈
𝜉𝜉𝜉𝜉𝑅𝑅𝑅𝑅 , cos 𝜉𝜉𝜉𝜉𝑅𝑅𝑅𝑅 ≈ 1. Then equations (3) takes the following form
1
𝑦𝑦𝑦𝑦1 (𝑡𝑡𝑡𝑡) = 𝑟𝑟𝑟𝑟𝑟𝑟𝑟𝑟𝑅𝑅𝑅𝑅 (𝑡𝑡𝑡𝑡) cos 𝛼𝛼𝛼𝛼 𝛼 𝑟𝑟𝑟𝑟𝑟𝑟𝑟𝑟𝑅𝑅𝑅𝑅 2 (𝑡𝑡𝑡𝑡) sin 𝛼𝛼𝛼𝛼 + 𝜉𝜉𝜉𝜉𝑆𝑆𝑆𝑆
� 2 . (4)
1 2
𝑧𝑧𝑧𝑧1 (𝑡𝑡𝑡𝑡) = 𝑟𝑟𝑟𝑟𝑟𝑟𝑟𝑟𝑅𝑅𝑅𝑅 (𝑡𝑡𝑡𝑡) cos 𝛼𝛼𝛼𝛼 + 𝑟𝑟𝑟𝑟𝑟𝑟𝑟𝑟𝑅𝑅𝑅𝑅 (𝑡𝑡𝑡𝑡) sin 𝛼𝛼𝛼𝛼 + 𝜉𝜉𝜉𝜉𝐻𝐻𝐻𝐻
2
Differentiating (4) in time, the velocity components of mass of lifeboat is obtained:
𝑦𝑦𝑦𝑦𝑦2 = 𝑦𝑦𝑦𝑦𝑦1 + 𝑣𝑣𝑣𝑣𝑦𝑦𝑦𝑦 sin(𝜑𝜑𝜑𝜑) + 𝑙𝑙𝑙𝑙 cos(𝜑𝜑𝜑𝜑) 𝜑𝜑𝜑𝜑𝜑
� . (5)
𝑧𝑧𝑧𝑧2𝑧 = 𝑧𝑧𝑧𝑧1𝑧 − 𝑣𝑣𝑣𝑣𝑦𝑦𝑦𝑦 cos(𝜑𝜑𝜑𝜑) + 𝑙𝑙𝑙𝑙 sin(𝜑𝜑𝜑𝜑) 𝜑𝜑𝜑𝜑𝜑
The equation of rotation of angles 𝜑𝜑𝜑𝜑 is derived from Lagrange’s equation:
𝑑𝑑𝑑𝑑 𝜕𝜕𝜕𝜕𝜕𝜕𝜕𝜕𝑘𝑘𝑘𝑘 𝜕𝜕𝜕𝜕𝜕𝜕𝜕𝜕𝑘𝑘𝑘𝑘 𝜕𝜕𝜕𝜕𝜕𝜕𝜕𝜕𝑝𝑝𝑝𝑝
� �− + =0 (6)
𝑑𝑑𝑑𝑑𝑡𝑡𝑡𝑡 𝜕𝜕𝜕𝜕𝜕𝜕𝜕𝜕𝜕 𝜕𝜕𝜕𝜕𝜑𝜑𝜑𝜑 𝜕𝜕𝜕𝜕𝜑𝜑𝜑𝜑
where 𝐸𝐸𝐸𝐸𝑘𝑘𝑘𝑘 is kinetic energy of considered system and 𝐸𝐸𝐸𝐸𝑝𝑝𝑝𝑝 is potential energy of the system. The kinetic
energy is as follows:
106
Ship seakeeping
1
𝐸𝐸𝐸𝐸𝑘𝑘𝑘𝑘 = 𝑚𝑚𝑚𝑚2 (𝑦𝑦𝑦𝑦̇22 + 𝑧𝑧𝑧𝑧2̇ 2 ) (7)
2
1
= 𝑚𝑚𝑚𝑚2 (𝑦𝑦𝑦𝑦̇12 + 𝑧𝑧𝑧𝑧1̇ 2 + 2𝑣𝑣𝑣𝑣𝑧𝑧𝑧𝑧 (𝑦𝑦𝑦𝑦̇1 sin 𝜑𝜑𝜑𝜑 − 𝑧𝑧𝑧𝑧1̇ cos 𝜑𝜑𝜑𝜑) + 2𝑙𝑙𝑙𝑙𝜑𝜑𝜑𝜑̇(𝑦𝑦𝑦𝑦̇1 cos 𝜑𝜑𝜑𝜑 + 𝑧𝑧𝑧𝑧1̇ sin 𝜑𝜑𝜑𝜑) + 𝑣𝑣𝑣𝑣𝑧𝑧𝑧𝑧2 + 𝑙𝑙𝑙𝑙 2 𝜑𝜑𝜑𝜑̇2 )
2
and the potential energy:
𝐸𝐸𝐸𝐸𝑝𝑝𝑝𝑝 = 𝑚𝑚𝑚𝑚2 𝑚𝑚𝑚𝑚𝑧𝑧𝑧𝑧2 (8)
Substituting 𝐸𝐸𝐸𝐸𝑘𝑘𝑘𝑘 and 𝐸𝐸𝐸𝐸𝑝𝑝𝑝𝑝 to (6) and performing appropriate differentiation, the following differential
equation if of lifeboat motion is obtained
𝑙𝑙𝑙𝑙 2 𝜑𝜑𝜑𝜑̈ + 𝑙𝑙𝑙𝑙𝑣𝑣𝑣𝑣𝑦𝑦𝑦𝑦 𝜑𝜑𝜑𝜑̇ + 𝑙𝑙𝑙𝑙(𝑦𝑦𝑦𝑦̈1 cos 𝜑𝜑𝜑𝜑 + 𝑧𝑧𝑧𝑧1̈ sin 𝜑𝜑𝜑𝜑) + 𝑚𝑚𝑚𝑚𝑚𝑚𝑚𝑚 sin 𝜑𝜑𝜑𝜑 = 0 (9)
Because rotation angles are small, (9) is reduced to damped harmonic, nonhomogeneous oscillator:
𝑣𝑣𝑣𝑣𝑦𝑦𝑦𝑦 1 1
𝜑𝜑𝜑𝜑̈ + 𝜑𝜑𝜑𝜑̇ + (𝑧𝑧𝑧𝑧1̈ + 𝑚𝑚𝑚𝑚)𝜑𝜑𝜑𝜑 = − 𝑦𝑦𝑦𝑦̈1 (10)
𝑙𝑙𝑙𝑙 𝑙𝑙𝑙𝑙 𝑙𝑙𝑙𝑙
where
𝑦𝑦𝑦𝑦̈1 = 𝑟𝑟𝑟𝑟𝜉𝜉𝜉𝜉𝑅𝑅𝑅𝑅̈ �cos 𝛼𝛼𝛼𝛼 − (𝜉𝜉𝜉𝜉𝑅𝑅𝑅𝑅̇ 2 + 𝜉𝜉𝜉𝜉𝑅𝑅𝑅𝑅 ) sin 𝛼𝛼𝛼𝛼� + 𝜉𝜉𝜉𝜉𝑆𝑆𝑆𝑆̈

