Return to search

Scale model validation of QUAYSIM and WAVESCAT numerical models of ship motions

Thesis (MSc)--Stellenbosch University, 2015. / ENGLISH ABSTRACT: Various numerical modelling software packages are available for predicting moored ship motions
and forces. The focus of this study was to validate the numerical models QUAYSIM and
WAVESCAT and how these models together form a procedure for predicting moored ship motions
and forces under the impact of high and low frequency waves.
The validation procedure applied in the study involved numerical modelling of a given physical
model situation in which moored ship motions and forces were measured under both high and low
frequency wave conditions. A physical model with built-in bathymetry was provided by the Council
for Scientific and Industrial Research (CSIR) Hydraulics Laboratory in Stellenbosch. The model
consisted of a moored container vessel at a jetty, with various mooring lines and fenders. A
JONSWAP spectrum, which combines high and low frequency wave components, was used to
simulate wave conditions for the modelling of ship motions. The wave periods and wave heights
were measured at observation stations located at specific points in the basin. Other measurements
such as those of the forces in the fenders and mooring lines were also determined.
A multi-step approach was used to numerically predict the ship motions and forces. Firstly, the
coastal processes occurring within the basin, which was set up to simulate the physical model
wave behaviour, were measured to calibrate the SWAN Delft3D-WAVE model. The wave heights
and periods for the respective observation stations were obtained and compared to the physical
model measurements. The Delft3D-FLOW SURFBEAT model was used to calculate the low
frequency waves in the coastal area. Low frequency waves are the main cause of larger ship
motions and forces, therefore it is important to investigate them as part of the ship motion
prediction procedure.
After the waves had been computed, wave forces acting on the vessel needed to be determined
for both high and low frequency waves. These wave forces were modelled with the combinations
SURFBEAT/LF-STRIP (low frequency waves) and SWAN/WAVESCAT (high frequency waves).
LF-STRIP provided the link between low frequency wave models and ship motion models,
converting the low frequency waves into long wave forces acting on the vessel. WAVESCAT
converted the high frequency waves to short wave forces. The calculated long wave forces and
short wave forces served as the input required to run the ship motion model QUAYSIM to
determine the movements of the moored ship as well as the restraining forces in the lines and
fenders. The ship motions and forces were compared to the physical model, with the intention of possibly validating the QUAYSIM/WAVESCAT approach for predicting moored ship motions.
The study provides an overview of both the setup and results of the physical and numerical model.
A description of each of the numerical models SWAN, SURFBEAT, LF-STRIP, WAVESCAT and QUAYSIM is provided, along with a comparison between the physical and numerical models for
each procedure. The validation procedure provided useful documentation of the quality of these
numerical modelling approaches, already in use in some design projects.
The numerical models WAVESCAT and QUAYSIM models of ship motion have shown to provide a
good correlation between the physical model and the numerical approach. However, improvements
are still required. Good comparisons were obtained for the long wave motions (horizontal
movements - surge, sway and yaw). The surge and sway motions were slightly overestimated by
QUAYSIM. The magnitude of the yaw was comparable but the not well represented in spectral plots. / AFRIKAANSE OPSOMMING: Daar is verskeie numeriese modellering-sagtewareprogramme beskikbaar waarmee