� (11)
𝑧𝑧𝑧𝑧1̈ = 𝑟𝑟𝑟𝑟𝜉𝜉𝜉𝜉𝑅𝑅𝑅𝑅̈ (sin 𝛼𝛼𝛼𝛼 + (𝜉𝜉𝜉𝜉𝑅𝑅𝑅𝑅̇ 2 + 𝜉𝜉𝜉𝜉𝑅𝑅𝑅𝑅 ) cos 𝛼𝛼𝛼𝛼) + 𝜉𝜉𝜉𝜉𝐻𝐻𝐻𝐻
̈
are the accelerations of davit crane.
3. Numerical solution of motion equation

The fourth-order Runge-Kutta method is used to solve numerically the second order equation of
lifeboat motion transformed to system of two ordinary differential equations (ODE). The 1st-order
ODEs:
𝑑𝑑𝑑𝑑
𝜑𝜑𝜑𝜑 = 𝜒𝜒𝜒𝜒 = 𝑓𝑓𝑓𝑓(𝑡𝑡𝑡𝑡, 𝜑𝜑𝜑𝜑, 𝜒𝜒𝜒𝜒) (12)
𝑑𝑑𝑑𝑑𝑡𝑡𝑡𝑡
can be converted into a system of two ODEs by using the following variable substitution:
𝑑𝑑𝑑𝑑 𝑑𝑑𝑑𝑑2 𝑑𝑑𝑑𝑑

𝜑𝜑𝜑𝜑 = 𝜒𝜒𝜒𝜒, 𝜑𝜑𝜑𝜑 = 𝜒𝜒𝜒𝜒 (13)
𝑑𝑑𝑑𝑑𝑡𝑡𝑡𝑡 𝑑𝑑𝑑𝑑𝑡𝑡𝑡𝑡 2 𝑑𝑑𝑑𝑑𝑡𝑡𝑡𝑡
with initial conditions 𝜑𝜑𝜑𝜑(0) = 0 and 𝜒𝜒𝜒𝜒(0) = 0. Substituting (13) to (10) the following system of ODE
of first order is obtained:
𝑑𝑑𝑑𝑑𝜑𝜑𝜑𝜑
= 𝜒𝜒𝜒𝜒
� 𝑑𝑑𝑑𝑑𝑡𝑡𝑡𝑡 (14)
𝑑𝑑𝑑𝑑𝜒𝜒𝜒𝜒 1 1 𝑣𝑣𝑣𝑣𝑦𝑦𝑦𝑦
= − 𝑥𝑥𝑥𝑥̈1 − (𝑦𝑦𝑦𝑦̈1 + 𝑚𝑚𝑚𝑚)𝜑𝜑𝜑𝜑 − 𝜒𝜒𝜒𝜒
𝑑𝑑𝑑𝑑𝑡𝑡𝑡𝑡 𝑙𝑙𝑙𝑙 𝑙𝑙𝑙𝑙 𝑙𝑙𝑙𝑙
The system of equations is solved in interval 0 ≤ 𝑡𝑡𝑡𝑡[𝑠𝑠𝑠𝑠] ≤ 𝑡𝑡𝑡𝑡𝑘𝑘𝑘𝑘 , 𝑡𝑡𝑡𝑡𝑘𝑘𝑘𝑘 = 20[𝑠𝑠𝑠𝑠] – is the lowering time of
lifeboat. The solution of equation (14) is the angular displacement 𝜑𝜑𝜑𝜑 and angular velocity 𝜑𝜑𝜑𝜑̇.
4. Results of computations of lifeboat motion

The figures in the Appendix present the results of the calculation for
- maximum loading condition;
- following four sea steady-states:
107
Ship seakeeping
- 𝐻𝐻𝐻𝐻𝑠𝑠𝑠𝑠 = 1.5[𝑚𝑚𝑚𝑚] , 𝑇𝑇𝑇𝑇𝑧𝑧𝑧𝑧 = 4.5[𝑠𝑠𝑠𝑠],

where 𝐻𝐻𝐻𝐻𝑠𝑠𝑠𝑠 is the significant wave height and 𝑇𝑇𝑇𝑇𝑧𝑧𝑧𝑧 is the zero up-crossing period.
An example of numerical solution of equation (14) is presented in Fig. 3. Other results of calculations
are presented in the Annexes.
Fig. 3. Horizontal displacement (red line) and velocity (green line) of a lifeboat (𝐻𝐻𝐻𝐻𝑠𝑠𝑠𝑠 = 1.5[𝑚𝑚𝑚𝑚], 𝑇𝑇𝑇𝑇𝑧𝑧𝑧𝑧 = 4.5[𝑠𝑠𝑠𝑠])
5. Conclusions
This paper presents mathematical model (differential equations) used to determine the motion of
the emergency lifeboat launched from the ship. The motion of a the lifeboat is induced by moving ship
in waves.
The motion of the lifeboat is complex, therefore, simplifying assumptions were introduced in order
to develop appropriate model. Despite the simplifications introduced, the essence of the phenomenon
has been maintained because the ferry operates in low sea states.
According to the criterion of impact velocity (LSA-Code (MSC.48(66))), each launched lifeboat
should withstand, when loaded with its full complement of persons and equipment, a lateral impact
against the ship's side at an impact velocity of at least 3.5[m/s]. According to the computations
performed, this criterion is fulfilled for sea states up to 𝐻𝐻𝐻𝐻𝑠𝑠𝑠𝑠 = 3.5[𝑚𝑚𝑚𝑚] , 𝑇𝑇𝑇𝑇𝑧𝑧𝑧𝑧 = 5.5[𝑠𝑠𝑠𝑠] which means that
the ferry can be operated up to such sea states.
6. Reference
[1] Jankowski J.: Ship facing the waves (in Polish), Polski Rejestr Statków, Gdańsk, 2007.
[2] Kowalski J.: A set of problems in mechanics with application to the calculation of machine elements (in
Polish). PWN, Warszawa, 1976.
[3] Jankowski A.: Analysis of lifeboat motions during its emergency launching from passenger ship Ro-
Ro/Pass Ferry rolling in waves, Polski Rejestr Statków, Gdańsk, 2019
108
Ship seakeeping
7. Appendix
109
Ship seakeeping
110
Ship seakeeping
111
Offshore structures.
Dynamics of the marine
environment.
Hydroacoustics.
112
Impact of steep waves underneath a horizontal deck