skipbewegings en -kragte voorspel kan word. Die fokus van hierdie studie was om die numeriese
modelle QUAYSIM en WAVESCAT te valideer. Saam vorm hierdie twee modelle ’n prosedure om
vasgemeerde skipbewegings en -kragte veroorsaak deur lang- en kortgolfaksie te bepaal.
Die validasieprosedure wat in hierdie studie gebruik is, behels ’n numeriese modelering van ’n
fisiese situasie waar ’n vasgemeerde skip se bewegings en kragte onder kort- en langgolfkondisies
gemeet is. ’n Fisiese model met ingeboude batimetrie is voorsien deur die Council for Scientific
and Industrial Research (CSIR) se hidroliese laboratorium in Stellenbosch. Die model bestaan uit
’n vasgemeerde houerskip by ’n pier met verskeie ankerlyne en bootbuffers. ’n JONSWAPspektrum,
wat kort- en langgolfkomponente kombineer, is gebruik om golfomstandighede vir die
modellering van skipbewegings te simuleer. Golfperiodes en golfhoogtes is by spesifieke
waarnemingstasies in die gesimuleerde hawe-area gemeet. Verdere opmetings, soos dié van die
kragte in die bootbuffers en ankerlyne, is ook gedoen.
’n Stap-vir-stap benadering is gevolg om die skipbewegings numeries te voorspel. Eerstens is die
kusprosesse wat in die gesimuleerde hawe plaasvind, gekalibreer met die numeriese paket SWAN
Delft3D-WAVE. Die golfhoogtes en golfperiodes vir elke waarnemingstasie is bereken en vergelyk
met die fisiese model se opmetings. Die SURFBEAT-module van Delft3D-FLOW is gebruik om die
lae-frekwensie golwe in die kusarea te bereken. Lae-frekwensie golwe is die hoofoorsaak van
skipbewegings en daarom is dit belangrik om dit te ondersoek gedurende die
voorspellingsprosedure van skipbewegings.
Na die golwe bereken is, moes die kragte wat beide kort en lang golwe op die skip uitoefen ook
bereken word. Hierdie golfkragte is gemodelleer deur middel van die kombinasies SURFBEAT/LFSTRIP
(langgolwe) en SWAN/WAVESCAT (kortgolwe). LF-STRIP het die skakel tussen
golfmodelle en skipbewegingsmodelle verskaf en die lae-frekwensie golwe omgeskakel in
langgolfkragte wat op die skip uitgeoefen is. WAVESCAT het die hoë-frekwensiegolwe
omgeskakel in kortgolfkragte wat op die skip uitgeoefen is. Die berekende langgolf- en
kortgolfkragte is ingevoer op die skipbewegingsmodel QUAYSIM om die skipbewegings en
inperkingskragte in die bootbuffers en ankerlyne te bepaal sodat dit vergelyk kon word met die
fisiese model, met die doel om moontlik die QUAYSIM/WAVESCAT-prosedure om gemeerde
skipbewegings te voorspel te valideer.
Die studie verskaf ’n oorsig van die opstel en resultate van die fisiese en numeriese modelle. Elk
van die numeriese modelle SWAN, SURFBEAT, LF-STRIP, WAVESCAT en QUAYSIM word
beskryf en vergelykings word getref tussen die numeriese en fisiese modelle vir elke prosedure. Die validasieprosedure verskaf nuttige dokumentasie van die kwaliteit van hierdie numeriese
modeleringsprosedures wat reeds in sekere ontwerpprojekte gebruik word.
Die numeriese WAVESCAT en QUAYSIM modelle van skipbewegings het ’n goeie korrelasie
tussen die fisiese model en die numeriese benadering gelewer. Verbeteringe is wel steeds nodig.
Goeie vergelykings is verkry vir langgolfbewegings (horisontale bewegings – stuwing (“surge”),
swaai (“sway”) en gier (“yaw”)). Die stu- en swaaibewegings was effens oorskat met QUAYSIM.
Die grootte van die gier was wel vergelykbaar maar is nie grafies goed uitgebeeld nie.

Identiferoai:union.ndltd.org:netd.ac.za/oai:union.ndltd.org:sun/oai:scholar.sun.ac.za:10019.1/96758
Date03 1900
CreatorsEigelaar, Lerika Susan
ContributorsToms, Geoff, Stellenbosch University. Faculty of Engineering. Dept. of Civil Engineering.
PublisherStellenbosch : Stellenbosch University
Source SetsSouth African National ETD Portal
Languageen_ZA
Detected LanguageEnglish
TypeThesis
Format154 pages : illustrations
RightsStellenbosch University

Page generated in 0.0029 seconds