Impact of steep waves underneath a horizontal deck
By
and Wojciech Sulisz
Dawid Majewski1By
1
Institute of Hydro-engineering, Polish
Dawid Academy
Majewski1ofandSciences,
WojciechKoscierska
Sulisz 7, 80328 Gdańsk, Poland,
[email protected]
1
Institute of Hydro-engineering, Polish Academy of Sciences, Koscierska 7, 80328 Gdańsk, Poland,
[email protected]
A theoretical approach was applied to investigate the impact of nonlinear waves underneath a horizontal deck. A
solution of an initial boundary-value problem was achieved by the Boundary Element Method. Numerical
simulations
A theoreticalwere carriedwas
approach outapplied
to predict impact pressure
to investigate underneath
the impact a deck.waves
of nonlinear The results show athat
underneath the wavedeck.
horizontal impact
A
is a very of
solution complex momentary
an initial process.
boundary-value An impact
problem was pressure
achievedincreases in a strongly
by the Boundary nonlinear
Element Method.manner with
Numerical
increasing a were
simulations wavecarried
height out
eventofor a linear
predict or weakly
impact pressurenonlinear wavea attack.
underneath Moreover,
deck. The with decreasing
results show that the wave distance
impact
of athe
is deck
very from the
complex still waterprocess.
momentary level, anAnimpact
impact pressure
pressure increases.
increasesThe in aanalysis
stronglyshows that purely
nonlinear mannersinusoidal
with
waves of very
increasing small
a wave amplitude
height may
even for cause or
a linear an weakly
impact pressure
nonlinearseveral orders Moreover,
wave attack. of magnitudewithhigher than a distance
decreasing pressure
arising fromfrom
of the deck typical
theapplications of wave
still water level, theories.
an impact Moreover,
pressure the results
increases. show that
The analysis a wavelength
shows that purelyhas a low effect
sinusoidal
on
wavesan impact
of verypressure despite strongly
small amplitude may cause dependence
an impactofpressure
a velocity on a wavelength.
several Additionally,
orders of magnitude higherthe
thanstudy shows
a pressure
that thefrom
arising nonlinear
typicalcorrection of velocity
applications of wave has practically
theories. no effect
Moreover, theonresults
an impact
showpressure what is a non-intuitive
that a wavelength has a low effect
result
on becausepressure
an impact this correction
despite substantially changes of
strongly dependence a velocity onon
a velocity thea free surface and
wavelength. is expectedthe
Additionally, to study
affect shows
an
impact
that thepressure.
nonlinearThe analysisofshows
correction thathas
velocity all practically
these non-intuitive
no effectoutcomes
on an impactarisepressure
from thewhat
complexity of a wave
is a non-intuitive
impactbecause
result processthis
andcorrection
its enormous sensitivitychanges
substantially to initiala conditions
velocity onwhat indicates
the free surfaceserious
and isdifficulties
expected toinaffect
a reliable
an
prediction
impact of a wave
pressure. The impact
analysisfor complex
shows that wave fields.
all these Laboratory
non-intuitive experiments
outcomes arisewere
fromconducted to verify
the complexity of a wave
theoretical results.
impact process andAitscomparison shows thattotheoretical
enormous sensitivity results are
initial conditions whatin reasonable agreement
indicates serious with experimental
difficulties in a reliable
data.
prediction of a wave impact for complex wave fields. Laboratory experiments were conducted to verify
theoretical results. A comparison shows that theoretical results are in reasonable agreement with experimental
data.
1
Inst. of Hydroengineering, Polish Acadademy of Sciences, Koscierska 7, 80328 Gdańsk, Poland ([email protected])
1
Inst. of Hydroengineering, Polish Acadademy of Sciences, Koscierska 7, 80328 Gdańsk, Poland ([email protected])
113
Numerical modeling results of harbor wave penetration versus in-situ measurements
T. Marcinkowski, P. Szmytkiewicz
Abstract
The aim of the research is to elaborate a methodology of analyzing both measured and
computed wave characteristics for seaports entered by waves simultaneously coming from
multiple sectors.
The study was done for the Northern Harbor in Gdańsk, which is located in the Gulf of
Gdańsk. This harbor is not fully screened from waves by offshore breakwaters and local wind
waves may propagate simultaneously from two slightly different directions.
The presented study deals with the investigation of the sea surface wave phenomena and
particularly its spectral analysis in terms of frequency as well as directionality. One of the
principal objectives was to introduce the tools utilized to analyze the directionality of wave
time series.
Results of long-term experience in numerical modeling of sea surface dynamics, dealing with
the calculation of wave fields transformation in the coastal zone and particularly in the harbor
areas, indicated a necessity of more accurate consideration of the wave spectrum
characteristics limited mostly to only one-peaked frequency distribution.
Nowadays, in the wind-wave numerical modeling, one strive to provide the most reliable
character of the wave spectrum with particular consideration of its directionality. Without the
mentioned above aspects, the numerical models are not able to interpret the physical
phenomena in proper way. In practice, it means that the user of the sea waves computational
tools is forced to determine not only the mean directions of the dominant wave frequencies,
but also its ranges of spreads.
The conclusions obtained based on the performed analysis of the measured wave data let to
extend the knowledge in the area of investigation and to formulate preliminary
recommendations regarding to the parameters describing the directional wave spectra
distribution.
114
Magdalena Kempa
MODELING THE DYNAMICS OF THE MARINE ENVIRONMENT FOR THE DESIGN OF OFFSHORE
STRUCTURES.
will be delivered
115
Paweł Dymarski, Ewelina Ciba
MODELING THE DYNAMICS OF A FLOATING WIND TURBINE IN THE SOUTHERN BALTIC SEA AREA
WITH A DEPTH OF 84 M. PRELIMINARY ANALYZES.
will be delivered
116
Katarzyna Warnke-Olewniczak
ANALYSIS OF THE RECORDED UNDERWATER HYDROACOUSTIC SIGNAL OF A CARGO SHIP.
will be delivered
117
Contact:
[email protected]
118

E Book Hydronav2021

Uploaded by

Copyright:

Available Formats

E Book Hydronav2021

Uploaded by

Document Information

Original Description:

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

E Book Hydronav2021

Uploaded by

Copyright:

Available Formats

1

Resistance and Propulsion

Advanced computational and experimental modeling

Quasi-real verification of autonomous navigation. Zbigniew Pietrzykowski. Piotr Wołejsza ........................................................ 65

Simulation of fishing vessel motion for seakeeping analysis.

Offshore structures. Dynamics of the marine environment. Hydroacoustics.

Resistance and Propulsion

22nd International Symposium on Hydrodynamics

HYDRODYNAMIC STATE OF ART REVIEW: ROTOR – STATOR

HYDRODYNAMIC STATE OF ART REVIEW: ROTOR – STATOR

2. Design methods for stator type Energy Saving Devices

3. Energy saving effect

Table 6 Propulsor operation in uniform flow, inviscid fluid

Table 10 Particulars of analyzed propeller and its operation

Table 15 Model test results with subsequent ESDs:

Table 17 Model test results KCS

Table 18 Full scale propulsion prognosis for KVLLC2

Table 19 Full scale propulsion prognosis for KCS

Table 20 Particulars of analyzed vessel

Table 21 Model test results

Table 22 Full scale prediction (WAFon)

Table 23 Full scale prediction (WAFon-D)

Table 28 Model test results (Design draught)

Table 29 Model test results (Ballast draught)

Table 35 Particulars of the propellers

Model tests were conducted for four configurations:

Table 37 Model test results

Table 38 Propeller loading conditions (model scale)

Table 39 Full scale prognosis

Table 40 Propeller loading conditions (full scale)

Table 42 Model test results (Full load)

Table 43 Model test results (Ballast condition)

Table 44 Particulars of 180m Diamond 34 Bulk Carrier and its propeller

Table 47 FVM simulation results for rudder ORIG

Figure 1 Propeller (“No ESD” conditions) versus declared efficiency increase

Resistance and Propulsion

PLASTEX Sprint Boats - Research Investigations 1998-2020

Figure 3 K1, K2 and K4 canoes – free surface elevation

2. PLASTEX Sprint Boats -Original research and design achievements

3. Hydrodynamics aspects of sprint boats design

1. Brief outline of origin of Ice Class Rules

3. Polar Class (PC)

PC1 Year round operation in all Polar Waters

6. Propulsion systems in ice

Ilawa Ship Handling Research and Training Centre:

The Foundation for Safety of Navigation and Environment Protection

Fig.2 Damage ship drift model tests

Currently, the Foundation for Safety of Navigation and Environment Protection is a

Numerical assessment of hull hydrodynamic derivatives in ship

22nd International Symposium

Influence of propulsion-steering system on the position of pivot point on a

Influence of propulsion-steering system on the position of pivot point on a

2.1 Investigated ship

Fig. 1. Twin-propeller twin-rudder version Fig. 2. Twin-pod version

2.1 Model test details

Zig-zag (°) 10/10, 20/20 10/10, 15/15, 20/20

2.2 IMO standard manoeuvres procedure

3. Model test results